Study On Theory And Applications Of Intelligent Optimization Algorithms Based On Immunity

As optimization problems exist widely in all domains of scientific research and engineering application,research on optimization methods is of great theoretical significance and practical value.Since traditional deterministic optimization methods show up a lot of shortcomings,they are difficult to solve more and more complex problems in modern society.The new intelligent optimization algorithms based on biological intelligence or natural phenomenon have characters of simple operations,generalization, robustness and parallelism,so they have become the powerful tool to solve complex optimization problems.Biological immune system is a highly parallel adaptive information learning system, which can identify and remove the antigen invading the body.The system has the ability of study,memory and adaptive adjustment.With comparison to other intelligent optimization algorithms,the algorithm base on immune can overcome effectively premature convergence and diversity scarcity.Therefore,developing new intelligent tools and establishing hybrid intelligent algorithms have become a new research focus of artificial intelligence with reference to immune mechanism.Inspired by the immune system,research on improved immune algorithms is carried through by combining with other intelligent methods in this dissertation.The hybrid optimization technology integrated with multiple intelligent methods is introduced to provide new practical technology for complex engineering problems.The main work can be summarized as follows:1.Simple genetic algorithm has a slow convergence velocity in late evolution and gets premature convergence easily.To solve these.problems,a genetic algorithm based on immunity learning mechanism is proposed on the basis of immune mechanism.The algorithm not only keeps the leading position of excellent antibody,but also develops the potential of rapidly growing antibody in seeking optimum.Under the action of excellent memory cell,the search of algorithm to global optimum is rapid and effective.Simulation results show that the algorithm has better global convergence ability and rapider convergence velocity through optimization tests of benchmark functions.Applying RBF network PID controller based on the optimization of ILGA,a 3D hover model system is realized steady-state control.Experiment results show the method possesses well control quality compared with LQR control.It has good adaptability,robustness and quick response speed.2.A new genetic algorithm based on specific immunity was proposed.The algorithm combines niche technique with specific immunity strategy in view of the mechanism of immune response.Simulation results show the algorithm can improve population diversity effectively and global convergence ability.Compared with homogeneous immune algorithm,the genetic algorithm based on specific immunity has higher convergence success probability and smaller average iterations.Applying the algorithm to the optimal design of T-S fuzzy neural network controller,the controller can control a double inverted pendulum system well.Experiment results demonstrate the method has ideal dynamic and steady performance,anti-disturbance and robustness.3.Conventional evolutionary programming gets premature convergence easily.To solve the problem,an immune programming based on double mutation operators is proposed in view of the mechanism of immune response.The key to the algorithm lies in using global Cauchy mutation operator and local Gauss mutation operator.In order to make the searching rapid and effective,the algorithm maintains the diversity of population and performs the strategy of memory protection,immature protection.The theoretical analysis and simulation results show the algorithm have higher convergence speed and solution precision than conventional evolutionary programming.4.Aimed at multimodal function optimization,an immune network algorithm is proposed based on double mutation operators.The algorithm refers to the clonal selection and immune network theory.Double mutation operators are adopted to improve global and local searching ability.The strategy of dynamic network suppression is used to maintain the diversity of population,and to adjust adaptively the scale of antibody population. Simulation results show that the algorithm can not only improve population diversity effectively,but also combine well global optimization with local optimization.Therefore,it has excellent optimization performance to multimodal function.5.Referred to the character of particle swarm optimization and immune network theory,an immune particle swarm network algorithm is proposed.By making use of the information sharing and memory function of particle swarm,the cognitive part based on its Own experience has been enhanced to improve local searching ability of the algorithm.The strategy of dynamic network suppression has been used to maintain diversity of population, and adjust adaptively the scale of particle swarm.Simulation results of typical test functions show the hybrid algorithm reduces the required iterations during optimization search,and effectively improves the optimization succeed probability.Aimed at the problem of wireless sensor node deployment,the optimal coverage control scheme based on immune particle swarm algorithm is introduced.The immune particle swarm algorithm is used to seek the optimal place of sensor node in different state,and realize maximum network coverage area in wireless sensor net.Experiment results demonstrate that the control scheme can implement mobile sensor deployment more efficiently and rapidly.