Research And Application Of Multi-Objective Optimization Algor Ithm Based On Three-Way Decision

In the field of practical engineering and scientific research,multi-objective optimization problems are increasing.Natural computational methods have emerged to provide new ideas for solving multi-objective problems,but many existing methods still have some defects and shortcomings in solving multi-objective optimization problems.For example,the algorithms have poor solution accuracy,slow convergence speed,and the existence of constraints,which lead to the traditional multi-objective optimization algorithms can not meet the needs of practical problems,and it is difficult to obtain the optimal solution quickly.The emergence of three-way decision provides new ideas for solving complex decision problems.Three-way decision is a decision making method that can simultaneously take into account both global and local criteria,and its essential idea is in line with the human thinking of solving uncertainty problems.To improve the efficiency and convergence speed of the method for solving multi-objective optimization problems,based on the theory of three-way decision,this paper proposes a new natural computing strategy for multi-objective optimization based on three-way decision,(3WD-MEA),which combines this strategy with the natural computing method to balance the convergence and diversity of the algorithm,obtain good solution set distribution and solution accuracy,and solve the problems that the multi-objective optimization algorithm is easy to fall into local optimum and poor initialized population diversity.The main research work of this paper is as follows.(1)Using the improved Tent mapping chaotic initialized population,combined with the idea of three-way decision,we propose a population partitioning strategy based on the idea of three-way decision.The three-way decision divide the whole into three parts:positive domain,negative domain,and boundary domain,and divide all solutions into three categories: accepted,rejected,and delayed decision individuals.By balancing different objectives and constraints through the division of the three domains,a more comprehensive solution to the multi-objective optimization problem is obtained.Under the guidance of heterogeneous partitioning strategy,the sub-domain population is subjected to a multi-strategy mixture of partitioning methods to find the optimal solution,thus guiding the population evolution,ensuring the pre-diversity while making the accepted individuals uniformly distributed and able to be searched accurately,effectively improving the convergence accuracy of the algorithm and further investigating the treatment of multi-objective optimization problems in the 3WD-MEA algorithm.(2)The strategy proposed in this paper is applied to the classical multi-objective natural computation method,and the simulation experiments are conducted on eight test functions to evaluate and verify the stability and comprehensive performance of the convergence speed of the algorithm,which is experimentally verified to have good convergence and diversity,and good generality and universality.Taking the optimal pump scheduling problem as an example,the overall objective of this optimization problem is to minimize the cost of water supply,effectively balance the constraints of power cost and maintenance cost,and obtain a more excellent cost solution for pump scheduling.Through simulation and experimental results analysis,it is proved that the strategy proposed in this paper has a good effect,which provides new ideas and directions for the research and application of multi-objective optimization algorithms and has important value and significance for practical engineering applications.