T/JES 001-2025 The technical specification for collaborative prediction of load in sub-stations across different levels of power systems in new China

Scope

Main technical content: The standard specifies the basic technical requirements for collaborative load forecasting of substations at the provincial, municipal, and county levels in the new power system. Based on historical data related to different load levels, a load forecasting model is established. Models can be set up for different load types, different load levels, different types of days, and different time periods, or a comprehensive load forecasting model can be established according to business needs. The selection is based on comprehensive consideration of the load forecasting model, load characteristics, forecasting accuracy requirements, seasons, weather, etc. Forecasting algorithms can include least squares method, linear regression, multiplicative smoothing method, fuzzy clustering method, time series algorithm, linear extrapolation method, frequency domain component method, artificial intelligence algorithm, expert system method, etc. One or more algorithms can be selected for combined forecasting. Load forecasting is divided into two stages: initial forecasting and forecasting correction. Initial forecasting is carried out through the load forecasting model, and forecasting correction takes into account relevant influencing factors (expert experience, social factors, weather factors, new energy power generation factors, annual natural growth rate) to correct the initial forecasting results. For load forecasting of social factors and weather factors on similar days, it is appropriate to consider combining similar days and superimposing the annual natural growth rate for load forecasting correction.

Development Information

Related Standards