Multi-objective probability optimiz

Multi-objective probability optimization has been widely used in engineering optimization design in specific fields, mainly focusing on the use of intelligent optimization algorithm in static sampling. For example, Zhang Guoxin et al. [3-7] built mopop model for the decision-making of air drop breakthrough point and reservoir resource scheduling, transformed the weighted objective function into a single objective probability optimization model, and then obtained the single objective hybrid genetic algorithm solution by means of random simulation or fuzzy simulation, neural network, compromise algorithm; many scholars [10-12] mopop optimized for the planning of automobile charging station and other projects In the model, membership function and weight coefficient are used to transform the model into a single objective programming model, and the improved single objective bat algorithm and discrete (or continuous) particle swarm optimization algorithm are designed to solve the problem. For the study of solving mopop with multi-objective intelligent optimization algorithm, virivinti et al. [8-9] used fuzzy simulation or random simulation and Latin hypercube sampling to deal with noise in industrial grinding and optimal load reduction of power grid with uncertain parameters, and then used NSGA-II to solve the problem; Yang Xinyu et al. [13-14] coordinated scheduling of MRO service resources By means of stochastic simulation and BP neural network, the noise is suppressed and solved by multi-objective particle swarm optimization or multi-objective difference algorithm. In the study of deterministic transformation model, Li Xiaona et al. [1-2] established a multi-objective chance constrained programming model for the joint scheduling of multiple water sources, and transformed the model into a single objective optimization model with analytical formula through complex transformation and linear weighting. In addition, for the optimization of the combined scheduling of wind and fire described as mopop, Li Zhuohuan et al. [15-16] used the dual interior point method, the normal boundary crossing method, the mixed interactive fuzzy programming method to solve the problem, and selected the compromise solution by the order preference approximate ideal solution technology.