Gravity matching navigation algorithm based on multiscale search and Hadamard transformed difference
Hui Liu,Lin Wu,Lifeng Bao,Qianqian Li,Panpan Zhang,Yong Wang
DOI: https://doi.org/10.1016/j.isatra.2021.10.013
IF: 7.3
2021-10-01
ISA Transactions
Abstract:A novel gravity matching navigation algorithm based on multiscale search and Hadamard transformed difference (MS-HTD) is proposed in this paper. As far as we know, this is the first time Hadamard transformation was used in gravity matching aided navigation.Owing to the orthogonality, duality, and energy conservation of MS-HTD, the proposed MS-HTD algorithm focused on both the spatial structure variation characteristics and the difference of the numerical value of the gravity sequences effectively, whereas other classical methods rarely focused well on these two aspects simultaneously. In addition, to improve the searching efficiency, we have designed a multiscale neighborhood search strategy based on contour constraint instead of original point-by-point global searching which dramatically reduced the computational complexity.Marine gravity maps of South China Sea were chosen to construct the simulation tests to evaluate the performance of MS-HTD and other algorithms under the influences of different numbers of measurement, measuring accuracies, or matching areas.Simulation results showed that under various conditions the proposed algorithm has lower requirements for the number of measurements, measuring accuracy, and matching areas while obtaining higher navigation accuracy, matching success rate, and matching efficiency. During all experiments, MS-HTD algorithm was more effective, robust, and efficient than other algorithms.The accuracy and stability of navigation algorithms are crucial preconditions for underwater gravity matching aided navigation. To improve the matching accuracy and robustness of the matching algorithm, this paper presents a novel gravity matching navigation algorithm based on multiscale search and Hadamard transformed difference. The Hadamard transformation process was first time introduced in the new algorithm for the Walsh–Hadamard kernel function with the property of energy conservation. The gravity measurement sequences could be converted to the Hadamard domain; thus, the difference in numerical values, tendency, and spatial structure of the gravity measurement sequence were also a focus in the new algorithm, whereas only gravity statistical values were considered in classical matching algorithms. Therefore, with the proposed algorithm, the number of measurements necessary for matching can be effectively reduced, while improving the matching accuracy and success rate. In addition, a multiscale neighborhood search strategy based on contour constraints was designed to improve the matching efficiency, whereas a point-by-point global search was widely used in classical matching algorithms. Marine gravity maps of the South China Sea were used to construct the simulation tests. Simulation results show that under various conditions, the new algorithm has lower requirements for the number of measurements, measuring accuracy, and matching areas while exhibiting a higher navigation accuracy, matching success rate, and matching efficiency. Thus, the proposed algorithm could provide a new option for future practical applications of gravity matching aided navigation.
automation & control systems,instruments & instrumentation,engineering, multidisciplinary