Ride-comfort-oriented Analysis And Optimization Of A Vehicle Using The Pseudo-excitation Method

With the progress and development of the society, vehicles have become indispensable tools of transport. However, drivers and passengers often have to stay in a vehicle for several hours or even one or two days, which may cause them uncomfortable or even very tired. Therefore, it is necessary to find appropriate design parameters, in order to improve the comfort of passengers.Vehicle vibrations induced by road irregularity are essentially random, however, optimum design associated with such random vibration has not been well developed and the existing methods are very time consuming due to the inefficient random vibration analysis. Moreover, it is previously difficult to obtain the gradients of the objective function and constraint functions with respect to the design variables, so the traditional gradient-based optimization methods are not suitable to such optimality problems. In this thesis, the pseudo excitation method has been successfully used in vehicle random vibration analysis, which reduces computation efforts considerably while retaining theoretical accuracy, and an efficient sensitivity analysis formula for optimizing the ride comfort of vehicle suspension system is derived. Based on the dynamic response analysis using PEM, the first and second orders of sensitivity formulas are calculated conveniently by differentiating these equations. At the same time, a non-linear fitting and sensitivity-based method, known as Method of Moving Asymptotes (MMA), was used in this thesis to solve the suspension optimization of a vehicle successfully.Because of the complexity and low efficiency of the conventional random vibration approach, previously, a vehicle was simplified into a model with very limited degrees of freedom typically ten or twenty, in order to perform the random vibration analysis or optimization. However, such crude models are sometimes not sufficient for some special analysis, e.g. when the fatigue life of some high stress areas is of special concern, constructing a FE model with many DOFs would be necessary. A real vehicle model is investigated in this thesis by means of a FE model, which agrees very well with those based on the realistic situatio. In the present thesis, the use of such a detailed FE model of a coach in conjunction with PEM has been well justified by its computational efficiency and the accuracy of the numerical results presented.