Implementation Of Narrowband Radar Air Target Classification Method Based On PowerPC

Since the micro-doppler effect can provide description of moving target characteristics,this unique features of moving targets can provide a new technical way when applied into the field of radar target recognition.It's of great significance to classify Jet aircraft,propeller aircraft and helicopters as they undertake different tasks and carry different threat in the battlefield.On the other hand,the classification of aircraft target can be based on the micro-doppler information,as the geometry size and moving characteristics of the three kinds of plane is obvious different,which will produce different micro-doppler modulation on radar transmitting wave when they are flying.In recent years,with Power PC processor with high performance,low power consumption,and Vx Works operating system wit high stability,good real-time having been widely used in various fields,the real-time signal processing system design based on Power PC processor has become an important research direction in radar signal processing system design.This thesis proposed some research on the narrowband radar target classification and system implementation,and the main contents included principle of airplane target classification,analysis of the engineering application,system implementation based on Power PC processor and so on.The details summarized as follows.First of all,this thesis introduced theory basis of narrowband radar aircraft target classification,provided the ideal rotor model based on the principle of JEM,analyzed the effect of the actual aircraft on an ideal parameter model,and described the method of target classification eliminating clutter and feature extraction,and the SVM classifier for classification was also introduced.Then,combining with the engineering application background of aircraft target classification,the thesis were studied from the following aspects: characteristics feasibility of engineering application,system parameter requirements,reducing the computational complexity and improving the recognition rate.Besides,by analyzing characteristics feasibility and selecting suitable feature feasibility method,the system parameter requirements can be analyzed,the computational complexity can be reduced by reducing the feature dimension,and the recognition rate can be improved by voting decisions.Finally,a detailed implementation of air target recognition and classification system based on Power PC hardware platform was discussed in this thesis.The target classification processing board and MPC8640 D processor were introduced,the system data path and memory allocation were illustrated,and the required computational complexity of the classification function was estimated.The thesis adopted multitasking classification system design scheme,and the availability of the software was validated by the simulated data.