scholarly journals A Novel Parameter Estimation Method Based on LSU-EKF for Polynomial Phase Signal

2017 ◽  
Author(s):  
Yi-xiong Zhang ◽  
Hua-wei Xu ◽  
Rong-rong Xu ◽  
Zhen-miao Deng ◽  
Cheng-Fu Yang
Keyword(s):  
Monte Carlo ◽  
Parameter Estimation ◽  
Kalman Filter ◽  
Estimation Method ◽  
Computation Complexity ◽  
Parameter Estimation Problem ◽  
Performance Loss ◽  
Polynomial Phase ◽  
Polynomial Phase Signals ◽  
Phase Signal

The parameter estimation problem for polynomial phase signals (PPSs) arises in a number of fields, including radar, sonar, biology, etc. In this paper, a fast algorithm of parameter estimation for monocomponent PPS is considered. We propose the so-called LSU-EKF estimator, which combines the least squares unwrapping (LSU) estimator and the extended Kalman filter (EKF). First, the coarse estimates of the parameters of PPS are obtained by the LSU estimator using a small number of samples. Subsequently, these coarse estimates are used to initial the EKF. Monte-Carlo simulations show that the computation complexity of the LSU-EKF estimator is much less than that of the LSU estimator, with little performance loss. Similar to the LSU estimator, the proposed algorithm is able to work over the entire identifiable region. Moreover, in the EKF stage, the accurate estimated results can be output point-by-point, which is useful in real applications.

Parameter Estimation of Polynomial Phase Signal by Unscented Kalman Filtering

2014 ◽  
Vol 644-650 ◽  
pp. 4253-4256
Author(s):  
Wan Ge Li ◽  
Jin Feng Hu ◽  
Hui Ai ◽  
Zhi Rong Lin ◽  
Ya Xuan Zhang
Keyword(s):  
Parameter Estimation ◽  
Kalman Filtering ◽  
State Space Model ◽  
Radar Signal ◽  
Signal Parameters ◽  
Polynomial Phase ◽  
Polynomial Phase Signals ◽  
Phase Signal ◽  
Stable Estimation ◽  
Core Issues

The parameter estimation of the Polynomial Phase Signals (PPS) is one of the core issues. In this paper, UKF-based algorithm is proposed to estimate the parameter of PPS embedded in Gaussian noise. The algorithm constructs an adequate state-space model to represent the PPS and the model can also be implied in real radar signal. Unscented Kalman filtering is applied to estimate the signal parameters. The method achieves the lower SNR threshold, the faster convergence speed, the higher accuracy and more stable estimation performance compared with the existing methods. Simulation also verifies the efficiency of the proposed method.

scholarly journals Simultaneous estimation of land surface scheme states and parameters using the ensemble Kalman filter: identical twin experiments

2011 ◽  
Vol 15 (8) ◽  
pp. 2437-2457 ◽  
Author(s):  
S. Nie ◽  
J. Zhu ◽  
Y. Luo
Keyword(s):  
Parameter Estimation ◽  
Soil Moisture ◽  
Kalman Filter ◽  
Ensemble Kalman Filter ◽  
Land Surface ◽  
Estimation Method ◽  
State Parameter ◽  
Hydraulic Parameters ◽  
Constrained Estimation ◽  
Estimated Parameters

Abstract. The performance of the ensemble Kalman filter (EnKF) in soil moisture assimilation applications is investigated in the context of simultaneous state-parameter estimation in the presence of uncertainties from model parameters, soil moisture initial condition and atmospheric forcing. A physically based land surface model is used for this purpose. Using a series of identical twin experiments in two kinds of initial parameter distribution (IPD) scenarios, the narrow IPD (NIPD) scenario and the wide IPD (WIPD) scenario, model-generated near surface soil moisture observations are assimilated to estimate soil moisture state and three hydraulic parameters (the saturated hydraulic conductivity, the saturated soil moisture suction and a soil texture empirical parameter) in the model. The estimation of single imperfect parameter is successful with the ensemble mean value of all three estimated parameters converging to their true values respectively in both NIPD and WIPD scenarios. Increasing the number of imperfect parameters leads to a decline in the estimation performance. A wide initial distribution of estimated parameters can produce improved simultaneous multi-parameter estimation performances compared to that of the NIPD scenario. However, when the number of estimated parameters increased to three, not all parameters were estimated successfully for both NIPD and WIPD scenarios. By introducing constraints between estimated hydraulic parameters, the performance of the constrained three-parameter estimation was successful, even if temporally sparse observations were available for assimilation. The constrained estimation method can reduce RMSE much more in soil moisture forecasting compared to the non-constrained estimation method and traditional non-parameter-estimation assimilation method. The benefit of this method in estimating all imperfect parameters simultaneously can be fully demonstrated when the corresponding non-constrained estimation method displays a relatively poor parameter estimation performance. Because all these constraints between parameters were obtained in a statistical sense, this constrained state-parameter estimation scheme is likely suitable for other land surface models even with more imperfect parameters estimated in soil moisture assimilation applications.

Sign in / Sign up

Export Citation Format

Share Document