Adaptive extended Kalman filter and point ahead angle prediction in the detection of gravitational waves in space

Haojie Li; Jianjun Jia; Jinke Yang; Liang Zhang; Wenlin Tang; Xindong Liang; Xue Wang; Yong Xie; Yun Kau Lau; Zhao Cui

REVIEW 3 major objections 5 minor 17 references

Adaptive extended Kalman filter and point ahead angle prediction in the detection of gravitational waves in space

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read An adaptive extended Kalman filter that predicts the point-ahead angle from the constellation's orbital motion can shrink the PAAM's dynamic range from micro-radians to tens of nano-radians in simulation.

desk verdict A promising idea—AEKF for point-ahead angle prediction—but the simulation's central claim is unsupported because the filter is fed the true noise realization, and the sampling rate contradicts the frequency band it claims to filter. read the letter →

arxiv 2411.17146 v1 pith:BTTFAX2U submitted 2024-11-26 astro-ph.IM

Jinke Yang , Yong Xie , Wenlin Tang , Xindong Liang , Liang Zhang , Zhao Cui , Xue Wang , Haojie Li

show 2 more authors

Jianjun Jia Yun Kau Lau

This is my paper · ORCID

classification astro-ph.IM PACS 04.80.Nn95.55.Ym

keywords adaptiveextendedKalmanfilterpointaheadanglemechanismpredictionspacegravitationalwavedetectioncoloredmeasurementnoiselaserpointingcontrolorbitdeterminationTTLcoupling

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper investigates whether an adaptive extended Kalman filter (AEKF) can be placed inside the point-ahead angle mechanism (PAAM) control loop of a space-based gravitational-wave detector. The idea is to predict the point-ahead angle from the near-circular orbital dynamics of the triangular spacecraft constellation rather than let the actuator chase the full annual pointing motion. In simulation with a colored measurement-noise model built from a LISA-type noise spectrum, the AEKF reduces the in-plane PAAM dynamic range from about 63 microradians to roughly 40 nanoradians, the out-of-plane range to about 6.9 microradians, and keeps prediction errors below 1 nanoradian. If these results hold on orbit, the PAAM needs much less stroke, which lowers tilt-to-length coupling and position noise and leaves room for the fine-pointing stage to be more accurate.

What carries the argument

The core mechanism is the adaptive extended Kalman filter with a colored measurement-noise model in which the measurement noise is appended to the state. The state is the 18-dimensional vector of positions and velocities of the three spacecraft; the dynamics are the Keplerian equations of motion under solar-system gravity; and the observation map $Z_k = [\mathrm{PAA}_{\mathrm{out},12}, \mathrm{PAA}_{\mathrm{in},12}]$ is linearized into the observation matrix $H_k$. The adaptive part updates the measurement covariance $R_k$ in real time from the noise and sets the colored-noise transition $\Psi_{k,k-1}=v_k/v_{k-1}$; the noise time series with the LISA-type power spectral density are generated by an FFT-based Monte Carlo method. This lets the filter track the slow annual pointing motion instead of letting the actuator chase it.

What would settle it

Repeat the simulation with $\Psi_{k,k-1}$ and $R_k$ estimated online from the filter's innovations, or from an independent noise draw, and check whether the maximum prediction errors stay below 1 nrad and the in-plane dynamic range stays near 40 nrad; a large degradation would show the reported numbers depend on knowing the noise rather than on the filter.

Watch

Extended reading notes

Core claim

The central claim is that the PAAM control problem can be reframed as a state-estimation problem: with the 18-dimensional positions and velocities of the three spacecraft propagated by Keplerian solar-system gravity, and with the in-plane and out-of-plane point-ahead angles as nonlinear measurements, an adaptive extended Kalman filter predicts the point-ahead angle one step ahead well enough that the PAAM only corrects the residual. The reported simulation numbers are a reduction of the in-plane dynamic range from about 63 microradians to roughly 40 nanoradians, the out-of-plane range to about 6.9 microradians, and maximum prediction errors below 1 nanoradian on all six links. Because the point-ahead angle varies on an annual timescale, one filter update per day suffices.

