Search Results

The stochastic properties of continuous time model parameters obtained through discrete least squares estimation are examined. Particular attention is given to the application of estimating the fuel evaporation dynamics of a V-8 SI engine. The continuous time parameter distributions in this case are biased. The bias is shown to be a function of both measurement noise and sampling rate selection. Analysis and experimental results suggest that for each particular model, there is a corresponding optimum sampling rate. A bias compensation formula is proposed that improves the accuracy of least squares estimation without iterative techniques.