The Research On Chaotic Neural Networks For Price Forcasting And Economic Load Dispatching For Combined Power And Heat System

With the development of the combined heat and power (CHP) system and the electric power market, the CHP enterprises will face to connect to the national network by price competition in the power market. On one hand, it is important to use the relative historic data to forecast the future market clearing electricity price to obtain more opportunity to improve the income. On the other hand, the enterprise profit means subtract the cost from the income. So in order to improve the profit, the enterprise needs to reduce the cost too. Because of the different characters of the generator, it is important to find the most effective combined way to reduce the cost. In this paper, we give an effective way to solve with the two problems.First, by calculating the chaotic character exponents-correlation dimension, we found that the electric price and load time series are not random series but chaotic ones. So it gives us a new acquaintance to the complexity of price and load time series, and the short-term electric price could be forecasted through chaos theory.Second, based on phase space reconstruction of chaotic time series, the short-term price forecasting approach on multi-variable phase space reconstruction theory and improved chaotic neural network is presented. It may adequately reflect the varying rule of the data itself, in order to improve the accuracy of price forecasting.Third, in order to find the distinguishing feature of the economic CHP dispatch, this paper researches on the different characters of the generator in the system. And we found there are two kinds of thermal system in different enterprises. They are header system and unit system. So this paper constructs the models respectively when dealing with different thermal system.Forth, it is because there are many design objectives coupled with tighter constraints need to be incorporated. We make the economic CHP dispatch be formulated as a multi-objective optimization problem. An improved particle swarm optimization algorithm is used to achieve the reasonable tradeoff between multiple objectives including the total generation cost, power generation deviation and heat generation deviation. The decision makers can choose different scheme in different environment.The decision support system is developed initially using programming language Visual C++ 6.0. The system has these merits such as better operational interface, easy operation and enough accuracy. It can provide suggestion for price bidding economic CHP dispatch.