1: %% IMPORTANT NOTE: Please read carefully all information preceded by % sign
2: %% This is a Sample File showing the right templates
3: %% for all AIMS Journals Prepared in AmS-LaTeX
4: %% You need to down load from http://aimsciences.org
5: %% the file "AIMS.cls"
6: %% and then replace "\documentclass{amsart}" with "\documentclass{AIMS}"
7: %% at the very beginning of your amslatex file
8:
9: \documentclass{AIMS}
10: \usepackage{amsmath}
11: \usepackage{epsfig} %% add this line and the next only if you have pictures
12: \usepackage{graphics} %% pictures should be in esp format
13: %For pictures: screened artwork should be set up with an 85 or 100 line screen
14:
15: \textheight=8.2 true in
16: \textwidth=5.0 true in
17: \topmargin 30pt
18: \setcounter{page}{1}
19:
20: \def\currentvolume{2}
21: \def\currentissue{2}
22: \def\currentyear{2007}
23: \def\currentmonth{June}
24: \def\ppages{193--210}
25:
26: \newtheorem{theorem}{Theorem}
27: \newtheorem{corollary}{Corollary}
28: \newtheorem*{main}{Main Theorem}
29: \newtheorem{lemma}{Lemma}
30: \newtheorem{proposition}{Proposition}
31: \newtheorem{conjecture}{Conjecture}
32: \newtheorem*{problem}{Problem}
33: \theoremstyle{definition}
34: \newtheorem{definition}{Definition}
35: \newtheorem{remark}{Remark}
36: \newtheorem*{notation}{Notation}
37: \newcommand{\ep}{\varepsilon}
38: \newcommand{\eps}[1]{{#1}_{\varepsilon}}
39:
40: \title[Self-Organized Network Flows]
41: {Self-Organized Network Flows}
42:
43: \author[Dirk Helbing, Jan Siegmeier, and Stefan L\"ammer]{}
44:
45: %\subjclass{Primary: 58F15, 58F17; Secondary: 53C35}
46: \keywords{Traffic flows, production, emergent oscillations, self-organized traffic lights, synchronization}
47:
48: \email{helbing@trafficforum.org}
49:
50: \begin{document}
51: \maketitle
52:
53: \centerline{\scshape Dirk Helbing,$^{1,2}$ Jan Siegmeier,$^1$ and Stefan L\"ammer$^1$}
54: \medskip
55: {\footnotesize
56: \centerline{$^1$ Institute for Transport \& Economics, Dresden University of Technology}
57: \centerline{Andreas-Schubert-Str. 23, 01062 Dresden, Germany} \centerline{$^2$ Collegium Budapest~-- Institute for Advance Study}
58: \centerline{Szenth\'{a}roms\'{a}g utca 2, 1014 Budapest, Hungary}
59: }
60:
61: \medskip
62:
63: \centerline{(Communicated by Aim Sciences)}
64: \medskip
65:
66: \begin{abstract}
67: A model for traffic flow in street networks or material flows in supply networks is presented, that takes into account the conservation of
68: cars or materials and other significant features of traffic flows such as jam formation, spillovers, and load-dependent transportation times.
69: Furthermore, conflicts or coordination problems of intersecting or merging flows are considered as well. Making assumptions regarding
70: the permeability of the intersection as a function of the conflicting flows and the queue lengths, we find self-organized oscillations in the
71: flows similar to the operation of traffic lights.
72: \end{abstract}
73:
74: \section{Introduction}
75:
76: Material flows are found in many places of the world. This concerns, for example, traffic flows in urban areas or flows of commodities in
77: logistic systems. There is also some similarity with material flows in production or biological systems, from cells over bodies upto
78: ecological food chains. Many of these material flows are not of diffusive nature or going on in continuous space. They are often directed and
79: organized in networks. In comparison with data flows in information networks, however, there are conservation laws, which can be used to set
80: up equations for material flows in networks. It turns out, however, that this is not a trivial task. While there is already a controversial
81: discussion about the correct equations representing traffic flows along road sections \cite{Daganzo,RMP,mitSchoenhof,Kernerbuch}, their
82: combination in often complex and irregular networks poses further challenges. In particular, there have been several publications on the
83: treatment of the boundary conditions at nodes (connections) of several network links (i.e. road sections)
84: \cite{Hilliges,CellTransmission,Lebacque,control,Klar,Piccoli1,Piccoli2,Herty3,Herty2,Herty1}.
85: In particular, the modelling of merging and intersecting flows
86: is not unique, as there are many possible forms of organization, including the use of traffic lights. Then, however, the question comes up
87: how these traffic lights should be operated, coordinated, and optimized. In order to address these questions, in Sec.~\ref{Sec2} we formulate
88: a simple model for network flows, which contains the main ingredients of material or traffic flows. Section \ref{Sec3} will then discuss the
89: treatment of diverges, merges, and intersections. Equations for the interaction-dependent permeability at merging zones and intersections
90: will be formulated in Sec.~\ref{Sec4}. We will see that, under certain conditions, they lead to spontaneous oscillations, which have features
91: similar to the operation of traffic lights. Finally, Sec.~\ref{Sec5} summarizes and concludes this paper.
92:
93: \section{Flows in Networks}\label{Sec2}
94:
95: The following section will start with a summary of the equations derived for traffic flows in networks in a previous paper. These equations
96: are based on the following assumptions:
97: \begin{itemize}
98: \item The road network can be decomposed into road sections of homogeneous capacity
99: (links) and nodes describing their connections.
100: \item The traffic dynamics along the links is sufficiently well described by the
101: Lighthill-Whitham model, i.e. the continuity equation for vehicle conservation and a flow-density relationship (``fundamental diagram'').
102: This assumes adiabatic speed adjustments, i.e. that acceleration and deceleration times can be neglected.
103: \item The parameters of vehicles such as the maximum speed $V_i^0$ and the safe time headway $T$
104: are assumed to be identical in the same road section, and who enters a road section first exits first (FIFO principle). That is, overtaking
105: is assumed to be negligible.
106: \item The fundamental diagram can be well approximated by a triangular shape, with an
107: increasing slope $V_i^0$ at low densities and a decreasing slope $c$ in the congested regime. This implies two constant characteristic
108: speeds: While $V_i^0$ corresponds to the free speed or speed limit on road section $i$,
109: \begin{equation}
110: - c = -\frac{1}{\rho_{\rm max} T}
111: \end{equation}
112: the dissolution speed of the downstream front of a traffic jam and the velocity of upstream propagation of perturbations in congested traffic.
113: While $\rho_{\rm max}$ denotes the maximum vehicle density in vehicle queues, $T\approx 1.8$s is the safe time gap between two successive
114: vehicles.
115: \item The vehicle density in traffic jams is basically constant.
116: \end{itemize}
117: These assumptions may be compensated for by suitable corrections \cite{RMP}, but already the model below displays a rich spectrum of
118: spatio-temporal behaviors and contains the main elements of traffic dynamics we are interested in here.
119:
120: \subsection{Flow Conservation Laws}\label{laws}
121:
122: In the following, we will introduce our equations for traffic flows in networks very shortly, as a detailed justification and derivation has
123: been given elsewhere \cite{JPhysA,TGF03,control}. These equations are also meaningful for pipeline networks \cite{pipelines} (if
124: complemented by equations for momentum conservation), logistic
125: systems \cite{logistics}, or supply networks \cite{supply}. Our notation is illustrated in Fig.~\ref{Fig1}.
126: \par
127: Compared to Ref.~\cite{control}, we will use a simplified notation,
128: here.\footnote{The arrival flow $A_j(t)$ has previously
129: been denoted by $Q_j^{\rm arr}(t)$, the potential arrival flow $\widehat{A}_j(t)$ by $Q_j^{\rm arr, pot}(t)$,
130: the departure flow $O_j(t)$ by $Q_j^{\rm dep}(t)$ and the potential departure flow $\widehat{O}_j(t)$ by
131: $Q_j^{\rm dep,pot}(t)$.} The {\em arrival flow} $A_j(t)$ denotes the actual inflow of vehicles into the upstream
132: end of road section $j$, while $O_j(t)$ is the actual {\em departure flow}, i.e. the flow of vehicles
133: leaving road section $j$ at its downstream end. The quantity
134: \begin{equation}
135: \widehat{Q}_j = \left( T + \frac{1}{V_j^0\rho_{\rm max}} \right)^{-1}
136: = \frac{\rho_{\rm max}}{1/c + 1/V_j^0}
137: \end{equation}
138: represents the maximum in- or outflow of road section $j$.
