Forecast with Seasonal Trend Model 

Purpose

With this process the system executes a forecast with the seasonal trend model. The seasonal trend model can be used if you want to create a forecast for a seasonal product, but the available demand history is less than 24 periods long.

Using this forecast model, the system does not smooth the seasonal factors by default; that is, the gamma factor is zero. You must enter the seasonal coefficients in Customizing.

Before calculating the forecast, the system executes model initialization and outlier correction. After calculating the forecast, the system calculates the standard deviation and the MAD.

The system does not consider the seasonal trend model during automatic model selection.

Prerequisites

●      In Customizing for Advanced Planning and Optimization under Supply Chain Planning → Service Parts Planning (SPP) → Forecasting, you have defined a seasonal group profile with seasonal factors in the IMG activity Define Seasonal Coefficients.

●      In the SAP Easy Access menu under Advanced Planning and Optimization → Service Parts Planning → Planning → Forecasting → Forecast Profile, you have assigned seasonal group profiles in the Profile Name of Smoothing Factor Set field on the Model Parameter tab page.

Process

●      The system uses the smoothed trend variable. The system calculates forecast values with the seasonal trend model using the following formula:

  

○       F(p+1) is the forecast for the next period.

○       V(p) is the final history of the last period.

This value is displayed in the Final History key figure on the Adjust Demand History screen. You get to this screen on the SAP Easy Access screen under Advanced Planning and Optimization → Service Parts Planning (SPP) → Environment → Data Administration → Adjust Demand History.

○       B(p) is the basic value.

○       T(p) is the trend value.

○       S(p) is the seasonal value.

○       S(p-L) is the seasonal value for the same period last season.

○       i is the period in the future for which a forecast is to be created.

○       L is the length of the season in periods (always 12).

○       Alpha is the smoothing constant for the basis value, beta for the trend value, and gamma for seasonal factors.

You define the smoothing constants alpha and beta in the forecast profile on the Model Parameter tab page in the Alpha Factor and Beta Factor fields. The smoothing constant is zero.