Open AccessFuzzy day‐ahead scheduling of virtual power plant with optimal confidence level
Author(s) -
Fan Songli,
Ai Qian,
Piao Longjian
Publication year - 2016
Publication title -
iet generation, transmission and distribution
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.92
H-Index - 110
eISSN - 1751-8695
pISSN - 1751-8687
DOI - 10.1049/iet-gtd.2015.0651
Subject(s) - virtual power plant , fuzzy logic , mathematical optimization , credibility theory , monte carlo method , computer science , scheduling (production processes) , fuzzy set , credibility , operator (biology) , mathematics , power (physics) , artificial intelligence , distributed generation , statistics , physics , biochemistry , chemistry , repressor , quantum mechanics , political science , transcription factor , law , gene
To obtain the optimal trade‐off between economy and reliability, this study presents a fuzzy chance constrained programming (CCP) approach to the day‐ahead scheduling of virtual power plant (VPP). In this model, uncertain factors in VPP are characterised by fuzzy parameters, and reserve requirements are formulated as fuzzy chance constraints. Considering that economy and risk of VPP are sensitive to different confidence levels (CLs), it is important to select a proper CL for the operator. Different from a pre‐given CL in most literatures, this study proposes a method to determine the optimal CL, hoping to provide references for the operator in optimisations involving CCP. A synthetic satisfaction function is introduced, which depicts the satisfaction degree of VPP under different probabilities. Meanwhile, the satisfaction function reflects VPP's distinct attitudes toward risk and profit. A matrix real‐coded genetic algorithm combined with Monte Carlo simulation is used to solve this model. To reduce computation burden, the fuzzy chance constraint is converted into its crisp equivalent by utilising credibility theory. Numerical tests are performed in a VPP system, and the best CL is determined through comparing VPP's satisfaction degree under different cases, which prove the validity of the proposed model.