Improved Algorithm For Complex Distribution System Reliability And Its Programming

In this paper, a new kind of recursive algorithm has been raised. It optimized the distribution network system reliability calculations, improved some efficiency, and met the user's needs closely. This algorithm needs no network equivalence process, has no complex principle and brings no difficulty to programming. Through its failure analysis, can be more easily identify key equipment of distribution systems for customers to adjust. This algorithm can easily cooperate with fault reconfiguration. No adjacency matrix and special chain table will be formed during searching procedure; it takes advantage of basic table of node data and branch data to do searching work, and easily to get access with field data. The algorithm uses recursive search also reduces programming greatly.In today's priority to efficiency, quality and quantity of the times, Electric power industry, especially for its distribution system reliability requirements are also increasing day by day. Therefore, a set of convenient, purposeful, targeted reliability analysis software is indispensable. Reliability evaluation of distribution network can be divided into two categories: simulation and analysis method. Simulation method is suitable for calculation of complex system, but this method is time-consuming and accuracy is not high. Analytical calculation is relatively simple and fast in calculation speed. It has been widely used in power distribution system reliability evaluation. It's including network equivalent method[l - 2], minimal path method[3 - 4], fault traversal algorithm[5 - 6] and failure mode and effect analysis[7 - 8] etc.Most existing algorithnis for system reliability high dependence on network topology and excessive dependence network association matrix or adjacency matrix [9] or lists [10-11] instead in the calculation process. And they have to do some additional processing on the network data. But the matrixes and the lists can only provide connected relation of the devices, nothing more. It could not give the useful information such as device type and device parameters during calculation.These algorithms focus on mathematical modeling of distribution network components for reliability evaluation, while paying little attention to distribution network reliability optimization method of computer. Network equivalent method and minimal path method all need network equivalent; most of the equivalent process by computer is more complicated, especially for the distribution of more complexes, the equivalent process sometimes impossible to achieve. And the fault traversal method is relatively convenient for computer calculation. It will be affected by faults and topology search enumeration of each component element classification calculation, finally get