Abstract: In order to optimize mining dump truck ride comfort for bad driving traffic, nonlinear stiffness and damping characteristics of hydro-pneumatic suspension are needed. Time-varying stiffness and damping with linear combination under a series scaling function is scattered, based on Daubechies wavelet’s compactness and regularization and least-square method. This means to turn “black problem” of identifying time-varying parameters into identifying invariant coefficients when the scaling function sequence and the input and output of system are known. The multi-rigid body dynamic model of whole truck is built in Adams/view for the sake of testing stiffness and damping of identifying. Time-acceleration and power spectral density under identifying parameter are found to be extremely close to the result of experiment, which implies validity of this method. The physical parameters of hydro-pneumatic suspension by genetic algorithm is optimized, which makes the identifying parameter as the initial condition and makes the root mean square value of the seat acceleration as optimization objective. After optimizing, the value is descended by 51.84%, which achieves the purpose of improving the ride comfort.

Key words: Daubechies wavelet, Genetic algorithm, Identification, Least-square method, Ride comfort, Scaling function