Limits...
Medium to Long Range Kinematic GPS Positioning with Position-Velocity-Acceleration Model Using Multiple Reference Stations.

Hong CK, Park CH, Han JH, Kwon JH - Sensors (Basel) (2015)

Bottom Line: In order to obtain precise kinematic global positioning systems (GPS) in medium to large scale networks, the atmospheric effects from tropospheric and ionospheric delays need to be properly modeled and estimated.The performance of the proposed algorithms is validated by analyzing and comparing the results with those from reference values.The results show that reliable and comparable solutions in both position and kinematic acceleration levels can be obtained using the proposed algorithms.

View Article: PubMed Central - PubMed

Affiliation: Department of Geoinformatics Engineering, Kyungil University, 50 Gamasilgil, Kyeongsan, Gyeongbuk 712-701, Korea. ckhong@kiu.ac.kr.

ABSTRACT
In order to obtain precise kinematic global positioning systems (GPS) in medium to large scale networks, the atmospheric effects from tropospheric and ionospheric delays need to be properly modeled and estimated. It is also preferable to use multiple reference stations to improve the reliability of the solutions. In this study, GPS kinematic positioning algorithms are developed for the medium to large-scale network based on the position-velocity-acceleration model. Hence, the algorithm can perform even in cases where the near-constant velocity assumption does not hold. In addition, the estimated kinematic accelerations can be used for the airborne gravimetry. The proposed algorithms are implemented using Kalman filter and are applied to the in situ airborne GPS data. The performance of the proposed algorithms is validated by analyzing and comparing the results with those from reference values. The results show that reliable and comparable solutions in both position and kinematic acceleration levels can be obtained using the proposed algorithms.

No MeSH data available.


Convergence of the covariance matrix, Pk: (a) whole trajectory; (b) beginning part of the trejectory.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4541913&req=5

sensors-15-16895-f002: Convergence of the covariance matrix, Pk: (a) whole trajectory; (b) beginning part of the trejectory.

Mentions: As described in Section 2, the kinematic accelerations together with positions and velocities are estimated through the Kalman filter approach. The multi-baseline data processing is performed with radial type of network configuration. To evaluate the filter efficiency, the convergence of the covariance matrix Pk is examined. Figure 2a presents the variations of the trace of Pk with respect to time. As shown in Figure 2a, the trace of Pk converges rapidly with respect to time, which indicates the reliability of the filter. It is also notable that two peaks in Figure 2a correspond to the epochs at which new satellites are observed. Then, new states for DD ionospheric delays and ambiguities are included in state vector, and predefined variance values, i.e., (0.5 m)2 for DD ionospheric delay and 1 × 1010 for DD ambiguity, are assigned to the new states.


Medium to Long Range Kinematic GPS Positioning with Position-Velocity-Acceleration Model Using Multiple Reference Stations.

Hong CK, Park CH, Han JH, Kwon JH - Sensors (Basel) (2015)

Convergence of the covariance matrix, Pk: (a) whole trajectory; (b) beginning part of the trejectory.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4541913&req=5

sensors-15-16895-f002: Convergence of the covariance matrix, Pk: (a) whole trajectory; (b) beginning part of the trejectory.
Mentions: As described in Section 2, the kinematic accelerations together with positions and velocities are estimated through the Kalman filter approach. The multi-baseline data processing is performed with radial type of network configuration. To evaluate the filter efficiency, the convergence of the covariance matrix Pk is examined. Figure 2a presents the variations of the trace of Pk with respect to time. As shown in Figure 2a, the trace of Pk converges rapidly with respect to time, which indicates the reliability of the filter. It is also notable that two peaks in Figure 2a correspond to the epochs at which new satellites are observed. Then, new states for DD ionospheric delays and ambiguities are included in state vector, and predefined variance values, i.e., (0.5 m)2 for DD ionospheric delay and 1 × 1010 for DD ambiguity, are assigned to the new states.

Bottom Line: In order to obtain precise kinematic global positioning systems (GPS) in medium to large scale networks, the atmospheric effects from tropospheric and ionospheric delays need to be properly modeled and estimated.The performance of the proposed algorithms is validated by analyzing and comparing the results with those from reference values.The results show that reliable and comparable solutions in both position and kinematic acceleration levels can be obtained using the proposed algorithms.

View Article: PubMed Central - PubMed

Affiliation: Department of Geoinformatics Engineering, Kyungil University, 50 Gamasilgil, Kyeongsan, Gyeongbuk 712-701, Korea. ckhong@kiu.ac.kr.

ABSTRACT
In order to obtain precise kinematic global positioning systems (GPS) in medium to large scale networks, the atmospheric effects from tropospheric and ionospheric delays need to be properly modeled and estimated. It is also preferable to use multiple reference stations to improve the reliability of the solutions. In this study, GPS kinematic positioning algorithms are developed for the medium to large-scale network based on the position-velocity-acceleration model. Hence, the algorithm can perform even in cases where the near-constant velocity assumption does not hold. In addition, the estimated kinematic accelerations can be used for the airborne gravimetry. The proposed algorithms are implemented using Kalman filter and are applied to the in situ airborne GPS data. The performance of the proposed algorithms is validated by analyzing and comparing the results with those from reference values. The results show that reliable and comparable solutions in both position and kinematic acceleration levels can be obtained using the proposed algorithms.

No MeSH data available.