Load-bearing premise

The load-bearing premise is that the filter can be given the true measurement-noise realization at every step, because Eq. (39) sets $\Psi_{k,k-1}=v_k/v_{k-1}$ and Eq. (41) sets $R_k=\mathrm{cov}[v_k,v_k]$ from the injected noise; on a real mission the noise is unknown and must be estimated from the measurements themselves.

Editorial extensions

If this is right

The PAAM stroke shrinks from tens of microradians to about 40 nanoradians in-plane, reducing tilt-to-length coupling and actuator position noise.
The fine-pointing stage inherits a much smaller residual range, so its accuracy can be improved without demanding more actuator travel.
The position-noise budget for the laser pointing chain as a whole can be reduced, directly easing one of the gravitational-wave sensitivity limits.
The filter needs only one update per day with simple orbital dynamics, so it is feasible to run on-board with modest computational resources.
The same predicted point-ahead angle can also shorten acquisition time in the acquisition-tracking-pointing phase of the mission.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The paper's use of the true injected noise for $\Psi_{k,k-1}$ and $R_k$ makes the simulation an optimistic bound; a practical filter would have to estimate both from innovations, and the drop in performance would measure the method's true margin.
If the approach transfers beyond the heliocentric case as the paper suggests, the same feedforward estimator would need a different force model; testing that transfer would be a direct next step.
Mapping the dynamic-range reduction as a function of orbit-determination error would show whether the assumed 20 km and 2 cm/s accuracies are a limiting requirement or just a conservative input.
The approximation that the constellation-plane normal is constant contributes about 0.45 nanoradians of error, suggesting that pushing below the nanoradian level will require coupling the filter to an attitude estimator.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The core idea here is worth a look: using an extended Kalman filter with colored measurement noise to predict the point-ahead angle (PAA) for LISA-like triangular constellations, potentially shrinking the PAAM dynamic range and easing the pointing control budget. That is a legitimate extension of known filtering techniques to a specific and relevant problem, and the paper is honest that the AEKF framework itself is standard. The orbit model with Keplerian dynamics and the PAA geometry derivation are plausible, and the authors cite the relevant colored-noise EKF literature.

The problem is that the quantitative claim—nrad-level dynamic range—rests on a simulation that gives the filter the answer. Equation (39) sets Ψk,k−1 = vk/vk−1 using the actual injected measurement noise samples, and Eq. (41) computes Rk = cov[vk,vk] from those same samples. That is not an adaptive filter; it is a clairvoyant filter. No real on-orbit system has access to the noise realization, only to noisy measurements. The paper gives no evidence that innovation-based estimates would preserve performance. This is a load-bearing flaw, not a minor tuning detail.

There are additional problems. The sampling frequency is chosen as one per day, yet the results show PSDs from 1 mHz to 10 Hz; the Nyquist frequency is below 1×10⁻⁵ Hz, so any noise above that is aliased. The abstract claims dynamic range reduced to nano-radians, but Table I lists out-of-plane PAA at 6.9×10⁻⁶ rad—micro-radians, not nano-radians. Equation (42) gives the measurement noise in m/√Hz and calls it displacement noise, then applies it as if it were an angle; the units do not match the PAA measurement. There is no code or data released, no ensemble statistics over multiple noise realizations, and Table III's R² = 1.000000 is too perfect to be a meaningful metric.

Credit where due: the paper clearly flags that Section III is review material, it does not oversell the algorithm as new, and the idea that slow orbital dynamics can be exploited for feedforward control is sensible. But as written, the simulation is not an independent validation of the AEKF's benefit.

The paper would benefit from a serious referee who can push the authors to re-run the simulation with honestly estimated noise parameters, fix the sampling inconsistency, and clarify the units. I would not desk-reject it, but it needs major revision before the central claim can be trusted.

