Transaction Behavior Surveillance And Design Of Surveillance System In Electricity Market

In China, the power industry is carrying on the reform of electricity market, which is attempted to change the system of the monopoly management in order to enhance the competition through market mode. The partition of power plant and power network shows that electricity market has made preliminary success.However, due to electricity market is still at an initial stage, a negative effect is that some market participants could control the market depending on their monopoly power. In this case, it is necessary to develop the effective surveillance to electricity market, in order to ensure that the competition is fair and just in electricity market.In this paper, the surveillance theory of electricity market is introduced. Referred to the transaction types of electricity market, the disadvantage that the generation companies manipulate market when they adopted market power is analyzed. And then, some effective surveillance approaches are put forward, such as surveillance indexes, bidding behavior surveillance, market simulation, etc.When we analyze market with surveillance indexes, analytic hierarchy process (AHP) is used to do a comprehensive evaluation, meanwhile, to score every generator we are concerned referred to the aspect of surveillance. The result is considered as an important reference to discriminate if there is market power within them.As to supervise if generators' bidding behavior is in regularity, the cluster analysis technique in data mining is introduced. Using self-organizing feature map (SOM) network as the cluster method, the generators' bidding dates are divided into different groups, which is considered as a reference to distinguish if every generator has rule-broken behavior in electricity market.Based on the requirement of surveillance management, there is a great deal of dates to be analyzed, a surveillance system of electricity market on transaction behavior is designed. The system has underlying function: information collection, model of surveillance, and information bulletin. Therefore, transaction behavior of rule-broken can be detected automatically, and will be warned on the website in public, and which will provides important support by means of technique measure to keep market competitive environment fair and reasonable.