Search

This article is devoted to the questions of the reliability and the accuracy of the invented algorithm. Background: The implementation of autonomous navigation system for unmanned airborne vehicle is required for the reliable navigation in case of the instability of the satellite navigation system. Traditionally this problem is solved by the terrain landform navigation system, also the synthetic aperture radar navigation systems could be used as the correction system for the inertial navigation system. Nevertheless, these correction systems do not allow designers to provide necessary stability and reliability of whole system. Therefore, we use lengthened objects as the reference objects, which are stable enough either in different weather seasons. Also these objects (such as roads, rivers, railways) are very often on flat landforms, where the terrain landform navigation system does not work. Methods: We process the amplitude distribution of the reflected signal in order to get the parameter of the algorithm. The obtained distributions could be compared with etalon distributions, which can be computed for various terrain types. Therefore, we can use the intersection area as the indicator of the underlying terrain. This algorithm could be found in previous papers. This paper describes in detail problem of choice necessary counts number for reliable detection of lengthened objects. The suggested solutions allow user to implement this algorithm in the real systems. Also the comparison of the developed algorithm with algorithm based on the KolmogorovSmirnov criterion briefly described there. Conclusion: The implementation of the designed algorithm contains resolving of many small but important questions such as choice of necessary counts number of the accuracy of the algorithm. These questions are described in this paper. [ABSTRACT FROM AUTHOR]

Background: This article is devoted to the question of correction system implementation. The correction system is necessary to provide the ability of autonomous navigation the unmanned airborne vehicle. We can see a lot of examples of terrain landform correction systems and “synthetic aperture radar” – based systems. But in the case of using information about lengthened objects, we have a great opportunity to provide robustness objects, which we can reliably cross during our flight. Nevertheless, these objects should be placed in some on-board system, which will process this information to make a suggestion about unmanned airborne vehicle position. Methods: Recently it was suggested such an algorithm, which is based on the histogram comparison of the reflected signal of the pulse radar altimeter. It is well known that the distribution of the reflected signal of the pulse radar is depend on the terrain from which this signal was reflected. Therefore, the idea of using this information as the parameter, which helps to define the terrain type, is described in this paper. The comparison of the histograms (one is etalon, other is current histogram) in this algorithm is based on the evaluation of the intersection area of these histograms. Results: The main part of this paper reveals how this idea could be implemented in the real navigation system structure. The question of performance is also discussed there. It is shown a useful case of implementation of this algorithm. Conclusion: The suggested algorithm could be implemented in the real on-board system and it allows us to increase the accuracy of autonomous navigation without implementing any additional hardware decisions. [ABSTRACT FROM AUTHOR]