LinkExchange Member
The paper [PS166] proposes a stochastic approach to power system Var planning with a GA. A stochastic voltage index obtained from the stochastic flow is defined to capture the behavior of nodal voltage variations for a given power system conditions. This paper optimizes the index with capacitor banks that may be expressed in integer. GA is used to obtain the better solutions from a standpoint of global minimization. The proposed method is tested in sample systems.
Reactive power planning of power systems using GA is presented in the paper [PS167]. The problem is formulated as the minimization of the real power loss flowing in transmission lines and improvement in the voltage profile (minimization of the total voltage deviation). The GA approach to reactive power planning is formulated as a search problem in the space of state variables consisting of generator voltages, transformer taps and shunt capacitors subjected to the minimization of objective functions and satisfaction of the constraint relations. The proposed approach is described in detail for the IEEE 30 Bus system.
The paper [PS168] presents an approach to improve power system voltage stability using the bus voltage characteristics in a transmission system based on Hopfield's ANN and GAs. The authors try to find the optimal allocation of reactive power (VAr) facilities considering bus voltage deviation and bus voltage characteristics. Traditional VAr facility planning is used to consider only the bus voltage deviation with local bus voltage information. The information of the bus voltages from the whole power system makes VAr allocation planning more efficient and economical. The bus voltage sensitivities to load deviations are added into the evaluation function to maximize the voltage stability of the power system. To overcome the discrete nature of Var facilities, Hopfield's ANN and GAs are employed. The proposed algorithms were tested on an 11 node power system and results show the validity and effectiveness of the algorithms.
The allocations of reactive sources and the control of discrete reactive facilities create a large scale nonlinear discrete VAr optimization problem, which is a nonconvex problem with multiple integer feasible solutions mathematically. The research [PS169] attempts to solve the discrete VAr optimization problem based on such a technique that a GA is employed to find an approximation solution near to the global optimum. To speed up the solution procedure, an expert system is utilized to provide power system information. A real distribution system with 18 changeable taps and 7 capacitors is used to demonstrate the validity of the approach.
Please send Questions, Comments, Suggestions to bencr@hotmail.com