Energetics
The application of forecasting methods (correlation, regression, time series analysis,
neural networks, SVM models) does not produce a sufficiently accurate result for each (special)
day or day type. The forecasting framework can predict with great precision, taking several different
aspects into account. Considering of certain specialties, various predictive methods can be created.
These together from the hybrid predictor.
- It can take special care of the types of days / seasons
- Learning algorithms are able to learn the noise, so the system can take the daily exclusions into account.
- When preparing the forecasting the program also pays attention to the weather conditions.
- It can recognize the hot and very cold days.
The Sound Center Central Control (HKV) devices can active certain zones of the energy provider The customers use electric boilers and stoves typically on these circuits. However, due to the operation of electric boilers, it is questionable how long and how big electric energy demand will be in the given zone. The solution of Dopti Ltd. is able to create a mathematical model that determines the energy consumption under different conditions.
The amount of purchased energy on the stock must be recorded and reported to the local energy authority. This must be monitored to see the amount of hedge covered and to react quickly to changing market conditions. With the help of the unique energy procurement software developed by Dopti Ltd. the user can keep track of existing contracts, energy needs, and the proper reports can be created quickly and efficiently.