突发地震灾害中无人驾驶救援车伺服控制系统研究

On the Servo Control System of Unmanned Rescue Vehicle in Sudden Earthquake Disasters

摘要:

针对当前采用PID控制器控制无人驾驶救援车伺服系统时存在的轨迹跟踪精度不高，误差控制性能较差，灵活性、平稳性和安全性能不佳等问题，提出并设计基于BP神经网络整定PID控制器的无人驾驶救援车伺服控制系统，建立突发地震灾害中无人驾驶救援车伺服控制系统驱动模型，并以此模型作为被控对象；根据系统期望输出值与实际输出值构成的控制偏差获得PID控制规律，并通过调节PID控制器控制参数实现系统控制，在此基础上，采用BP神经网络通过对无人驾驶救援车伺服控制系统性能的学习，构建基于BP神经网络整定的PID控制器，并采用梯度下降法修正控制器加权系数，通过在线调整BP神经网络加权系数即可实现控制器的自适应调整，控制突发地震灾害中无人驾驶救援车实施救援。实验结果表明，设计的基于BP神经网络整定PID控制器的无人驾驶救援车伺服系统可有效提高轨迹跟踪精度，具有较好的灵活性，且能够保证驾驶员的安全和车辆平稳行驶。

Abstract:

Today, the PID controller is used to control the servo system of unmanned rescue vehicles, but it has a number of problems, i.e., the trajectory tracking accuracy is not high, the error control performance is poor, and the flexibility, stability, and safety performance are not good.In this work, we propose and design a servo system for unmanned rescue vehicles based on the BP-neural-network-tuning PID controller.Then, we establish a drive model for the servo control system of the unmanned rescue vehicle in sudden earthquake disasters.We obtained the PID control law based on the control deviation and realized system control by adjusting the parameters of the PID controller.On this basis, we constructed a PID controller based on BP-neural-network-tuning, and use the gradient descent method to correct the weighting coefficient of the controller.By adjusting the weighting coefficient of the BP neural network online, the controller can be adaptively adjusted to control an unmanned rescue vehicle in sudden earthquake disasters.The experimental results show that the servo system of the designed unmanned rescue vehicle based on a BP-neural-network-tuning PID controller can effectively improve the trajectory tracking accuracy, improve flexibility, and ensure the safety of the driver and smooth operation of the vehicle.