Referee Report

3 major / 5 minor

Summary. The paper proposes inserting an adaptive extended Kalman filter (AEKF) into the point-ahead angle mechanism (PAAM) control loop of a LISA-like triangular constellation. The state vector contains the positions and velocities of the three spacecraft, the measurement equation outputs in-plane and out-of-plane point-ahead angles, and the measurement noise is modeled as colored noise augmented into the state. Using this setup, the authors simulate three years of data and report that the AEKF reduces the in-plane PAA dynamic range from about 63 µrad to about 40 nrad and the out-of-plane dynamic range from about 63 µrad to about 6.9 µrad, with prediction errors below 1 nrad. They conclude that the PAAM stroke and position noise budget could be reduced accordingly.

Significance. If the claimed reductions were validated, the result would be significant for space-based gravitational-wave detector design: a smaller PAAM dynamic range reduces TTL coupling noise, relaxes fine-pointing requirements, and lowers the position noise budget. The paper also usefully lays out the PAA kinematics and a colored-noise EKF architecture for this application, and it presents simulation results for all three arms. However, the central feasibility claim rests on a simulation that gives the filter access to the true injected measurement noise, and there are additional inconsistencies in the sampling rate and noise units. As presented, the quantitative conclusions are not independently validated.

major comments (3)

[§V Eq. (39), §VII Eq. (41)] The simulation gives the AEKF oracle access to the measurement noise. Eq. (39) sets Ψk,k−1 = vk/vk−1 using the actual injected noise samples, and Eq. (41) sets Rk = cov[vk, vk] from those same samples. In a real on-board filter the noise realization is not observable; Ψ and R must be estimated from the innovation sequence or from a pre-calibrated model. The paper provides no evidence that such estimates preserve the reported performance, so the sub-nanoradian prediction errors in Table II and the dynamic-range reductions in Table I are not demonstrated for an operational filter. This directly undermines the abstract's central claim.
[§IV, §V, §VIII] The sampling frequency is stated as one day in Section IV and again in Section V, but Section VIII reports filtering performance in the band 1 mHz–10 Hz. For daily sampling the Nyquist frequency is about 5.8 µHz, so 1 mHz–10 Hz signals are not even representable at that sampling rate. Either the simulation actually uses a much higher sampling rate, or the frequency-domain claims in Figs. 8 and 9 are not supported by the described time step. This inconsistency must be resolved before the PSD-reduction results can be evaluated.
[§VII Eq. (42)] Eq. (42) specifies the measurement noise as a displacement noise PSD with units m/√Hz, while the measurement equation (35) is for PAA in radians. The text does not describe any conversion from displacement noise to angular noise, such as division by the arm length or an optical-lever factor. It is therefore unclear what noise amplitude is actually added to the PAA observables, and the numerical values in rad/√Hz reported in Section VIII cannot be traced back to the stated 10×10−12 m/√Hz input without such a conversion.

minor comments (5)

[§III Eq. (22)] In Eq. (22), Zk* is defined as Zk+1 − Ψk,k−1·Zk+1, but Eq. (23) is derived as if the second term were Ψk,k−1·Zk. Please correct this typo, as the subsequent derivation depends on the intended form.
[§V Eq. (35)] The phrase 'measurement nosie' below Eq. (35) should read 'measurement noise'.
[§VII] The sentence 'position and velocity of SCs are generated by an the orbit integrator' contains a grammatical error and should be revised.
[§VIII] There are several typos in this section: 'AKEF' should be 'AEKF' and 'PDS' should be 'PSD'.
[§VIII.1, Table III] Table III reports R-square and adjusted R-square as exactly 1 for all cases. Reporting a residual plot or more significant figures would help confirm that this is a meaningful fit result rather than an artifact of the fitting procedure.

Circularity Check

1 steps flagged · score 6.0 of 10

AEKF noise parameters are set from the true injected measurement noise (Eqs. 39 and 41), so the reported nrad-level PAA prediction errors and dynamic-range reduction are an oracle-noise result, not a demonstrated on-orbit capability.

fitted input called prediction [Section V, Eq. (39); Section VII, Eq. (41)]
"As shown in Eq. (39), we approximately take the ratio of the measurement noise amplitudes at the previous moment and the current moment as input to the coefficient transfer matrix Ψ k,k−1 of the colored measurement noise. ... Ψk,k−1 = vk/vk−1. ... At the same time, the covariance matrix Rk, which characterizes the measurement noise within the AEKF, is dynamically computed from the generated noise and updated in real-time [17] as Rk = cov[vk, vk]."