139: All the above quantities refer to flows {\it per lane}.
140: $I_j$ is the number of lanes and $L_j$ the length of road section $j$.
141: $l_j(t)\le L_j$ is the length of the congested area on link $j$ (measured from the
142: downstream end),
143: and $\Delta N_j$ is the number of stopped or delayed vehicles, see Eqs. (\ref{shock}) and (\ref{delayed}).
144: With these definitions, we can formulate constraints for the {\em actual} arrival and departure flows,
145: which are given by the {\em potential arrival flows} $\widehat{A}_j(t)$ and the {\em potential
146: departure flows} $\widehat{O}_i(t)$, respectively.
147: \par
148: The actual arrival flow $A_j(t)$ is limited by the maximum inflow $\widehat{Q}_j$, if
149: road section $j$ is not fully congested ($l_j(t) < L_j$). Otherwise (if $l_j=L_j$) it is limited
150: by the actual departure flow $O_j(t-L_j/c)$ a time period $L_j/c$ before,
151: as it requires this time period until the downstream flow value has propagated upto the upstream end
152: of the road section by forward movement of vehicles under congested traffic conditions.
153: This implies
154: \begin{equation}
155: 0 \le A_j(t) \le \widehat{A}_j(t) := \left\{
156: \begin{array}{ll}
157: \widehat{Q}_j & \mbox{if } l_j(t) < L_j \\
158: O_j(t-L_j/c) & \mbox{if } l_j(t) = L_j.
159: \end{array}
160: \right. \label{one}
161: \end{equation}
162: Moreover, the potential departure flow $\widehat{O}_i(t)$ of road section $i$
163: is given by its permeability $\gamma_i(t)$ times the
164: maximum outflow $\widehat{Q}_i$ from this road section,
165: if vehicles are queued up ($\Delta N_i >0$) and waiting to leave.
166: Otherwise (if $\Delta N_i =0$) the outflow is limited by the permeability times the
167: arrival flow $A_i$ a time period $L_i/V_i^0$ before,
168: as this is the time period that entering vehicles need to reach the end of road section $i$
169: when moving freely at the speed $V_i^0$.
170: This gives the additional relationship
171: \begin{equation}
172: 0 \le O_i(t) \le \widehat{O}_i(t) := \gamma_i(t) \left\{
173: \begin{array}{ll}
174: A_i(t-L_i/V_i^0) & \mbox{if } \Delta N_i(t) = 0 \\
175: \widehat{Q}_i & \mbox{if } \Delta N_i(t) > 0 \, .
176: \end{array}
177: \right. \label{two}
178: \end{equation}
179: The permeability $\gamma_i(t)$ for traffic flows at the downstream end of section $i$
180: can assume values between 0 and 1. In case of a traffic light, $\gamma_i(t) = 1$ corresponds
181: to a green light for road section $i$, while $\gamma_i(t) =0$ corresponds to a red or amber
182: light.
183: \par
184: Alternatively and shorter than Eqs.~(\ref{one}) and (\ref{two}) one can write
185: \begin{equation}
186: \widehat{A}_j(t) = \max\Big[ \widehat{Q}_j \Theta(l_j(t) < L_j),O_j(t-L_j/c)\Big]
187: \label{easy1}
188: \end{equation}
189: and
190: \begin{equation}
191: \widehat{O}_i(t) = \gamma_i(t) \max \Big[\widehat{Q}_i\Theta(\Delta N_i > 0), A_i(t-L_i/V_i^0)\Big] \, ,
192: \label{easy2}
193: \end{equation}
194: where the Heaviside function $\Theta$ is 1, if the argument (inequality) has the logical value ``true'', otherwise it is 0. Note
195: that the above treatment of the traffic flow in a road section requires the specification of the boundary conditions only, as we have
196: integrated up Lighthill's and Whitham's partial differential equation over the length of the road section. The dynamics in the inner part of
197: the section can be easily reconstructed from the boundary conditions thanks to the constant characteristic speeds. However, a certain point
198: of the road section may be determined either from the upstream boundary (in the case of free traffic) or by the downstream boundary (if lying
199: in the congested area, i.e. behind the upstream congestion front). Therefore, we have a switching between the influence of the upstream and
200: the downstream boundary conditions, which makes the dynamics both, complicated and interesting. This switching results from the maximum
201: functions above and implies also that material flows in networks are described by hybrid equations. Although the dynamics is determined by
202: linear ordinary differential equations in all regimes, the switching between the regimes can imply a complex dynamics and even deterministic
203: chaos \cite{Peters}.
204: \par
205: Complementary to the above equations, we have now to specify the constraints for the nodes, i.e. the connection, merging, diverging or
206: intersection points of the homogeneous road sections. Let the ingoing links be denoted by the index $i$ and the outgoing ones by $j$. To
207: distinguish quantities more easily when we insert concrete values $1,2,\dots$ for $i$ and $j$, we mark quantities of outgoing links
208: additionally by a prime (${}'$).
209: \par
210: Due to the condition of flow conservation, the arrival flow into a road section $j$ with $I'_j$ lanes must agree with the sum of the
211: fractions $\alpha_{ij}$ of all outflows $I_iO_i(t)$ turning into road section $i$. Additionally, the arrival flows are limited, i.e. we have
212: \begin{equation}
213: I'_jA'_j(t) = \sum_i I_iO_i(t) \alpha_{ij} \le I'_j\widehat{A}'_j(t)
214: \label{inequal}
215: \end{equation}
216: for all $j$. Of course, the turning fractions $\alpha_{ij}\ge 0$ are normalized due to flow conservation:
217: \begin{equation}
218: \sum_j \alpha_{ij}(t) = 1 \, .
219: \end{equation}
220: \par
221: In cases of no merging flows, Eq.~(\ref{inequal}) simplifies to
222: \begin{equation}
223: I'_jA'_j(t) = I_iO_i(t) \alpha_{ij} \le I'_j\widehat{A}'_j(t)
224: \end{equation}
225: for all $j$. At the same time, $0 \le O_i(t) \le \widehat{O}_i(t)$ must be fulfilled for all $i$. Together, this implies
226: \begin{equation}
227: O_i(t) \le \min \left[ \widehat{O}_i(t) , \min_j \left( \frac{I'_j \widehat{A}'_j}{I_i\alpha_{ij}} \right) \right] \label{nomerge}
228: \end{equation}
229: for all $i$.
230: \par
231: The advantage of the above model is that it contains the most important elements of the traffic dynamics in networks. This includes the
232: transition from free to congested traffic flows due to lack of capacity, the propagation speeds of vehicles and congested traffic, spillover
233: effects (i.e. obstructions when entering fully congested road sections) and, implicitly, load-dependent travel times as well.
234:
235: \subsection{Two Views on Traffic Jams}
236:
237: Let us study the traffic dynamics on the road sections in more detail.
238: Traffic jams can be handled in two different ways: First by
239: determining the number of cars that are delayed compared to free traffic or, second, by determining fronts and ends of traffic jams. The
240: former method is more simple, but it cannot deal correctly with spill-over effects, when the end of a traffic jam reaches the end of a road
241: section. Therefore, the first method is sufficient only in situations where the spatial capacity of road sections is never exceeded.
242:
243: \subsubsection{Method 1: Number of Delayed Vehicles}
244:
245: The first method just determines the difference between the number $N_i^{\rm in}(t)$ of vehicles that would reach the end of road section $i$
246: upto time $t$ and the number $N_i^{\rm out}(t)$ of vehicles that actually leave the road section upto this time. $N_i^{\rm in}(t)$ just
247: corresponds to the number of vehicles which have entered the road section upto time $t-L_i/V_i^0$, as $L_i/V_i^0$ is the free travel time.
248: This implies
249: \begin{equation}
250: N_i^{\rm in}(t) = \int\limits_0^t dt' \; A_i(t'-L_i/V_i^0) \, ,
251: \end{equation}
252: while the number of vehicles that have acually left the road section upto time $t$ is
253: \begin{equation}
254: N_i^{\rm out}(t) = \int\limits_0^t dt' \; O_i(t') \, .
255: \end{equation}
256: Hence, the number $\Delta N_i(t)$ of delayed vehicles is given by
257: \begin{equation}
258: \Delta N_i(t) = \int\limits_0^t dt' \; [A_i(t'-L_i/V_i^0) - O_i(t')] \ge 0 \, .
