基于PSO算法优化的模糊PID异步电动机控制系统研究

发布时间：2018-04-19 21:31

本文选题：异步电动机 + 矢量变频控制　；参考：《湖南科技大学》2014年硕士论文

【摘要】：交流异步电动机以其结构简单、价格低廉、坚固耐用、并能在恶劣条件具有良好工作状态等优点被广泛应用于生产、生活的各个领域，如何提高异步电动机的调速性能已经越来越受到学者们的关注。随着电力电子技术的快速发展，使得矢量控制变频调速技术得到了更大的应用空间，近年来，大量的智能控制理论也逐渐被运用到异步电动机矢量控制变频调速系统中。其中，模糊PID控制因其控制效果优良且不依赖被控对象的模型等优点，使异步电动机调速系统的动、静态性能得到进一步提高。但是，想要获得优良的模糊控制器要求设计者必须要有足够的经验去选择隶属度函数和控制规则表，并且在设定后无法在线更改，使得模糊PID控制存在一定的局限性。针对上述问题本文将模糊控制器中的隶属度函数和控制规则以十进制的方式进行编码，定义两种不同的粒子分别表示隶属度函数和控制规则，在粒子群算法中进行寻优，并以ITAE规则作为适应度函数。完成改进PSO算法与模糊PID控制相结合，实现模糊控制器的在线整定以获得更强的适应能力。粒子群算法具有简单、易于实现等优点在科学与工程领域得到了很好的验证，但是粒子群优化算法与其他进化算法一样存在容易陷入局部最优和早熟收敛等缺点。文章分析了其存在缺点的主要原因，并标准粒子群算法的基础上提出了一种改进的粒子群优化（DZIA-PSO）算法。首先通过matlab对算法中粒子的运行轨迹进行可视化处理，更加直观的观察粒子的整个寻优过程，并总结算法出现早熟现象特点。然后，针对粒子群局部收敛时全局最优位置的更新停滞的现象，利用sharing函数对局部最优范围内的粒子进行重新初始化并标定为死区，这种方法显著的提高了粒子群算法在寻优过程中的种群多样性。最后，利用标准测试函数对本文的改进算法进行仿真分析，结果表明相比于LDW-PSO算法改进后的算法具有更好的克服局部最优能力。本文首先介绍了异步电动机的控制原理及数学模型，并在此基础上介绍异步电动机的矢量控制变频调速系统。然后，本文完成了基于PSO算法优化的模糊PID异步电动机的控制系统设计，包括软件部分和硬件部分，其中硬件设计是由以DSP芯片为核心的控制部分和以IPM为核心的驱动部分组成，，软件算法中包括了磁链环与转矩环的PI控制算法和速度环的基于粒子群优化的模糊PID控制算法。最后，利用Matlab/Simulink搭建基于粒子群优化的模糊PID控制异步电动机调速系统的仿真模型，在电机的空载启动、突加负载以及相同负载下改变转速等情况进行了仿真。仿真实验验证了相对常规的模糊PID控制，基于粒子群优化的模糊PID控制在调速过程中具有响应快、超调低、带负载启动能力强等优点。
[Abstract]:Ac asynchronous motor is widely used in many fields of production and life because of its simple structure, low price, strong durability, and good working condition in bad conditions. How to improve the speed regulation performance of asynchronous motor has been paid more and more attention by scholars. With the rapid development of power electronics technology, vector control frequency conversion speed control technology has a greater application space, in recent years, a large number of intelligent control theory has been gradually applied to the asynchronous motor vector control frequency conversion speed control system. Among them, the fuzzy PID control has the advantages of excellent control effect and independent of the controlled object model, which makes the dynamic and static performance of the asynchronous motor speed regulation system further improved. However, in order to obtain a good fuzzy controller, the designer must have enough experience to select membership function and control rule table, and can not be changed online after setting, which makes fuzzy PID control have some limitations. In this paper, the membership function and control rule in fuzzy controller are coded in decimal form, and two kinds of particles are defined to represent membership function and control rule respectively, which are optimized in particle swarm optimization algorithm. The ITAE rule is used as fitness function. The improved PSO algorithm is combined with the fuzzy PID control to realize the on-line tuning of the fuzzy controller to obtain better adaptability. Particle swarm optimization (PSO) has many advantages, such as simple and easy to implement, which are well verified in the fields of science and engineering. However, PSO has the same disadvantages as other evolutionary algorithms, such as local optimization and premature convergence. This paper analyzes the main reasons for its shortcomings and proposes an improved particle swarm optimization (PSO) algorithm based on the standard particle swarm optimization (PSO) algorithm. Firstly, the particle trajectory in the algorithm is visualized by matlab, and the whole process of particle optimization is observed more intuitively, and the precocious phenomenon of the algorithm is summarized. Then, aiming at the phenomenon that the global optimal position stagnates when the particle swarm is locally convergent, the particle in the local optimal range is reinitialized by sharing function and demarcated as a dead zone. This method significantly improves the population diversity of PSO in the process of optimization. Finally, the improved algorithm is simulated by the standard test function. The results show that the improved algorithm has better ability to overcome the local optimum than the improved LDW-PSO algorithm. This paper first introduces the control principle and mathematical model of asynchronous motor, and then introduces the vector control variable frequency speed regulating system of asynchronous motor. Then, the design of fuzzy PID asynchronous motor control system based on PSO algorithm is completed, including software and hardware. The hardware design is composed of the control part with DSP chip as the core and the driving part with IPM as the core. The software algorithm includes the Pi control algorithm of magnetic chain loop and torque loop and the fuzzy PID control algorithm based on particle swarm optimization of speed loop. Finally, the simulation model of fuzzy PID control asynchronous motor speed control system based on particle swarm optimization (PSO) is built by using Matlab/Simulink. The simulation is carried out under the condition of no-load starting, sudden loading and changing speed under the same load. The simulation results show that compared with the conventional fuzzy PID control, the fuzzy PID control based on particle swarm optimization (PSO) has the advantages of fast response, overturning, and strong starting ability with load.
【学位授予单位】：湖南科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM343

【相似文献】