一级倒立摆的单神经元PID控制器的设计毕业设计.docVIP

余黄伟：一级倒立摆的单神经元PID控制器的设计 安徽工程大学毕业设计（论文） PAGE PAGE II PAGE PAGE 1 一级倒立摆的单神经元PID控制器的设计 摘 要 倒立摆装置被公认为自动控制理论中的典型试验设备，是控制理论教学和科研中最典型的物理模型。通过对它的研究不仪可以将控制理论中的许多分析问题加以验证，还能将控制理论涉及的三个主要基础学科:力学、数学和电学(包含计算机)进行有机的综合应用。倒立摆可为多种具有实际工程背景的物理系统提供一个很好的实验平台，从而可以将所设计的控制器应用于此平台上。 首先利用牛顿一欧拉方法对倒立摆系统进行建模，从LQR控制方法，传统的PID控制方法与神经网络控制理论出发，学习了LQR算法和单神经元PID控制器算法，并根据控制模型设计LQR控制器和单神经元PID控制器。通过MZTLAB仿真和倒立摆实验装置实现倒立摆系统的稳定控制。文中采用两种控制方法：LQR控制和单神经元PID控制法。 通过仿真可以看出采用单神经元PID控制方法的控制效果比LQR控制效果要好，超调量小或无超调量，调节时间比LQR控制短。 关键词：倒立摆 PID控制 单神经元网络控制 Design of single neuron PID controller for an inverted pendulum Abstract Nverted pendulum device is regarded as the typical test equipment in the theory of automatic control, and it is the most typical physical model in the teaching and scientific research of control theory. Through the research instrument can control many of the analytical problems in theory is verified, but also the control theory involves three main basic disciplines: mechanics, mathematics and electrical (including computer) for organic synthesis application. Inverted pendulum can provide a good experimental platform for a variety of physical systems with practical engineering background, so that the designed controller can be applied to the platform. Firstly, the Newton Euler method of inverted pendulum system modeling, the LQR control method, the traditional PID control method and neural network control theory, to study the related learning algorithm, the LQR algorithm and single neuron PID controller algorithm, and according to the control model design of LQR control system with the single neuron PID controller. The stability control of inverted pendulum system is realized by the MZTLAB simulation and the inverted pendulum experiment device. Two control methods are used in this paper: LQR control and single neuron PID control. Through the simulation, we can see that the control effect of single neuron PID control method is better than the LQR control effect, the overshoot