Research On Orthogonal Genetic Algorithm And Its Applications

Many problems can be converted into all kinds of optimization problems in almost every field of science,engineering,business,etc.. Genetic Algorithms(GAs)are random optimization searching method based on the mechanics of natural selection and genetic theory,and are effective algorithms for solving all kinds of optimization problems. However,the results of a large number of researches indicated that the traditional genetic algorithms have many deficiencies and limitations,such as they are prone to premature convergence,high computational cost and weak local search abilities.To overcome the disadvantages mentioned above,this paper combines orthogonal experimental design methods and local search scheme to genetic algorithm,and two new algorithms are proposed to solve global optimization problems and constrained optimization problems separately.At last,the new global optimization algorithm is used for optimizing the parameters of the new PID immune feedback controller proposed in this paper.The main contribution and work are described as following:(1)Proposed a hybrid self-adaptive orthogonal genetic algorithm (HSOGA).The new algorithm can self-adaptively adjusts the numbers and location of orthogonal array's factors when the two parent chromosome are divided,according to the similarity of the two parent chromosomes,and the selected orthogonal array is used to redesign the crossover operator and a new self-adaptive orthogonal crossover operator(SOC)is proposed.In addional,in the purpose of enhancing the local search ability and speeding up the convergence of the algorithm,a clusting local search scheme,whish is based on population splitting and simplex crossover is proposed.The new algorithm is test on fourteen benchmark functions,and the result demonstrates that the performance of the new algorithm is supper to the other algorithms.(2)Proposed a new constrained optimization algorithm based on orthogonal design method.In search scheme,the orthogonal design method is used to arrange the crossover operation of multi-parent,and a new multi-parent othogonal crossover is proposed.In addition,the simplex crossover is used to enrich the exploratory and exploitive abilities of the algorithm proposed.In constraint-handing technique,a new comparison criterion of comparing and selecting individuals is incorporated to the new algorithm.The new approach is tested on 13 well-known benchmark functions,and the empirical evidence demonstrates that the new approach is generic and effective.(3)Considering the P-type immunity feedback controller can't get over dynamical disturbance and eliminate static error,according to the universal approximation capability of the fuzzy control system,a new design of fuzzy immune PID controller based on the mixed connection of conventional PID and P-type immunity feedback controller is put forward in the paper.In order to gain optimal controller parameters,the parameters of the controller are offline optimized by HSOGA.The simulation results demonstrate that the control performance of the this Immune controller is superior to that of conventional controller and this control method with higher dynamic and static performance...