Research On Unit Commitment Problem Considering Price

Unit commitment (UC) problem, which is a large scale mixed integerprogramming problem with lots of constraints, is one of the most important problemsto be solved in the electric power systems. In monopoly regime, researches on the UCproblem can bring significant economic benefits, while in the electricity marketenvironment, researches on the UC problem play a very important role in makingbidding strategies for generation companies and in making transaction schedules forpower exchange centers or independent system operators.This paper mainly focuses on the UC problem considering the electricity prices:Firstly, this paper presents the background and the state-of-art of researches onthe UC problem, giving the organization of this paper. Secondly, this paperformulates the mathematical models of the traditional UC problem and the stochasticunit commitment (SUC) problem in the monopolistic power systems, and the pricebased unit commitment (PBUC) problem and the stochastic price based unitcommitment (SPBUC) problem in the electric power market environment,respectively. The uncertainties in electric power demand and electricity prices aremodeled using a set of scenarios, which is introduced by scenario analysis. Then, afterintroducing the principle and the convergence property of the particle swarmoptimization (PSO) algorithm, this paper gives a detailed solution procedure of thesetwo kinds of UC problem in electric power market, which is based on an improvedPSO algorithm. This paper also defines two risk indicators, namely volatility riskindicator which can measure the volatility of the profit, and the expected downsiderisk indicator which can measure the loss of the target profit, to evaluate the risksthese two kinds of UC problems face. Finally, test results on a10-unit system with a24-hour scheduling period show the validity of the proposed improved PSO algorithmbased solution method to the PBUC problem and the SPBUC problem. Test resultsalso show that, with consideration of the risk, generation companies with differentrisk attitudes should adopt different mathematical models of the UC problem, andmake their decisions accordingly.