Multi-objective Optimization Strategy Of Controllable Load In Distribution Grid

Power demand side management (DSM) is point to improve the efficiency ofelectricity by adopting effective incentives to guide or auxiliary power users tochange the way of power, It have many other benefits like reduce the power cost,smooth the load curve, reduce network loss, improve power supply reliability. It isgreat projects which have great significance to the environment, power companies andpower users. As the most important aspects of demand side management, loadmanagement fundamentally changed the situation of cope with the rapid growth loaddemand from increasing the power generator unit capacity single in the past, alsowould encourage the enthusiasm of power users to participate in the security andstability of reasonable operation of grid. Power companies and power userscollaborate to stabilize the stability of power system and reduce the cost of powersupply and utilization, maximize the interests of both sides. This paper studies animportant component of the load management-controllable load and it’smulti-objective optimization control strategy in distribution grid, and applies theimproved multi-objective particle swarm optimization algorithm to the optimizationmodel, the corresponding simulation waveforms show the effect iveness and reliabilityof the proposed multi-object control strategy.There are various problems existing in the current distribution grid, such as lowload rate, larger peak with shorter duration generally, system capacity is insufficientto meet the rapid growth of the load, the volatility of output power due large-scalerenewable energy access and randomness characteristics of load and frequency,voltage fluctuation even collapse in the turn, correspond with the deficiency ofexisting measures are also be analyzed, the advantages of controllable load on thesolutions to these problems is expounded at last.The paper focuses on the operating characteristics of typical controllable load,air conditioning, water heater, refrigerator, electric cars, accordin g to the theory of"black box", paper work only consider its external features, a simplified mathematicalmodel for the controlled load is established. Process simplified and parameters of theprobability of choice, the simplicity and accuracy of the practical model are improved.Based on the proposed model, multi-objective optimization strategy of controllableload in the low voltage distribution grid is proposed in this paper, the strategy would optimize the working condition of different kinds controllable load in different nodesand at different times using heuristic algorithm, so as to decrease network loss, poorpeak valley and reduce the power cost of controllable load owner, the controlstrategy flexible and does not need to make substantial transformation of existingdistribution grid which make it more facilitate in the practical application.Due the multidimensional and multi-constrained characteristics multi-objectiveoptimization problems of controllable load, multi-objective particle swarm algorithmis proposed in this paper, aiming at cope with the existing problem of multi-objectiveparticle swarm optimization algorithm, such as excellent individual selection methodsare not clear, the constraint processing are not flexible, a particle selection strategy isproposed based on the normalized function values and constraints penalty, thecomparison strategy considering the value of every objective function and the degreeof every constraint violation, which reflect the fitness of the particles in turn andaccelerate particles towards the Pareto frontier.At last of the paper improved multi-objective particle swarm optimization(IMOPSO) algorithm procedures is compiled in the Matlab, with take the network lossand electricity charge as objective function of optimization problem, the simulation ofthe multi-objective optimization strategy of controllable load in the IEEE14node isanalyzed under different weather conditions, the simulation results show that themulti-objective optimization strategy of controllable load under different weatherconditions cannot only meet the user’s comfort but also decrease the network loss andelectricity cost, which make it has very great prospect on application. Based on theMonte Carlo method, the sensitivity of the control strategy optimization results to thepermeability of various electric cars and controllable load control ratio, provides adirection for further research on controllable load applications.