Entropy compromise programming method for parameter identification in the seated driver biomechanical model

Abstract

The paper proposes a method for identifying principal parameters of the n-DOF ( n degrees-of-freedom) biomechanical model of a seated human driver. The model is considered as a structure represented by masses of human body segments, mechanical springs and dampers, and also as a biomechanical system with dynamic performances described in terms of response functions in the frequency domain, such as the driving-point mechanical impedance, and seat-to-head transmissibility function. The problem addressed is identifying the structural components that provide the best dynamic model performance for the related driver-seat system. The method proposed is based on the Shannon entropy concept for identifying the intrinsic information contained in a stochastic process, which describes the dynamic error in model response to randomly generated sets of structural model parameters and vibrational excitations in the frequency domain. The appropriate use of error functions shifts the problem of identification into multicriteria framework. This makes it possible to determine the objective weights of errors and identify the solution closest to an ideal point. The compromise programming method is used for ranking various solutions and pointing to the best. The backward procedure locates the parameter combination for which the response of the model is most accurate. The proposed method is comprehensive in approach and efficient in application. Relevance to industry The mechanical response of the seated human body to the vibrational excitation can be predicted reasonably well by the n-DOF model if the parameters are identified and used correctly. The method suggested in this paper, which is based on the entropy evolution of the information data obtained by measuring and statistical analyses, enables an easy and objective identification of the parameters in a straightforward manner.