Energy Efficiency Management Of Commercial Load Based On SCADA System

With the increasing competition in the traditional industries, in order to get moreprofits, companies have to strengthen cost management, as the third largest item coststhat electricity costs have get more and more attention by the managers. Especiallywith the outbreak of energy crisis, the strengthening macro-control of countries,energy efficiency management of power have become a problem of enterprises thatcan’t be ignored. Currently, most of the research in energy management systems is forthe power supply section, a few scholars have expounded energy saving method ofcommercial building from the angle of architectural design. Based on the review andstudy of existing research, energy efficiency management for commercial load hadbeen studied, an energy consumption monitoring platform with the core of a loadforecasting model, the alarm of the method of statistical process control had beenestablish.Detailed and accurate analysis of the power load characteristic is the necessarypremise to improve the forecasting accuracy and the precondition to make sure thatenergy consumption monitoring platform can run normally. Therefore, thecommercial load characteristics were analyzed firstly. The result pointed out thatcommercial load had obvious business cyclicity, smaller base load and biggervolatility relatived to electricity power system. According to the difference ofcyclicity and volatility, loads could be divided into three kinds of loads that theirregular cyclicity and big volatility load, the regular cyclicity and big volatility loadand the regular cyclicity and big volatility load. Considering noise exist in the dataand can be easily drowned out by the burr of load curve, history data was denoised bywavelet soft threshold denoising method based on the assumptions of that the spreadfrequency of noise and the signal is different. To improve the prediction accuracy ofthe load with large volatility, a combined model was used to prediction the load.Firstly, the load denoised was decomposed into components with different frequencyusing of wavelet multi-resolution analysis method; Secondly, each component wasrespectively predicted using the support vector regression model; Finally, the predicted values of each component were obtained by the wavelet reconstruction asthe predicted values of load. The feasibility and effectiveness of the combined modelwas verified by project.At last of the article, SPC method applying to quality control of enterprise wasintroduced to the energy consumption control of commercial load. An energyconsumption monitoring platform was established by the SPC method and thecombined model. It could give early warning to the abnormal load, keep energyconsumption in a reasonable area, and provide a feasible method to energy efficiencymanagement of commercial load.