Robust adaptive unscented Kalman filter for attitude estimation of pico satellites (original) (raw)
Related papers
The Kalman filter requires knowledge of the noise statistics; however, the noise covariances are generally unknown. Although this problem has a long history, reliable algorithms for their estimation are scant, and necessary and sufficient conditions for identifiability of the covariances are in dispute. We address both of these issues in this paper. We first present the necessary and sufficient condition for unknown noise covariance estimation; these conditions are related to the rank of a matrix involving the auto and cross-covariances of a weighted sum of innovations, where the weights are the coefficients of the the minimal polynomial of the closed-loop system transition matrix of a stable, but not necessarily optimal, Kalman filter. We present an optimization criterion and a novel six-step approach based on a successive approximation, coupled with a gradient algorithm with adaptive step sizes, to estimate the steady-state Kalman filter gain, the unknown noise covariance matrices...
In this paper is study the role of Measurement noise covariance R and Process noise covariance Q. As both the parameter in the Kalman filter is a important parameter to decide the estimation closeness to the True value , Speed and Bandwidth [1]. First of the most important work in integration is to consider the realistic dynamic model covariance matrix Q and measurement noise covariance matrix R for work in the Kalman filter. The performance of the methods to estimate and calculate both of these matrices depends entirely on the minimization of dynamic and measurement update errors that lead the filter to converge[2]. This paper evaluates the performances of Kalman filter method with different adaptations in Q and in R. This paper perform the estimation for different value of Q and R and make a study that how Q and R affects the estimation and how much it differ to true value to the estimated value by plot in graph for different value of Q and R as well as we give a that give the brief idea of error in estimation and true value by the help of the MATLAB.
ISA Transactions, 2010
In the normal operation conditions of a pico satellite, a conventional Unscented Kalman Filter (UKF) gives sufficiently good estimation results. However, if the measurements are not reliable because of any kind of malfunction in the estimation system, UKF gives inaccurate results and diverges by time. This study introduces Robust Unscented Kalman Filter (RUKF) algorithms with the filter gain correction for the case of measurement malfunctions. By the use of defined variables named as measurement noise scale factor, the faulty measurements are taken into consideration with a small weight, and the estimations are corrected without affecting the characteristics of the accurate ones. Two different RUKF algorithms, one with single scale factor and one with multiple scale factors, are proposed and applied for the attitude estimation process of a pico satellite. The results of these algorithms are compared for different types of measurement faults in different estimation scenarios and recommendations about their applications are given.
Adaptive Kalman Filter based on Posteriori Variance-Covariance Component Estimation
There are different ways to construct adaptive Kalman filtering (AKF) algorithms. This paper proposes an innovative way to simultaneously estimate the variance matrix R of the measurement vector and the variance matrix Q of the process noise vector based on the variance-covariance component estimation by taking the advantages of the measurement residuals and the process noise residuals (Wang, 1997, 2009; Wang et al, 2009) and the measurement redundancy contribution (Ou, 1989). The core of the novel AKF algorithm lies in the projection of the system innovation vector into the three groups of residuals: the residuals of the measurement vector, the residuals of the process noise vector and the residuals of the predicted state vector exclusive of the effect of the process noise. The simulated and real GPS data in kinematic relative positioning mode were used to demonstrate the performance of the proposed adaptive Kalman filter. The results from the simulated datasets confirm to the simulated variance-covariance components well. The results from real kinematic GPS datasets are also provided and discussed.
Measurement Noise Covariance-Adapting Kalman Filters for Varying Sensor Noise Situations
Sensors
An accurate and reliable positioning system (PS) is a significant topic of research due to its broad range of aerospace applications, such as the localization of autonomous agents in GPS-denied and indoor environments. The PS discussed in this work uses ultra-wide band (UWB) sensors to provide distance measurements. UWB sensors are based on radio frequency technology and offer low power consumption, wide bandwidth, and precise ranging in the presence of nominal environmental noise. However, in practical situations, UWB sensors experience varying measurement noise due to unexpected obstacles in the environment. The localization accuracy is highly dependent on the filtering of such noise, and the extended Kalman filter (EKF) is one of the widely used techniques. In varying noise situations, where the obstacles generate larger measurement noise than nominal levels, EKF cannot offer precise results. Therefore, this work proposes two approaches based on EKF: sequential adaptive EKF and p...