259: \end{equation}
260: Alternatively, one can use the following differential equation for the temporal change in the number of delayed vehicles:
261: \begin{equation}
262: \frac{d\,\Delta N_i}{dt} = A_i(t-L_i/V_i^0) - O_i(t) \, .
263: \label{delayed}
264: \end{equation}
265: In contrast, the number of {\em all} vehicles on road section $i$ (independently of whether they are delayed or not) changes in time
266: according to
267: \begin{equation}
268: \frac{dN_i}{dt} = A_i(t) - O_i(t) \, .
269: \end{equation}
270:
271: \subsubsection{Method 2: Jam Formation and Resolution}
272:
273: In our simple macroscopic traffic model, the formation and resolution of traffic jams is described by the shock wave equations, where we have
274: the two characteristic speeds $V_i^0$ (the free speed) and $c$
275: (the jam resolution speed). According to the theory of shock waves \cite{LW,Whitham}, the upstream end of a traffic jam, which is located at
276: a place $l_i(t)\ge0$ upstream of the end of road section $i$, is moving at the speed
277: \begin{equation}
278: \frac{dl_i}{dt} = - \frac{A_i\big(t-[L_i-l_i(t)]/V_i^0\big) - O_i\big(t-l_i(t)/c\big)
279: }{\rho_1(t) - \rho_2(t)} \, \label{shock}
280: \end{equation}
281: with the (free) density
282: \begin{equation}
283: \rho_1(t) = A_i\big(t-[L_i-l_i(t)]/V_i^0\big)/V_i^0
284: \end{equation}
285: immediately before the upstream shock front and the (congested) density
286: \begin{equation}
287: \rho_2(t) = [1-TO_i\big(t-l_i(t)/c\big)]\rho_{\rm max}
288: \end{equation}
289: immediately downstream of it. This is, because free traffic is upstream of the shock front, and congested traffic downstream of it (for
290: details see Eqs. (1.6) and (1.4) in Ref.~\cite{control}). In contrast, the downstream front of a traffic jam is moving at the speed
291: \begin{equation}
292: - \frac{0 - O_i\big(t-l_i(t)/c\big)
293: }{\rho_{\rm max} - O_i\big(t-l_i(t)/c\big)/V_i^0 } = \frac{O_i\big(t-l_i(t)/c\big) }{\rho_{\rm max} - O_i\big(t-l_i(t)/c\big)/V_i^0 }
294: \, ,
295: \end{equation}
296: since congested traffic with zero flow is upstream of the shock front and free traffic flow occurs downstream of it.
297:
298: \subsubsection{Comparison of the Two Methods}\label{compa}
299:
300: Let us discuss a simple example to make the differences of both descriptions clearer. For this, we assume that, at time $t=0$, traffic flow
301: on the overall road section $i$ is free, i.e. any traffic jam has resolved and there are no delayed vehicles. The flow shall be stopped by a
302: red traffic light for a time period $t_0$. At time $t=t_0$, the traffic light shall turn green, and the formed traffic jam shall resolve. For
303: the arrival flow, we simply assume a constant value $A_i$, and the road section shall be long enough to take up the forming traffic jam.
304: Moreover, the departure flow shall be $O_i$. Then, according to method 1, the number of delayed vehicles at time $t_0$ is
305: \begin{equation}
306: \Delta N_i(t_0) = A_i t_0 \, ,
307: \end{equation}
308: and it is reduced according to
309: \begin{equation}
310: \Delta N_i(t) = A_i t_0 - (O_i - A_i) (t-t_0) \, .
311: \end{equation}
312: Therefore, any delays are resolved after a time period
313: \begin{equation}
314: t-t_0 = \frac{A_it_0}{O_i - A_i} = \frac{\Delta N_i(t_0)}{O_i -A_i} \, ,
315: \label{dis1}
316: \end{equation}
317: i.e. at time
318: \begin{equation}
319: t_2 = t_0 \frac{O_i}{O_i - A_i} \, .
320: \label{acco}
321: \end{equation}
322: Afterwards, $\Delta N_i(t) = 0$.
323: \par
324: In contrast, the end of the traffic jam grows with the speed
325: \begin{equation}
326: \frac{dl_i}{dt} = - \frac{A_i - 0} {A_i/V_i^0 - (1-0)\rho_{\rm max}}
327: = \frac{1}{\rho_{\rm max}/A_i - 1/V_i^0} =: C_i \, .
328: \end{equation}
329: Therefore, we have $l_i(t_0) = C_i t_0$. Surprisingly, this is greater than $\Delta N_i(t_0)/\rho_{\rm max}$, i.e. the expected length of the
330: traffic jam based on the number of delayed vehicles. The reason is that the delay of a vehicle joining the traffic jam at location $x_i = L_i
331: - l_i$ is noticed at the downstream end of the road section only after a time period $l_i/V_i^0$.
332: \par
333: The resolution of the traffic jam starts from the downstream end with the speed
334: \begin{equation}
335: \frac{0-\widehat{Q}_i}{\rho_{\rm max} - \widehat{Q}_i/V_i^0} = \frac{-1}{\rho_{\rm max}/\widehat{Q}_i - 1/V_i^0} = -c \, ,
336: \end{equation}
337: if the outflow is free (i.e. $O_i = \widehat{Q}_i$), otherwise with the speed
338: \begin{equation}
339: \frac{0-O_i}{\rho_{\rm max} - (\rho_{\rm max} - O_i/c)} = - c \, ,
340: \end{equation}
341: since congested traffic with zero flow and maximum density is upstream of the shock front.
342: \par
343: Obviously, the jam resolution has reached the further growing, upstream jam front when $C_it = c(t-t_0)$.
344: Therefore, the jam of density $\rho_{\rm max}$ has
345: disappeared after a time period $t-t_0 = C_it_0/(c - C_i)$, i.e. at time
346: \begin{equation}
347: t_1 = ct_0/(c - C_i) \, .
348: \end{equation}
349: Surprisingly, it can be shown that $t_1 < t_2$, i.e. the traffic jam resolves before the number of delayed
350: vehicles reaches a value of zero. In fact, it still
351: takes the time $C_it_1/\widehat{V}_i^0$ until the last delayed vehicle has left the road section, where
352: \begin{equation}
353: \widehat{V}_i^0 = \frac{A_i -O_i}{A_i/V_i^0 - (\rho_{\rm max} - O_i/c)}
354: \end{equation}
355: is the shock front between free upstream traffic flow and the congested outflow $O_i$,
356: which usually differs from the speed $V_i = O_i/[(1-TQ_i)\rho_{\rm max}]$ of outflowing vehicles.
357: For $O_i = \widehat{Q}_i$, we have $\widehat{V}_i^0 = V_i^0$ because of
358: $1/c = \rho_{\rm max}/\widehat{Q}_i - 1/V_i^0$.
359: \par
360: Undelayed traffic starts when this shock front reaches the end of the road section, i.e. at time
361: \begin{equation}
362: t_2 = t_1 \left(1 + \frac{C_i}{\widehat{V}_i^0}\right)
363: = \frac{t_0}{1 - C_i/c} \left( 1 + \frac{C_i(A_i/V_i^0-\rho_{\rm max}) + C_iO_i/c}{A_i - O_i} \right) \, .
364: \end{equation}
365: Inserting $C_i(A_i/V_i^0-\rho_{\rm max}) = - A_i$ eventually gives $t_2 = t_0 O_i/O_i - A_i)$.
366: This agrees perfectly with the above result for the first method (based on vehicle delays rather than traffic jams).
367: \par
368: In conclusion, both methods of dealing with traffic jams are consistent, and delayed vehicles occur as soon as traffic jam formation begins.
369: However, according to method 1, a queued vehicle at position $x_i = L_i - l_i$
370: is counted as delayed only after an extra time period
371: $l_i/V_i^0$, but it is counted as undelayed after the same extra time period. This is because method 1 counts on the basis of vehicle
372: arrivals at the downstream end of road section $i$.
