Impact Monitoring Of Composite Plate Based On Sensor Distribution Optimization

Since the last century,composite materials have been widely used in aerospace,high-speed rail and automotive industries because of their light weight and strong ability.However,the impact resistance of composites is poor,and the impact damage concealment is strong,which has caused great potential safety hazards to the manufacture and usability of composites.Therefore,it is important to accurately identify the location and size of impact load for the health monitoring of composite materials.The purpose of this study is to obtain better impact monitoring effect of composite plates by optimizing the distribution of sensors.Next,the main contents of this study are introduced.(1)The finite element model of composite plate was established.By applying two different degrees of impact to 64 nodes with uniform distribution on the surface of composite plate,128 kinds of shock response signals in the whole region of composite plate were obtained.(2)Optimizing sensor distribution.The purpose of sensor distribution optimization is to improve the accuracy and economy of monitoring,that is,to ensure that the impact recognition rate of different types can meet the requirements when the number of sensor response points is as small as possible.In this study,the objective functions of sensor distribution optimization are as follows: 1)the number of sensors,2)characteristic performance parameters.In order to quantify the second objective function,the wavelet band energy analysis is used to extract the characteristics of the shock response signal,then the principal component analysis is used to reduce the dimension,and then the minimum distance of the eigenvectors between different shock categories is used as the characteristic performance index parameters.Since the number of sensors is as small as possible and the accuracy of monitoring is as accurate as possible as the contradictory objective function,this is a two-objective optimization problem.In this study,a variety of multi-objective optimization algorithms,including improved non-inferior hierarchical genetic algorithm(NSGA-?)and non-inferior hierarchical monkey swarm algorithm(NSMA),are used to compare and study in order to obtain better sensor distribution optimization results.The results show that the two optimization algorithms can obtain the same optimal sensor distribution,but for the problems in this study,the convergence speed of NSGA-? is faster than that of NSMA.(3)To verify the effectiveness of impact condition monitoring of composite plate based on sensor distribution optimization,an experimental system for impact condition monitoring of composite plate was designed.According to the four groups of optimization results given by the above sensor distribution optimization,the corresponding sensor networks were constructed and compared.The results show that a)when the optimal sensor network arrangement is 2,3 and 4 sensors respectively,the recognition accuracy is 25.27%,59.14% and 76.95%,respectively.The results show that the number of sensors is positively correlated with the impact recognition accuracy.b)The recognition accuracy of the second non-inferior sensor network arrangement with four sensors is 75.55%,which is lower than that of the fourth optimal sensor network arrangement.That is,the performance parameters are positively correlated with the impact recognition results.It shows that the optimal sensor distribution mode can be obtained by using the optimization method in this study,and the recognition accuracy of the condition monitoring system can be improved.It has certain theoretical significance and practical value.