单亲遗传算法对于在梯级水力发系统中多级电站的经济分配方法的改进作用外文翻译.docx

外文资料原文An improved partheno genetic algorithm for multi-objective economic dispatch in cascaded hydropower systemsAbstract：The multi-objective economic dispatch (MOED) problem in cascaded hydropower systems is a complicated nonlinear optimization problem with a group of complex constraints. In this paper, an improved partheno genetic algorithm (IPGA) for resolving the MOED problem in hydropower energy systems based on the non-uniform mutation operator is proposed. In the new algorithm, the crossover operator is removed and only mutation operation is made, which makes it simpler than GA in the genetic operations and not generate invalid offspring during evolution. With the help of incorporating greedy selection idea into the non-uniform mutation operator, IPGA searches the solution space uniformly at the early stage and very locally at the later stage, which makes it avoid the random blind jumping and stay at the promising solution areas. Finally, the proposed algorithm is applied to a realistic hydropower energy system with two giant scale cascaded hydropower plants in China. Compared with other algorithms, the results obtained using IPGA verify its superiority in both efficienc.Key words: multi-objective optimization; Improved partheno genetic algorithm; Non-uniform mutation operator. Cascade hydropower station group of. Introduction Optimization of multi-objective economic dispatch (MOED) in cascaded hydropower systems is one of the most complicated issues in water resources management as it typically involves trade-offs. For example, a single multipurpose reservoir, which not only serves hydropower but also navigation, its dispatcher may wish to maximize benefits from hydropower generation, while releasing sufficient water for navigation to satisfy the demands. However, a higher profit from hydropower generation would conflict with the navigation releases, that is to say, any improvement of one objective can be achieved only at the expense of another. The curve or surface (for more than 2 objectives), describing the optimal trade-off solutions between the objectives, is known as the Pareto front. In real life, most of reservoir systems serve multiple purposes and they are multi-objective in nature. Due to the dispatch rules for a joint operation of cascade reservoirs enable to develop the capacity of hydropower generation, the MOED problem in cascaded hydropower systems becomes an active research area in recent years. The goal of MOED in cascaded hydropower systems is to determine the water discharge process of all hydropower stations during the scheduling period in order to maximize the total benefit while fulfilling various actual water demands and other complicated constraints simultaneously. Because of the complex power and hydraulic relations between cascaded hydropower systems, the multi-objective optimal operation of cascade hydropower stations is a large scale, dynamic, and strong coupling nonlinear problem, which involves many variables, such as, inflow, storage, discharge, water level, water head, output and generated energy. So far, different methods to solve the MOED problem have been proposed and discussed by many researchers. The traditional methods, such as, mixed integer linear programming (MILP), Lagrange relaxation (LR), nonlinear programming (NLP), dynamic programming (DP) and progressive optimality algorithm (POA), have been widely applied in the past. And many achievements have also been obtained. Nevertheless, all of the methods listed above exist some shortcomings which make them less efficient and even difficult in searching for the optimal solution. When MILP is employed to solve the MOED problem, linearization of the multi-objective functions and constraints could deviate the original problem, which makes the non-inferior solution inaccuracy. For NLP methodology, some approximate approaches are adopted to deal with discontinuous, non-differentiable and non-convex multi-objective functions, and these methods are computational expensive. Lagrange multipliers updating strategy has an unfavorable influence on the efficiency of LR, which makes the stability of solution poor. Though DP method can resolve the optimal operation of a single multipurpose reservoir, it suffers from the ‘‘curse of the dimensionality’’ which makes the computation time increase dramatically when the dimension of cascaded hydropower systems increases. POA is widely used in the multi-objective optimal operation of cascaded hydropower plants, but it is sensitive with the initial solution, which generally shrinks the searching area and traps it in the local Pareto optimal front easily. Recently, there has been an increasing interest in adaptive heuristic search algorithms modeled from the biologically motivated adaptive systems, for solving the MOED problem, because of their powerful global searching capacity, such as genetic algorithm (GA), ant colony optimization (ACO) algorithm, parti。