373: \par
374: As it is much simpler to use the method 1 based on determining the number of delayed vehicles than using method 2 based on determining the
375: movement of shock fronts, we will use method 1 in the following. More specifically, in Eq.~(\ref{one}) we will replace $l_j(t) < L_j$ by
376: $\Delta N_j(t) < N_j^{\rm max} := L_j \rho_{\rm max}$ and $l_j(t) = L_j$ by
377: $\Delta N_j(t) = N_j^{\rm max}$. This corresponds to a situation in which the vehicles would not queue up along the
378: road section, but at the downstream end of the road section, like in a wide parking lot or on top of each other. As long as road section $j$
379: is not fully congested, this difference does not matter significantly. If it is fully congested, the dynamics will potentially be
380: different, defining a modified model of material network flows. However both, the original and the
381: modified model fulfill the conservation equation and show spillover effects.
382:
383: \subsubsection{Calculation of Cumulative and Maximum Individual Waiting Times}
384:
385: In Ref. \cite{JPhysA}, we have derived a delay differential equation to determine the travel time $T_i(t)$ of a vehicle entering road section
386: $i$ at time $t$ (see also Ref. \cite{Astarita1995,Astarita2002,Carey2003}):
387: \begin{equation}
388: \frac{dT_i(t)}{dt} = \frac{A_i(t)}{O_i(t+T_i(t))} - 1 \, .
389: \label{travtime}
390: \end{equation}
391: According to this, the travel time $T_i(t)$ increases with time, when the arrival rate $A_i$ at the time $t$ of entry exceeds the departure
392: rate $O_i$ at the leaving time $t+T_i(t)$, while it decreases when it is lower. It is remarkable that this formula does not explicitly depend
393: on the velocities on the road section, but only on the arrival and departure rates.
394: \par
395: Another method to determine the travel times is to integrate up over the number of vehicles arriving in road section $i$,
396: \begin{equation}
397: N_i^A(t) = \int\limits_0^t dt' \; A_i(t') = N_i^{\rm in}(t+L_i/V_i^0) \, ,
398: \end{equation}
399: and over the number of vehicles leaving it,
400: \begin{equation}
401: N_i^O(t) = \int\limits_0^t dt' \; O_i(t') = N_i^{\rm out}(t) \, ,
402: \end{equation}
403: starting at at time $t=0$ when there are no vehicles in the road. If $T'_i(t)$ denotes the time at which $N_i^O(t+T'_i(t)) = N_i^A(t)$, then
404: $T'_i(t)$ is the travel time of a vehicle entering road section $i$ at time $t$ and
405: \begin{equation}
406: T_i(t) = T'_i(t) - L_i/V_i^0
407: \end{equation}
408: is its waiting time.
409: \par
410: Another interesting quantity is the cumulative waiting time $T_i^{\rm c}(t)$, which is determined by integrating up over the number $\Delta
411: N_i$ of all delayed vehicles. We obtain
412: \begin{eqnarray}
413: T_i^{\rm c}(t) &=& \int\limits_0^t dt' \; \Delta N_i(t') = \int\limits_0^t dt'\; [N_i^{\rm in}(t'-L_i/V_i^0) - N_i^{\rm out}(t')]
414: \nonumber \\
415: &=& \int\limits_0^t dt' \int\limits_0^{t'} dt^{\prime\prime} \; [A_i(t^{\prime\prime}-L_i/V_i^0) - O_i(t^{\prime\prime}) ]
416: \end{eqnarray}
417: and the differential equation
418: \begin{equation}
419: \frac{dT_i^{\rm c}(t)}{dt} = \Delta N_i(t) = \int\limits_0^t dt' \; [A_i(t'-L_i/V_i^0) - O_i(t') ] \, .
420: \end{equation}
421: For a constant arrival flow $A_i$ and a red traffic light from $t=0$ to $t=t_0$ (i.e. $O_i(t) =0$), we find
422: \begin{equation}
423: T_i^{\rm c} = \frac{A_i t^2}{2} \, .
424: \end{equation}
425: In this time period, a number of $N_i(t) = A_i t$ vehicles accumulates, which gives an average waiting time of
426: \begin{equation}
427: \frac{T_i^{\rm c}(t_0)}{\Delta N(t_0)} = \frac{t_0}{2}
428: \end{equation}
429: at the end of the red light. The first vehicle has to wait twice as long, namely, a time period $t_0$.
430:
431: \section{Treatment of Merging, Diverging and Intersection Points}\label{Sec3}
432:
433: While the last section has given general formulas that must be fulfilled at nodes connecting two or more different links, in the following we
434: will give some concrete examples, how to deal with standard elements of street networks. For previous
435: treatments of traffic flows at intersections see, for example, Refs.~\cite{Piccoli1,Hilliges,CellTransmission,Lebacque}.
436: \begin{figure}[htbp]
437: \begin{center}
438: \includegraphics[width=13cm, angle=0]{crossing}
439: \end{center}
440: \caption[]{Schematic illustration of the (a) diverging, (b) merging, and (c) intersecting flows
441: discussed in this paper.\label{Fig1}}
442: \end{figure}
443: \subsection{Diverging Flows: One Inflow and Several Outflows}
444:
445: In the case of one road section $i$ diverging into several road sections $j$ (see Fig.~\ref{Fig1}a),
446: Eqs.~(\ref{nomerge}) and (\ref{easy1}) to
447: (\ref{inequal}) imply
448: \begin{eqnarray}
449: O_i(t) &\le& \min \left\{ \gamma_i(t) \max \left[\widehat{Q}_i\Theta(\Delta N_i > 0), A_i\left(t-\frac{L_i}{V_i^0}\right)\right] ,\right.
450: \nonumber \\
451: & & \qquad\quad \left. \min_j \left[ \frac{I'_j}{I_i\alpha_{ij}} \max \left( \widehat{Q}_j \Theta(l_j < L_j),
452: O_j(t- L_j / c) \right) \right] \right\}
453: \end{eqnarray}
454: for all $i$. If we assume that downstream road sections are never completely congested, this simplifies to
455: \begin{equation}
456: O_i(t) = \min \left\{ Q_i, \gamma_i \max \left[\widehat{Q}_i\Theta(\Delta N_i > 0),
457: A_i\left(t-L_i/V_i^0 \right)\right]
458: \right\}
459: \end{equation}
460: with
461: \begin{equation}
462: Q_i = \min_j \left( \frac{I'_j\widehat{Q}_j}{I_i\alpha_{ij}} \right) \, .
463: \end{equation}
464: Otherwise
465: \begin{equation}
466: Q_i(t) = \min_j \left[ \max\left( \frac{I'_j\widehat{Q}_j}{I_i\alpha_{ij}}\Theta(l_j < L_j) ,
467: \frac{I'_jO_j(t-\frac{L_j}{c})}{I_i\alpha_{ij}} \right)\right] \, . \label{otherwise}
468: \end{equation}
469:
470: \subsection{Merging Flows: Two Inflows and One Outflow}
471:
472: We assume a flow $I_1O_1(t)$ that splits into two flows $I_1O_1(t)\alpha_{11} $ (going straight) and $I_1O_1(t)\alpha_{12} $ (turning right),
473: but a right-turning flow $I_2O_2(t)$ merging with flow $I_1O_1(t) \alpha_{11}$, as in turn-right-on-red setups (see Fig.~\ref{Fig1}b). For
474: this situation, we have the equations
475: \begin{eqnarray}
476: I'_1A'_1(t) &=& I_1O_1(t) \alpha_{11} + I_2O_2(t) \le I'_1\widehat{A}'_1(t) \, , \\
477: I'_2A'_2(t) &=& I_1O_1(t) \alpha_{12} \le I'_2\widehat{A}'_2(t) \, .
478: \end{eqnarray}
479: One can derive
480: \begin{equation}
481: 0 \le O_1 = \min\Big[ \widehat{O}_1(t), \frac{I'_1\widehat{A}'_1(t)-I_2O_2(t)}{I_1\alpha_{11}},
482: \frac{I'_2\widehat{A}'_2(t)}{I_1\alpha_{12}}\Big]
483: \end{equation}
484: and
485: \begin{equation}
486: 0 \le O_2 = \min\Big[ \widehat{O}_2(t), \frac{I'_1\widehat{A}'_1(t)-I_1O_1(t)\alpha_{11}}{I_2} \Big] \, .
487: \end{equation}
488: Let us set
489: \begin{equation}
490: O_1 = \min\Big[\widehat{O}_1, \frac{I'_1\widehat{A}'_1(t)}{I_1\alpha_{11}},
491: \frac{I'_2\widehat{A}'_2(t)}{I_1\alpha_{12}}\Big] \label{speci}
492: \end{equation}
493: and
494: \begin{equation}
495: O_2(O_1) = \min\Big[ \widehat{O}_2(t), \frac{I'_1\widehat{A}'_1(t)-I_1O_1\alpha_{11}}{I_2} \Big] \, .
