The Design Of Assistant Power Balance System In Power Station

As the Information Technology and Artificial Intelligence developing, more and more power plant begins to use Energy Management System to manage the enterprise, for improving the producing effect and saving cost.This paper came from Enterprise Project of Daqing Hongwei thermal power plant, associated by Northeast Petroleum University, introducing a Power Balance Assistant Project. Power Balance System plays an important role in Hongwei thermal power plant, one of the main power plant in Daqing, daily management because electronic has the feature that it cannot be saved in high-capacity, and the electronic system has high requirement of real-time and stability. For avoiding waste and preventing pollution of the environment, the Power Balance System in power plant became very important.For dispatching the electric generators reasonably and economize on bunkers and assisting related personnel to balance the electric power. This paper designed a Power Balance Assistant Decision System based on Convolutional Neural Network, by establishing an semi-auto balancing system based on the data in intelligent electric meters. The main function in this system is predicting short-term power-load information. Compared to traditional predicting model, Back Propagation Neural Network, it is showed that CNN model has more accuracy, converging faster and has more advantage in data spread.This system could help relative principal in Hongwei power plant to balancing the power load via short-term power load predicting and combined with original information management system in Hongwei power plant. It is proved that, for power balancing problem, with help of this system, can be solved in a better way considering this system using quantization model instead of human experience.At last, this paper gives a data analysis for Hongwei power plant’s producing data based on Convolutional Neural Network and the evaluating index of correspondent index.