Page 46 - Industrial Process Plant Construction Estimating and Man Hour Analysis

16 Industrial process plant construction estimating and man-hour analysis

1.6.3 Simple exponential smoothing
A statistical technique for working out averages while allowing for recent
changes in values by moving the period under consideration at regular intervals.
In simple exponential smoothing, the forecast value for Y at time ¡ + 1 that is
made at time t equals the simple average of the most recent m observations. In
particular, for some α where 0<¼ α <¼1 for all i >1, define

^ y ¼ y ^ y þ1 ¼ ^ y¡þae¡
1 1 i
Iteration is expressed as
^ y ¼ y ^ yiþ1 ¼ ay¡þ 1 að Þ ^ y¡
1 1
Using a simple exponential smoothing, forecast the next value(s) with
α ¼0.2.
Most recent period’s demand multiplied by the smoothing factor, plus the
most recent period’s forecast multiplied by (one minus the smoothing factor)
Formula:
∗
F t ¼ α Dþ 1 αð Þ∗F
where
D¼most recent period’s demand
α ¼smoothing factor in decimal form
F¼most recent period’s forecast
Enter formula in cell E371 is ¼D371, and the formula in cell E372 is
¼$D$382*D371+(1 $D$382)*E371 copy formula in cell E373 to cell E378.
Enter formula in cell F372 is ¼D372-E372, and the formula in cell G372 is
¼F372^2; copy formula in cell F373 to cell F378 and formula in cell G372 to
cell G378.