航空学报 > 2015, Vol. 36 Issue (4): 1348-1358   doi: 10.7527/S1000-6893.2014.0130

考虑运动学约束的不规则目标遗传避碰规划算法

张智, 林圣琳, 朱齐丹, 王开宇   

  1. 哈尔滨工程大学 自动化学院, 哈尔滨 150001
  • 收稿日期:2014-05-08 修回日期:2014-06-18 出版日期:2015-04-15 发布日期:2014-07-09
  • 通讯作者: 张智Tel.: 0451-82519594 E-mail: neverbadzz@163.com E-mail:neverbadzz@163.com
  • 作者简介:张智 男, 博士, 讲师, 硕士生导师。主要研究方向: 机器感知与智能控制。Tel: 0451-82519594 E-mail: neverbadzz@163.com;林圣琳 男, 硕士研究生。主要研究方向: 机器感知与智能控制。Tel: 0451-82519594 E-mail: lin_44627079@yeah.net朱齐丹 男, 博士, 教授, 博士生导师。主要研究方向: 机器感知与智能控制。Tel: 0451-82519594 E-mail: zhuqidan@hrbeu.edu.cn;王开宇 男, 博士研究生。主要研究方向: 机器感知与智能控制。Tel: 0451-82519594 E-mail: wangkaiyu@hrbeu.edu.cn
  • 基金资助:

    国家自然科学基金(61104037, 61304060);国家国际科技合作专项(2013DFR10030);中央高校基本科研业务费专项资金(HEUCFX41304)

Genetic collision avoidance planning algorithm for irregular shaped object with kinematics constraint

ZHANG Zhi, LIN Shenglin, ZHU Qidan, WANG Kaiyu   

  1. College of Automation, Harbin Engineering University, Harbin 150001, China
  • Received:2014-05-08 Revised:2014-06-18 Online:2015-04-15 Published:2014-07-09
  • Supported by:

    National Natural Science Foundation of China (61104037, 61304060); International S & T Cooperation Program of China (2013DFR10030); Fundamental Research Funds for the Central Universities (HEUCFX41304)

摘要:

针对复杂环境下不规则目标的路径规划问题,提出了一种带有运动学约束的遗传避碰规划算法。以舰载机在航母甲板上的路径规划问题作为研究对象,并且该算法可推广至其他具有此类约束的路径规划问题中,它较好地解决了目标形状复杂、障碍环境复杂、目标运动时带有回转半径约束等特殊问题。在传统遗传路径规划算法的基础上,针对性地设计了三维位置和姿态混合编码、三段法路径解码、轨迹包围盒的碰撞检测及距离计算等方法,并在遗传操作中引入惩罚项和修补策略来辅助算法寻优。最后,为得出复杂环境下的最优路径,基于VC++平台对算法进行了仿真验证。结果表明,在复杂障碍环境下,本文提出的算法可求得最优避碰路径,并满足预先设定的目标回转半径约束,能够有效地解决此类目标的避碰路径规划问题。

关键词: 避碰路径规划, 遗传算法, 不规则形状, 运动学约束, 舰载机, 甲板调运

Abstract:

To deal with the path planning problems of irregularly shaped objects in complex environment, a genetic collision avoidance algorithm with kinematics constraint is developed. This algorithm is then applied to the path planning operations on carrier-based aircraft scheduling on carrier flight deck. Moreover, it can be extended to solve other path planning cases under such constraints. For the problems resulting from these objects, which are characterized by complex shape and the bending radius constraint while moving in complicated obstacle situations, the technique proposed is proved to be effective. Based on the traditional genetic path planning algorithm, a three-dimensional position and orientation coding method, a three-stage path decoding method and an approach specific to the collision detection and distance calculation of a track bounding box are presented. Also, a penalty term and a gene repairing strategy are brought into the genetic process to seek the optimum. Finally, simulated verifications are conducted using VC++ platform to obtain the optimal paths. The results show that the optimal collision avoidance paths in complex obstacle environment are achieved utilizing the proposed algorithm, with the pre-set bending radius constraints satisfied. It is indicated that the design yields effective solutions to the collision avoidance path planning problems correlated with this kind of objects.

Key words: collision avoidance path planning, genetic algorithms, irregular shape, kinematics constraint, carrier-based aircraft, aircraft scheduling on flight deck

中图分类号: