The Research Of Comprehensive Optimization For Distribution Network Power Supply Capability Based On Multi-strategy Technique

The reliability of power supply and network loss are two important indicators for measuring the power supply capability of distribution network. They decide whether the distribution network can distribute the electrical energy to the consumer safely and economically. In order to enhance the power supply capacity of distribution network, the two indicators should have a comprehensive optimization in this paper. Optimal switches device and network reconfiguration are two important methods for comprehensive optimization of power supply capability, switches device can improve the reliability of network and reconfiguration is a useful way to network loss reduction, so the two methods are used in this paper.Compared to a single-objective optimization, comprehensive optimization of power supply capacity is a multi-stage strategy process and complex enough.Based on optimization theory, multi-strategy technique is proposed to reduce the optimization difficulty, the technique devides the comprehensive optimization problem into number of single-stage optimization problem by indicators whice various optimization methods can optimize best, at the same time, each stage of the optimization process would be added constraints to optimize the result and the result is more access to the actual requirements.Switching device is a non-linear, multi-objective and complicated combinatorial optimization problem, usually solved by intelligent optimization algorithm. Based on binary transformation of differential evolution (DE) algorithm, the algorithm's important control parameters are adaptability improved.The new algorithm is applied to optimize switching device, in the optimization process, through simplifying the complex distribution system into a simple model consisted with buses and switch-lines by "bus-line" method, reliability calculation speed of distribution network is improved.A actual distribution systems in a district is chosen to perform the above approach, the results show that the improved DE has much advantage on getting the global optimal solution and convergence rate. Distribution network reconfiguration has the "curse of dimensionality" problem as the network scale expansion. In this paper, through the inertia factor of particle swarm optimization (PSO) algorithm is dynamic adaptability improved and two important operators of genetic algorithms (GA) are redesigned, the paper proposes a novel artificial algorithm - hybrid Particle Swarm Intelligence genetic algorithm (PGA) whice combined with the respective advantages of both algorithms; in the calculation of process, the solutions of coding scheme and calculation of power flow are designed for adapting to fast computing.The new algorithm is used in IEEE typical network reconfiguration calculation for testing its feasibility.In the Actual appliance, the analytic hierarchy process (AHP) method is used to calculate the weight of each optimization indicators and form the mathematical model in stages in multi-strategy technique, according to the restrictions of various stages and the optimization order of reliability and loss of distribution network, the calculation result of test system verifies the effectiveness and practicality of multi-strategy technique.