Research On Shock Wave Test System And Dynamic Characteristic Compensation

Shock wave over-pressure is an important parameter for weapon damage assessment.The accuracy of test data directly affects the authenticity of damage assessment.As a warhead explosion shock wave signal test equipment,shock wave over-pressure storage tester and its dynamic performance of the system composed of sensors is the key to accurate signal testing.Expanded shock wave overpressure storage tester performance optimization design and test system off-line compensation research dynamic characteristics,which is of practical significance to improve the accuracy of weapon damage assessment.Starting from the optimization design of the tester,the design points of the signal conditioning circuit and the key device selection method are expounded.In order to improve the data storage rate of negative latency storage space,a high-speed storage structure composed of two-level Flash memory array is proposed,and the design idea is elaborated from hardware structure,storage rate analysis and software work flow.In addition,a detailed analysis of the argument about the negative delay storage space data storage mode,Flash memory array bad block detection,and shock wave over-pressure key data wireless extraction to achieve the principle be given.On the basis of the developed shock wave over-pressure tester,the dynamic characteristic compensation of the test system composed of the tester and the sensor be carried out.Firstly,the dynamic characteristic description equation,the system dynamic characteristic response and related indexes of the test system are introduced.The principle of system identification and system dynamic characteristic compensation based on BP neural network is deduced and the effectiveness of the above method is proved by the experimental simulation.Then,through the step pressure signal generated by the shock wave tube encourage test system to obtain the system time domain response data,and the dynamic characteristic compensation model of the BP neural network system is off-line trained by the system input and output data.Aiming at the problem of insufficient compensation network training data,a method of generating a large number of training samples under the excitation of the step signal cluster by the system model identified by neural network is proposed to improve the generalization ability of the dynamic compensation network of the test system.Finally,by comparing the error relationship between the measured shock wave data,off-line compensation data and the oretical calculation data,the validity of the compensation algorithm to improve the dynamic characteristics of the test system can be verified.