I. Introduction

In general, parameter estimation methods assume availability of measured data containing adequate information about the cause effect relationship with a minimum amount of corruption caused by systematic errors like scale factor, zero shift biases, and time lags. The process to generate experimental data with adequate information contents has been dealt with in Chapter 2 on data gathering. In this chapter we deal with the means of checking and improving the quality of the recorded data. The presence of noise is treated differently by the various methods. As discussed in Chapter 6, the least squares estimates are sensitive to such systematic errors and noise in the independent variables. The output error method of Chapter 4 and filter error method of Chapter 5 principally allow estimation of noise statistics and incorporation of corrections for systematic errors as unknown parameters, but it may lead to correlations among the initial conditions or other aerodynamic derivatives, which affects the convergence and accuracy of the estimates. Thus, a data check, independent of system parameter estimation, is usually desirable. Furthermore, in the context of aerodynamic characterization of an aircraft, a large number of variables is usually measured and recorded during flight test programs. Before using the raw flight test data, it is often necessary and time-saving to verify whether the recorded data are compatible or not. The basis for verifying the compatibility of measured data is the use of kinematic relationships. Such a general procedure, which in addition allows for estimation of systematic instrument errors,1 ^,