Research Of Bidding Strategy For Generators In Electricity Market

At the end of twenty century eighty years, In American and many western countries, a profound reform in power system has begun. The key of the reform is to smash monopoly and to lead into competition in power system. The goal of the reform is to improve the efficiency and to optimize resources disposition. Replacing administrative method with market method rules the running of power system. In China, the former National Power Company was vertically split since 1998, so the generation part and transmitting part were separated. At the same time, Shandong, Shanghai, Zhejiang, Liaoning, Jilin and Heilongjiang were selected as first 6 experimental units which involved in the generation side competition. The market-restructuring of power industry in China was started.In the electricity market environment, the bidding strategy of generators always be an important research field in the electricity market. The generation companies as the providers of power must bid based on the regulation of electricity market. To them, the last goal is to accomplish the profit maximization and risk minimization. In the past time, many research about this problem only focus on the profit maximization. Hence, price has inherent uncertainties in electricity market. Consequently, the plants will then encounter with some extent of risk. This reveals obviously the limitations of the bidding strategy for pursuing the profit maximization simply, and it also shows that the profit and risk is a pair of mutually contradictory index. High profit causes usually high risk. How to accomplish the profit maximization under controlling the risk is a typical risk decision making problem.This dissertation has conducted a series of research about the bidding strategy of generation companies, from building up the bidding model to risk analysis and bidding risk decision and optimization algorithm etc. The built bidding models all based on price prediction in this paper and suppose all the generation companies are price takers. The companies have no market power.The main contents of this dissertation including: 1. Firstly, This paper presents a power plant bidding model based on market-clearing price distribution. Based on forecasting price with Lognormal distribution, the probability density functions of next-day hourly market-clearing prices are obtained. Depending on the probability density functions, the authors get the bid acceptance probability functions. Based on the bid acceptance probability functions, the power plant bidding model is built up that is used to accomplish the profit maximization. At last, results from a realistic case study are discussed in detail. This model can give the plants some guidance in the realistic bidding process. Hence, the method did not considerate the factors of risk, so there are some limitations in it.2. The profit's probability density function of a bidding strategy was developed based on forecasting price with Lognormal distribution, so that the probability of the obtaining specified profit under the scenario can be known, a profit-probability curve can be drawn. The risk curve provides an effective tool for risk analysis. The risk of the bidding strategy obtained by the expected profit maximization was discussed using the profit-probability curve presented in this paper. It showed that risk was implied in this bidding strategy, the risk analysis is necessary.3. Price has inherent uncertainties in electricity market. Consequently, the plants will then encounter with some extent of risk. This paper presented a new risk decision making method of suppliers based on profit probability formula. The proposed algorithm can exactly calculate the profit of different scenarios when the decision-maker gave the profit probability. Some examples shows that the proposed method can pick out the best decision quickly and exactly based on the decision-maker's risk attitude. This kind of research method of bidding strategy is the main direction at present.4. In this paper, the optimization method of bidding model is particle swarm optimization. The algorithm is a kind of evolution computational method based on swarm intelligence. The merit of this algorithm is simple and easy to accomplish, and there are not many parameter to be adjusted. At present, many scholars applied particle swarm optimization to the solution of electricity power system problem. And the research of this paper also shows that the algorithm can get to the overall situation restrains. The conclusion is also very precise.