A Recognizing Method For Weak Buses For The Infeasibility Of Reactive Optimization Problem

The reactive power optimization computing refers to the process of reducing the real power loss and improving system voltage quality by means of using voltage-adjusting devices’ control ability, and further optimized the reactive power of system, which is of significance for power system security, reliability and economical operation.However, since the reactive power installment and planning are unbalanced or unreasonable, some voltage- monitoring buses for substations and power generation have broken the voltage security ranges seriously, and even for data error caused by power operators, power system operation has emerged an infeasibility of reactive optimization problem. In fact, the most common and serious problem of the infeasibility is about voltage violations, however, expanded reactive power compensation and control ability of all voltage-adjusting devices, have significant influence on voltage magnitudes in power systems. Under the background of practical power systems, if using a recognizing method for weak buses on a random power flow state, this is definitely leading to computing being time-consuming and low-efficient. Thus, how to adjust power flow is the fundamental research for recognizing for weak buses of the infeasibility of reactive optimization problem.This thesis is conducted for the infeasibility of reactive optimization problem research in depth. Finally, what caused to the infeasibility, where the weak reactive power compensation of the substations is located, and how to expanded reactive power compensation capability or planning are given respectively. The main research content in the thesis is below:①Since power flow effectively adjusting is the primarily condition of the recognizing method for weak buses of the infeasibility of reactive optimization problem, this thesis is to propose a feasible power flow adjusting method considering network partitioning of reactive power in power systems. The current literatures’ methods are used for adjusting reactive power and voltage on the entire layer, and the adjusting reactive power to satisfy its hierarchical and partitioned balance by a heuristics method. This thesis is further taking the partitioned balance of reactive power into consideration in the reactive power / voltage adjusting method. Furthermore, this thesis is to propose a new concept of a partitioned layer load ratio and an optimal adjusting strategy based on the hierarchical and the partitioned balance of reactive power in power systems. The main research framework is: a partitioned network method of reactive power and voltage is used for dividing the power system into serval areas. And after that, with the nearest balance principle of reactive power and reverse voltage-adjusting principle, and the limitation of substation incoming line’s power factor, how much reactive power compensation could be connected and how much the central bus’ s voltage is adjusted are conducted in all system. But the power system still has reactive power unbalance problem for power generations, substations and partitioned layers. Then, we hire a heuristic method for adjusting the generations that have reactive power violations. If it is not still adjusted successfully, we apply our optimal adjusting strategy based on the hierarchical and the partitioned balance of reactive power, for dealing with this issue for each area by using GA. Finally, some voltage-violated buses are handled by a local voltage-adjusting strategy. A 63-bus test system and Chongqing grid were used for simulating and verifying that the proposed method was effective.②For the infeasibility of reactive optimization problem caused by the unqualified voltage security problem,this thesis is to propose a recognizing method for weak buses for the infeasibility of reactive optimization problem. Based on the physical features caused infeasibility, the two concepts of weak buses with installing reactive power compensation and with low voltage magnitude were proposed. According to three existing possibilities of weak buses in the system with installing reactive power compensation exclusively, with low voltage magnitude, or in a hybrid of those two types, a recognizing plan for weak buses’ type and a recognizing method for weak buses were proposed. The homogeneous interior point method is employed for detecting the weak buses’ type, and three recognizing models are established in accordance with the weak buses’ type. The objectives were the minimum total number of weak buses, the least quantitation of virtual reactive power compensation or voltage magnitude violations. The set of constraints were composed of the conventional reactive optimization constraints and the expanded capacity of virtual reactive power compensation constraints and the relaxed specific voltage security constraints. The immunity genetic algorithm(IGA) was hired for solving the proposed recognizing models with an initial population created by a customized strategy. Finally, a 63-bus test system and Chongqing grid were used for simulating and verifying that the proposed method was effective.