The filter's colored-noise transition matrix and measurement-noise covariance are set from the exact noise realization vk that is added to the true PAA to form the measurement. Ψk,k−1 = vk/vk−1 uses the instantaneous ratio of actual noise samples, and Rk = cov[vk, vk] uses the same generated-noise samples. This makes the AEKF 'adaptive' in name only: rather than estimating noise statistics from innovations, it is handed the true noise. The reported sub-nanoradian prediction errors and Table I dynamic-range reductions are therefore an optimal-filtering result for a model that knows the noise, not a demonstration that an on-board filter can achieve them.

full rationale

The main load-bearing numerical claim—that AEKF reduces the PAAM dynamic range to nanoradian levels—depends on the filter being given the true measurement noise realization. Equation (39) sets the colored-noise transition matrix to the ratio of actual noise samples, and Eq. (41) sets the measurement-noise covariance from the generated noise. These are not estimates from observable data; they are the exact noise statistics used to create the measurements. Consequently, the reported prediction errors are an upper bound on what a filter with perfect noise knowledge can do, not evidence that the proposed 'adaptive' loop would work on orbit. The orbital-dynamics part of the paper is self-contained and not circular: the PAA formulas, the Keplerian state model, and the observation matrix are derived from first principles and external constants. No load-bearing self-citation chain was found: reference [4] is by a co-author but supplies only orbit integrator constants and light-time variation values, and the core filtering claim does not reduce to that citation. The circularity is therefore partial but real: the adaptive filter is evaluated under a clairvoyant noise model, so the main feasibility conclusion is not independently demonstrated.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim rests on an assumed orbit model, a simplified noise model, and the availability of true noise statistics inside the simulation. The free parameters are hand-chosen tuning values, not fitted to external data. No new physical entities are introduced.

free parameters (4)

AEKF sampling interval = 1 day
Chosen by hand because the PAA varies slowly; sets the filter bandwidth and the maximum disturbance frequency the loop can reject. No sensitivity study is reported.
Colored noise autoregressive coefficient = time-varying (v_k/v_{k-1})
Equation (39) sets the colored noise transition to the ratio of actual injected noise samples, so the filter is given the ground-truth noise correlation rather than estimating it.
Measurement noise amplitude = 10 x 10^-12 m/sqrt(Hz)
Adopted from a LISA-like displacement PSD (Eq. (42)) but applied to an angular PAA without stating the conversion, so the effective angular noise level is ambiguous.
Orbit determination error bounds = 20 km position, 2 cm/s velocity
Used to set system noise via orbit prediction errors; taken from reference [13] and not varied. The resulting PAA system noise is quoted as about 0.2 nrad.

assumptions (5)

domain assumption Keplerian N-body dynamics with Sun and eight major planets is an adequate model of each SC orbit
Eq. (31)-(32) use only solar system gravitational point masses; non-gravitational forces and higher-order effects are neglected, which is acceptable for a feasibility study but not verified.
domain assumption The constellation plane normal e_z is constant over the mission
Section II.1 approximates e_z as constant based on an annual inclination variation of about 0.08 degrees and quotes a 0.45 nrad error; this is small but not exactly zero.
ad hoc to paper Measurement noise is purely colored with no white component
Section V states only colored noise is considered and the white noise term xi is ignored; the colored noise transition is then chosen as Psi = v_k/v_{k-1}.
ad hoc to paper The filter may use the true measurement noise realizations to set R_k and Psi
Eq. (39) and Eq. (41) compute filter parameters from the injected noise; this is valid for the simulation but not for an on-orbit implementation.
ad hoc to paper Orbit prediction errors are linearly superposed as white system noise with about 0.2 nrad PAA error
Section VI: 'The system noise is linearly superimposed onto the genuine PAA value... in the form of white noise.' This is a modeling simplification.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Adaptive extended Kalman filter and point ahead angle prediction in the detection of gravitational waves in space." pith.science (2026). https://pith.science/paper/BTTFAX2U

