改进的萤火虫算法及其在PID控制器参数整定中的应用

发布时间：2018-04-05 14:20

本文选题：PID控制器　切入点：参数整定　出处：《安徽大学》2017年硕士论文

【摘要】：在当今的工业过程控制中,PID(Proportional Integral Differential)控制器由于结构简单、易实现、可靠性高、鲁棒性能高等特点,因此被广泛的应用。根据相关的统计资料显示,实际的工业过程中,使用PID控制思想的控制器占95%以上。众所周知,PID控制器性能的优劣直接与PID控制器的参数相关联。然而随着现代工业技术的飞速发展,工业控制系统变的越来越复杂,传统的整定PID控制器参数的算法已经不能很好的适应于现代越来越复杂的控制问题。但随着人工智能领域的不断发展和计算机技术不断向自动化领域的渗透,涌现了大量智能算法整定P1D控制器参数的方法。萤火虫算法属于智能优化算法中的一种,因其并行性、自组织、易实现、分布式和鲁棒性等特点被广泛使用。本文将用改进的萤火虫算法(Glowworm Swarm Optimization,GSO)整定PID控制器参数,研究的主要内容概述如下:1、本文提出了一种基于导向机制和自适应步长的萤火虫算法(Based on Directed mechanism and Adaptive-step mechanism GSO,D-AGSO)。一方面,本文在算法中引入了自适应步长机制,使步长可以在算法迭代前期保持较大值,以便可以全局内搜索最优解,防止算法过早成熟,陷入局部最优。在迭代后期,使步长可以保持较小值,防止算法跳过最优解或者出现震荡现象,更加有利于精确寻找最优解;另一方面,对于基本的萤火虫算法,如果萤火虫个体在其自身动态决策域半径内没有找到比自己更亮的萤火虫,则这类萤火虫将不确定随机移动。因此,这类萤火虫虽然付出了大量计算代价,但却没有发生较好的位置移动,并且存在导致算法陷入局部最优的危险且不利于算法的快速收敛。为了减小这种算法缺陷,本文针对这类的萤火虫个体,采取了导向性的移动策略,以加快迭代速度和求解精度。为了验证改进算法的有效性与可行性,本文算法与基本的萤火虫算法、自适应步长的萤火虫算法(Enhanced Glowworm Swarm Optimization Algorithm,EGSO)、基于荧光因子与觅食行为的萤火虫算法(Foraging-behavior Adaptive-step Glowworm Swarm Optimization Algorithm,FA-GSO)进行实验比较,结果证明本文改进的萤火虫算法具有较优的寻优性能。2、本文提出了基于一种新的改进的萤火虫算法整定PID控制器参数的方法,并且使用MATLAB中的Simulink工具建立PID控制器仿真模型,对本文提出的方法进行仿真实验。本文使用四种不同类型的被控对象来验证改进的萤火虫算法整定PID控制器参数的方法的性能,并且本文引入经典Z-N公式整定方法、粒子群算法整定方法、差分进化整定方法与本文的算法进行比较,仿真结果证明本文提出的算法取得了较好的控制效果。3、本文使用改进的ITAE评价函数来验证本文提出的算法,仿真实验证明通过改变评价函数中的性能权重系数,本文算法可有效的使PID控制器偏重于某一特定指标性能。因此本文算法可以整定出工业控制过程中所需要的具有针对性的PID控制器。
[Abstract]:In the industrial process control, PID (Proportional Integral Differential) controller has the advantages of simple structure, easy realization, high reliability, high robustness, it has been widely applied. According to the statistics show that the real industrial process, the controller can control the thought of PID accounted for more than 95%. As everyone knows, associated the performance of the PID controller and the PID controller parameters directly. However, with the rapid development of modern industrial technology, industrial control systems are becoming more and more complex, the traditional PID controller parameter tuning algorithm has not well adapted to the control problem more complex in modern more. But with the continuous development of computer technology and artificial intelligence the field continues to permeate the field of automation, the emergence of a large number of intelligent algorithm method for tuning parameters of P1D controller. The firefly algorithm belongs to intelligent optimization algorithm In one, because of its parallelism, self-organization, easy to realize, the characteristics of distributed and robust is widely used. This paper will use the firefly algorithm (Glowworm Swarm Optimization, GSO) PID controller tuning parameters, the main contents of the study are as follows: 1, this paper presents a firefly the guide mechanism and algorithm based on adaptive step (Based on Directed mechanism and Adaptive-step mechanism GSO, D-AGSO). On the one hand, this paper introduces adaptive step mechanism in the algorithm, the algorithm in the early iteration step can keep larger values, so that you can search global optimal solution, to prevent the algorithm premature, falling into the local optimum. In the iteration later, the step can maintain a smaller value, to prevent the algorithm skip the optimal solution or shock phenomenon, more conducive to the exact optimal solution; on the other hand, the firefly algorithm basic, if Firefly individuals did not find more than their bright fireflies in its own dynamic decision domain radius, then the firefly will not determine the random movement. Therefore, this kind of firefly although pay a lot of computation cost, but did not move a good position and fast convergence of the algorithm into a local danger guide the best and not conducive to the algorithm. This algorithm in order to reduce defects, according to this kind of firefly individual, take mobile strategy oriented, to accelerate the convergence speed and accuracy and feasibility. In order to verify the effectiveness of the improved algorithm, the firefly algorithm in this paper and the basic of the firefly algorithm adaptive step (Enhanced Glowworm Swarm Optimization Algorithm, EGSO), firefly algorithm based on fluorescence factor and foraging behavior (Foraging-behavior Adaptive-step Glowworm Swarm Optimization Algorit HM, FA-GSO) were compared. Results show that the improved firefly algorithm has better optimization performance of.2, this paper proposed a new method to improve the firefly algorithm tuning parameters of PID controller based on MATLAB, and use the Simulink tools to build PID controller simulation model, the simulation experiment of the the method proposed in this paper. The performance of the four different types of objects to verify the firefly algorithm to improve the method of tuning the parameters of PID controller, and introduced the classical Z-N formula tuning method, particle swarm optimization tuning method, compare the differential evolution setting method and the simulation results show that this algorithm. The proposed algorithm has.3 better control effect, this paper uses ITAE evaluation function to verify the improved algorithm proposed in this paper, simulation results show that by changing the performance evaluation function in the Weight coefficient, this algorithm can effectively make the PID controller bias the performance of a specific index. Therefore, this algorithm can set the pertinent PID controller needed in the process of industrial control.

【学位授予单位】：安徽大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TP273

【参考文献】