基于遗传算法的模糊控制器在二级倒立摆系统中的应用

发布时间：2018-06-21 09:55

本文选题：遗传算法 + 模糊控制　；参考：《华北电力大学》2017年硕士论文

【摘要】：遗传算法简称GA(Genetic Algorithm),是一种全局优化算法。遗传算法是通过研究自然界中生物“优胜劣汰、适者生存”的自然法则,来搜索最优解的一种方法。基于遗传算法的自适应性、高度并行性以及鲁棒性等特点,模糊控制器参数的优化,常常用到遗传算法。但是遗传算法仍存在不足,如早熟收敛,局部搜索效率差等问题。本文针对基本遗传算法的不足,作了如下工作:引进了两个影响交叉概率和变异概率的指标,通过对给定指标的分析,分别给出相应的交叉概率和变异概率增量表,从而指导交叉概率和变异概率的动态变化,使得交叉和变异概率适应性的调整。然后通过对典型测试函数实例计算和与其他文献方法结果对比,证明了改进遗传算法(IGA)是一种有效的全局搜索算法。模糊控制器参考人们的经验并不依赖具体的数学模型,它被广泛地应用于具有复杂非线性系统的控制领域当中。在模糊控制规则一定时,模糊变量的隶属函数在模糊控制器的控制效果中起主要作用。但是隶属函数的确定一般需要试凑的方法,得不到较理想的控制效果。本文用改进的遗传算法来优化模糊控制器中各语言变量的隶属函数,通过Matlab仿真实验,用优化的模糊控制器控制二级倒立摆的稳定性,控制结果表明了算法的有效性。
[Abstract]:Genetic algorithm is a global optimization algorithm. Genetic algorithm (GA) is a method to search for the optimal solution by studying the natural law of "survival of the fittest and survival of the fittest" in nature. Based on the characteristics of self-adaptability, high parallelism and robustness of genetic algorithm, genetic algorithm is often used to optimize the parameters of fuzzy controller. However, genetic algorithm still has some shortcomings, such as premature convergence, poor local search efficiency and so on. Aiming at the shortcomings of the basic genetic algorithm, this paper introduces two indexes that affect the crossover probability and the mutation probability, and gives the corresponding increment table of the crossover probability and the mutation probability by analyzing the given index. Therefore, the dynamic variation of crossover probability and mutation probability can be guided, and the adaptability of crossover and mutation probability can be adjusted. Then, it is proved that the improved genetic algorithm (GA) is an effective global search algorithm by calculating the typical test function examples and comparing with the results of other literature methods. Fuzzy controller is widely used in the field of control with complex nonlinear systems, which does not depend on specific mathematical models for reference to people's experience. The membership function of fuzzy variable plays an important role in the control effect of fuzzy controller when the fuzzy control rules are fixed. However, the determination of membership function usually needs the method of trial and error, so it can not get ideal control effect. In this paper, an improved genetic algorithm is used to optimize the membership function of each language variable in the fuzzy controller. Through the Matlab simulation experiment, the stability of the two-stage inverted pendulum is controlled by the optimized fuzzy controller. The control results show that the algorithm is effective.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TP13

【参考文献】