基于神经网络的刮板输送机链动态自动张紧控制策略研究

Research on dynamic automatic tension control strategy for scraper conveyer chain based on neural

以刮板输送机链自动张紧系统为研究对象,通过控制机尾电机转速和伸缩液压缸的位移,使得机尾链轮上下链张紧力始终恒定,实现刮板链动态自动张紧控制,这是一个时变的两输入两输出耦合系统。本文根据解耦和神经网络的思想,在线系统辨识和调整两个PID控制器的参数,实现了不依赖于对象模型的自适应PID解耦控制。计算机仿真结果表明,该系统具有较强的抗干扰能力和自适应能力,系统鲁棒性增强,为今后进一步研究奠定了基础。

With automatic tensioning system of the scraper conveyor chain as a study object , by controlling the tail motor speed and displacement of telescopic hydraulic cylinder , the tensions from upper and lower tail chain are constant to ensure dynamic tensioning control of the scraper chain , which is a time-varying two-input two-output coupling system.In the paper , parameters of both PID controllers are identified and adjusted on line based on the idea of decoupling and neural network, to realize the self-adaption PID decoupling control independent of the object model.The simulation result shows that the system is provided with strong capacity of resisting disturbance and self-adaption, robustness, which will establish foundation for further research.