496: \label{maxi}
497: \end{equation}
498: Then, it can be shown that $O_2(t) \ge 0$ and $O_1(t) \le [I'_1\widehat{A}'_1(t)-I_2O_2(t)]/(I_1\alpha_{11})$, as demanded. If $O_1(t)$ is
499: chosen a value $\Delta O_1$ smaller than specified in Eq.~(\ref{speci}), but $O_2$ is still set to the maximum related value $O_2(O_1-\Delta
500: O_1)$ according to Eq.~(\ref{maxi}), the overall flow
501: \begin{equation}
502: F=I_1O_1+I_2O_2
503: \end{equation}
504: is reduced as long as $\alpha_{11} < 1$, since this goes along with additional turning flows (while the number of lanes does not matter!).
505: Therefore, it is optimal to give priority to the outflow $O_1(t)$ according to Eq.~(\ref{speci}) and to add as much outflow $O_2(t)$ as
506: capacity allows. This requires suitable flow control measures, otherwise the optimum value of the overall flow $F$ could not be reached. In
507: fact, the merging flow would ``steel'' some of the capacity reserved for the ``main'' flow ($i=1$), which would reduce the possible outflow
508: $O_1(t)$ and potentially cause a breakdown of free traffic flow, as it is known from on-ramp areas of freeways \cite{mitTreiber} .
509:
510: \subsection{A Side Road Merging with a Main Road}
511:
512: Compared to the last section, the situation simplifies, if we have just a side road or secondary turning flow merging with a the flow of a
513: main road without any turning flow away from the main road. In this case, we have $\alpha_{11}=1$ and $\alpha_{12} =0$, which leaves us with
514: the relationships
515: \begin{equation}
516: O_1 = \min\Big[\widehat{O}_1, \frac{I'_1\widehat{A}'_1(t)}{I_1}\Big]
517: \end{equation}
518: and
519: \begin{equation}
520: O_2(O_1) = \min\Big[ \widehat{O}_2(t), \frac{I'_1\widehat{A}'_1(t)-I_1O_1}{I_2} \Big] \, .
521: \end{equation}
522: according to Eqs.~(\ref{speci}) and (\ref{maxi}).
523:
524: \subsection{Intersection-Free Designs of Road Networks}
525:
526: With the formulas for the treatment of merges and diverges in the previous sections, it is already possible
527: to simulate intersection-free designs of urban road networks, which do not need any traffic light control.
528: The most well-known design of intersection-free nodes are roundabouts (see the upper left illustration in Fig.~\ref{free}). It is,
529: however, also possible to construct other intersection-free designs based on subsequent merges and
530: diverges of flows with different destinations. Two examples are presented in Fig.~\ref{free}b and c.
531: \par
532: \begin{figure}[htbp]
533: \begin{center}
534: \includegraphics[width=5.5cm, angle=0]{free1}\hspace*{1cm}
535: \includegraphics[width=6cm, angle=0]{cheese}\\
536: \includegraphics[width=6cm, angle=0]{hybrid}
537: \end{center}
538: \caption[]{Three examples for intersection-free designs of urban road networks.\label{free}}
539: \end{figure}
540: Although intersection-free designs require the driver to take small detours, such a road network will
541: normally save travel time and fuel, given that the traffic volume is not too low. This is because intersections then
542: need to be signalized in order to be safe and efficient.\footnote{Of course, a first-come-first-serve or right-before-left
543: rule will be sufficient at small traffic volumes.}
544: Traffic signals, however, imply that vehicles will often be stopped for considerable time intervals.
545: This causes significant delays, at least for vehicles not being served by a green wave. Intersection-free
546: designs, in contrast, do not necessarily require vehicles to stop. Therefore, the average speeds are expected to be
547: higher and the travel times lower than for road networks with intersections. This has significant
548: implications for urban transport planning, if intersections cannot be avoided by bridges or tunnels.
549:
550: \subsection{Two Inflows and Two Outflows}
551:
552: The treatment of intersecting flows is more complicated than the treatment of merges and diverges.
553: Moreover, the resulting flows are only uniquely defined, if additional rules are introduced such as
554: the optimization of the overall flow.
555: Let us here treat the case of an intersection with two inflows and two outflows (see Fig.~\ref{Fig1}c).
556: Equation (\ref{easy1}) implies the inequalities
557: \begin{eqnarray}
558: & & 0 \le I'_1A'_1(t) = I_1O_1(t) \alpha_{11} + I_2O_2(t) \alpha_{21} \le I'_1 \widehat{A}'_1(t) \, , \nonumber \\
559: & & 0 \le I'_2A'_2(t) = I_1O_1(t) \alpha_{12} + I_2O_2(t) \alpha_{22} \le I'_2 \widehat{A}'_2(t)
560: \label{constraints}
561: \end{eqnarray}
562: with the constraints
563: \begin{eqnarray}
564: & & 0 \le O_1(t) \le \widehat{O}_1(t) \, , \nonumber \\
565: & & 0 \le O_2(t) \le \widehat{O}_2(t) \, , \label{rectangle}
566: \end{eqnarray}
567: so that $I'_jA'_j(t) \ge 0$ is automatically fulfilled. The constraints (\ref{rectangle}) define an rectangular area of possible $O_i$-values
568: in the $O_1$-$O_2$ plane, where the size of the rectangle varies due to the time-dependence of $\widehat{O}_i(t)$. The inequalities
569: (\ref{constraints}) can be rewritten as
570: \begin{equation}
571: O_2(t) \le \frac{I'_1\widehat{A}_1(t) - I_1O_1(t)\alpha_{11}}{I_2\alpha_{21}}
572: =: a_1 - b_1 O_1(t) \, , \label{const1}
573: \end{equation}
574: and
575: \begin{equation}
576: O_2(t) \le \frac{I'_2\widehat{A}_2(t) - I_1O_1(t)\alpha_{12}}{I_2\alpha_{22}}
577: =: a_2 - b_2 O_1(t) \, . \label{const2}
578: \end{equation}
579: They potentially cut away parts of this rectangle, and the remaining part defines the convex set of feasible points $(O_1,O_2)$ at time $t$.
580: We are interested to identify the ``optimal'' solution $(O_1^*,O_2^*)$, which maximizes the overall flow
581: \begin{equation}
582: \sum_j I'_j A'_j(t) = \sum_i I_i O_i(t) \, .
583: \end{equation}
584: As this defines a linear optimization problem, the optimal solution corresponds to one of the corners of the convex set of feasible points,
585: namely the one which is touched first by the line
586: \begin{equation}
587: O_2 = \frac{Z - I_1 O_1}{I_2} \, ,
588: \label{goal1}
589: \end{equation}
590: when we reduce $Z$ from high to low values.
591: \par
592: Let us, therefore, determine all possible corners of the convex set and the conditions, under which they correspond to the optimal solution.
593: We will distinguish the following cases:
594: \begin{itemize}
595:
596: \item[(a)] {\it None} of the boundary lines (\ref{const1}) and (\ref{const2}) corresponding to the equality signs
597: cuts the rectangle defined by $0 \le O_1(t) \le \widehat{O}_1(t)$ and $0 \le O_2(t) \le \widehat{O}_2(t)$ in more than 1 point. This case
598: applies, if $a_1 - b_1\widehat{O}_1 \ge \widehat{O}_2$ and $a_2 - b_2\widehat{O}_1 \ge \widehat{O}_2$, as $a_i\ge 0$ and $b_i \ge 0$ implies
599: that both lines are falling or at least not increasing. Since the line (\ref{goal1}) reflecting the goal function is falling as well, the
600: optimal point is
601: \begin{equation}
602: (O_1^*,O_2^*) = (\widehat{O}_1,\widehat{O}_2) \, ,
603: \end{equation}
604: i.e. the outer corner of the rectangle corresponding to the potential or maximum possible departure flows (see Fig.~\ref{Fig2}).
