Research On Optimization Problem Based On Cloud-based Adaptive Genetic Algorithm

The genetic algorithm is one kind of to draw biological naturalselection and natural genetic mechanisms stochastic optimization searchalgorithm, an efficient algorithm to solve various function optimizationproblem. Coding techniques and genetic manipulation is relatively simple,the requirements are very low, with a parallel and global search ability ofthe restrictive conditions of the optimization problem. Already inmachine learning, pattern recognition, image processing, optimal control,portfolio optimization and management decision-making and other fieldshas been very good application.Traditional genetic algorithm to solve the complex problems thattraditional genetic algorithm is often difficult to balance development andthe ability to explore the search space, there is a greater randomness andblindness, prone to premature convergence and local search ability and convergence slow speed. Adaptive genetic algorithm proposed a certainextent, improve the performance of the algorithm, but also increases thealgorithm into a local optimum possible. To solve the above problem,this paper cloud adaptive genetic algorithm is introduced in thefoundation of the traditional genetic algorithm cloud theory, by thecrossover and mutation probability the X condition generator adaptiveadjustment, the randomness and stable tendency cloud model clouddroplets crossover and mutation probability of both a traditional AGAtrend, the meet fast optimization, and a random, when a the populationsfitness is not an absolute value of zero, help to improve the diversity ofthe population, greatly improving avoid falling into local the optimalcapacity. This paper improves the CAGA algorithm with shrinking searchin the Simulating Fisher fishing Optimization algorithm (SFOA) andproposes the Improved Cloud-based Adaptive Genetic Algorithm (ICAGA), experiments show that, compared with the traditional GA andCAGA,ICAGA has better convergence speed and robustness.Finally,the Improved Cloud Adaptive Genetic Algorithm forreactive power optimization, validation and analysis through typicalexample, and the simulation results of GA and CAGA,ICAGA algorithmwere compared to show that the algorithm can get better global optimalsolution, in order to verify the correctness of the proposed model andalgorithm, the applicability and economy.