@misc{pith2026241117146,
  author       = {Pith},
  title        = {Pith review of: Adaptive extended Kalman filter and point ahead angle prediction in the detection of gravitational waves in space},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BTTFAX2U}},
  note         = {Machine review of arXiv:2411.17146}
}

read the original abstract

In the detection of gravitational waves in space, during the science phase of the mission, the point ahead angle mechanism (PAAM) serves to steer a laser beam to compensate for the angle generated by the relative motion of the two spacecrafts (SCs) during the approximately 10 seconds of flight time a laser beam will take from one SC to reach a distant SC of three million kilometers away. The common practice for pointing stability control of a laser beam is to first do a coarse tracking by the PAAM to steer a laser beam to compensate for the relative motion between two SCs, to be followed by a fine pointing stability control. In the present work, by exploiting the near-circular orbit structure of individual SC in the triangular constellation, the feasibility of inserting an adaptive Kalman filter (AEKF) into the PAAM control loop is investigated. By adopting a colored measurement noise model that closely resembles the prospective on orbit situation, numerical simulation suggests that the dynamic range of the PAAM may be reduced to the level of nano-radians using the prediction of the pointing head angle (PAA) by the AEKF. This will cut down on the TTL coupling noise and the position noise budget allocated to the PAAM. This in turn reduces the dynamic range of the fine pointing control and leaves room to improve its accuracy, thereby offers the prospect of reduction of the position noise budget allocated to the laser pointing instability as a whole.

Figures

Figures reproduced from arXiv: 2411.17146 by the authors.

**Figure 2.** FIG. 2: Dynamic range of PAAs [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** FIG. 3: The angle variation between the constellation plane and the Heliocentric Ecliptic [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (20 more)

**Figure 4.** Figure 4: FIG. 4: PAAM control framework [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

**Figure 5.** Figure 5: FIG. 5: PAA calculation flow chart [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]

**Figure 6.** Figure 6: FIG. 6: Flow chart of time domain noise acquisition. [PITH_FULL_IMAGE:figures/full_fig_p016_6.png]

**Figure 7.** Figure 7: FIG. 7: PAA noise model [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]

**Figure 8.** Figure 8: FIG. 8: PSD curves of the out-of-plane PAA between SC1 and SC2. [PITH_FULL_IMAGE:figures/full_fig_p017_8.png]

**Figure 9.** Figure 9: FIG. 9: PSD curves of the in-plane PAA between SC1 and SC2 [PITH_FULL_IMAGE:figures/full_fig_p018_9.png]

**Figure 10.** Figure 10: FIG. 10: Out-of-Plane PAA between SC1 and SC2 before filtering and after filtering. [PITH_FULL_IMAGE:figures/full_fig_p019_10.png]

**Figure 11.** Figure 11: FIG. 11: Out-of-plane PAA between SC1 and SC2 after filtering. [PITH_FULL_IMAGE:figures/full_fig_p019_11.png]

**Figure 12.** Figure 12: FIG. 12: Prediction error of the out-of-plane PAA between SC1 and SC2. [PITH_FULL_IMAGE:figures/full_fig_p019_12.png]

**Figure 13.** Figure 13: FIG. 13: In-plane PAA between SC1 and SC2 before filtering and after filtering . [PITH_FULL_IMAGE:figures/full_fig_p020_13.png]

**Figure 14.** Figure 14: FIG. 14: In-plane PAA between SC1 and SC2 after filtering. [PITH_FULL_IMAGE:figures/full_fig_p021_14.png]

**Figure 15.** Figure 15: FIG. 15: Prediction error of the in-plane PAA between SC1 and SC2. [PITH_FULL_IMAGE:figures/full_fig_p021_15.png]

**Figure 16.** Figure 16: FIG. 16: Out-of-plane PAA after filtering. [PITH_FULL_IMAGE:figures/full_fig_p024_16.png]

**Figure 17.** Figure 17: FIG. 17: Prediction error of the out-of-plane PAA. [PITH_FULL_IMAGE:figures/full_fig_p025_17.png]

