A Study On Power Supplier's Bidding Strategy Based On Agent Simulation

Based on the monthly agreement and the real time operation rules of powermarket, this thesis develops a monthly market bidding strategy and a real timemarket bidding strategy respectively from a power supplier's view. Public marketdata is used to support algorithms of those two previous strategies. The thesisissues an ANN and chance constrained programming combined approach in themonthly bidding strategy and a two-population evolutionary algorithm method inthe real time bidding strategy. Besides the strategies, a bidding decision system isdeveloped in the thesis as an important part of the study. Results conforms thesystem is able to simulate the strategies successfully.