| Recently,the problem of energy shortage becomes increasingly serious around the world as well as there is a surge demand in electricity.Distributed generation evolve as a major routine for clean energy access to the power grid with its characteristics,such as economic benefits,environmental protection and flexible control.On the one hand,with gradually higher permeability of DG accessed,the distribution network status is badly affected.Inappropriate DG allocation usually led to the reverse of power flow,reduction of power reliability and islanding effect which will also create security risks.On the other hand,the operation of distribution network has new challenge for a large scale grid-connected of wind power and photovoltaic increase the randomness of both source and load side.Thus,the research of optimal allocation and operation of DG considering the temporal characteristics of source-load side plays a necessary role in the planning of distribution network.In this paper,the main work and conclusions are as follows:An optimization model of distributed generation is established basing the principle of hierarchical coordination.The primary model is used to confirm the initial location and capacity of DG in distribution network through calculating the sensitivity of power loss.In the senior model,the optimal operation of DG is further researched considering economic benefits,technical and environment index comprehensively.Multi-objective optimization problems are always converted to single objective optimization through weighting strategy,when dealing with the uncertainty of the wind power generator(WPG)and the photovoltaic generator(PVG)output applying chance constraints.A multi objective opportunity constraints method is applied in this thesis to eliminate the subjective effects bought by weight factors,which considering whole operation benefits of DG,system stability and environmental portection to optimize the operation of DG.The probabilistic model of WPG and PVG is built according to real data,further more,the probabilistic load follow is adopted to the interval estimation of objective function as well as the testing of chance constraint conditions.The application of the proposed model in PG&E69 simulation system indicates its effectiveness.An improved multi-objective particle swarm optimization algorithm based on step by step eliminate strategy(MOSEPSO)with dynamic inertia gene and acceleration coefficient is proposed in this article to solve the problem of prematurity and falling into local optimal solution easily,which in the meantime,improved the accuracy and variety of the Pareto set.By programming and simulating with MATLAB,the model is solved by using the MOSEPSO algorithm above,and the validity of the proposed algorithm is verified through the analysis. |
PDF Full Text Request