Multiple tuned model approach for the analysis of nonlinear dynamics of the long term blood pressure regulation

Abstract

This paper presents a modified parameter tuning approach that enables a better quantitative analysis of the nonlinear dynamics of long term arterial blood pressure regulation. A previously developed nonlinear model of blood pressure regulation is linearized about its equilibrium point. Instead of any traditional fixed parameter approach this linearized version utilizes multiple model approach. Fractional adjustments have been made in the parameters of the basal transfer function or state model of the seventh order system to fit with the nonlinear system data. A detailed sensitivity study has been performed on the basal linear model to identify the tuning parameters, their basal values and the realistic bounds. Traditional least square estimation technique has been used for parameter identification. Thus a set of approximate linear models have been developed under various physiological conditions. By combining these linear models spanning in the range of expected operation of the nonlinear system, the actual behavior can be approached. The dynamic responses of the tuned models under various conditions such as aldosterone loading are closely matching to those of the actual nonlinear model. The eigen value study of the linearized model shows that the model is stable under normal conditions. Thus the proposed frame work for the parameter tuning will be useful for the nonlinear physiological model to get a better structural analysis using the well known linear system techniques.