自動停車路徑規劃與決策系統開發

隨著當前汽車無人駕駛技術日新月異的發展，國內外車廠紛紛由停車輔助系統，轉往開發全自動駕駛技術的研發，其中，自動停車技術更是當前重要發展焦點之一。至目前為止，市場上的自動停車系統，多採用距離感測器偵測停車空間，或是需要目標停車車格兩旁需有鄰車以定位停車空間，使用上的諸多限制反而造成停車不易的現象。自動停車系統同時也是發展全自動駕駛系統的重要指標之一。本文將介紹團隊發展全自動駕駛停車系統，從過去的完善的視覺輔助偵測以及多階段停車系統為基礎，在面對停車空間受到限制的情況下，透過發展Exploration-improved Bidirectional RRT*停車路徑規劃演算法，可搜索停車空間中的有效區域，考量車輛動態，產生一條安全且符合駕駛習慣的停車路線。同時開發基於多階段停車系統經驗，開發出停車決策系統，引領自駕車於停車場內找尋空閒車位，控制車輛的制動器，使自動駕駛車能在空間受限的環境下，安全、有效的完成停車任務。

With the rapid development of current car driverless technology, domestic and foreign automakers have shifted from conservative parking assistance systems to investing in fully automatic driving technology research and development. Among them, automatic parking technology is one of the current important development focuses. So far, most of the automatic parking systems on the market use distance sensors to detect the parking space or rely on neighboring cars to locate the parking space. Many restrictions on use have made parking difficult. The automatic parking system is also one of the important indicators for the development of a fully automatic driving system. This article will introduce the team's development of a fully automated parking system based on the past complete visual assisted detection and multi-stage parking system to overcome the limited parking space's bottleneck. in the face of limited parking space, through the development of Exploration-improved Bidirectional RRT* parking path planning The algorithm can search the effective area in the parking space, consider the dynamics of the vehicle, and generate a safe parking route that conforms to driving habits.. At the same time, based on the experience of the multi-stage parking system, a parking decision-making system was developed to guide self-driving cars in the parking lot to find free parking spaces, and to control the brakes of the vehicles, so that the self-driving cars can safely and effectively complete the parking task in a limited space environment.