FPGA (Field Programmable Gata Array)-based unscented kalman filter system and parallel implementation method

Abstract

The invention discloses an FPGA (Field Programmable Gata Array)-based unscented kalman filter system mainly solving the problem that the traditional unscented kalman filter hardware has great implementation difficulty and poor instantaneity and comprising a covariance matrix Cholesky decomposition model A, a Sigma point generation module B, a time updating module C, an observation and prediction module D, a part-mean value and covariance matrix computation module E, a population mean value computation module F, a population covariance matrix computation module G, an observation and predictioncovariance matrix inversion module H, a gain computation module I and a state quantity and state covariance matrix estimating module J, wherein the module A generates K group of column vector to the module B and the B, C, D and E modules are connected in series and respectively comprise K submodules adopting a parallel arithmetic modular construction; the F and G modules receive and process the Kgroup of results of the module E and the processed results pass through the modules H, I and J in sequence to obtain the present result. The invention has the advantages of quick filter speed and easy hardware implementation and can be used for target tracking and parameter estimation.

Description

technical field [0001] The invention belongs to the technical field of signal processing, relates to a nonlinear filtering method, and can be used for target tracking and parameter estimation. Background technique [0002] Filtering theory is a theory and method for estimating the state of the system based on the measurement of the observable signals of the system and according to certain filtering criteria. According to Bayesian theory, the optimal estimate is only an idealized solution, and its analytical form cannot be obtained in general. Kalman filter is currently the most classic and widely used linear filter. When the system is a linear Gaussian distribution, it can obtain the minimum mean square error solution of recursive Bayesian estimation. However, the actual system is often nonlinear and non-Gaussian, which cannot be solved by Kalman filter. Therefore, a large number of nonlinear filtering methods have been proposed, which can be mainly divided into two categor...

AI Technical Summary

Problems solved by technology

[0006] Although the real-time performance of UKF has been significantly improved compared with the particle filter, complex operations such as matrix inversion and Cholesky decomposition are still involved in UKF. When the system state dimension and observation dimension are large, the operation speed is obvious. Reduced, hardware implementation is much more difficult than EKF

Images