Design And Application Of Quantum Genetic Algorithm For Stochastic Optimization

Quantum genetic algorithm(QGA)is an efficient intelligent optimization algorithm,which can not only increase the diversity of the population,but also make the results more accurate in determining the optimization problem with its unique quantum coding method and revolving gate updating strategy.However,in stochastic optimization problems,the solution often deviates from its theoretical value seriously due to small probability events caused by random factors,resulting in significant errors,or even no optimal solution can be found.Therefore,before comparing the advantages and disadvantages of different solutions,it is more necessary to evaluate the accuracy of the solutions through multiple independent simulation experiments.The optimal calculation amount allocation technology can not only greatly save the calculation amount through order comparison,but also improve the probability of obtaining satisfactory solutions.Therefore,in this thesis,the quantum genetic algorithm(QGA)is improved by combining the optimal computational load allocation technology,so that it can still maintain good performance in solving stochastic optimization problems.The main work is as follows:First of all,the characteristics and principles of the problem are analyzed on the theoretical basis.Aiming at the migration rules between populations,the optimal computation amount allocation technology is introduced,and the optimal computation amount allocation model for multi population quantum genetic algorithm is established.A optimal computation amount allocation rule based on multi population quantum genetic algorithm is proposed,which reduces the calculation times of the algorithm and enables the limited computing resources to be used efficiently and reasonably.At the same time,the multi population is initialized by taking the entropy of the population as the reference,the global search space is rapidly compressed,and the evolutionary goal is updated by the global optimal individual and the sub population optimal individual,thus maintaining the diversity of the population and the search scope.Secondly,the proposed algorithm is tested by stochastic optimization function.In order to avoid the algorithm falling into local optimization,elite population is introduced into multiple populations,and the calculation amount allocation rules of elite population are given.The efficiency of the algorithm and the accuracy of the results are further improved by changing the adjustment strategy of quantum rotation angle.The experimental results show that the improved algorithm has higher convergence speed and smaller error.Finally,the multi population quantum genetic algorithm is applied to the project partner selection risk problem.In order to correctly describe the risk of project partner selection,a mathematical model for risk assessment of partner selection problem is established based on mathematical expectation,with the minimum completion cost as the goal,and with time and risk as constraints.Aiming at the characteristics that random variables obey normal distribution,the model is transformed into a partner selection optimization problem model under a certain environment,and it’s solved by using a multi population quantum genetic algorithm based on the optimal computation amount allocation technology.The results of an example show that the mathematical model and risk assessment method in this thesis are feasible,and the improved algorithm has better effect in simulation.