Short-term Power Load Forecast Of Office Park

Short-term load forecasting is the basis of smart energy system operation.Accurate forecasting can provide strong data support for the formulation of scheduling plan,and ensure the safe and stable operation of microgrid in smart energy system.At the same time,wind power and photovoltaic power generation can be absorbed locally to the maximum extent,and the operation efficiency of intelligent energy system can be improved.Generally,researchers predict the power load according to the power consumption area or users.These studies are based on the whole,and lack of research on the power load characteristics of a single electrical equipment and their impact on the overall power load.In this paper,the short-term power load prediction of an office park is taken as the research object,combined with the power consumption characteristics and laws of the office park,the following research work is carried out for the specific power consumption equipment:(1)For the electric load of central air conditioning,a prediction model based on the operation mechanism of central air conditioning and the method of cooling load coefficient is established.Secondly,the combined prediction model of variable weight coefficient based on Long Short-Term Memory neural network(LSTM),Support Vector Regression(SVR)and Extreme Learning Machine(ELM)is established.According to the historical load and environmental temperature data,the simulation experiment is carried out.The experimental results show that the average absolute error and other performance indicators meet the actual project requirements.(2)For the electric load of the charging station,according to the initial charging time and initial charging capacity distribution of the electric vehicle in the office park,Monte Carlo simulation method is used to simulate the charging load of the charging station during the working day,and the simulation results can provide reference for the formulation of scheduling.Secondly,according to the historical load data of charging station,a combined forecasting model is established.The experimental results show that the performance of the combination prediction is stable compared with the single algorithm,and it can provide stable and reliable prediction data for the intelligent energy system scheduling in the office park.