Research On Probability Equivalent And Network Partition Coordination Model For Power System Reliability Evaluation

Reliability evaluation of power system has been one of the research hot spots in electric power industry. The increasement of the network size which is raised by the interconnection between power grids, increases the reliability and economical efficiency of power supply, but also to bring the difficulty of reliability assessment. Because that the number of components and the network states which are considered in large power system reliability evaluation increase rapidly, the assessment time increases also, which brings great inconvenience to relevant scientific research and operation planning. How to calculate quickly in reliability assessment of bulk power system becomes a problem urgently to be solved.Through dividing the network into a plurality of sub areas, based on the method of non parametric kernel density estimation and the concept of maximum load capacity, the probabilistic Ward equivalent model and the network partition and coordination model for reliability evaluation of bulk power system are researched.The main contents of this thesis include the following:Firstly, while using the conventional Ward equivalent technology to shrink the scale of power system for reliability assessment, this paper regard the external network(EA) equivalent parameters as continuous random variable dependent on external regional network topology and equipment fault. The non parametric kernel density estimation method is adopted to construct the joint probability density distribution function of EA equivalent parameters. We also present the calculation method of equivalent model of probability Ward equivalent model. Examples in two regional RBTS system, RTS79 and RTS96 system are simulated and verify that the probability equivalent model can effectively improve the accuracy of equivalent network, and improve the computational efficiency of equivalent power system reliability assessment.Secondly, the paper proposes a new network partition and coordination idea for power system reliability assessment. The idea considers power support and dynamic equivalent progress between sub regions. Through regarding the contact line between sub regions as equivalent load at regional boundary node, and based on the concept of maximum load capacity, the power support ability and the optimal load reduction of each sub domain are solved successively. Through successive coordination calculation of sub area, we achieve active power correction and optimization of load curtailment in full network. The author compiled reliability calculation program of network partition and coordination model using Matlab, and through the test of IEEE-RTS96 system of 48 nodes and 73 nodes, the validity and efficiency of the proposed model are verified. At last, this paper discusses the influence on the computational efficiency when different rules of network division are adopted.Finally, based on the idea of the sensitivity analysis, network partition and coordination model and coordinated decomposed evaluation algorithm are further improved, in order to improve the calculation precision of reliability evaluation using the network partition and coordination model. The physical meaning of Lagrange multipliers in the optimal power flow calculation is used to analysis sensitivity between the equivalent power and load shedding reduction of sub region including equivalent power. According to situation whether equivalent power reachs the limit in the optimal power flow calculation of sub area, iterative correction measures is decided to take in order to improve the model accuracy. An example based on the modified RTS96 network is conducted to verify the effectiveness of the proposed method. Only a small increase appears in calculation time using iteration based on sensitivity analysis, thus this method can further improve the calculation accuracy of the network partition and coordination model.