Warship Localization Algorithm Based On Magnetic Gradient Tensor And It's Implementation In FPGA

In this paper, the magnetic field of a warship on the surface of water was simulated with a magnetic dipole model based on magnets to analyze the applicable conditions. Besides, the localization based on magnetic gradient tensor at a single point and a magnetic dipole model was examined. Meanwhile, the impacts of different conditions upon localization error were simulated through Matlab. The results suggested that a warship could be passively localized at a single point by the magnetic gradient tensor and the localization precision was mainly influenced by following factors, such as magnetic field of a warship, distance from measuring point to target, baseline length of a measuring system and sensor precision.To measure and apply the magnetic localization algorithm, an algorithm processing platform was designed. Furthermore, a high performance platform was proposed for processing parallel algorithms, which was composed of three FPGAs, 4 DSPs and high-speed communication interfaces, so as to analyze the demands for the algorithm while meeting subsequent needs for functions. Being advantageous in high bandwidth, high-rate data processing, high-speed communication and parallel processing, this algorithm processing platform may be used for realizing most complicated data processing algorithms.Based on the algorithm processing platform, the localization algorithm based on magnetic gradient tensor was realized in FPGA with hardware description language and the operational efficiency was analyzed. After comparing the number of clocks used for implementing this algorithm by FPGA with that utilized for realizing DSP, FPGA was suggested to be highly advantageous in speed in realizing the algorithm, of which the speed was approximately 50 times as high as that of the DSP from the perspective of number of clocks.The hardware system of the algorithm platform was debugged and verified. The performance and the functions of FPGA board were verified based on the test for electric properties of hardware and software functions. The results suggested that the hardware system achieved expected design objective and could work properly in stable manners.