Research On Fuzzy Comprehensive Risk Assessment For Distribution Network Fault Based On Rough Set Theory

As a complex large-scale dynamic network system, the basic task of the power system is to provide users with safe, reliable, economical and high quality power energy. Modern power system is continuous developing to an ultra-high voltage, large capacity, automation, intelligent system. Its topology is more and more complex, especially distribution network, which is in an important connection position between transmission and load. So, the safe and reliable operation of the distribution network is directly related to the power system safe and stable operation and control.Due to the complex topology, uncertainty and inconsistent information, the distribution network system brings challenges to the research of fault risk assessment. The traditional distribution network fault risk assessment often has some shortcomings such as the evaluation index is too simple; unable to effectively deal with the distribution of fuzzy and uncertain information. Therefore, it is how to effectively analyze the distribution network data during its operation and fault, to correctly choose the risk indicators and to establish a more rational and comprehensive risk evaluation system that is the key to reduce fault loss caused by distribution network and improve the power supply reliability and power quality.Rough set theory and fuzzy comprehensive evaluation method can effectively deal with uncertainty and fuzzy characteristics of information, reveal the potential rules of data, which is extremely effective theoretical basis and methods to assess the system risk. This paper combines advantages of rough set theory and fuzzy comprehensive evaluation method in processing uncertainty and inconsistent information. For unique operation mode and fault attributes of distribution network system, by rough and fuzzy parallel method for attribute reduction and data processing, a new method is proposed based on rough set theory and fuzzy comprehensive evaluation method, and effectively applied to the risk assessment of distribution network fault. Mainly in the following aspects were studied:Firstly, introduce the basic risk assessment theory this paper uses of risk assessment theory and fuzzy comprehensive evaluation theory. Based on rough set theory, this paper does the characteristic statistical classification analysis and fault risk quantitative analysis of distribution network fault information. In order to solve the inaccurate, incomplete, inconsistent problems of fault data and the problem of unable use of continuous data in system, before the risk assessment of the distribution network fault, based on rough set theory and corresponding improvement method, distribution network original fault data is to be filtered, completed and discretized. On the basis of the original algorithm, it is combined with rough set theory and the distribution network fault practical to improve the algorithm.Secondly, the fault information of distribution network is deeply studied, and the risk assessment index system of distribution network is established. By using the fault record of the distribution network, the fault attribute of distribution network is analyzed, classified and organized in detail. Taken records of distribution network fault information as the data source, from the relevant multiple aspects such as fault time, fault occurrence region, fault causes, fault-affected equipment, fault level, statistical characteristics are classified and analyzed. And, quantification of failure risk, uncertainty and cross impact of qualitative and quantitative research are done. Based on risk assessment theory and reliability analysis, select indicators to assess the failure rate, and establish a distribution network fault risk index system, in which the top layer are comprehensive failure probability and failure consequences and the remainder of the risk sources improve the top layer.Thirdly, a fuzzy comprehensive risk evaluation method based on rough set theory is proposed, which is affected by risk factors corresponding to fault attributes. To evaluate risk factors, the method chooses discernibility matrix attribute frequency of the attribute reduction method to make attribute reduction of key risk factors for distribution network. Also, risk index weight is therefore obtained. After this, use fuzzy comprehensive evaluation method to improve upper risk assessment model.Fourthly, use the proposed risk assessment method parallelly based on rough set and fuzzy comprehensive evaluation to perform a distribution network fault cause risk assessment without considering DG and a multi-factor distribution network fault risk assessment considering DG. The results show that the method proposed which is integrated and affected by a variety of factors can be reasonably analysis contrast of distribution network outage risk and a true reflection of the different risk factors. Also the rationality and validity of the method can be verified.