Quantum Swarm Intelligence Algorithm And Its Application In Controller Optimization Design;quantum Swarm Intelligence Algorithm And Its Application In Controller Optimization Design

Intelligent optimization algorithms developed by simulating or revealing some phenomena or processes of nature, afford new ways for solving complex problems, as the algorithms don't need the accurate mathematical models of the objects. Swarm intelligence algorithm is an important branch of intelligent optimization algorithms. Because of its superior abilities, swarm intelligence algorithm, especially the Ant Colony Optimization algorithm (ACO) and Particle Swarm Optimization algorithm (PSO), has already been widely used in many fields. However, some shortcomings still unavoidably exist in ACO and PSO. In order to improve the shortcomings of ACO and PSO, one method is to introduce quantum mechanism to ACO and PSO respectively. Meanwhile, this method is a new way to solve optimized problems.The paper introduces the basic principle of ACO and PSO, and sums up the features of the two algorithms firstly, and than analyzes several improved ACO and PSO. All of this is the theory basis of the following research.The main disadvantage of ACO is the slow search speed, and the other is stagnation exists in the research process. In order to improve these two disadvantages and make ACO solve continuous domain optimization problems more available, the continuous quantum ant colony algorithm (CQACA), including the basic idea and principle, is introduced in this paper. Several improved strategies of the CQACA are also proposed. First, according the research process, the parameters including global selection factor and pheromone volatilization coefficient will change adaptively. Second, add a quantum cross operation to the algorithm. Finally, the quantum rotating gate was calculated in a new method. The improved quantum ant colony algorithm (IQACA) is used in function optimization and neural network weights optimization. Simulation shows the optimization results of IQACA are better than CQACA.Parameters of a controller are very important to the control effect. Always the parameters of a fuzzy controller and a PID controller are very difficult to determine. In the paper, a novel design method based on IQACA for fuzzy controller and PID controller is proposed. The IQACA is employed in this method and the optimal control parameters could be founded automatically to make the selected fitness function minimized. Simulation results also indicate that the tuning method is available and feasible.Aim at PSO can not research the global optimal solution with probability 1, quantum particle swarms optimization algorithm (QPSO) for continuous space problems is presented. Simulation by test functions shows the search ability and optimal efficiency of QPSO are better than PSO.Finally, a novel design method for active-disturbances rejection controller (ADRC) is proposed. The structure and principle of ADRC are giving, and than giving the mathematic description of ADRC. In matlab the ADRC simulation system is completed by system function. The results of experimental prove the validity and rationality of the design method.