605: \begin{figure}[htbp]
606: \begin{center}
607: \includegraphics[width=6cm, angle=0]{simplex}
608: \end{center}
609: \caption[]{Illustration of the possible optimal solutions for two intersecting flows (see text for details).\label{Fig2}}
610: \end{figure}
611: \item[(b)] Only {\it one} of the two boundary lines/border lines,
612: $O_2(t) = a_1-b_1O_1(t)$ or $O_2(t) = a_2 - b_2O_1(t)$, cuts the rectangle in {\it more} than one point. Let us assume, this holds for line
613: $i$, i.e. $a_i - b_i\widehat{O}_1 < \widehat{O}_2$. Then, the left cutting point
614: \begin{equation}
615: \qquad (O_1^{i{\rm l}},O_2^{i{\rm l}}) = \left\{
616: \begin{array}{ll}
617: \Big((a_i - \widehat{O}_2)/b_i,\widehat{O}_2\Big) & \mbox{if } a_i > \widehat{O}_2 \, ,\\
618: (0,a_i) & \mbox{otherwise}
619: \end{array}\right.
620: \end{equation}
621: is the optimal point if $I_1/I_2 < b_i$, i.e. if the slope $I_1/I_2$ of the goal function (\ref{goal1}) is smaller than the one of the
622: cutting border line. Otherwise, if $I_1/I_2 > b_i$, the optimal point is given by the right cutting point
623: \begin{equation}
624: \qquad (O_1^{i{\rm r}},O_2^{i{\rm r}}) = \left\{
625: \begin{array}{ll}
626: (\widehat{O}_1,a_i-b_i\widehat{O}_1) & \mbox{if } a_i > b_i\widehat{O}_1 \, , \\
627: (a_i/b_i,0) & \mbox{otherwise}
628: \end{array}\right.
629: \end{equation}
630: (see Fig.~\ref{Fig2}).
631:
632: \item[(c)] If {\it both} border lines cut through the rectangle, but one of them lies above the other line,
633: then only the lower line determines the optimal solution, which can be obtained as in case (b). Case (c) occurs if $a_2-b_2O_1^{1{\rm l}} >
634: a_1-b_1O_1^{1{\rm l}}$ and $a_2-b_2O_1^{1{\rm r}} > a_1-b_1O_1^{1{\rm r}}$ (line 1 is the lower one) or if $a_2-b_2O_1^{1{\rm l}} <
635: a_1-b_1O_1^{1{\rm l}}$ and $a_2-b_2O_1^{1{\rm r}} < a_1-b_1O_1^{1{\rm r}}$ (line 2 is the lower one).
636:
637: \item[(d)] The boundary lines cut each other in the inner part of the rectangle. This occurs if
638: $a_1 - b_1\widehat{O}_1 < \widehat{O}_2$ and $a_2 - b_2\widehat{O}_1 < \widehat{O}_2$. Then, the left-most cutting point $ (O_1^{i{\rm
639: l}},O_2^{i{\rm l}})$ is the optimal solution, if the slope $I_1/I_2$ of the goal function is smaller than the smallest slope of the two
640: boundary lines, while it is the lower right cutting point $ (O_1^{i{\rm r}},O_2^{i{\rm r}})$, if $I_1/I_2$ is greater than the steepest slope
641: of the two boundary lines, otherwise, the cutting point of the two boundary lines,
642: \begin{equation}
643: (O'_1,O'_2) = \left( \frac{a_2 - a_1}{b_2-b_1} , \frac{a_1b_2 - b_1a_2}{b_2-b_1} \right)
644: \end{equation}
645: is the optimal point (see Fig.~\ref{Fig2}). Mathematically speaking, we have
646: \begin{equation}
647: (O_1^*,O_2^*) = \left\{
648: \begin{array}{ll}
649: (O_1^{\rm 1l},O_2^{\rm 1l}) & \mbox{if } I_1/I_2 < b_1 < b_2, \\
650: (O'_1,O'_2) & \mbox{if } b_1 < I_1/I_2 < b_2, \\
651: (O_1^{\rm 2r},O_2^{\rm 2r}) & \mbox{if } b_1 < b_2 < I_1/I_2, \\
652: (O_1^{\rm 2l},O_2^{\rm 2l}) & \mbox{if } I_1/I_2 < b_2 < b_1, \\
653: (O'_1,O'_2) & \mbox{if } b_2 < I_1/I_2 < b_1, \\
654: (O_1^{\rm 1r},O_2^{\rm 1r}) & \mbox{if } b_2 < b_1 < I_1/I_2,
655: \end{array}\right.
656: \end{equation}
657: \end{itemize}
658: \par
659: It is astonishing that the simple problem of two intersecting traffic flows has so many different optimal solutions, which sensitively depend
660: on the parameter values. This can reach from situations where both outgoing road sections experience the maximum possible outflows upto
661: situations, where the outflow in the system-optimal point becomes zero for one of the road sections. A transition from one optimal solution
662: to another one could easily be triggered by changes in the turning fractions $\alpha_{ij}$ entering the parameters $a_i$ and $b_i$, for
663: example due to time-dependent turning fractions $\alpha_{ij}(t)$.
664:
665: \subsection{Inefficiencies due to Coordination Problems}
666:
667: An interesting question is how to actually establish the flows corresponding to the system optima that were determined in the previous sections on
668: merging and intersecting flows. Of course, zero flows can be enforced by a red traffic light, while maximum possible flows can be established
669: by a node design giving the right of way to one road (the ``main'' road). However, it is not so easy to support an optimimum point
670: corresponding to mixed flows, such as $(O'_1,O'_2)$. That would need quite tricky intersection designs or the implementation of an
671: intelligent transportation system ensuring optimal gap usage, e.g. based on intervehicle communication. Only in special cases, the task could
672: be performed by a suitable traffic light control.
673: \par
674: In normal merging or intersection situations, there will always be coordination problems \cite{mitJohansson} when entering or crossing
675: another flow, if the traffic volumes reach a certain level. This will cause inefficiencies in the usage of available road capacity, i.e. mixed flows will not
676: be able to use the full capacity. Such effects can be modelled by specifying the corresponding permeabilities $\gamma_i(t)$ as a function of
677: the merging flows, particularly the main flow or crossing flow. The deviation of $\gamma_i(t)$ from 1 will then be a measure for the
678: inefficiency. A particularly simple, phenomenological specification would be
679: \begin{equation}
680: \gamma_2(t) =
681: \frac{1}{1+ a\mbox{e}^{b(O_1-O_2)}} \, ,
682: \label{phen}
683: \end{equation}
684: where the own outflow $O_2$ supports a high permeability and the intersecting outflow $O_1$ suppresses it.
685: However, rather than using such a phenomenological approach, the permeability could also be calculated analytically, based on
686: a model of gap statistics, since large enough vehicle gaps are needed to join or cross a flow. Such kinds of calculations have been carried
687: out in Refs.~\cite{analyticJiang,Biometrica,Troutbeck1986,Troutbeck1997}.
688: \section{Towards a Self-Organized Traffic Light Control}\label{Sec4}
689:
690: In Ref.~\cite{control}, it has been pointed out that, for not too small arrival flows, an oscillatory service at intersections reaches higher
691: intersection capacities and potentially shorter waiting times than a first-in-first-out service of arriving vehicles. This is due to the fact that
692: the outflow of queued vehicles is more efficient than waiting for the arrival of other freely flowing vehicles, which have larger time gaps. For
693: similar reasons, pedestrians are passing a bottleneck in an oscillatory way \cite{mitMolnar}, and also two intersecting flows tend to
694: organize themselves in an oscillatory way \cite{analyticJiang,TransportationScience}.
695: \par
696: Therefore, using traffic lights at intersections is natural and useful, if operated in the right way. However, how to switch the traffic lights
697: optimally? While this is a solvable problem for single traffic lights, the optimal coordination of many traffic lights
698: \cite{Papadimitriou1999} is a really hard (actually NP hard) problem \cite{Schutter2002}. Rather than solving a combinatorial optimization problem, here, we
699: want to suggest a novel approach, which needs further elaboration in the future. The idea is to let the network flows self-organize
700: themselves, based on suitable equations for the permeabilities $\gamma_i(t)$ as a function of the outflows $O_i(t)$ and the numbers $\Delta
701: N_i(t)$ of delayed vehicles.
702: \par
703: Here, we will study the specification
704: \begin{equation}
705: \gamma_1(t) = \frac{1}{1+ a\mbox{e}^{b(O_2-O_1) - cD}}
706: \label{pre1}
707: \end{equation}
708: and
709: \begin{equation}
710: \gamma_2(t) = \frac{1}{1+ a\mbox{e}^{b(O_1-O_2) + cD}} \, ,
711: \label{pre2}
712: \end{equation}
713: which generalizes formula (\ref{phen}). While the relative queue length
714: \begin{equation}
715: D(t)=\Delta N_1(t) - \Delta N_2(t)
716: \end{equation}
717: quantifies the pressure to increase the permeability $\gamma_1$ for road section 1,
718: the outflow $O_2(t)$ from the road section 2 resists this tendency, while the flow
719: $O_1(t)$ on road section 1 supports the permeability.
