| When studying and analyzing the relevant characteristics of the projectile,a resistive strain sensor is attached to the outer wall of the barrel to obtain the situation of the projectile inside the barrel.However,the actual collected signal features are not obvious,contain a large amount of noise,and due to the inherent properties of the sensor,it is not possible to directly obtain ideal data for analysis and processing.Therefore,it is necessary to analyze and study the collected array strain signals.Not only can the relevant hardware module composition of the analysis system be optimized and improved,but also the data can be further processed and corrected to obtain relevant information and features of the projectile,Finally,obtain the ideal precise data.In this regard,this article will conduct error and impact analysis research on the composition of the strain collection system,and process and correct the measured data based on the aforementioned analysis.The main content is as follows:For the high g-value strain collection system,a brief overview of the various modules of its system composition is provided.Then,starting from the core components of the resistive strain sensor and the various modules of the collection system,possible errors and influencing factors are analyzed,and the degree of influence of each influencing factor on the measurement system is judged.Finally,improvement and treatment plans are proposed for some influencing factors.For the collection of nonlinear and non-stationary signals by the high g-value strain collection system,the following processing is carried out based on the aforementioned analysis.Firstly,in response to the harsh experimental environment of the system,where high temperatures can affect sensor data measurement,a BP neural network algorithm based on improved Tianniu swarm optimization is studied.The algorithm optimizes the parameters of the BP neural network through the improved Tianniu Qun optimization algorithm,which avoids the BP neural network from falling into the local optimum.Through the comparison of simulation experiments,the method has a smaller fitness,so the search ability is better,and finally highprecision correction of strain values.Secondly,an improved CEEMDAN algorithm was studied to address the presence of certain noise in the system.This method first improves on the shortcomings of EMD methods,and then analyzes them using three methods to select the required components and determine the screening criteria.Therefore,it can quickly and accurately adaptively process array strain signals.Through simulation experiments,it was found that this method is superior to traditional methods and avoids the misjudgment caused by component selection using one method,effectively extracting the effective components of the signal.Moreover,due to microprocessors,there may be energy leakage when analyzing the spectral characteristics of data in practice.Therefore,using the spectral center correction method can perform spectral analysis more accurately.Finally,the system finally needs to fit the trajectory curve of the projectile motion through the position information of the sensor and the time information of the projectile arriving at the sensor.Therefore,three commonly used methods in engineering are adopted,namely,the least square method,cubic B-spline interpolation method,and cubic Hermite interpolation interpolation method to fit the curve of the simulation experimental data,and the fitting effect is evaluated by calculating the relevant indicators: residual range,sum variance,root mean square error,etc.Based on the actual measured array strain data,the effective components of the signal are extracted using the aforementioned method and spectrum correction and analysis are performed.Then,by obtaining the same feature points of the signal,curve fitting is performed on the discrete data to reproduce the motion parameters of the projectile. |
