Adaptive Iterated Extended KALMAN Filter for Relative Spacecraft Attitude and Position Estimation

A FPGA implementation for a model-based state of charge (SOC) estimation is described in this paper. A Thevenin equivalent circuit model is designed for SOC estimation. The extended Kalman filter (EKF) is designed to complete the SOC estimation, and the error is within 1 % . The FPGA is chosen to achieve realtime SOC estimation. A fast matrix method is proposed to improve the calculation speed of the EKF in FPGA because the EKF algorithm requires many matrix operations. In addition, the embedded system based on the FPGA with a system on a programmable chip (SOPC) technique is built using the Qsys platform in Quartus II. Based on the embedded system, an online testing platform is established to monitor the terminal voltage and load current of the experimental battery in real time; experimental results show that the online SOC estimation is successful. The measurement results show that the FPGA embedded scheme of the EKF allows for successful implementation of the SOC estimation with accuracy and speed. The fast matrix method requires 0.00007 s to implement the SOC estimation and is four times faster than the conventional matrix method.

Section: Introductionmentioning

confidence: 99%

FPGA implementation of extended Kalman filter for SOC estimation of lithium‐ion battery in electric vehicle

Duan

et al. 2019

“…However, localization based on the EKF has some limitationx. First, it is prone to divergence as it is based on linearization of nonlinear dynamics and measurement functions . Second, it is only suitable for mobile localization with Gaussian noise models.…”

Section: Introductionmentioning

confidence: 99%

Mobile robot localization based on PSO estimator

Havangi

2019

Localization is fundamental to autonomous operation of the mobile robot. A particle filter (PF) is widely used in mobile robot localization. However, the robot localization based PF has several limitations, such as sample impoverishment and a degeneracy problem, which reduce significantly its performance.Evolutionary algorithms, and more specifically their optimization capabilities, can be used in order to overcome PF based on localization weaknesses. In this paper, mobile robot localization based on a particle swarm optimization (PSO) estimator is proposed. In the proposed method, the robot localization converts dynamic optimization to find the best robot pose estimate, recursively. Unlike the localization based on PF, the resampling step is not required in the proposed method. Moreover, it does not require noise distribution. It searches stochastically along the state space for the best robot pose estimate. The results show that the proposed method is effective in terms of accuracy, consistency, and computational cost compared with localization based on PF and EKF.

“…However, they are not quite effective in real time, since EKF performance degrades due to imperfect initialization of system states . In a recent work , an adaptive iterated EKF has been designed and applied successfully on nonlinear navigation system model for spacecraft.…”

Section: Introductionmentioning

confidence: 99%

Tuning of evolutionary particle filtering approach for estimation of trajectory deviation

Mukherjee

Chattaraj

Chakraborty³

2019

An autonomous vehicle launched from moving platform is subjected to flexure-like disturbance and needs to stabilize itself before embarking on the target tracking phase. Control and guidance remains inactive during stabilization phase. Hence, estimating the deviation from the intended trajectory is important in the absence of control and guidance commands for trajectory correction. This estimation problem is here mapped to the misalignment estimation problem, with the intended trajectory data resembling mother data and the followed trajectory data resembling daughter data. The main requirement is fast convergence for early commencement of target tracking phase. Kalman filter and its variants fail to converge due to the inability to model time varying nonlinearities and flexure in the system model. Conventional particle filter converges, but computation time is high. This work explores evolutionary particle filter variants that may provide faster convergence through adaptive tuning of the range from where particle support points are selected. Such heuristics is based on the observed residual trend. Monte Carlo simulation results underline the importance of designing estimator based on adaptive tuned evolutionary particle filter algorithm.