| This paper is derived from a underwater target positioning system,aiming at detect weak single frequency periodic pulse signal.Starting from the traditional single-frequency signal detection method,this paper studies the technology of TBD based on Hough transform,analyses the performance and characteristics of the algorithm,and puts forward improvement measures.A weak single frequency periodic pulse signal detection algorithm is designed and implemented on the platform of TMS320C6678.Its performance and time-consuming are analyzed.Firstly,the traditional single-frequency signal detection algorithms such as matched filtering,narrowband filtering and short-time Fourier transform are analyzed,and the principle and theoretical processing gain of TBD are studied.Considering the characteristics of TBD technology,considering both time resolution and frequency resolution,this paper chooses the pretreatment method of short-time Fourier transform and adjusts the parameters of the algorithm to obtain higher processing gain.Then,aiming at the preprocessing results,this paper deeply studies the TBD algorithm based on Hough transform.Firstly,the principle and performance of Standard Hough Transform(SHT-TBD)and Random Hough Transform(RHT-TBD)are studied.Then a Fast Ergodicity Hough transform(FEHT-TBD)algorithm is designed to solve the problem of poor detector performance caused by random sampling in RHT-TBD algorithm.To overcome the quantization defect of SHT-TBD algorithm,a new algorithm based on NLHT-TBD is proposed,and the voting mean and variance are defined according to the voting distribution in parameter space.Advanced Form(INLHT-TBD).The simulation results show that the NLHT-TBD algorithm can achieve the detection performance of FEHT-TBD algorithm while the computational complexity is much less than FEHT-TBD algorithm.As an improved form of NLHT-TBD,INLHT-TBD has better detection performance than NLHT-TBD.After completing the theoretical analysis,this paper designs and implements a weak single frequency periodic pulse signal detection algorithm based on the platform of TMS320C6678.It includes algorithm implementation,network interface data transmission,multi-core and multi-task parallel processing and so on.After that,the network interface module,preprocessing module and time accumulation module are tested,and the algorithm time,the use of DSP memory and the load of each core are analyzed.Finally,the two groups of measured data are compared and analyzed,and the processing results are consistent with the theoretical research,which verifies the practical value of the scheme. |