Data mining on crashworthiness data of occupant restraint system

The applicability of data mining on the finite element data and the crashworthiness simulation data of occupant restraint system was demonstrated.According to the characteristics of CAE data,a framework for data preparation was developed based on object-oriented ideology.Decision tree and attribute selection algorithms were used to perform data mining on the crash simulation data of 98 instances.The results of the data mining give us a deeper insight into the interrelations between the key parameters of occupant restrain system in design and the performances in crash simulation.Finally,the learned rules have been tested with the real data sets.The results of the testing show that the learned rules are proper,and can be used in design of occupant restraint systems.