An Improved Immune Genetic Algorithm And A Pid Controller Optimized Design

Up to now, Genetic Algorithms (GAs) have been applied in many industrial areas. But, some major drawbacks of GAs, such as premature convergence, easily getting trapped in local optima, low local search ability, and slow convergence speed, have been found. The premature convergence and easily getting trapped in local optima are mainly related to the selection pressure of GAs and can be overcome by using the technique of fitness scaling, which is usually problem-dependent.The essential difference between a GA and an Artificial Immune-Algorithm (AIA) is that an AIA mimics the antibody reproduction strategy in the natural immune system and employs concentration regulation mechanism, which makes the antibodies with high fitness and low concentration proliferate and at the same time restrain the antibodies with high concentration. Thus, an AIA can regulate the selection pressure efficiently, keep the diversity of solution set, and overcome some drawbacks of GAs such as premature convergence and easily getting trapped in local optima. On the other hand, there exist drawbacks in AIAs such as slow running speed and convergence speed.In this thesis, new definitions and formulas of antibody similarity, expected reproduction probability, and selection probability are proposed. Based on these definitions and the elitism strategy, a novel immune-genetic algorithm is presented, which is called the immune-genetic algorithm with elitism (IGAE). IGAE has two important properties. The first is that the similarity and expected reproduction probability of antibody can be adjusted dynamically in the evolutionary process of the antibody population to balance the diversity of the population and the convergence speed of the algorithm, which helps the algorithm find the high-quality solutions rapidly. The second is that the algorithm is able to find the globally optimal solution because of the use of elitism strategy.Based on IGAE, a novel optimal design method for PID controller is proposed. In this method, the ITAE criterion is adopted as the objective function of the optimization problem and IGAE is used to optimize the three gain parameters of PID controller so as to obtain the optimal gains. The designed controller is called IGAE-PID controller. The results of the simulation experiments on the four typical control objects show that IGAE-PID controller has good control performance and robustness. Compared with the standard differential evolution algorithm, the canonical genetic algorithm with elitism strategy, and the standard simulated annealing algorithm, IGAE-PID controller exhibited better or equivalent control performance. Furthermore, the simulation results also verified that the proposed IGAE algorithm has better performance in convergence speed and dynamic convergence behavior.