Information Fusion Technology And Its Application In Power System Studies

Information fusion is an uncertain information processing theory coming into being in the 1970s, the main idea is obtaining more accurate and credible information through integrated processing of multiple information sources to provide decision support. Uncertain information processing is involved in many aspects of power system, such as equipment monitoring and fault diagnosis, load forecasting, power protection and so on, information fusion plays an indispensable role in improving the reliability and the the level of intellectualization of power system and a whole new stride has been hit in the building path of smart grid.This thesis focuses on the fault diagnosis of power equipment to describe information fusion applications in this area. Based on the study of existing fault diagnosis methods, combining the intelligent information fusion method with traditional ones, a multi-level diagnosis model based on D-S theory and ANN is proposed and applied to fault diagnosis module of GVMS. Since the lack of D-S evidence theory in dealing with conflict evidences, the concept of distance is introduced to amend the evidence combination rule. Considering the shortcomings of ANN, such as poor reasoning ability,slow convergence speed and easy to fall into local minimum, fuzzy theory is introduced, BP is combined with particle swarm optimization to adjust and optimize network parameters, an improved PSO-BP FNN algorithm model is established ultimately. System data show an absolute advantage of the proposed multi-level diagnosis model in improving the accuracy and reliability of diagnosis. Finally, according to the preestablished evaluation index system, a certain evaluation method is used to synthetically assess the process performance of fusion, for the problem of inaccurate weight and gravity shift in traditional AHP, Fuzzy Comprehensive Evaluation is combinated with AHP, and fuzzy hierarchical analysis comprehensive evaluation method which can quantitatively conduct an comprehensive assessment for the processing performance of the fusion module is proposed.