| With the continuous development of society,countries pay more and more attention to environmental issues.With the massive consumption of coal、oil and other resources,some resources are facing depletion,which brings a lot of environmental problems and restricts the economic development of all countries.The low-carbon economic development mode of low energy consumption and low pollution is valued by all countries in the world.This means that under the premise of ensuring a certain productivity,it is necessary to reduce production energy consumption.Therefore,this dissertation reduces the loss rate of job and reduces the production energy as the optimization goal,study the multi-objective optimization of CSPS system considering energy consumption.The main work is as follows:Firstly,considering the variable service rate,the physical model and working mechanism of CSPS are introduced,and this dissertation models the system as a semi-Markov decision process.Then,the policy iterative algorithm is used to solve the theoretical optimal solution of the problem.However,the policy iterative algorithm is slow to solve and the optimization objective is sum all the objectives by weighting,it needs a certain prior knowledge to determine the weights between the objectives,and the solution set obtained is easily affected by the shape of Pareto front.In order to improve the efficiency and optimization effect,CMOPSO and NSGAⅡ are used to solve the optimization model,and compares with the solution set obtained by policy iterative algorithm.Simulation results show that the nondominated solution set obtained by the multi-objective evolutionary algorithm has a significant improvement in many indicators,but the degree of exploration in some regions is slightly insufficient.Therefore,this dissertation adds an indicator-based population to the original algorithm,and proposes a two population co-evolution algorithm.The two populations influence each other on the basis of independent evolution,which makes up for the shortcomings of their respective algorithms and improves the overall optimization effect of the algorithm.In addition,in order to improve the local exploration ability of the algorithm,a divergence strategy is proposed.The simulation results show that the proposed algorithm has a significant improvement in convergence speed and optimization effect compared with the original multi-objective evolutionary algorithm.According to the characteristics of the algorithm in the optimization,the dissertation also proposes a multi-population two-strategy algorithm,and applies it to CMOPSO and NSGAⅡ algorithm respectively,using different search operators and communication between populations to improve the quality of the Pareto solution set.Multi-population means that the algorithm contains multiple populations with different search operators and two-strategy means that the algorithm adopts different strategies at different stages of evolution.Finally,the effectiveness of the proposed algorithm is verified by experiments. |
