Research On De-Noising Algorithm Of Millimeter Wave Short-Range Detecting System Target Signals Based On Stochastic Resonance

Owing to its advantage in bad weather conditions,coupled with the development of millimeter wave solid state devices technologies,the millimeter wave short-range detecting technology has been widely used in many fileds.However,when it worked in the rainy,snowy,smog,haze and other harsh conditions,the performance of millimeter wave short-range detection system is greatly affected by noise.Thus,the signal denoising technology becomes to be an indispendsable part of millimeter wave precision detection systems and the effective noise removing technique can improve its performance.The existing linear denoising technique is not effective for millimeter wave short-range detection system when working in low singal-to-noise radio conditions.Stochastic resonance is a nonlinear denoising technology,and can obtain better singal-to-noise radio compared with linear denoising methods.This paper focuses on detected signal denoisng methods of millimeter wave short-range detecting system,and the stochastic resonance method is applied to weak signal detected from millimeter wave detecting system,mainly including the follows:A weak signal denoising method based on stochastic resonance of bistable system model is studied.Accoding to the high frequency characteristics of the millimeter wave detection system,an improved method of frequency compression denoising algorithm based on stochastic resonance is proposed,the power spectrum gain and output signal to noise ratio(SNR)is analyzed.The results show that the proposed method is able to denoise the millimeter-wave detecting signal in real time.The millimeter-wave detecting signal denoising method based on the stochastic resonance denosing model of the over damped fractional oscillator driven by the double valued noise is studied.The experimental results show that the proposed algorithm can get better SNR than the integer order differential system on the weak detection signal of the millimeter wave linear frequency modulation.The de-noising method of asymmetric monostable system and nonlinear bistable system with multiplicative noise and additive noise is studied.And the aforementioned methods are applied to denoise the output signal of active millimeter wave detecting system.The effect of multiplicative and additive noise is analyzed and the experimental results show that the signal to noise ratio of active millimeter wave detecting system can be improved.The denoisng method of passive millimeter wave detecting system based on stochastic resonance model with improved Langevin equation.To overcome of the problem of adaptive stochastic resonance method which only performs the optimization of signle parameter,an optimized particle swarm optimization based adaptive parameter adoption of stochastic resonance is proposed.To achieve the optimal signal to noise ratio of the bistable stochastic resonance which is considered as the fitness function of particle swarm optimization algorithm,taking advantage of the global search optimization ability,the multiple paramters of stochastic resonance are optimized in parallel and the optimal signal to noise ratio of stochastic resonance is achived.Experiments show that structure parameters,noise intensity,degree of stochastic resonance system can be adaptive adopted.8mm LFM sytem is designed.The proposed denoising methods are used to process real echoes of designed system and the performance is compared to classical linear denoising method.The indoor calibration and complex background based 8mm millimeter wave radiometer systems are designed.Improved Langevin equation based stochastic resonance is performed to denoise signal of indoor point source and tank model.The effectiveness of the proposed algorithm is verified.