Energy Load Forecasting System Architecture

The following diagram illustrates the flow of data throughout the Energy Load Forecasting system, and the CygNet services that are involved in storage and calculation.

The diagram shows how external historical gas load data, forecast gas load data, historical input data, and forecast input data must be delivered into the system, which system data must be created, maintained and be stored internally, and the processes by which the various categories of data must be utilized to generate energy load forecasts.

CygNet’s implementation of the ELF system involves several logical processes, each of which is represented by the circular symbols on the diagram above. The logical processes in the ELF system include the following:

• external input data import
• neural network training
• energy load forecasting
• forecast variance analysis.

While each process may be depicted as distinct in the diagram above, a single logical process may be implemented as more than one actual process. Each logical process is described in the following topics.