720: The increasing pressure eventually changes the resistance threshold and
721: the service priority. An analogous situation applies to the permeability $\gamma_2$ for road section 2,
722: where the pressure corresponds to $-D$, which is again the difference in queue length.
723: $a$, $b$, and $c$ are non-negative parameters. $a$ may be set to 1, while $c$ must be large enough
724: to establish a sharp switching. Here, we have assumed $c=100$.
725: The parameter $b$ allows to influence the switching frequency $f$,
726: which is approximately proportional to $b$. We have adjusted the frequency $f$ to the cycle time
727: \begin{equation}
728: T^{\rm cyc} = \frac{2\tau}{1-(A_1+A_2)/\widehat{Q}} \, ,
729: \label{cycl}
730: \end{equation}
731: which results if the switching (setup) time (``yellow traffic light'') is $\tau = 5$s and
732: a green light is terminated immediately after a queue has dissolved after lifting the red
733: light.\footnote{If $\Delta T_1$ and $\Delta T_2$ denote the green time periods for the intersecting
734: flows 1 and 2, respectively, the corresponding red time periods for a periodic signal control
735: are $\Delta T_2$ and $\Delta T_1$, to which
736: the switching setup time of duration $\tau$ must be added. From formula (\ref{acco}) and with
737: $O_i = \widehat{Q}$ we obtain $\Delta T_1 = (\Delta T_2 +\tau)\widehat{Q}/(\widehat{Q}-A_1)$
738: and $\Delta T_2 = (\Delta T_1 +\tau)\widehat{Q}/(\widehat{Q}-A_2)$. Using the definition
739: $T^{\rm cyc} = \Delta T_1 + \tau + \Delta T_2 + \tau$ for the cycle time, we finally arrive at Eq.~(\ref{cycl}).}
740: The corresponding parameter value is
741: \begin{equation}
742: b = \frac{500}{\widehat{Q} - (A_1+A_2)} \, .
743: \label{choice}
744: \end{equation}
745: Figure \ref{ill} shows a simulation result for $A_1/\widehat{Q} = 0.3$ and $A_2/\widehat{Q} = 0.4$.
746: \par\begin{figure}[htbp]
747: \begin{center}
748: \includegraphics[width=\textwidth, angle=0]{osc}
749: \end{center}
750: \caption[]{Illustration of the dynamics of self-organized oscillations in the permeabilities and
751: the resulting flows for a single intersection with constant inflows (see text for details). Note that
752: the road section with the higher inflow (arrival rate) is served longer, and
753: its queues are shorter (see solid lines).\label{ill}}
754: \end{figure}
755: The properties of the corresponding specification of the permeabilitities $\gamma_i(t)$ are as follows:
756: \begin{itemize}
757: \item $\gamma_i(t)$ is non-negative and does not exceed the value 1.
758: \item For the sum of permeabilities and $a\ge 1$, we have
759: \begin{equation}
760: \gamma_1 +\gamma_2 = \frac{2+a(\mbox{e}^E + \mbox{e}^{-E})}{1+ a^2 + a(\mbox{e}^E + \mbox{e}^{-E})}
761: \le 1 \, ,
762: \end{equation}
763: where we have introduced the abbreviation
764: \begin{equation}
765: E = b(O_1 - O_2) + c(\Delta N_1 - \Delta N_2) \, .
766: \end{equation}
767: The sum is close to 1 for large absolute values of $E$, while for $E \approx 0$
768: the overall permeability $\gamma_1 + \gamma_2$ is small.
769: \item For large enough values of $ab$ and for $c, A_1, A_2 >0$,
770: the equations for the permeability do not have a stable stationary solution.
771: This can be concluded from
772: \begin{equation}
773: \frac{dE}{dt} = b \left( \frac{dO_1}{dt} - \frac{O_2}{dt}\right)
774: + c \left( \frac{d\Delta N_1}{dt} - \frac{d\Delta N_2}{dt}\right)
775: \end{equation}
776: together with
777: \begin{equation}
778: \frac{d\Delta N_i}{dt} = A_i - O_i(t)
779: \end{equation}
780: and
781: \begin{equation}
782: O_i(t) = \gamma_i(t) \max[\widehat{Q}\Theta(\Delta N_i> 0), A_i] \, ,
783: \end{equation}
784: see Eqs. (\ref{delayed}) and (\ref{easy2}). As
785: $dD/dt = d\Delta N_1/dt - d\Delta N_2/dt$ varies around
786: zero, the same applies to $D(t)$, which leads to oscillations of the permeabilities $\gamma_i(t)$.
787: \item With the specification (\ref{choice}) of parameter $b$,
788: the cycle time is approximately proportional to the overall inflow $(A_1+A_2)$.
789: \item The road section with the higher flow gets a longer green time period
790: (see Fig. \ref{ill}).
791: \end{itemize}
792: If the above self-organized traffic flows shall
793: be transfered to a new principle of traffic light control, phases with $\gamma_i(t) \approx 1$ could be
794: interpreted as green phases and phases with $\gamma_i(t)\approx 0$ as red phases. Inefficient,
795: intermediate switching time periods for certain choices of parameter values could be translated
796: into periods of a yellow traffic light.
797:
798: \section{Summary and Outlook}\label{Sec5}
799:
800: We have presented a simple model for conserved flows in networks. Although our specification has been illustrated for traffic flows in urban
801: areas, similar models are useful for logistic and production system or even transport in biological cells or bodies. Our model considers
802: propagation speeds of entities and congestion fronts, spill-over effects, and load-dependent transportation times.
803: \par
804: We have also formulated constraints for network nodes. These constraints contain several minimum and maximum functions, which implies a
805: multitude of possible cases even for relatively simple intersections. It turns out that the arrival and departure flows of diverges have
806: uniquely defined values, while merges or intersections have a set of feasible solutions. This means, the actual result may sensitively
807: depend on the intersection design. For mathematical reasons, we have determined flow-optimizing solutions for two merging and two intersecting
808: flows. However, it is questionable whether these solutions can be established in reality without the implementation of intelligent transport
809: systems facilitating optimal gap usage: In many situations, coordination problems between vehicles in merging or intersection areas cause
810: inefficiencies, which reduce their permeability.
811: \par
812: In fact, at not too small traffic volumes, it is better to have an oscillation between minimum and maximum permeability values. Therefore, we
813: have been looking for a mechanism producing emergent oscillations between high and low values. According to our proposed specification (which
814: is certainly only one of many possible ones), the transition between high and low permeability was triggered, when the difference between the
815: queue lengths of two traffic flows competing for the intersection capacity exceeded a certain value. The resulting oscillatory service could
816: be used to {\it define} traffic phases. One potential advantage of such an approach would be that the corresponding traffic light control would be based on
817: the self-organized dynamics of the system. Further work in this direction seems very promising.
818:
819: \section*{Acknowledgements}
820:
821: The authors are grateful for partial financial support by the German Research Foundation (research projects He 2789/5-1,
822: 8-1) and by the ``Cooperative Center
823: for Communication Networks Data Analysis'', a NAP project sponsored by
824: the Hungarian National Office of Research and Technology under grant
825: No.\ KCKHA005.
826:
827:
828: \begin{thebibliography}{99}
829: \bibitem{Astarita1995}
830: V. Astarita, {\it Flow propagation description in dynamic network loading models},
831: in Y. J. Stephanedes and F. Filippi (eds.), Proc. IV Internat. Conf. Appl. Adv. Tech. Transportation Engrg (AATT), 1995, pp. 599--603.
832:
833: \bibitem{Astarita2002}
834: V. Astarita, {\it Node and link models for traffic simulation}, Math. Comput. Model., {\bf 35} (2002), no. 5, 643--656.
835:
836: \bibitem{pipelines}
837: M. K. Banda, M. Herty and A. Klar, {\it Gas flow in pipeline networks},
838: NHM, {\bf 1} (2006), no. 1, 41--56.
839:
840: \bibitem{Piccoli2} G. Bretti, R. Natalini and B. Piccoli,
841: {\it Numerical approximations of a traffic flow model on networks},
842: NHM, {\bf 1} (2006), no. 1, 57--84 .
