Purchasing Portfolio Model Based On Robust Optimization For Load Serving Entities

Aiming at macroeconomic environment and market environment of power market, this paper fully consider the the complexity of the load serving entities(LSE) in more than one investment in electricity market and the sensitivity of the electricity purchasing portfolio model to parameters, and apply the robust optimization method to purchasing portfolio model of load serving entities(LSE) electricity, then use it to express and solve the problems of data and parameter uncertainty, establish Power company's multi-market purchase of electricity distribution model based on Robust Mean-Semi Deviation modle and Robust Worst-Case VaR model, which provide new ideas for risk controlling and multi-market electricity purchasing of load serving entities(LSE). The main contents of this article are summarized below.The first chapter, the introduction part of the paper analyze the research background,significance and current research situation of load serving entities(LSE) multi-market purchase portfolio model in electricity market.Then lists the outline of this paper.The second chapter, we introduce the mathematical theory background, including several common forms of robust optimization, duality theory, and Worst-case VaR and its uncertainties.The third Chapter, it is based on the random fluctuation characteristics of the electricity-purchasing price of the power company in the multi-market, and A Robust Mean-Semi Deviation modle is presented for developing the optional electricity purchasing strategy in multiple markets,which quantize the risk of LSES with the Mean-semi deviation and apply Robust optimization processe price uncertainty.Finally using the GX Power Grip Corp(GXPGR)'s date for empirical analysis and it has demonstrated the validity and applicability of the model, indicating that this model has a certain reference value for the power company's investment portfolio decisions.The fourth chapter, we consider the two uncertain robust investment choice problem of purchasing electricity for load serving entities(LSE) in Multiple real-time electricity market:the uncertainty of distribution at exit time and the uncertainty of investment income conditional distribution at exit time. In case part of the information of the distribution at exit time and the conditional distribution of ROI is availiable, we use Worst-Case VaR model as the multi-market electricity-purchasing and investment strategy model for load serving entities(LSE), and then express correspond issues as an semi-definite programming problem which can be effectively solved. At last, we use Guangxi Power Grid data for empirical analysis, the results demonstrate the effectiveness and applicability of the model, which have certain guiding significance to the Portfolio decisions for load serving entities(LSE).