欢迎访问《汽车安全与节能学报》，
汽车安全与节能学报

JASE ›› 2018, Vol. 9 ›› Issue (4): 449-455.DOI: 10.3969/j.issn.1674-8484.2018.04.012

• 汽车节能与环保 • 上一篇    下一篇

基于UTMD 的汽车自动驾驶的路径规划寻优算法

李学鋆1,2   

  1. （1. 湖北汽车工业学院 汽车工程学院，十堰 442002，中国；2. 汽车动力传动与电子控制湖北省重点实验室，湖北汽车工业学院，十堰 442002，中国）
  • 收稿日期:2018-07-29 出版日期:2018-12-31 发布日期:2019-01-02
  • 作者简介:李学鋆(1991—)，男（土家族），湖北，助教。E-mail: leexueyun@126.com。

Path planning optimization-algorithm for a self-driving vehicle based on UTMD principle

LI Xueyun1,2   

  1. (1. School of Vehicle Engineering, Hubei University of Automotive Technology, Shiyan 442002, China; 2. Key Laboratory of Automotive Power Train and Electronics, Hubei University of Automotive Technology, Shiyan 442002, China)
  • Received:2018-07-29 Online:2018-12-31 Published:2019-01-02

PDF

254

摘要：

       为在汽车自动驾驶中路径规划中能兼顾运算的实时性和可靠性，设计了一种智能仿生算法的路径寻优算法。该算法基于超声靶向微泡破坏(UTMD) 算法的原理。迭代运算分为靶标圈定、微泡迭代、微小核糖核酸(miRNAs) 迭代。圈定靶标，以便有效减小路径搜索范围。利用非线性函数Rastrigin 对该算法进行验证。分析了迭代次数的设定方式，并采用程序自检的方式进行自行跳转。将该算法运用到二维路径规划中，并在Matlab 中进行模拟。结果表明：与Dijkstra 算法相比，该UTMD 算法所规划的路径长度缩减4.52%。因此，该算法可有效地应用于汽车自动驾驶。

关键词: 自动驾驶汽车, 路径规划, 超声靶向微泡破坏(UTMD) 算法, 算法设计, 靶标圈定, 微小核糖核酸(miRNAs)

Abstract:

A path-planning optimization-algorithm for self-driving vehicles was designed by using some bionic intelligent algorithms to take account of the real-time and the reliability in computation. The path planning algorithm based on the basic principles of ultrasound-targeted microbubble destruction (UTMD). Divided the
iterative operation into target circle, microbubble iteration and micro-RNAs (miRNAs) iteration with the target delineation to reduce effectively the path search range. Verified the algorithm by using the Rastrigin function. Analyzed the setting of iteration numbers when they were set artificially. And used a self-test program to jump by itself. The algorithm was applied to 2-D path planning and simulated in Matlab. The results show that the path length planned by the UTMD algorithm is reduced by 4.52% compared with the length by Dijkstra algorithm. Therefore, the path-planning algorithm is effective for self-driving vehicles.

Key words: self-driving vehicle, path-planning, ultrasound-targeted microbubble destruction (UTMD), algorithm design, target delineation, micro-RNAs (miRNAs)