Deep reinforcement learning-based long-range autonomous valet parking for smart cities

Abstract

In this paper, to reduce the congestion rate at the city center and increase the traveling quality of experience (QoE) of each user, the framework of long-range autonomous valet parking is presented. Here, an Autonomous Vehicle (AV) is deployed to pick up, and drop off users at their required spots, and then drive to the car park around well-organized places of city autonomously. In this framework, we aim to minimize the overall distance of AV, while guarantee all users are served with great QoE, i.e., picking up, and dropping off users at their required spots through optimizing the path planning of the AV and number of serving time slots. To this end, we first present a learning-based algorithm, which is named as Double-Layer Ant Colony Optimization (DLACO) algorithm to solve the above problem in an iterative way. Then, to make the fast decision, while considers the dynamic environment (i.e., the AV may pick up and drop off users from different locations), we further present a deep reinforcement learning-based algorithm, i.e., Deep Q-learning Network (DQN) to solve this problem. Experimental results show that the DL-ACO and DQN-based algorithms both achieve the considerable performance.