Adaptive Kalman Filtering Based on Posteriori Variance-Covariance Components Estimation
This paper proposes an innovative way to simultaneously estimate the variance matrix R of the measurement vector and the variance matrix Q of the process noise vector based on the variance-covariance component estimation by taking the advantages of the measurement residuals and the process noise residuals (Wang, 1997, 2009; Wang et al, 2009) and the measurement redundancy contribution (Ou, 1989). The core of the novel AKF algorithm lies in the projection of the system innovation vector into the three groups of residuals: the residuals of the measurement vector, the residuals of the process noise vector and the residuals of the predicted state vector exclusive of the effect of the process noise. The simulated and real GPS data in kinematic relative positioning mode were used to demonstrate the performance of the proposed adaptive Kalman filter. The results from the simulated datasets confirm to the simulated components well. The results from real kinematic GPS datasets are also provi...
Online Sensor Noise Covariance Identification Using a Modified Adaptive Filter
Volume 2: Mechatronics; Estimation and Identification; Uncertain Systems and Robustness; Path Planning and Motion Control; Tracking Control Systems; Multi-Agent and Networked Systems; Manufacturing; Intelligent Transportation and Vehicles; Sensors and Actuators; Diagnostics and Detection; Unmanne..., 2017
Modern control systems heavily relay on sensors for closed-loop feedback control. Degradation of sensor performance due to sensor aging affects the closed-loop system performance, reliability, and stability. Sensor aging characterized by the sensor measurement noise covariance. This paper proposes an algorithm used to identify the slow varying sensor noise covariance online based on system sensor measurements. The covariance-matching technique, along with the adaptive Kalman filter is utilized based on the information about the quality of weighted innovation sequence to estimate the slow time-varying sensor noise covariance. The sequential manner of the proposed algorithm leads to significant reduction of the computational load. The covariance-matching of the weighted innovation sequence improves the prediction accuracy and reduces the computational load, which makes it suitable for online applications. Simulation results show that the proposed algorithm is capable of estimating the...
Adaptive Unscented Kalman Filter at the Presence of Non-additive Measurement Noise
This paper proposes an Adaptive Unscented Kalman Filter (AUKF) for nonlinear systems having non-additive measurement noise with unknown noise statistics. The proposed filter algorithm is able to estimate the nonlinear states along with the unknown measurement noise covariance (R) online with guaranteed positive definiteness. By this formulation of adaptive sigma point filter for non-additive measurement noise, the need of approximating non-additive noise as additive one (as is done in many cases) may be waived. The effectiveness of the proposed algorithm has been demonstrated by simulation studies on a nonlinear two dimensional bearing-only tracking (BOT) problem with non-additive measurement noise. Estimation performance of the proposed filter algorithm has been compared with (i) non adaptive UKF, (ii) an AUKF with additive measurement noise approximation and (iii) an Adaptive Divided Difference Filter (ADDF) applicable for non-additive noise. It has been found from 10000 Monte Carlo runs that the proposed AUKF algorithm provides (i) enhanced estimation performance in terms of RMS errors (RMSE) and convergence speed, (ii) almost 3-7 times less failure rate when prior measurement noise covariance is not accurate and (iii) relatively more robust performance with respect to the initial choice of R when compared with the other nonlinear filters involved herein.
Estimation of unknown noise covariances in a Kalman filter is a problem of significant practical interest in a wide array of applications. This paper presents a single-pass stochastic gradient descent (SGD) algorithm for noise covariance estimation for use in adaptive Kalman filters applied to non-stationary systems where the noise covariances can occasionally jump up or down by an unknown magnitude. Unlike our previous batch method or our multi-pass decision-directed algorithm, the proposed streaming algorithm reads measurement data exactly once and has similar root mean square error (RMSE). The computational efficiency of the new algorithm stems from its one-pass nature, recursive fading memory estimation of the sample cross-correlations of the innovations, and the RMSprop accelerated SGD algorithm. The comparative evaluation of the proposed method on a number of test cases demonstrates its computational efficiency and accuracy.