UAV State Estimation Using Adaptive Complementary Filters

Marantos, Panos; Koveos, Yannis; Kyriakopoulos, Kostas J.

doi:10.1109/tcst.2015.2480012

Cited by 85 publications

(58 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The raw data is processed here with the proposed filter and representative complementary ones including the LMA-Complementary Observer (LMA-CO) by Fourati et al [17] and Wahba's Complementary Filter (WCF) by Marantos et al [19] The factor of LMA in LMA-CO is set to 0.01 according to related literature while its complementary gain is set to 0.01 denoting the motion's change rate. The parameters of WCF follow the original author's determination that are given by 푤 푎 = 0.9, 푤 푚 = 0.8, 푐 1 = 0.7, 푐 2 = 0.3, 푐 3푎 = 8500.0, 푐 3푚 = 5500.0.…”

Section: Experiments 2: Comparisons With Representative Complementarymentioning

confidence: 99%

See 1 more Smart Citation

Novel MARG-Sensor Orientation Estimation Algorithm Using Fast Kalman Filter

Guo

Wang

et al. 2017

Journal of Sensors

View full text Add to dashboard Cite

Orientation estimation from magnetic, angular rate, and gravity (MARG) sensor array is a key problem in mechatronic-related applications. This paper proposes a new method in which a quaternion-based Kalman filter scheme is designed. The quaternion kinematic equation is employed as the process model. With our previous contributions, we establish the measurement model of attitude quaternion from accelerometer and magnetometer, which is later proved to be the fastest (computationally) one among representative attitude determination algorithms of such sensor combination. Variance analysis is later given enabling the optimal updating of the proposed filter. The algorithm is implemented on real-world hardware where experiments are carried out to reveal the advantages of the proposed method with respect to conventional ones. The proposed approach is also validated on an unmanned aerial vehicle during a real flight. Results show that the proposed one is faster than any other Kalman-based ones and even faster than some complementary ones while the attitude estimation accuracy is maintained.

show abstract

Section: Experiments 2: Comparisons With Representative Complementarymentioning

confidence: 99%

“…Mahony et al tried different way where the attitude representation on special orthogonal groups is studied [18]. Using transformation of Wahba's problem, Marantos et al designed one efficient complementary filter with singular value decomposition (SVD) as well [19].…”

Section: Introductionmentioning

confidence: 99%

Novel MARG-Sensor Orientation Estimation Algorithm Using Fast Kalman Filter

Guo

Wang

et al. 2017

Journal of Sensors

View full text Add to dashboard Cite

show abstract

“…In [3][4][5], frequency domain linear filter is used to estimate attitude. In [6][7], Extended Kalman Filter (EKF) is widely applied to aerospace. Particle filter is used for special attitude estimation application with multi information fusion [8].…”

Section: Introductionmentioning

confidence: 99%

“…Similar to Linear complementary filter, Nonlinear complementary filter can also provide fusing strategy for gyroscope and accelerometer. In [2][3], [6][7][8][9][10][11][12][13], Nonlinear complementary filter is designed by referring to reference physical quantity of geomagnetic or local gravity, then passed by proportion and integral estimator, compensates angular velocity last. Nonlinear complementary filter is usually used when angular motion is stable and linear acceleration is small enough, or provided wrong angular velocity compensation otherwise [13][14][15].…”

Section: Introductionmentioning

confidence: 99%

Complementary Filter Design Based Error Allocation for Attitude Estimation with Low-Cost Sensor

Yang¹,

Ren²,

Song³

et al. 2018

Proceedings of the 2017 2nd International Conference on Electrical, Control and Automation Engineering (ECAE 2017)

View full text Add to dashboard Cite

Abstract-This paper investigates complementary filter designed for attitude estimation for low-cost inertial measurement unit and magnetometer. An error allocation strategy which focuses on two weight parameter matrices regulated according to output modulus of accelerometer and magnetometer automatically, is proposed. The strategy of error allocation fuses gravity reference error angle comes from accelerometer and geomagnetic reference error angle comes from magnetometer to get the reference angle error which is used to correct the angular velocity drift of gyroscope after passed proportion and integral estimator (PI estimator). Attitude is obtained with quaternion solution. Experimental results show that this error allocation strategy can get reference angle error when output modulus of accelerometer is under the acceleration threshold or the changing horizontal magnetic component is below the changing threshold.

show abstract

“…Although CF is much faster than KF, it owns different convergence speed for different motions and thus lowers the overall attitude estimation accuracy. In this case, related adaptive schemes are introduced by Tian et [18], and Vasconcelos et al [19], which provide adaptive performances. Nonlinear CF is also developed by Mahony et al using special orthogonal groups [20].…”

Section: Introductionmentioning

confidence: 99%

Linear Kalman Filter for Attitude Estimation from Angular Rate and a Single Vector Measurement

Shan

Hou

2017

Journal of Sensors

View full text Add to dashboard Cite

In this paper, a new Kalman filtering scheme is designed in order to give the optimal attitude estimation with gyroscopic data and a single vector observation. The quaternion kinematic equation is adopted as the state model while the quaternion of the attitude determination from a strapdown sensor is treated as the measurement. Derivations of the attitude solution from a single vector observation along with its variance analysis are presented. The proposed filter is named as the Single Vector Observation Linear Kalman filter (SVO-LKF). Flexible design of the filter facilitates fast execution speed with respect to other filters with linearization. Simulations and experiments are conducted in the presence of large external acceleration and magnetic distortion. The results show that, compared with representative filtering methods and attitude observers, the SVO-LKF owns the best estimation accuracy and it consumes much less time in the fusion process.

show abstract

UAV State Estimation Using Adaptive Complementary Filters

Cited by 85 publications

References 24 publications

Novel MARG-Sensor Orientation Estimation Algorithm Using Fast Kalman Filter

Novel MARG-Sensor Orientation Estimation Algorithm Using Fast Kalman Filter

Complementary Filter Design Based Error Allocation for Attitude Estimation with Low-Cost Sensor

Linear Kalman Filter for Attitude Estimation from Angular Rate and a Single Vector Measurement

Contact Info

Product

Resources

About