The Research And Implementation Of Energy Management System Based On Energy Consumption Prediction Model

Manufacturing is the pillar industry which holds the balance of national economy. So far,manufacturing enterprises have been confronting with the problems of large energyconsumption and high cost, and the energy consumption influences the benefits as well as thesocial duty of enterprises. With the lack of energy and the rising of the price in domestic andoverseas, it’s very important to carry out the research on energy consumption predictionmodel. In order to improve the efficiency of energy management, not only need the statisticsand analysis of energy historical data, also need a good forecasting for data in the future,which can improve the level of intelligence analysis and meet the enterprises’ different needs.The main works are shown as follows:1. A comprehensive review of energy management system and energy consumptionprediction model is carried out in this dissertation. Several significant development statusesare summarized, including energy management system and energy consumption predictionmodel. Meanwhile, the problems of energy consumption prediction model are discussed.2. Aiming at the energy waste problems resulting from the contradiction between energydemand and supply in manufacturing enterprises, an energy demand prediction model basedon support vector machine is built. And a manufacturing enterprise demand for electricity, forexample, to validate the model, simulation results show that this model has a good prediction.3. Aiming at the energy consumption data lag issues in the traditional energy-savingeffect evaluation method, an energy-saving effect evaluation model based on support vectormachine is built. This model is comparing the energy consumption values of the actual andthe predicted to assess the energy-saving effect, and simulation results verify the feasibility ofthe model.4. Aiming at the energy waste problems resulting from energy consumption anomalies inhigh energy consumption process, an abnormal energy consumption monitoring model basedon BP neural network is built. Tire curing process, for example, the interval of normal curingenergy efficiency is predicted using BP neural network algorithm, and all the observed valuesbeyond the predicted interval are considered as energy anomaly. Simulation results show thatthis model can effectively found the energy consumption anomalies. 5. One energy management system based on energy consumption prediction model isdesigned, including system architecture design, function modules design, database design andso on. And an energy consumption prediction model library is built. This energy managementsystem achieves good results in the initial application.