Improved Moea/D Approach To Many-Objective Day-Ahead Scheduling With Consideration Of Adjustable Outputs Of Renewable Units And Load Reduction In Active Distribution Networks

Jingrui Zhang,Xiaoqing Zhu,Tengpeng Chen,Yanlin Yu,Wendong Xue
DOI: https://doi.org/10.1016/j.energy.2020.118524
IF: 9
2020-01-01
Energy
Abstract:Within the increasing concerns on environment, economy and security and requirements on power quality, the conventional economic dispatching scheme will face challenges in the optimizing objectives and constraints of the day-ahead optimal scheduling of active distribution network (ADN). This paper proposes a many-objective (the number of objectives is more than three) day-ahead optimal scheduling model for the ADN. With consideration of adjustable outputs of renewable energy units and reducible load, four objectives including the minimization of total operating cost, minimization of active network loss, minimization of voltage deviation and minimization of the total output reduction rate of renewable energy are involved satisfying various constraints such as component constraints, load constraints and network constraints. An improved multi-objective evolutionary algorithm based on decomposition (MOEA/D) is put forward to solve this many-objective optimization problem. In the proposed approach, normalizing objective functions is firstly implemented before calculating the aggregation function in order to reduce effects of different magnitude orders of objectives. Then, a new aggregation method is utilized instead of the conventional Tchebyshev aggregation method to calculate the improvement rate, which is the basis for allocations of computing resources. The new aggregation function utilizes the weighted sum of horizontal distance and vertical distance between the subproblem and the corresponding weight factor, in which not only convergence but also diversity is considered. Moreover, a dynamic neighborhood replacement strategy is also employed to avoid the mismatch between generated solutions and subproblems and to balance the population diversity and convergence. The effectiveness of the proposed model and the improved MOEA/D are verified by the simulation analysis of the IEEE33-bus power system. (C) 2020 Elsevier Ltd. All rights reserved.
What problem does this paper attempt to address?