Multi-Objective Decision Making Method For Hydrothermal Power Systems Based On Data Envelopment Analysis

Electric power industry is the leading industry to support the development of the national economy, and its characteristics such as high energy consumption, large emissions also have become the key areas of energy saving and emission reduction. At present, as the main energy production in our country, hydroelectric and thermal power play an important role in the power structure. Therefore, through the coordination of hydrothermal power system operation mode, it can improve energy resource utilization efficiency, realize the economic operation of power system, alleviate the contradiction between supply and demand of energy and promote environmental protection.To coordinate the multiple optimization index of hydrothermal power system operation, taking into account the coordination in saving energy, water, and emission reduction, a coordinated multi-objective optimization scheduling model for hydrothermal power systems is constructed in this paper. Aiming at the conflicting and incommensurable characteristics of all sub-goals for multi-objective problem, the basic theory of multi-objective problem and data envelopment analysis (DEA) is elaborated on. Based on DEA effectiveness and multi-objective optimization problem pareto solution, a multi-objective decision making problem solution method with particle swarm optimization algorithm and the C2R model of data envelopment analysis is presented. In solving process, the constraint violation is treated as the fitness function to guide the algorithm converges to a feasible solution domain. Because of the defects such as the premature convergence of basic particle swarm algorithm, easy to fall into local optimum and low accuracy, it’s necessary to improve the efficiency of algorithm. So a improved strategies based on the dynamic adjustment of learning and social factors is presented to improve the convergence of the algorithm performance and accuracy.Using two hydrothermal power systems simulation analysis, the simulation results show that the constraint conditions as algorithm fitness function can move towards the direction of the constraint violation decrease and overcome the drawback of selection of appropriate penalty factor by using the penalty function method in constraints. In the iteration process, constraint violation and DEA efficiency value of each optimization results evaluated by data envelopment analysis as optimization criterion, can make the algorithm move towards the Pareto optimal front of multi-objective optimization problem effectively. This method for solving multi-objective optimization problem of hydrothermal power system doesn’t need consider factors such as difference of properties and dimension in multiple targets. And there is no need to set the weight coefficient for each index. So it not only avoids the subjective randomness of artificial preference weight impacting on optimization results, but also provides decision makers with abundant feasible scheme for the decision. In this paper, the correctness of the proposed model and the feasibility and effectiveness of the proposed method were demonstrated by the simulation results.