The Research On RLG Adaptive Filtering Algorithm And The Realization Based On FPGA

How to reduce the group delay of RLG filter at the premise of high accuracy is one ofthe most important works that many relative universities and the research departments areundertaking. The reduced of group delay can directly improve the response speed of RLG, andimprove the navigation accuracy at the high dynamic circumstances. The frequently-usedfiltering ways high order FIR、IIR and the IIR wave trap all have their personal defects, can notafford the high accuracy, low group delay and the liner response in the through band. Due to thevariance of the dither frequency，the available filtering system hardly have the universal properties.The adaptive filter based on LMS algorithm have advantage of zero group delay，and canadaptive deprive some frequency we want，so the LMS adaptive filter is the ideal scheme torealize the RLG dither signal removal. In the article, we analysis the composition of RLG outputsignal firstly, then determined the stop band and the pass band of the filter. After it, according tothe relationship of the reference signal and the input signal to the filter, we decided to choose thefeedback signal as the LMS filter input signal, at the same time we established a model inMATLAB. After the simulation，we get the value of parameters of the LMS filter, and through theanalysis on the results, we make some optimizes on the basic LMS model.The realize ways of the LMS filter is also a very important part in the article，we make mostof the effort to accomplish it. Firstly, we introduced the circuit design briefly, and then emphaticallydescribed the specific design process of realizing the adaptive filter based on FPGA, and then wegives the block diagram, flow chart and ISE simulation results. In order to improve the accuracyand reduce the dither of the filtering system’ outputs, we add a20order FIR filter after it. Throughthis way, we make good performance on realizing the high accuracy and low time delay to filteringthe RLG output signal.At last, we make some testing on the filtering model combine the RLG. After the static andthe dynamic test on it, we get the conclusion that the adaptive filtering system has very high filtering accuracy and wide dynamic response range. Due to the different RLG we used betweenthe static and dynamic test, it proved that the system has very good universal properties.