Multivariable regression - Manual - Energy Manager - Industrial Edge - Industrial Edge App

• With multivariable regression, a parameter (KPIs or variables) is placed in context with other parameters and mapped as a linear regression model. In this way, the value of the output parameter can be determined using the input parameters and the regression model. (Model result) example: The energy consumption of a plant is affected by certain influencing factors such as ambient temperature, the quantity produced and the quantity filled. The theoretical energy consumption can be determined if the regression model and the input variables are known. This can serve as a basis for the actual energy consumption. In addition to the model, qualitative parameters are also output, which determine the quality of the model. It may be that input variables are used which have no relation to the output variable. In this case, the correlation coefficient is 0. This means that the model is not trustworthy.
  Different models can be calculated and saved per asset and per parameter. The saved models are available to the user in an automatically generated dashboard, where the result of the model is compared to the actual value. In order to better recognize the deviation, the deviation and is also displayed in addition to the model result and the actual measured values.
  After the KPIs for the model result and the deviation are created for each model, a limit value can be defined via the standard functionality, at which the user receives a notification via the notifier.