Reliability Assessment Of Power System Based On Bayesian Network Method

The reliability assessment is one of the most important parts of power systems analysis, it plays a vital role in the system overall planning, design and operation process. Its purpose is to provide the reliability index basis for system decision-makers, and provide decision support for the planning and operation of power systems. After obtaining the reliability index, from which we might find some measures to improve the reliability level of the system.The research for algorithms of power system reliability assessment is an important issue on the study of power systems reliability. Finding an efficient, practical and feasible evaluation method, has great significance to the safe and stable operation of power systems. Firstly, this paper makes a comparative study on the reliability evaluation methods, finds out that the main widely used algorithms today in power system reliability assessment comprise Analytical method, Monte Carlo method, ANN evaluation method, Bayesian network method, etc. By comparison, the Bayesian network method can not only calculate the conventional reliability index, but also identify the reliability weak points,it is more suitable for study of power system reliability assessment.Secondly, by learning and summarizing its research results at home and abroad, the author finds out that in the reliability index calculation process of the previous Bayesian network methods, the occurrence probabilities of system elements joint fault events are usually not considered. To solve this problem, this paper builds a new Bayesian network model which considers the joint fault events in the C++ programming software platform. The fault states and their probability distribution of the system are then obtained by using the GeNIe software. In the next state analysis process, with the target of minimum load cutting, the OPF mathematical model is established. The optimal power flow algorithm is then used to solve the model,it can effectively save our simulation time. The simulation results of a test system example show that the method proposed in this paper is feasible and effective.Thirdly, this paper makes a further research for the application of Bayesian network method in composite generation and transmission system reliability assessment, and proposed a new method to build the Bayesian network model. This method takes the load nodes as the middle layer nodes into the element nodes and system node, and establish a completely Bayesian network model of the generation and transmission system. Since it includes the load nodes, the proposed approach provides the possibility of assessing the components importance in view of different load points.Composite generation and transmission systems with many components always have complicated physical structure, this makes it disadvantage for system decision-makers to analyze the results visually. In order to solve the problem, this paper calculates the connection dependence of different nodes based on the mutual information theory, and canceled the components whose reliability importance in the network is low, simplified the structure of Bayesian network. Through the probability analysis in the example simulation, it separately obtains the components failure probability under different angles, and identifies the weak points of system reliability. So system decision-makers can specifically analyze the important components, greatly promotes the working efficiency. This method also provides a theoretical reference for Bayesian network method applied to generation and transmission systems reliability assessment.With the development trend of high voltage level, cross-regional interconnection of power system, the loss caused by blackout accidents is even larger. Customers now have a higher demand for the system reliability level,research about the reliability evaluation methods in the future will become more and more popular. In order to evaluate the system reliability more reasonably, more effectively and more comprehensively, we need a deeper and further research.