Multiobjective Scheduling Strategy With Genetic Algorithm and Time-Enhanced A* Planning for Autonomous Parking Robotics in High-Density Unmanned Parking Lots

With the process of urbanization, the problem of insufficient parking spaces has become prominent. Adopting a high-density parking lot with parking robots can greatly improve the land utilization rate of the parking lot. This article tackles the multiple parking robots scheduling problem of high-density layout parking lots, including task execution sequence decision, robot allocation, and cooperative path planning. First, we mathematically describe the parking robot scheduling problem. Existing approximation algorithms are often far from the optimal solution. This article proposes an improved genetic algorithm and a time-enhanced A* path planning algorithm for high-density parking lots. The improved genetic algorithm can efficiently search task execution sequence and robot allocation and converge to the optimal solution even in large-scale complex scenarios. Meanwhile, the time-enhanced A* algorithm takes a new dimension "the time" into consideration, together with the distance, and security factors, to solve the multi-parking-robot path planning problem. Simulation experiments show that our algorithm can improve scheduling performance in many aspects such as task execution time, driving distance, and security in large-scale high-density parking lots. This article provides an efficient and convenient scheduling solution for the implementation of the high-density unmanned parking lot.