843:
844: \bibitem{Carey2003}
845: M. Carey, Y.E. Ge and M. McCartney, {\it A whole-link travel-time model with desirable properties},
846: Transport. Sci., {\bf 37} (2003), no. 1, 83--96.
847:
848: \bibitem{Daganzo} C. F. Daganzo,
849: {\it Requiem for second-order fluid approximations of traffic flow}, Transpn. Res. B, {\bf 29} (1995), 277--286.
850:
851: \bibitem{CellTransmission} C. F. Daganzo,
852: {\it The cell transmission model, {P}art {II}: {N}etwork traffic}, Transpn. Res. B, {\bf 29} (1995), 79--93.
853:
854: \bibitem{Schutter2002}
855: B. De Schutter, {\it Optimizing acyclic traffic signal switching sequences through an extended linear complementarity problem formulation},
856: Eur. J. Oper. Res., {\bf 139} (2002), no. 2, 400--415.
857:
858: \bibitem{Piccoli1} M. Garavello and B. Piccoli,
859: Traffic Flow on Networks, American Institute of Mathematical Sciences, Springfield, 2006.
860:
861: \bibitem{Klar} S. Goettlich, M. Herty and A. Klar, {\it Modelling and optimization of supply chains on complex networks},
862: Commun. Math. Sci., {\bf 4} (2006), no. 2, 315--330.
863:
864: \bibitem{RMP}
865: D. Helbing, {\it Traffic and related self-driven many-particle systems},
866: Rev. Mod. Phys., {\bf 73} (2001), 1067--1141.
867:
868: \bibitem{JPhysA} D. Helbing,
869: {\it A section-based queueing-theoretical traffic model for congestion and travel time analysis in networks}, J. Phys. Math. Gen. {\bf 36} (2003), L593--L598.
870:
871: \bibitem{TGF03} D. Helbing,
872: {\it Production, supply, and traffic systems: A unified description},
873: in S. P. Hoogendoorn, S. Luding, P. H. L. Bovy, M. Schreckenberg and D. E. Wolf (eds.), Traffic and Granular Flow '03, Springer,
874: Berlin, 2005, pp. 173--188.
875:
876: \bibitem{control} D. Helbing, S. L\"ammer and J.-P. Lebacque,
877: {\it Self-organized control of irregular or perturbed network traffic},
878: in C. Deissenberg and R. F. Hartl (eds.), Optimal Control and Dynamic Games, Springer, Dordrecht, 2005, pp. 239--274.
879:
880: \bibitem{analyticJiang} D. Helbing, R. Jiang and M. Treiber,
881: {\it Analytical investigation of oscillations in intersecting flows of pedestrian and vehicle traffic}, Phys. Rev. E, {\bf 72} (2005), 046130.
882:
883: \bibitem{mitJohansson} D. Helbing, A. Johansson, J. Mathiesen, M. H. Jensen and A. Hansen,
884: {\it Analytical approach to continuous and intermittent bottleneck flows}, Phys. Rev. Lett.,
885: {\bf 97} (2006), 168001.
886:
887: \bibitem{TransportationScience}
888: D. Helbing, L. Buzna, A. Johansson, and T. Werner,
889: {\it Self-organized pedestrian crowd dynamics: Experiments, simulations, and design solutions},
890: Transpn. Science, {\bf 39} (2005), 1--24.
891:
892: \bibitem{patent} D. Helbing and S. L\"ammer, {\it Method for the coordination of competing processes
893: or for the control of the transport of mobile entities within a network}, Patent DE 10 2005 023 742.8.
894:
895: \bibitem{supply}
896: D. Helbing, S. L\"ammer, T. Seidel, P. Seba and T. Platkowski,
897: {\it Physics, stability and dynamics of supply networks}, Phys. Rev. E, {\bf 70} (2004), 066116.
898:
899: \bibitem{mitMolnar} D. Helbing and P. Moln\'{a}r,
900: {\it Social force model for pedestrian dynamics}, Phys. Rev. E, {\bf 51} (1995), 4282--4286.
901:
902: \bibitem{logistics}
903: D. Helbing, T. Seidel, S. L\"ammer and K. Peters,
904: {\it Self-organization principles in supply networks and production systems},
905: in: B. K. Chakrabarti, A. Chakraborti, and A. Chatterjee (eds.) Econophysics and Sociophysics,
906: Wiley, New York, 2006.
907:
908: \bibitem{Herty3}
909: M. Herty and A. Klar, {\it Modeling, simulation, and optimization of traffic flow networks}, SIAM Appl. Math., {\bf 64} (2003), no. 2, 565--582.
910:
911: \bibitem{Herty2}
912: M. Herty and A. Klar, {\it Simplified dynamics and optimization of large scale traffic flow networks}
913: Math. Mod. Meth. Appl. Sci., {\bf 14} (2004), no. 4, 579--601.
914:
915: \bibitem{Herty1} M. Herty, S. Moutari and M. Rascle,
916: {\it Optimization criteria for modelling intersections of vehicular traffic flow}, NHM, {\bf 1} (2006), no. 2, 275--294.
917:
918: \bibitem{Hilliges} M. Hilliges and W. Weidlich,
919: {\it A phenomenological model for dynamic traffic flow in networks}, Transpn. Res. B, {\bf 29} (1995), 407--431.
920:
921: \bibitem{Kernerbuch}
922: B. Kerner, The Physics of Traffic, Springer, Berlin, 2004.
923:
924: \bibitem{synchronization} S. L\"ammer, H. Kori, K. Peters and D. Helbing,
925: {\it Decentralised control of material or traffic flows in networks using phase-synchronisation}, Phys. Stat. Mech. Appl., {\bf 363} (2006), 39--47.
926:
927: \bibitem{Lebacque}
928: J.-P. Lebacque and M. M. Khoshyaran, {\it First-order macroscopic traffic flow models: Intersection modeling, network modeling},
929: in H. S. Mahmasani (ed.), 16th International Symposium on Transportation and Traffic Theory, Elsevier, 2005, pp. 365-386.
930:
931: \bibitem{LW} M. J. Lighthill and G. B. Whitham,
932: {\it On kinematic waves: {II. A} theory of traffic on long crowded roads}, Proc. Roy. Soc. London, Ser. A, {\bf 229} (1955), 317--345.
933:
934: \bibitem{Biometrica}
935: A. J. Mayne,
936: {\it Some further results in the theory of pedestrians and road traffic},
937: Biometrika {\bf 41} (1954), pp. 375--389.
938:
939: \bibitem{Papadimitriou1999}
940: C. H. Papadimitriou and J. N. Tsitsiklis, {\it The complexity of optimal queuing network control}, Math. Oper. Res.,
941: {\bf 24} (1999), no. 2, pp. 293--305.
942:
943: \bibitem{Peters} K. Peters and U. Parlitz,
944: {\it Hybrid systems forming strange billiards}, Int. J. Bifurcation and Chaos, {\bf 13} (2003), no. 9, 2575--2588.
945:
946: \bibitem{mitSchoenhof}
947: M. Sch\"onhof and D. Helbing,
948: {\it Empirical features of congested traffic states and their implications for traffic modelling},
949: Transpn. Sci, accepted (2006).
950:
951: \bibitem{mitTreiber} M. Treiber, A. Hennecke and D. Helbing,
952: {\it Congested traffic states in empirical observations and microscopic simulations}, Phys. Rev. E, {\bf 62} (2000), 1805--1824.
953:
954: \bibitem{Troutbeck1986}
955: R. J. Troutbeck, {\it Average delay at an unsignalized intersection with two major each having a dichotomized headway distribution},
956: Transport. Sci., {\bf 20} (1986), no. 20, pp. 272--286.
957:
958: \bibitem{Troutbeck1997}
959: R. J. Troutbeck and W. Brilon, {\it Unsignalized intersection theory}, in
960: N. Gartner, H. Mahmassani, C. H. Messer, H. Lieu, R. Cunard and A. K. Rathi (eds.),
961: Traffic Flow Theory: A State-of-the-Art Report, Transportation Research Board, 1997, pp. 8.1--8.47.
962:
963: \bibitem{Whitham} G. B. Whitham,
964: Linear and Nonlinear Waves, Wiley, New York, 1974.
965:
966: \end{thebibliography}
967: %
968: %\medskip%
969: %
970: %Received ...........; revised ...............
971: %
972: \end{document}
973: