Multi-Objective Energy Consumption Scheduling in Smart Grid Based on Tchebycheff Decomposition.

Hui Lu,Mengmeng Zhang,Zongming Fei,Kefei Mao

DOI: https://doi.org/10.1109/tsg.2015.2419814

IF: 10.275

2015-01-01

IEEE Transactions on Smart Grid

Abstract:Demand response is an essential issue in smart grid. The central problem is balancing the user cost and the social utility. We focus on the multi-objective energy consumption scheduling problem based on the third-party management. The aim is to provide diverse, uniformly-distributed, and accurate solutions to the third-party decision-maker. The novel contribution of this paper is that it provides an exact choice in energy consumption scheduling. First, we investigate the mathematical model, which dispatches the power consumption for different users in different time slots considering the users' preferences. Then, we propose a matrix-encoding scheme. The energy matrix and the demand matrix are the key factors. The constraints are handled based on the dot product of the two matrixes. In addition, we adopt a scheduling algorithm based on Tchebycheff decomposition. We define several metrics to evaluate the quality of the solutions for the decision-maker. The neighbor generation distance is proposed to reflect the convergence. The metric S and the metric C are used to represent the diversity and coverage, respectively. The metric HV is used to give a comprehensive evaluation. The simulation illustrates that the proposed algorithm outperforms the non-dominated sorting genetic algorithm (NSGA)-II in convergence, diversity, and coverage. It obtains a wider search region at a faster search speed than the NSGA-II algorithm.