1: \begin{abstract}

2: %	This article is motivated by improving the performance of RRT* algorithm, but the proposed algorithm also applies to other sampling-based path planning algorithms.

3: %	The existing RRT* algorithm is a variant of RRT, inheriting the uniform sampling strategy of the RRT algorithm.

4: %	However, this sampling strategy limits the convergence speed of the RRT* algorithm.

5: %	To overcome the limitation, we propose a novel algorithm(CGAN-RRT*), which combines a non-uniform sampling strategy and an uniform sampling strategy together to obtain the optimal path.

6: %	The non-uniform sampling strategy is guided by the predicted possibility distribution of feasible paths generated from the CGAN model, while uniform sampling strategy is same as the strategy adopted by conventional RRT* algorithm. 

7: %	Experimental simulations illustrate that the proposed CGAN-RRT* algorithm is effective and achieves better performance than the conventional RRT* algorithm.

8: 	

9: %\end{abstract}

10: