Research On Optimization And Decision-making Of Extended Black-start Schemes Of Power Systems

With the continuous development of power systems, the voltage level of the modern large-scale interconnected power grids is becoming much higher while the network scale is becoming larger, which have a profound influence to improving economy and reliability of system operation, but many new challenges also emerge. The power network structure is becoming more complex with integration of large-scale new energy generation and interconnection of AC lines and DC lines. The operation safety of the power grids is also threatened by various natural disasters, and the cascading failure caused by improper handling of local faults may lead to widespread power outages. The recent multiple blackouts occurred worldwide have alarmed the secure operation of the power systems. As an important part of power system security and defense, the research on restoration strategy and optimization decision-making technology of power systems after blackout is of great importance to speeding up the restoration process, improving the restoration efficiency, reducing the economic loss and social influence caused by blackout.After thoroughly investigating the relevant research work about the black-start issue domestic and abroad, the optimization and decision-making technique about extended black-start has been studied systematically in this thesis, with the main work as:1. The extended black-start restoration strategy and optimization method coordinating unit restoration with load restoration is proposed. According to the specific practice of system restoration during the black-start stage, considering the great cranking power provided by the black-start source and the excellent operation regulating characteristics, the restoration strategy coordinating unit restoration with load restoration is investigated, where the corresponding optimization model is established by defining the respective objective functions of unit restoration and load restoration. Based on the different relative importance of unit restoration and load restoration during the black-start stage, the units and loads are coordinating restored in the manner of different priorities, and a hybrid algorithm combing lexicographic method with improved bacterial foraging algorithm is employed to solve the model.2. The optimal decision-making of unit restarting sequence for extended black-start is investigated. In view of the actual need of unit restoration, the factors which influence the restarting sequence of units are analyzed from the perspective of the entire system restoration process, and the indexes including qualitative and quantitative are extracted. The qualitative index is quantized by the triangular fuzzy number, and the fuzzy multiple attribute group decision-making model for unit restarting sequence is established. The comprehensive weight of experts is calculated by the static weight and the dynamic weight which reflect the authority of experts and the decision-making consistency of individual and group. The fuzzy multiple attribute group decision-making problem of unit restarting sequence is transformed into a fuzzy multi-attribute decision-making problem by integration of expert individual decision information, and then the TOPSIS based on the triangular fuzzy number is adopted to optimize the unit restarting sequence.3. The investigation on aftereffect of power system extended black-start scheme is carried out. In order to evaluate the effect on subsequent restoration of the extended black-start scheme, an assessment method based on multi-objective optimization model for network reconstruction and super efficiency data envelopment analysis model is proposed. The subsequent network reconstruction is optimized by multiple time step reconstruction strategy coordinated with the overall restoration after discretizing the network reconstruction process into a series of successive time step, and a multi-objective network reconstruction optimization model coordinating unit restarting with load restoration alternately is established, where each objective function is hierarchically solved. A solving method combining the improved bacterial foraging algorithm with the Dijkstra algorithm is proposed to solve the model, in which the units, loads and restoration paths in each time step are optimized from the perspective of the global optimum of the network reconstruction. By defining a group of input and output indexes reflecting the multiple elements of reconstruction process, the relative efficiency of the subsequent reconstruction, which represents the aftereffect of the extended black-start scheme, is evaluated by using the super efficiency data envelopment analysis model.4. A dynamic comprehensive assessment method for extended black-start schemes based on time series dynamic theory is proposed. The dynamic three-dimensional comprehensive assessment framework is introduced, where the restoration process is divided into two periods consisting of the black-start stage and the subsequent restoration stage, and an assessment index set is established by defining indexes reflecting the actual elements of restoration process. The comprehensive weights of indexes are determined by a hybrid weight assignment method in which the subjective weight is combined with the objective weight. The indexes of each period are integrated by using the linear weighted model to get respective static assessment value, and the optimal time weight of each period is optimized by nonlinear entropy programming based on time degree. The assessment value of all periods are then reintegrated by using mixed operator model consisting of time ordered weighted averaging operator and time ordered weighted geometric averaging operator, and the final assessment result of the schemes is given.