**Figure 18.** Figure 18: FIG. 18: In-plane PAA after filtering [PITH_FULL_IMAGE:figures/full_fig_p025_18.png]

**Figure 19.** Figure 19: FIG. 19: Prediction error of the in-plane PAA. [PITH_FULL_IMAGE:figures/full_fig_p025_19.png]

**Figure 20.** Figure 20: FIG. 20: Out-of-plane PAA PSD curve before and after filtering. [PITH_FULL_IMAGE:figures/full_fig_p026_20.png]

**Figure 21.** Figure 21: FIG. 21: PSD curve of the out-of-plane PAA error before and after filtering. [PITH_FULL_IMAGE:figures/full_fig_p026_21.png]

**Figure 22.** Figure 22: FIG. 22: In-plane PAA PSD curve before and after filtering. [PITH_FULL_IMAGE:figures/full_fig_p026_22.png]

**Figure 23.** Figure 23: FIG. 23: PSD curve of the in-plane PAA error before and after filtering. [PITH_FULL_IMAGE:figures/full_fig_p027_23.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 16 canonical work pages

[1]

Z. Luo, Z. Guo, G. Jin, Y. Wu, and W. Hu, A brief analysis to Taiji: Science and technology , Results in Physics 16, 102918 (2020)

work page 2020
[2]

Danzmann, L

K. Danzmann, L. S. Team, et al. , LISA: laser interferometer space antenna for gravitational wave measurements, Classical and Quantum Gravity 13, A247 (1996)

work page 1996
[3]

Luo, L.-S

J. Luo, L.-S. Chen, H.-Z. Duan, Y.-G. Gong, S. Hu, J. Ji, Q. Liu, J. Mei, V. Milyukov, M. Sazhin, et al., TianQin: a space-borne gravitational wave detector , Classical and Quantum Gravity 33, 035010 (2016)

work page 2016
[4]

X. Han, X. Peng, W. Tang, Z. Yang, X. Ma, C. Gao, L.-e. Qiang, Y. Zhang, M. Zhao, J. Zhang, and B. Liu, Effect of celestial body gravity on Taiji mission range and range acceleration noise, Physical Review D 106, 102005 (2022)

work page 2022
[5]

Houba, S

N. Houba, S. Delchambre, T. Ziegler, G. Hechenblaikner, and W. Fichter, LISA Point-Ahead Angle Control for Optimal Tilt-to-Length Noise Estimation , arXiv preprint arXiv:2208.11033 (2022)

arXiv 2022
[6]

Simon, Optimal state estimation: Kalman, H ∞, and nonlinear approaches

D. Simon, Optimal state estimation: Kalman, H ∞, and nonlinear approaches. Hoboken , NJ: John Wiley and Sons, Jg 10, 0470045345 (2006)

work page 2006
[7]

Ljung, Asymptotic behavior of the extended Kalman filter as a parameter estimator for linear systems, IEEE Transactions on Automatic Control 24, 36 (1979)

L. Ljung, Asymptotic behavior of the extended Kalman filter as a parameter estimator for linear systems, IEEE Transactions on Automatic Control 24, 36 (1979)

work page 1979
[8]

Y. Xu, Y. S. Shmaliy, S. Bi, X. Chen, and Y. Zhuang, Extended Kalman/UFIR Filters for UWB-based Indoor Robot Localization under Time-Varying Colored Measurement Noise, IEEE Internet of Things Journal (2023)

work page 2023

Show all 17 references

[9]

Chang, On kalman filter for linear system with colored measurement noise , Journal of Geodesy 88, 1163 (2014)

G. Chang, On kalman filter for linear system with colored measurement noise , Journal of Geodesy 88, 1163 (2014)

2014
[10]

Y. Wang, G. Heinzel, and K. Danzmann, First stage of LISA data processing: Clock synchro- nization and arm-length determination via a hybrid-extended Kalman filter , Physical Review D 90, 064016 (2014)

