Andrashitov D.S.1, Kostoglotov A.A.2, Lazarenko S.V.3, Shupletsov A.M.4

1 Military Academy of Strategic Rocket Troops after Peter the Great (Balashikha, Moscow Region, Russia)

2,3 Don State Technical University (Rostov-on-Don, Russia)

4 Rostov State University of Railway Engineering (Rostov-on-Don, Russia)

1dima-andrahitov@rambler.ru, 2kostoglotov@icloud.com, 3lazarenkosv@icloud.com, 3shupletsov83@mail.ru

Currently, there is widespread use of tracking systems, which include traffic safety systems for land, sea and air transport, as well as radar, radio navigation, optoelectronic complexes, etc. Regardless of the purpose, they solve a set of tasks to assess the parameters of the state of the observed object in order to form a control effect.

As a rule, tracking systems must provide estimates of motion parameters in real time, which does not allow the use of algorithms demanding computing power due to limitations on the processing time of measuring information. However, there are information and measurement systems with a posteriori processing of measurement results. This allows the use of algorithms of any complexity, but the scope of such solutions is research tasks and tests.

At the same time, the algorithm for estimating motion parameters must adapt to different modes of motion, each of which is characterized by a set of parameters such as speeds, accelerations, trajectory, etc. The low efficiency of the corresponding computational procedure can be compensated by high-precision measuring instruments, which must provide a sufficiently high frequency of measurement information. Thus, the use of inexpensive satellite positioning systems with an error in determining the coordinates of an object up to 5 meters and a sampling rate of measuring information of 5 Hz does not guarantee the autonomy of vehicle movement even at low maneuvering intensities. The best results can, for example, be obtained using optical sensors from DATRON Sensor Systems GmbH, which, as part of the information and measurement system, give a sampling frequency of measuring information less than 10 Hz, which, as noted in, is also insufficient for high-speed maneuvering of ground vehicles. Further increase in the efficiency of measuring instruments is associated with a sharp increase in their cost, which in some cases is economically impractical.

The sampling frequency of 1×103 Hz, which, for example, is provided by acceleration sensors of inertial navigation systems, can be considered a guideline for the necessary resolution when evaluating the motion parameters of ground targets. This is due to the fact that in this case, at relatively low speeds, it is possible to consider the curvilinear motion of an unmanned vehicle as a set of linear trajectories. At the same time, it should be noted that inexpensive accelerometers have low accuracy and are used to roughly determine the distance traveled and speed. An additional reason for the deterioration of the quality of measuring information is the degradation of measuring instruments due to the influence of various factors, which leads to a deviation of their parameters from the nominal (certified) values. As a result, the conversion factor changes and the measurement error increases. This, in turn, can lead to metrological failure. Therefore, the task of calibrating MEMS accelerometers is relevant. Its solution will ensure the necessary level of accuracy and reliability of such sensors.

Usually, accelerometer calibration is carried out by the method of test turns, however, in the process of operation of the tracking system, as a rule, this is not possible. In this case, calibration can be carried out using identification algorithms synthesized using the principles of regression analysis, the theory of statistical synthesis and gradient methods. Well-known studies show that constructive results of solving the problem of identifying the parameters of mathematical models can be obtained using the regularization principle of A.N. Tikhonov. However, then a relatively high amount of computational cost is required, which prevents the practical implementation of such algorithms in smart sensors. Therefore, it is necessary to resolve the contradiction between the computational complexity and the accuracy of the synthesized parametric identification algorithms. In this paper, it is proposed to use an approximate solution of the equation of the sensitivity function. The effectiveness of the synthesized identification algorithm is estimated based on mathematical modeling.

Andrashitov D.S., Kostoglotov A.A., Lazarenko S.V., Shupletsov A.M. Parametric identification of the MEMS-accelerometer model during its calibration using a numerical solution of the sensitivity equation. Information-measuring and Control Systems. 2024. V. 22. № 1. P. 24−32. DOI: https://doi.org/10.18127/j20700814-202401-04 (in Russian)

1. Kostoglotov A.A., Lazarenko S.V., Andrashitov D.S. Regulyarizirovannyi algoritm mnogoparametricheskoi variatsionnoi identifikatsii dinamicheskikh sistem. Servis v Rossii i za rubezhom. № 8(27). http://www.rguts.ru/electronic_journal/number27/contents). (in Russian)
2. Kiku A.G., Kostyuk V.I. i dr. Adaptivnye sistemy identifikatsii. Kiev: Tekhnika. 1975. 288 c. (in Russian)
3. Gropp D. Metody identifikatsii sistem. M.: Mir. 1975. 302 s. (in Russian)
4. Spravochnik po teorii avtomaticheskogo upravleniya. Pod red. A.A. Krasovskogo. M.: Nauka. 1987. 712 s. (in Russian)
5. Tikhonov A.N., Arsenin V.Ya. Metody resheniya nekorrektnykh zadach. M.: Nauka. 1986. 288 s. (in Russian)
6. Seidzh E.P., Mels D.L. Identifikatsiya sistem upravleniya. M.: Nauka. 1974. 246 c. (in Russian)
7. Kostoglotov A.A., Deryabkin I.V., Andrashitov D.S., Lazarenko S.V., Pugachev I.V. Synthesis of algorithms for estimation of parameters and state of dynamic systems using additional invariants. Proceedings of 2016 IEEE East-West Design and Test Symposium. EWDTS. 2016. P. 7807637.
8. Kostoglotov A.A., Kuznetsov A.A., Lazarenko S.V., Andrashitov D.S. Analiz effektivnosti innovatsionnykh reshenii na baze ob'edinennogo printsipa maksimuma v zadachakh obrabotki informatsii i upravleniya. Innovatika i ekspertiza: nauchnye trudy. 2015. № 2 (15). S. 38−51. (in Russian)
9. Andrashitov D.S., Deryabkin I.V., Kostoglotov A.A., Lazarenko S.V. Analiz funktsionirovaniya algoritmov parametricheskoi identifikatsii informatsionno-upravlyayushchikh sistem, udovletvoryayushchikh printsipu Gamiltona-Ostrogradskogo. Dinamika slozhnykh sistem - XXI vek. 2014. T. 8. № 2. S. 90−95. (in Russian)
10. Pavlov D.V., Lukin K.G., Petrov M.N. Razrabotka matematicheskoi modeli MEMS-akselerometra. Vestnik Novgorodskogo gosudarstvennogo universiteta. № 8(91). 2015. S. 22−25. (in Russian)