Frequency Response Analysis of Optimized ARX Model of Physiological System (A Case Study of Dynamic Cerebral Autoregulation Mechanism)

Quantitative modeling is an important tool for research on complex physiological systems. Mathematical model are playing important role in analyzing complex system. In order to create an accurate model and significant interpretations, system identification techniques are needed. An optimized auto-regressive model based on Sequential Genetic algorithm (Sequential GA) is investigated in this study to identify and analyze the physiological system. The results of optimized ARX model and simple ARX are compared here in terms of their fitness, and linear dynamic behaviour of the cerebral autoregulation is analyzed using frequency response, bode plot. Results based on a case study show that the optimized ARX model has, in general a higher accuracy of prediction and it can also be easily interpreted to provide physiologically meaningful information. Key Words-cerebral autoregulation, frequency response, mathematical modelling, system identification.