Research On Ignition Control And Failure Evaluation Method Of Pyrotechnic Based On DSP

With the development of electronic technology,the higher safety and reliability of pyrotechnics are required in aerospace and military fields.As the core component of weapon system,pyrotechnics have contributed greatly to ammunition launch and target damage.Once the initiating explosive devices fail,weapon system failure,economic loss or even casualties will be caused.Therefore,it is significant in strategy to evaluate the failure of pyrotechnics.An improved classification method based on Hilbert Huang Transform(HHT)and Support Vector Machine(SVM)is proposed to evaluate the failure of the pyrotechnics in this paper.The main contents of this paper are as follows:Firstly,aiming at the problem of low accuracy when SVM model is adopted with default parameters,an optimization algorithm based on Levy-PSO is proposed to obtain the best values of SVM model parameters.When Particle Swarm Optimization(PSO)is used to optimize SVM model,there are some defects such as slow convergence speed and prone to local optimum.In this paper,an improved optimization algorithm which combined Levy flight with PSO algorithm is proposed.The simulation results show that the accuracy of SVM model based on Levy-PSO is greatly improved.Secondly,aiming at the shortcomings of the traditional initiating device,such as large volume,poor versatility,poor mobility and so on,a device with super capacitor as energy storage element and DSP TMS320F28335 as controller is designed.The current and voltage signals of initiating explosive device are collected during initiation,and the data needed in the subsequent failure evaluation algorithm is provided.Thirdly,the characteristics of initiating current signal are extracted.The initiation process of initiating explosive device involves complex physical and chemical changes.The characteristics of each stage in initiation is characterized by current.According to the non-stationary and non-linear complex characteristics of initiation current signal,HHT method is adopted to extract the characteristics of current signal and the characteristic energy matrix is constructed as the input of SVM.Through learning and training,the simulation results show that the evaluation accuracy can reach 96.66%.Finally,the experimental verification of failure evaluation is completed in the system.The trained SVM model is transplanted to the initiation system to complete the failure evaluation.The results show that the failure evaluation accuracy of initiating explosive device can reach 86.6%,which meets the system design index and requirements.and the basis for the failure location and follow-up research of initiating explosive devices are provided.