A scalar Kalman filter is applied for a Least-Square estimated value Hs at s. The filter has an input for receiving Hs , a filter equation and an out for the corrected estimated value Hk for the kth variable. The filter equation is Hk = Kgain Sn [k] wherein: correction Sn [k] = S + Kn ( Hs -S); prediction of the correction S = Ka Sn [k]; Kalman filter gain Kn = P / ( 1 + P); minimum predication MSE P = Ka2 Pn [k] + Kb; minimum MSE Pn [k] = P ( 1 - Kn ); and Ka , Kgain and Kb are constants.