An Autonomous Parking System using the Hybridization of the Rapidly-Exploring Random Trees Star and Ant Colony System

Abstract

One key autonomous driving car application that can be utilized to address the problem of parking spaces is autonomous parking. The goal of this study is to develop a path planning algorithm that can swiftly create a path for autonomous parking in a variety of car parking settings. The method employed is a mix of Rapidly Exploring Random Tree Star and Ant Colony Systems (RRT-ACS). The RRT-ACS method can quickly provide an ideal path. The reed sheep planner technique is also used to provide a smooth curving path that non-holonomic vehicles can follow. The performance of the RRT-ACS algorithm-based autonomous parking system is compared to that of the RRT*-connect and informed RRT*-connect algorithms. The simulation tests were performed in common parking settings such as parallel parking and vertical parking. The test results reveal that the suggested autonomous parking system outperforms other comparative algorithms. It is possible to infer that a route planning algorithm capable of rapidly designing a path for automatic parking in a variety of car parking circumstances has been successfully built. These findings may have ramifications for self-driving cars equipped with advanced driver support systems.

Selecting Dynamic Path Planning Algorithm Based-Upon Ranking Approach for Omni-Wheeled Mobile Robot

“…It uses a cost function to identify the parent as the node with the lowest cost in the domain's collection of enlarged nodes. As stated by Karaman et al, [19] and Aria et al, [20], it additionally reconnects the nodes on the current tree after each iteration, ensuring computational difficulty and an asymptotically optimum solution. The algorithm's approach is similar to that of RRT up to the point of finding 𝑞 425 , with two stages following its discovery as shown in Figure 4(b).…”