Research On Adaptive Signal Processing Algorithm Of Array Antenna Soccer Robot

As a typical application of artificial intelligence,vision and sensing,communication,control science and other interdisciplines,large-field soccer robot matches need to simultaneously monitor the football trajectory in the air and on the ground to solve the problems of traditional robots that are not strong enough to adapt to the environment or are greatly affected by light.In addition,it is necessary to satisfy both up-looking and down-looking working conditions due to different antenna attitudes to face complex electromagnetic interference and clutter environment.The array antenna has the unique advantages of flexible beam control,and can meet the requirements of multi-target tracking and recognition,strong adaptive anti-interference ability and high reliability.This paper mainly studies the research on adaptive signal processing algorithm of soccer robot based on array antenna.Adaptive beamforming(ADBF)technology can enhance the target signal and suppress the interference signal through adaptive control of weights in the case of deterioration of electromagnetic working environment and the existence of large amount of radio frequency interference,so as to achieve spatial filtering.It's of great theoretical significance and engineering value to study and improve the real-time capability and robustness of adaptive processing algorithm.Focusing on the anti-interference performance and clutter suppression in the practical application of the antenna array,this paper studies the signal model of soccer robot array antenna,fast ADBF and STAP,eigenanalysis interference cancellation beamforming(EC)algorithm based on source estimation of sensitive function,and Fast Fourier Transform(FFT)algorithm with super large number of points using SMP(symmetric multiprocessor)system.The specific research contents are summarized as follows.Firstly,the signal model and adaptive processing principle of soccer robot array antenna are analyzed in detail from the mathematical model.Three kinds of criterion are studied under certain condi-tions..It is found that these different optimization criteria are equivalent and provide theoretical conditions for the mutual transformation of different adaptive structures and ADBF methods.The clutter distribution characteristics of soccer robot array antenna,the principle of STAP and the evaluation system of clutter suppression performance are proposed.The motion of antenna platform makes the clutter echo signal have serious space-time coupling characteristics,the sampling matrix inversion algorithm can solve it.At the same time,the performance evaluation system of clutter suppression is studied.The system improvement factor and output SINR measure the clutter suppression performance.Secondly,the fast ADBF and STAP are studied.In this paper,the RGS algorithm which carries out GS orthogonalization on sampling covariance matrix are deeply studied and discussed.It is found that RGS algorithm can make full use of the sampling number data information,effectively suppress the interference when the sampling number is small,achieve the desired performance output.In addition,the application of STAP technology in the actual system has the phenomenon of clutter subspace leakage,this study focuses on the robust partial adaptive STAP algorithm based on two-stage dimensionality reduction.This algorithm can effectively improve the performance of STAP algorithm in the case of clutter subspace leakage.Thirdly,this paper also studies the EC algorithm based on the source estimation of the sensitive function.The traditional ADBF algorithm has a large amount of computation and requires high hardware equipment,and has poor robustness to system errors.In this paper,a robust eigenanalysis interference cancellation beamforming(REC)algorithm using sensitive function source estimation is proposed to solve the problem that the performance of ADBF is seriously degraded when the traditional EC algorithm can not prepare to estimate the number of sources and the training data contain the desired signals.The algorithm can effectively estimate the number of sources in the signal interference subspace by using array sensitivity function,and can judge and eliminate the influence of the target signal in the main lobe on the calculation of adaptive weight vector by means of spatial spectrum search.Therefore,whether the target signal is included in the training sample or not,the anti-interference performance of the algorithm is well reflected.Fourthly,the FFT algorithm with super large number of points using SMP system is studied.FFT calculation is faced with two major problems of limited processor resources and reduced execution efficiency.In the STAP system,in order to maximize the efficiency of FFT algorithm,the implementation process of FFT should be combined with the characteristics of the processor,and written for the specific platform.In this paper,by analyzing the architecture characteristics of SMP parallel processing system,a FFT algorithm with super large number of points for SMP is proposed.Three problems caused by the improved FFT algorithm of super large number of points are solved.Based on the antenna signal model of soccer robot array,the interference and clutter signal model,the optimal beamforming criterion and the evaluation system of clutter suppression performance are proposed.A EC algorithm based on the source estimation of sensitive function,which can achieve good an-ti-interference performance no matter whether there is target signal in the training sample or not.This paper also proposes a fast ADBF algorithm and STAP technology to estimate the interference subspace more ac-curately and improve the performance under clutter subspace leakage,and proposes a FFT algorithm of super larger number of points based on SMP,which can save nearly half storage resources,improve the processing speed of FFT of super larger number of points,and improve the real-time performance of signal processing.Therefore,the soccer robot array antenna system can greatly improve the ability of target detection and field monitoring,and meet the requirements of multi-target tracking and recognition,strong adaptive anti-interference ability and high reliability.Therefore,the research and application of related technology will be an important part of the current field of artificial intelligence.