Research And Application On Biogeography-based Optimization

Biogeography-based optimization(BBO)proposed in 2008 is a novel efficient swarm intelligence optimization algorithm,which searches the optimization through simulating the species migration among habitats in nature.In view of numerous advantages of BBO(e.g.higher global search capability,fast convergence,higher utilization capability for current population information,etc.),BBO opens the new direction for the research field of intelligent computing due to its unique evolution mechanism,and has gradually attracted the attention of numerous experts and researchers.At present,BBO,which has been successfully applied for many scientific researches and engineering applications,such as image recognition,motion estimation,diagnosis of heart disease,production dispatch,synthesis of antenna arrays,etc.,has been proved its feasibility and superiority.However,because of being in its infancy stage,BBO,whose studies are more dispersed and lacking in systematicness,still suffers from the problems of premature convergence,slow convergence rate for high-dimensional multi-modal complex optimization.Furthermore,researches on BBO mostly have focused on single-objective complex nonlinear optimization problems,there are fewer studies focusing on constrained single-objective optimization problems and multi-objective optimization problems with constraint or not.This drawback of BBO mentioned above limits the field of applications to some extent.Therefore,not only can the studies dedicated to solve various types of optimization problems based on BBO complete the theoretical system of BBO,but also can expand the application field of BBO to some extent.This thesis is theoretically devoted to explore and analyze thoroughly the internal mechanism of BBO.In view of the drawbacks of solving single-objective and multi-objective optimization problems with constraints or not based on BBO,abundant researches and simulations have been carried on and a series of improvements have been proposed to enhance comprehensively solving performance on varied optimization problems.Concrete proposed algorithms are as follows.Firstly,in view of the problems of premature convergence,slow convergence rate of BBO in solving high-dimensional multi-modal complex optimization problems,a new modified BBO algorithm with hybrid migration strategy,in which a dynamic selection mechanism to the individuals for emigrating,a migration mechanism based on hybrid migration strategy and a chaotic mutation mechanism is designed,is proposed.Secondly,in view of lacking the capability for solving complex constrained single-objective optimization problems,a new s constrained biogeography-based optimization with dynamic migration is proposed.In this proposed algorithm,the ? level is set based on the constraint violation of the colony.Furthermore,according to the feature of ?constrained method,a new ordering rule based on ? constrained is utilized to obtain the immigration rate and the emigration rate and a new dynamic migration strategy is shown to increase the diversity of population and enhance the search ability of the migration mechanism.Thirdly,in order to improve the performance of BBO in solving multi-objective optimization problems,a new multi-objective optimization algorithm based on hybrid biogeography-based optimization,in which the model of MOEAs applied to BBO is built,and the new self-adaptive method of determining the migration rate and dynamic migration strategy are designed based on the feature on multi-objective as well as the internal mechanism of BBO,is proposed.Fourthly,in view of the drawbacks that constrained multi-objective optimization algorithm cannot be solved by current BBO,a new ? constrained multi-objective optimization based on biogeography-based optimization is proposed.On one side,a new ? constraint handling strategy for multi-objective optimization,in which the constraint violation is redefined and the self-adapting ? level is set,is designed.On the other side,the model of constrained multi-objective optimization which applies to BBO is built.Additionally,not only has BBO been applied in varied numerical optimization problems mentioned above,but also has been applied to design orthogonal waveform of polyphase sequences for MIMO radar.Simultaneously,in order to enable BBO to solve such problem,the method of determining the migration rate,the migration mechanism and the mutation mechanism of species in BBO are improved.The modified biogeography-based optimization algorithm of designing the polyphase sequences based on BBO has achieved better effect,and meanwhile the application field of BBO is expanded.