Research On Variable Time-step Modeling And Coordinative Optimization For Network Reconfiguration Of Power Systems

In recent years,the interconnection of power grids is a demonstrable trend.In spite of the high energy-usage efficiency that power system has brought to us,a lot of problems also with it.The complex network structure results in a high risk of fault and the scale of influence may be expanding.More seriously for a power system,power outages will be caused by a vicious chain failure.The extensive blackout occurred in the world has proved that the widespread blackout is a serious threat to the modern power systems.In this context,as one of self-rescue measures,reliable scheme for power system restoration is of great importance to relieve restoration burdens,speed up the system restoration process,and reduce the economic loss and social influence resulted from the faults.Therefore,the research on power system restoration and its optimization technique is essential.On the basis of analyzing previous work about the power system restoration issue,the optimization technique of network reconfiguration has been studied systematically,main contents of this thesis are described as following:1.Optimization for unit restoration sequence considering the decrease of the unit start-up efficiency.The units' start-up efficiency may decline with time elapse after a major blackout,which has an adverse effect on unit start-up time and rapidity to connect to the grid.According to decreasing trend of unit start-up efficiency with time,a fuzzy membership function is designed to describe the start-up time of the units in different conditions,and the decreasing level of unit's start-up efficiency is estimated.In terms of decreasing rates of units' start-up efficiency,a ranking principle is defined to select the units to be started up currently.The optimization model of unit restoration sequence considering the unit's start-up efficiency decrease is established,and the multiple population genetic algorithm is employed to obtain the optimal unit restoration sequences.2.Variable time-step modeling and coordinative optimization of network reconfiguration.According to the reconfiguration strategy of “modeling on discrete time steps,optimizing on the whole process”,the restart time of electrical equipments is evaluated based on the characteristic of electrical equipment.By analyzing the serial and parallel restoration patterns of units,a multiple time-step coordinative optimization approach for decision making of the unit restoration sequences is proposed.Adopting the global optimization strategy,an algorithm in the overall process is researched to obtain the optimal unit restoration sequences.By reasonably controlling the number of the reserved solutions for each time step,the diversity of solutions is remained to avoid the local traps,and the computation complexity is well controlled in an acceptable degree.3.Coordinative optimization of load recovery and network reconfiguration.Load restoration in the early stage of network reconfiguration plays a role in maintaining power balance and stable operation of the power system.According to the operation characteristics of units and the pickup pattern of load,corresponding model for load recovery is established.The objective functions of the optimization problem are selected from recovery benefit,important degree,power balance and subsequent influence.Moreover,a method for load pickup considering system security is presented to acquire the optimal scheme.Load recovery optimization is beneficial to stability of the power system.Espically,in the weak period of recovery process,systematic load recovery scheme is essential.It is the major promise for in subsequent recovery proess.4.Optimization for transition schemes considering the constraints of resource.According to the characteristic of common black start units and emergency power supply in substations,an optimization method is proposed to determine transition schemes for sustainable output,by considering resource constraints.A limited power source is firstly employed to crank some neigboring unrestricted thermal units with good start-up performance.The restarted thermal units are then used as power sources for subsequent restoration instead of the limited power source.The objective functions of the optimization problem are defined reasonably and the constraints are taken into account.An efficient solution algorithm for the model is studied.Continuous power generation is assured by this alternative scheme,and the feasibility of subsequent restoration plan is improved as well.