采用改进BP-PID控制的机器人避障仿真研究  

Research on robot obstacle avoidance simulation using improved BP-PID control

作　　者：吴静松[1] 耿振铎[2] WU Jingsong;GEN Zhenduo(College of Computer Science and Information Engineering,Anyang Institute of Technology,Anyang 455000,Henan,China;School of Physics,Henan Normal University,Xinxiang 457300,Henan,China)

机构地区：[1]安阳工学院计算机科学与信息工程学院,河南安阳455000 [2]河南师范大学物理学院,河南新乡457300

出　　处：《中国工程机械学报》2024年第4期437-441,共5页Chinese Journal of Construction Machinery

基　　金：2021年河南省科技厅科技攻关资助项目(212102210560);2020校科研培育基金资助项目(YPY2020013)。

摘　　要：针对移动机器人避障过程中行驶路径长、寻路速度慢等问题,提出了一种改进反向传播-比例-积分-微分(BP-PID)控制器,并对移动机器人避障效果进行仿真验证。利用移动机器人在二维坐标系的避障简图,得出了移动机器人运动方程式。引用比例-积分-微分(PID)控制器和3层BP神经网络结构,利用BP神经网络的学习能力调整PID控制器参数。引用粒子群算法进行改进,通过改进粒子群算法在线优化BP-PID控制器,确保移动机器人BP-PID控制器收敛于全局最优值,从而使移动机器人避障效果更好。在不同环境中,采用Matlab软件对移动机器人避障效果进行仿真,比较改进前和改进后的移动机器人避障效果。结果显示:在不同环境中,改进前和改进后的BP-PID控制器均能使移动机器人安全地躲避障碍物;但是采用改进的粒子群算法优化BP-PID控制器,可以使移动机器人运动路径更短,迭代次数更少,搜索时间更短。采用改进BP-PID控制器,能够提高移动机器人避障过程中寻路速度,缩短行驶路径,效果更好。An improved(back propagation proportional integral differential,BP-PID)controller is proposed to address the shortcomings of long path and slow pathfinding speed during obstacle avoidance for mobile robots,and the obstacle avoidance effect of mobile robots is simulated and verified.The obstacle avoidance diagram of the mobile robot in the two-dimensional coordinate system is given,and the motion equation of the mobile robot is given.The PID controller and threelayer BP neural network structure were cited,and the learning ability of BP neural network was utilized to adjust the parameters of the PID controller.Referencing particle swarm optimization algorithm and making improvements,by improving the particle swarm algorithm to optimize the BPPID controller online,it ensures that the mobile robot BP-PID controller converges to the global optimal value,thereby improving the obstacle avoidance effect of the mobile robot.Simulate the obstacle avoidance effect of mobile robots in different environments using Matlab software,and compare the obstacle avoidance effects of mobile robots before and after improvement.The results show that in different environments,both the pre improved and post improved BP-PID controllers enable mobile robots to safely avoid obstacles.However,using improved particle swarm optimization to optimize the BP-PID controller results in shorter motion paths,fewer iterations,and shorter search times for mobile robots.Adopting an improved BP-PID controller can improve the path finding speed of mobile robots during obstacle avoidance,shorten the travel path,and achieve better results.

关 键 词：移动机器人 BP神经网络 PID控制器 改进粒子群算法 避障 仿真 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]