The ELF device configuration, neural network training, and forecasting engine is implemented as an import/export EIE driver in the UIS.
The ELF EIE is responsible for importing historical gas load data, forecast gas load data, historical input data, and forecast input data, and storing that data in CygNet (see External Input Data Import).
The EIE then uses this data to perform neural network definition training (see Neural Network Training) and energy load forecasting tasks (see Energy Load Forecasting).
The analysis of forecast variance is a background processes in the ELF driver. Variance analysis processing adds retraining tasks to the training priority queue for neural network definitions whose forecasts are not within an acceptable range of the actual gas load (see Forecast Variance Analysis).
Configuration of the ELF system is made possible by using the property pages of the Energy Load Forecasting EIE editor. From the EIE editor an administrator can control all aspects of the forecasting process, including: the configuration of system default attributes, specification of file input locations, input data category setup, neural network template and definition creation, and viewing of the statuses of neural network definitions (see CygNet ELF EIE Editor).
A 64-bit version of this EIE is available. See 64-bit Device Drivers.
|
|
|