Fast Reliability Evaluation And Reconfiguration Of Distribution Networks

Distribution system is the bridge linking the transmission system and consumers. So its operation state directly relates to the power quality and economic benefits of power supply enterprises. Distribution system reliability evaluation is an important guarantee to system reliable operation, and its reconfiguration is the basis of system economic operation. Recently, the increasing connection of distributed generations to distribution systems has brought in many new issues about reconfiguration and operation. It generates inevitabley limitations for the conventional methods for reliability evaluation and reconfiguration. Therefore, the research of methods for reliability evaluation and network reconfiguration of distribution systems with distributed generations is of great importance in both theory and engineering.Based on the analysis of differences between the conventional distribution networks and the present ones with distributed generations, fast and accurate methods for reliability evaluation and reconfiguration of distribution networks with arbitrary configurations are studied in this thesis.A fast evaluating method based on fault delivering features of distribution system reliability is proposed. In this method, the distribution system is divided into main and branch networks. At first, the up-stream reliability indices of each point in branch networks are determined by up-stream delivering of each comoponent's reliability indices. Each branch network is replaced by an equivalent component in the main network. Then the failure-mode-effect-analysis (FMEA) method is employed to calculate the reliability indices of each point in the main-network, and the down-stream reliability indices of each point in the branch-network. Then the down-stream reliability indices of each point are determined by down-stream merging of reliability indices of the up-stream-network. Synthesizing the up-and down-stream reliability indices yielded the reliability indices of each point in branch-neworks. The proposed method takes into account the difference in operation time of breakers and section switches. It is not only as accurate as the FMEA method, but faster than the existing algorithms. Simulation results verified the effectiveness of the proposed novel algorithm.A novel network model and fast algorithm for distribution system reliability evaluation are presented, which considers the random energy output of wind power generators (WPGs). The novel model first simplifies the distribution network to a Switch-Section network, and then divided it into two kinds of parts named as the main- and branch-networks. After analyzing the energy output feature of WPGs and the fault characteristics in branch-networks, the supply area of WPGs is built. Then the order and probability of power supply to the supply area of WPGs are determined, which successfully implemented the mapping from the random output of WPGs to the reliability index of the power supply. The fast reliability evaluation alogrithm calculates the reliability indices of the main-network using the FMEA method, and those of branch-networks using the properties of elements in up-and down-stream reliability delivering. Meanwhile, the impact of WPGs on its supply area is counted in computing the downstream reliability indices. Simulation results verfied the effectiveness of the proposed method.The minimum feasible analysis unit and fast algorithm for distribution network reconfiguration are presented. Firstly, the conception of basic loop was presented based on network simplification. And then, by combining the characteristics of the nodal impedance matrix of the radial distribution network and the quantitative relationship between load and branch power loss, we verified that the basic loop was just the minimum feasible analysis unit of the physical reconfiguration optimization. Based on this conclusion, load dissipation component and path dissipation factor were constructed to reflect the power loss in a basic loop, and then a novel heuristic rules and an improved branch-exchange method were formed by analysis the properties of the path dissipation factor. In this improved method, the processing sequence of all the basic loops was firstly defined by the load dissipation component of the disconnecting switch in each basic loop, and then the alternative disconnecting branch was determined in the processing basic loop based on the new heuristic rules which can help to get a better solution, optimal reconfiguration scheme was simply obtained by repeating the above operation until there is no branch need to exchange. Test results of IEEE 69 buses system show that the proposed method produces the optimal reconfiguration scheme quickly and reliably, and can be applied to complex distribution network reconfiguration.A novel scenario distribution network reconfiguration model is presented. In this model, the scenario analysis method is applied to describe the random output of the WPG and its influence through the scenario selection and scenario voltage. Multiple WPGs and wind farms connected with a network is also considered in this model. And then, an efficient genetic algorithm is presented for solving the scenario distribution network reconfiguration model. Using the no unfeasible coding rule in the initial population strategy, cross strategy and eugenic strategy, individuals in the evolution are always formed feasible solutions which are satisfied the actual distribution network. Physical optimization based on scenario voltage in the process of evolution reduces the optimization time and the dependence of the initial population. The simulation results verify the feasibility and efficiency of the proposed model and algorithm.A fast algorithm for distribution network reconfiguration considering system realibility is presented. Firstly, the realibility indices to express the system realibility are confirmed and normalized. And then, a novel reconfiguration model considering system realibility is presented. In this model, power loss, average service availability index and energy not supplied index are included in the objective functions. A judgment matrix approach is employed to convert multiple objectives to one objective, and a fast algorithm is proposed to solve this problem. In this algorithm, realibility indices are not computed in the earlier reconfiguration stage, it is computed only in the the later reconfiguration stage. Simulation results verified the effectiveness of the proposed novel algorithm.A software package has been designed and developed based on the above proposed methods. It is fast in reliability evalution and reconfiguration of distribution systems. The software is applicable to both conventional distribution systems and the ones with distribuited generations.