| Swarm intelligence optimization algorithm by simulating or reveal certain natural phenomena and processes, developed from biological group of intelligent behavior. Since there is no requirement about continuous and differentiable for the optimization objective function, the result is not depend on the selection of initial values, and it has a global, parallel efficiency, strong robustness and generality, etc.Therefore, the study of swarm intelligence algorithm has important theoretical significance and practical value. This paper studied several kinds of typical swarm intelligence optimization algorithm in recent years for global function optimization and applications, specific work is as follows:First of all, fireworks algorithm(FA) is proposed as a simulation of the fireworks explode new type search algorithm in parallel, by adjusting the way of fireworks explosion to balance global exploration and local search ability of the algorithm.Through setting parameters of the initial fireworks and son sparks to verify the optimization performance of FA. By introducing hybrid leapfrog algorithm grouping strategy, FA- SFLA hybrid algorithm is proposed, the algorithm has better jumped out of local optimal and accelerated the global search ability. Simulation results show that the parameters have great influence on the performance of FA, the hybrid algorithm greatly increases the accuracy of solving the function optimization and optimization ability.Secondly, put forward a kind of the gravitational search algorithm(GSA), which based on the law of gravity. Detailed discusses the basic principle and process,analysis for reasonable setting of core parameters,Through simulation to optimize test functions to verify its performance. For GSA existing problem of slow convergence speed and poor local optimization ability, four improved GSAPSO hybrid algorithm is proposed which based on the literature, introduce a small constant update strategy,enhanced the speed, acceleration and the best individual position update feature,moreover, using PSO algorithm to optimize the speed and position of GSA, thus improve the convergence speed and optimization precision. Simulation results show that the optimization settings of GSA parameters provide the flexibility to improve the inhibition rate of the algorithm and improve the accuracy of the solution, compared with the simple algorithm, the hybrid algorithm has better optimization ability.Finally, Biogeography-based optimization algorithm(BBO) is put forward, it according to species migration among habitats to complete information circulation and sharing, by improving the adaptability of habitat to achieve global optimization.Design principle, the algorithm flow of the BBO has carried on the comprehensive analysis, put forward function optimization based on BBO algorithm and comparing with other swarm intelligence algorithm. Based on the adaptive population migration mechanism of BBO algorithm, given seven mobility model by reference related literature, complete different migration methods on analysis and comparison of the influence of the algorithm performance. In order to verify the BBO optimize performance of the mobility of high order nonlinear model, put forward eight kinds of mixed migration model. The optimal chaotic mapping and optimal migration model of BBO combined, the Chaotic Biogeography-based Optimization Algorithm(CBBO)is proposed.Comparing the model in the mapping of performance for function optimization, the simulation results show that the BBO algorithm has a good optimization performance, close to the nature of the mobility model both high order and combined with the chaos map, all have the optimization of high precision and fast convergence.In a word, through the simulation, it shows that the three species of intelligent algorithm parameters to set or combined with other algorithms to optimize functions were all achieved good optimization effect, it is of great significance for solving complex problems and its practical application. |
PDF Full Text Request