DYNAMIC MODELING AND SYSTEM IDENTIFICATION OF AN INVERTED PENDULUM APPLIED TO TALOCRURAL JOINT BIOMECHANICS

Abstract

Peripheral Neuropathy (PN) is associated with dysfunctions of the peripheral nervous system, potentially compromising postural control and increasing fall risks, especially in elderly individuals. In this context, the biomechanical modeling of human balance becomes relevant for developing analysis and rehabilitation strategies. This study proposes a methodology for estimating representative models of the talocrural joint behavior using the inverted pendulum analogy. The approach integrates mathematical modeling, computer simulation, and experimental validation through a low-cost electromechanical plant instrumented with an inertial sensor. It estimates two mathematical models using system identification techniques based on the acquired data. Results indicate that the model obtained from the acquisition strategy considering the entire operational range presents better performance, yielding a Mean Squared Error (MSE) of 0.0389 and a Mean Absolute Error (MAE) of 0.0129. This result demonstrates high agreement with the real system behavior. The findings evidence the feasibility of the proposed methodology for applications in postural control studies and the development of assistive technologies for motor rehabilitation.