The Detection On Bistatic Sea Clutter Based On Intelligent Learning

Bistatic radar is in the hot area of research in the modern radar technology.Despite this, current understanding of the properties of bistatic sea clutter is limitedat best. Different from monostatic sea clutter spectrum, bistatic sea clutterspectrum has more complex characteristics of spectral with biststic geometry.Meanwhile, bistatic sea clutter plays a very important role whether in targetdetection or measurement of sea surface characteristic parameter. Whether largely,faster and efficiently extraction of sea clutter characteristics or not influence on thefollowing study of radar technology.The paper is about bistatic sea clutter detection based on the intelligentlearning. First, we build a geometric model of bistatic radar so that we can easilyand theoretically analysis of the causes of bistatic sea clutter. Analyzed andsimulated sea clutter Doppler shift by the factors of bistatic angle, range sum binand ocean surface information. Discuss the inherent spectral characteristics bistaticsea clutter.In part two. Based on the spectral characteristics of bistatic sea clutter and realdata we can detect Doppler shift on rang bin of first-order sea clutter effectively.First of all, we design an artificial sea clutter detection system so that easily pickthe real Bragg peak on each batch and beam. In order to provide perform ancestandards for Self-Supervised network for post training and testing, the paper builda feature database on Bragg peak position returned by artificial detection. Then wegive a detailed study of the detection performance of local peak method, themaximum gradient method and the wavelet transform modulus maxima method in astable and non-stationary environment.Finally, we introduce intelligent learning based on database and the overallsystem design. At begin the paper introduce adaptive resonance theor y neuralnetwork and its main algorithm flow. Using a self-supervised ARTMAP artificialneural network try to train and learn sea clutter database. We analyze the detectioneffect of first-order sea clutter under various environments such as strong/weakcurrents mutation, strong/weak signal to noise ratio, with/without interferenceionosphere. Given sea clutter detection accuracy under various environment situation, and examine the robustness of the network. Complete the entire systemsoftware design under Matlab software GUI environment.