摘 要
PID控制器现在仍然是应用最广泛的工业控制器，其关键在于对PID参数的优化整定,而采用常规的手工整定方法已经难以满足要求，目前需要解决对PID参数的高效优化问题。
本论文首先介绍了PID控制方法和常规PID参数整定方法，编程实现了数字PID控制器和基于遗传算法的参数自整定程序，然后将遗传算法用于不同的被控对象进行PID参数的整定及优化，为了对比控制效果，应用MATLAB软件进行了仿真验证，并用仿真曲线进行直观的对比。
结果表明遗传算法能够在对所求解问题一无所知的情况下,快速从全局搜索出优化的控制参数,是一种高效的PID参数整定方法。
关键词: 遗传算法;PID控制器;参数整定;仿真
中国石油大学（华东）本科毕业设计（论文）
ABSTRACT
PID controller now it is still used widely in practical industrial control, and the key of PID controller lies in the tuning of PID parameters. The normal application of traditional PID controller cannot satisfy the request. What need to be resolved now is how to optimize PID parameters efficiently.
First of all, this paper introduces the PID control method and conventional method of PID parameter tuning, programs a digital PID controller and a parameter self-tuning procedure based on genetic algorithm, and then makes use of genetic algorithm to different objects for PID parameters tuning and optimization. In order to contrast the effects of control, we carried out simulations using MATLAB software, and contrast them directly with visual simulation curve.
The results show that the genetic algorithm can quickly search for a set of control parameters from the overall optimization knowing nothing about the circumstances, and so that it is a highly effective method of tuning PID parameters.
Keywords: Genetic algorithm; PID controller; parameter tuning; simulation
中国石油大学（华东）本科毕业设计（论文）
目 录
第1章 前言 ········································································································1
1.1 PID控制发展概况···················································································1 1.2 PID参数整定方法概述···········································································2 第2章 PID控制算法及参数整定 ····································································4
2.1 PID控制算法···························································································4 2.1.1 PID控制器的基本原理 ···································································4 2.1.2数字PID控制器 ··············································································6 2.2 常规PID参数整定方法 ·········································································9 2.2.1 Ziegler-Nichols整定方法·································································9 2.2.2 改进的Ziegler-Nichols整定方法·················································10 2.2.3 ISTE最优设定方法的经验公式 ··················································· 11 2.2.4 Haalman法的计算公式 ·································································12 2.2.5 KT整定法·······················································································14
第3章 基于遗传算法的PID参数寻优 ·························································16
3.1 遗传算法概述 ·······················································································16 3.2 标准遗传算法操作 ···············································································17 3.2.1 编码方式 ························································································18 3.2.2初始种群的设定 ·············································································19 3.2.3适应度函数 ·····················································································19 3.2.4遗传操作 ·························································································20 3.2.5 收敛性 ····························································································25 3.2.6遗传算法中关键参数的确定 ·························································25 3.3 遗传算法的主要步骤 ···········································································26 3.3.1 准备工作 ························································································26
中国石油大学（华东）本科毕业设计（论文）
3.3.2 基本遗传算法的步骤 ····································································27 3.4遗传算法PID参数整定的编程实现 ···················································27 3.4.1初始群体 ·························································································27 3.4.2 编码 ································································································28 3.4.3 基本操作算子 ················································································29 3.4.4 目标函数 ························································································32 3.4.5 画图 ································································································33
第4章 PID整定方法的仿真应用 ··································································34
4.1 一阶对象 ·······························································································34 4.2 二阶对象 ·······························································································36 4.3 三阶对象 ·······························································································38 第5章 结论 ······································································································40 致谢 ······················································································································41 参考文献 ··············································································································42 附录 ······················································································································43
中国石油大学（华东）本科毕业设计（论文）
第1章 前 言
1.1 PID控制发展概况
PID控制是最早发展起来的控制策略之一，是迄今为止最通用的控制方法，大多数反馈回路用该方法或其较小的变形来控制。PID调解器及其改进型是在工业过程控制中最常见的控制器，至今在全世界过程控制中用的80%以上仍是纯PID调解器，若改进型包括在内则超过90%。我们今天所熟悉的PID控制器产生并发展与1915--1940年期间，尽管自1940年以来，许多先进控制方法不断推出，但PID控制器以其结构简单，对模型误差具有鲁棒性及易于操作等优点，仍被广泛应用于冶金、化工、电力和机械等工业过程控制中。但是同其它控制方法一样，近几十年来，PID控制方法和技术也处于不断发展中，出现过以下几种控制思想:
(1) 自适应控制思想和常规PID控制器相结合的自适应PID控制或自校正PID控制。它既能自动整定控制器参数、能够适应被控过程参数的变化，也具有常规PID控制器结构简单、鲁棒性好、可靠性高等优点。
(2) 智能控制与常规PID控制相结合形成的智能PID控制。它具有不依赖系统精确数学模型的特点，对系统的参数变化具有较好的鲁棒性。
(3) 模糊控制和PID控制器两者的结合具有模型控制灵活而适应性强的优点，又具有PID控制精度高的特点。适用于工业控制过程中大滞后、时变、非线性的复杂系统，它可以不要求掌握受控对象的精确数学模型，而根据人工控制规则组织控制决策表，然后由该表决定控制量的大小。
(4) 神经网络PID控制。神经网络具有自学习能力和大规模并行处理能力，在认知处理上比较擅长。
(5) 预测PID控制。预测控制算法采用非参数模型，不通过复杂的系统辨识来建立过程的数学模型，而是通过检测到的过程响应根据某一优化性能指标设计控制系统，确定一个控制量的时间序列，是未来一段时间内的被调
1