中文摘要：

针对视觉导引AGV在纠偏过程中，由于视觉导引实时性不足和常规PID中积分积累所引起的系统超调与振荡问题，提出以位置偏差作为控制系统输入，利用积分分离PID纠偏算法求得左右轮速度差．通过下位机控制左右驱动电机实现差速纠偏的控制策略。首先根据上述控制策略建立AGV小车运动学模型．分析小车在纠偏过程中的曲线运动轨迹特征．建立纠偏方程，然后通过系统辨识获得控制系统参数，设计积分分离PID纠偏算法，最后通过MATLAB进行数值仿真。仿真结果表明，积分分离PID控制器在调节时间、抑制超调与振荡方面性能优于常规PID控制器．仿真曲线具有较好的快速性和平稳性，并能够更平稳地对AGV轨迹偏差进行校正。

英文摘要：

In order to solve the problem of system overshoot and oscillation caused by detention and integration accumulation during vision-guided AGV rectifying deviation, the paper put forward a control strategy that uses position deviation as the input of deviation- rectifying system, then calculates the speed difference between left and right wheels based on integral separation PID deviation-rectifying algorithm, finally, PLC controls the motors to execute and deviation can be rectified. First of all, we established AGV kinematic model and deviation-rectifying equation based on the control strategy. Then we used the method of system identification to calculate parameters of control sys- tems and design integral separation PID controller. The simulation results show that the integral separation PID can effectively suppress the overshoot and system oscillation, and significantly improve the stability of AGV line-tracking.