2014
[11]

Y. Wang, G. Heinzel, and K. Danzmann, First stage of LISA data processing. II. Alternative filtering dynamic models for LISA , Physical Review D 92, 044037 (2015)

2015
[12]

P. Yang, Y. Huang, P. Li, H. Wang, M. Fan, S. Liu, Q. Shan, S. Qin, and Q. Liu, Orbit determination of China’s first mars probe Tianwen-1 during interplanetary cruise , Advances 28 in Space Research 69, 1060 (2022)

2022
[13]

Li and J

Z. Li and J. Zheng, Orbit determination for a space-based gravitational wave observatory, Acta Astronautica 185, 170 (2021)

2021
[14]

Fattah, W.-P

S. Fattah, W.-P. Zhu, and M. Ahmad, Identification of autoregressive moving average systems based on noise compensation in the correlation domain , IET signal processing 5, 292 (2011)

2011
[15]

Mohamed and K

A. Mohamed and K. Schwarz, Adaptive Kalman filtering for INS/GPS , Journal of geodesy 73, 193 (1999)

1999
[16]

Hewitson, M

M. Hewitson, M. Armano, M. Benedetti, J. Bogenstahl, D. Bortoluzzi, P. Bosetti, N. Brandt, A. Cavalleri, G. Ciani, I. Cristofolini, et al. , Data analysis for the LISA Technology Package , Classical and Quantum Gravity 26, 094003 (2009)

2009
[17]

Kaba and H

U. Kaba and H. Temeltas, Generalized bias compensated pseudolinear Kalman filter for colored noisy bearings-only measurements, Signal Processing 190, 108331 (2022). 29

2022

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Z. Luo, Z. Guo, G. Jin, Y. Wu, and W. Hu, A brief analysis to Taiji: Science and technology , Results in Physics 16, 102918 (2020)

work page 2020

[2] [2]

Danzmann, L

K. Danzmann, L. S. Team, et al. , LISA: laser interferometer space antenna for gravitational wave measurements, Classical and Quantum Gravity 13, A247 (1996)

work page 1996

[3] [3]

Luo, L.-S

J. Luo, L.-S. Chen, H.-Z. Duan, Y.-G. Gong, S. Hu, J. Ji, Q. Liu, J. Mei, V. Milyukov, M. Sazhin, et al., TianQin: a space-borne gravitational wave detector , Classical and Quantum Gravity 33, 035010 (2016)

work page 2016

[4] [4]

X. Han, X. Peng, W. Tang, Z. Yang, X. Ma, C. Gao, L.-e. Qiang, Y. Zhang, M. Zhao, J. Zhang, and B. Liu, Effect of celestial body gravity on Taiji mission range and range acceleration noise, Physical Review D 106, 102005 (2022)

work page 2022

[5] [5]

Houba, S

N. Houba, S. Delchambre, T. Ziegler, G. Hechenblaikner, and W. Fichter, LISA Point-Ahead Angle Control for Optimal Tilt-to-Length Noise Estimation , arXiv preprint arXiv:2208.11033 (2022)

arXiv 2022

[6] [6]

Simon, Optimal state estimation: Kalman, H ∞, and nonlinear approaches

D. Simon, Optimal state estimation: Kalman, H ∞, and nonlinear approaches. Hoboken , NJ: John Wiley and Sons, Jg 10, 0470045345 (2006)

work page 2006

[7] [7]

Ljung, Asymptotic behavior of the extended Kalman filter as a parameter estimator for linear systems, IEEE Transactions on Automatic Control 24, 36 (1979)

L. Ljung, Asymptotic behavior of the extended Kalman filter as a parameter estimator for linear systems, IEEE Transactions on Automatic Control 24, 36 (1979)

work page 1979

[8] [8]

Y. Xu, Y. S. Shmaliy, S. Bi, X. Chen, and Y. Zhuang, Extended Kalman/UFIR Filters for UWB-based Indoor Robot Localization under Time-Varying Colored Measurement Noise, IEEE Internet of Things Journal (2023)

work page 2023