Identification of the separately excited DC motor with extended instrumental variables

Abstract

The problem of identifying the parameters of a DC motor is considered. The presence of measurement errors of currents and voltages leads to errors in both input and output signals. Existing methods for identifying the parameters of a DC motor do not take into account the presence of errors in both the output and input signals. A discrete model of a DC motor designed for use in digital systems is explored. One of the advantages of this model is the reduction of the problem of parameter estimation to two reduced dimension problems. Reducing the dimension of the problems being solved improves their conditionality and reduces sensitivity to measurement errors. A new method for identifying the parameters of DC motor models is proposed. Parameter estimates are found based on the extended method of instrumental variables. To find estimates, the use of extended equivalent systems of equations is proposed. Extended equivalent systems make it possible to reduce the condition numbers. The conducted numerical experiments showed the advantage of the proposed algorithm over the least squares. The results of the work can be applied in the development of predictive diagnostic systems. The use of the proposed method for identifying the parameters of a DC motor also makes it possible to reduce the requirements for measuring devices due to less sensitivity to measurement errors.