A Bi-level Programming Approach of Economic Dispatch of Electricity Retailers

Abstract

In order to solve the economic dispatch problem of a typical retailer coping with the stochastic risk brought by the distributed generation and demand-side fluctuation, a bi-level optimization model with the introduction of random variables is proposed. In this model, the upper-level sub-problem aims to simulate the market clearing process of electricity and reserve auxiliary services, where market clearing price and market clearing electricity are the decision variables, while the lower-level sub-problem is formulated to minimize the costs of purchasing under time-of-use pricing given by the simulation of market clearing. The model is converted into a deterministic model by chance-constrained programming, and the equilibrium solution is obtained by iterations between the upper and lower layers, where the CPLEX solvers are employed to address the upper and lower level sub-problems, respectively. Finally, the impact of different confidence level on the decision carried out by retailers is clarified through the analysis of the numerical simulation of the bi-level optimization model. It is verified that the model can effectively reflect the impact of the decision-making behavior of retailers on the market clearing process.