Reliability Optimization Design Of Automobile Crashworthiness Structure Based On Probabilistic-interval Hybrid Model

At present,safety design has become a research hotspot in the field of modern automobile industry.Therefore,consideration of structural optimization design to improve vehicle crash safety has been extensively studied.However,there are a lot of uncertainties in the optimization design of automobile crashworthiness structures,such as geometric dimensions,material properties,loads and boundary conditions.These uncertainties lead to a certain tolerance between the structural parameters and nominal values,and the tolerance of the uncertain variables directly affects the safety performance of the automobile structure,resulting in a lack of reliability in the whole design process.Therefore,it is very important to the structural reliability optimize design for the safety of vehicle crashworthiness.Traditional reliability optimization methods mainly use probability model to describe uncertain variables.However,the probability model relies on the exact probability distribution of random variables,and it is often very difficult to obtain the exact probability distribution of random variables.In many cases,the upper and lower bounds of uncertain variables are easier to obtain,and this kind of uncertain variables is suitable to be described by interval model.In practical engineering,we often encounter the following situation:part of the uncertain parameters is enough information,so it is suitable to use probability model to describe it;the other part of the uncertain parameters only knows the upper and lower bounds due to the lack of sufficient sample information,which is suitable for interval model description.Therefore,the study of the hybrid reliability optimization design with probability variables and interval uncertain variables has important practical engineering significance.In this paper,the reliability optimization design of automobile crash structure based on probability-interval hybrid model is explored and studied in order to improve the efficiency of optimization and the accuracy of solution.The main contents of the study are as follows1.A probability-interval hybrid reliability optimization method is proposed.The reliability optimization model is constructed by combining probability model and interval model to describe the uncertain parameters.The probability-interval hybrid reliability optimization model is solved by Sequence Optimization and Reliability Assessment method(SORA),and the nesting optimization problem is decoupling into a single-level optimization problem.2.A hybrid reliability optimization design method based on sequential Kriging model is proposed.The Kriging approximation model is used to construct the approximate objective function and constraint function.At each iteration step,the minimum function measure points and the most disadvantageous points of the interval are obtained based on the Kriging approximation model.The Kriging approximation model is updated by the improved maximum expectation criterion sampling,and the deterministic optimization and reliability evaluation are carried out based on the current optimal functional measurement points and the most disadvantageous points of the interval.Three numerical examples and one engineering example are used to verify the effectiveness of the method.3.A reliability optimization design method for vehicle front collision based on probability-interval hybrid model is proposed.According to the distribution types of uncertain parameters such as thickness and the yield strength of the material of automobile structure are described by probability model and interval model,respectively.Considering the optimization goal of vehicle safety with maximum collision force,the energy absorption index of the whole vehicle is selected as the reliability constraint.The hybrid reliability optimization model of automobile collision is established.The sample data of the finite element simulation model of automobile front collision are generated by Latin hypercube design of experiment,and the Kriging approximate model of objective and constraint functions are constructed according to the finite element simulation results.SORA method is used to decoupling and optimize each iterative step.Based on the sequence update Kriging model,the function measure method is used to evaluate the probability constraints.The results show that the proposed reliability optimization method based on probability-interval hybrid model meets the requirements of vehicle safety and reliability design,and the maximum energy absorption of the whole vehicle is increased by 14.92%,and the collision peak force of the whole vehicle is reduced by 8.11%.