Multiple model adaptive control (MMAC) is a model-based control strategy which incorporates a set of model/controller pairs rather than relying on a single model and controller to handle all possible operating conditions. A bank of of models is used with a weighting function which chooses the single model or combination of models that best represents the current plant input--output behaviour. A weighted-sum of the controller outputs is then used to supply the control action to the plant.
Although MMAC is model-based, it relaxes the requirement for a single, precise, nonlinear model. Often such a nonlinear model is not available either due to lack of process knowledge or insufficient ``non-production" time for development of detailed models. Many initial MMAC applications involved aircraft control where models for different flight conditions were used in the model bank. More recently, MMAC has been used in drug infusion systems where plant (i.e. patient) parameter variability is the largest concern.
In this chapter we review MMAC theory and applications. We present a specific single-input--single-output example based on a drug infusion application to highlight some of the strengths and weaknesses of the MMAC method.