ANOVA-Based Assessment of the Model-Free Adaptive Controllers Tuning Parameters

In this paper, a statistical approach for the assessment of the effectiveness level for the tuning parameters of the model-free adaptive control (MFAC) method is presented with the compact form dynamic linearization (CFDL) data model in single-input single-output unknown nonlinear systems. This is a first step in the parameter tuning of the MFAC data-driven methodology. The statistical approach is accomplished using the tool of N-way analysis of variance (ANOVA), the result of which is the determination of the most statistically significant parameters on the system performance. The process of assessing the effective parameters is data-based and no prior mathematical model assumptions are required. Two linear and nonlinear sample systems are considered and ANOVA tests are performed for different control performance assessment indexes. Finally, the simulation results are summarized and comparison results are provided by trial and error variation of the MFAC tuning parameters.