摘要：AbstractIndirect inference has been widely used in the field of econometrics for time series data, and only introduced to the field of system identification in recent years. It is especially useful as a parameter estimation technique when the log-likelihood of the real system is difficult to optimise or, for some systems, impossible to express explicitly. In this paper, we apply indirect inference to the cascaded tanks benchmark problem. The log-likelihood function of the grey-box model of the cascaded tanks, obtained based on some physical laws, is difficult to express in terms of the parameters to be estimated. Indirect inference provides an alternative way for parameter estimation of complex systems through the use of misspecified models. The misspecified model structure may not necessarily represent the true system but its likelihood function is easier to optimise. The results we obtained suggest that indirect inference is an appropriate method for the cascaded tanks benchmark problem.
关键词：KeywordsSystem identificationIndirect inferenceNonlinear systems
