Optimization Of DSM Time-of-use Power Price Based On Genetic Algorithm

Time-of-use(TOU) power price turns out to be an important economic means inthe field of demand side management(DSM). It works well in improving powerconsumption behavior and offsetting contradicts between power demand and supply.As an effective means of DSM, TOU has been applied more broadly and a lot ofresearch has been taken. The practice of an effective time-of-use electricity pricescheme could help saving power construction investment, increasing the electricpower efficiency of the customers, improving power quality, and shifting load. Thesignificance obviously is there for economic efficiency, security, and reliability.In this paper, the application and research situation of TOU at home and abroadare introduced in detail. After analyzing and summarizing the existing research, itsums up some problems and deficiencies being in the application of TOU, such as:strong subjectivity in time-period partitioning, imperfection in the theory of settingprice, unfulfilment in multi-objective TOU optimization. Aiming at the aboveproblems, it develops a set of TOU optimization scheme based on intelligentalgorithm, which could not only partition the time-period according to whole societyload, but also set power price on the basis of single-objective and multi-objective.This new scheme could help saving investment, decreasing network loss and performbetter than the traditional means.Based on intelligent algorithm, a new time-period partitioning method, whichcould realize single-objective and multi-objective optimization perfectly, is developedaccording to whole society load. On the basis of time-period partition, power priceoptimizing scheme is built for different industries on both fixed flat period powerprice and floating flat period power price. Specifically, as to the multi-objectiveoptimization, a new intelligent algorithm named rapid non-dominated sorting geneticalgorithm with elite strategy (NSGA-II) is applied in price optimizing scheme withwhich could help acquiring evenly distributed Pareto optimum set. By applyingNSGA-II, there are many objective solutions to be chosen as policymakers’inclination, besides we could get optimal compromise solution according to theprinciple of multiple attribute decision, which realizes the real sense of the TOU multi-objective optimization by overcoming the strong subjectivity disadvantage ofweighting method.By integrating the research above, this paper set up a brand new TOUoptimization scheme which could partition time-period properly and set price optimal.This scheme is based on abundant theoretic evidence and proved to be a goodguidance for policymakers to partition time-period and set power price in differentplace at different time. Examples prove this new scheme works better than thetraditional one. The uses of classical genetic algorithm and NSGA-II in this paper areinnovative applications and tentative exploration in the field of DSM.