Метод розпізнавання перешкод на шляху руху роботизованої збиральної техніки
Abstract
UDC 004.94:658.01
METHOD FOR RECOGNITION OF OBSTACLES TO MOVEMENT UNMANNED ROBOTIC CLEANING MACHINES
V. Lysenko, S. Shvorov, D. Komarchuk, D. Chirchenko
Now, to improve the efficiency of field work extensively apply the latest information technology system and precise positioning technology in the field. However, in this technique are the people who perform the functions of management to monitor the work it out automatically.
The aim is to improve research methods and technical principles of construction of the system of recognition of obstacles to the movement of robotic cleaning technology.
At the stage of solving the problem of building a system of recognition of obstacles to the movement of robotic cleaning technique there is a need to develop a portable interface and software for the PC that will allow the user to provide the planning and control of field work.
Existing means of identification of obstacles to the movement of robotic cleaning technology inherent in the following main drawbacks: low recognition accuracy in dynamic conditions of uncertainty, the high dependence of lighting facilities at night-time of the obstacles, significant capital and operating costs for the creation and use of equipment. One of the areas of eliminating these shortcomings are widely used neural networks.
To pre-processing (filtering) input images appropriate to apply Wavelet analysis is based on the use of wavelets that are mathematical functions and allow you to analyze different frequency components. In general, this analysis is in the plane, wavelet factor - time - level. Most wavelet coefficients determined by the integral transformation signal. The obtained wavelet spectrograms are fundamentally different from those of Fourier series, giving a clear binding signal range of features for a time. The third and fourth steps of pattern recognition is usually combined with pattern recognition system, which is the main element of the predictive complex.
For the synthesis and study of relevant neural networks used demo software package Statistica Neural Networks. Criterion training - minimizing mistakes neural network. In the context of this problem advantage of this package of similar developments is the implementation of the functional unit neuromodels optimization architecture that uses a linear approach and method "annealing" based on the probability distribution Gibbs. To implement the learning algorithm of neural network in the form of multilayer perceptron advisable to use a special genetic algorithm.
There were the synthesis and study of neural networks with application package STATISTICA allowing improved method for detection of obstacles in the path of cleaning technique, which is based on the use of the apparatus of neural networks and genetic pattern recognition algorithm for its training.
References
Anthony Stentz; et al. (2012). "A System for Semi-Autonomous Tractor Operations". Autonomous Robots. The Robotics Institute. 13 (1): 87–103.
Bill Howard (2013). "Google: Self-driving cars in 3-5 years. Feds: Not so fast". Extreme Tech.
Havrylov, A. (2002). Hybrydny intellektual'ny systemy [Hybrid intelligent systems]. Novosybyrsk, NHTU, 142.
David Hest. (2012). New driverless tractor, grain cart systems coming this year. Farm Industry News. Available at: http://farmindustrynews.com.
Shvorov, S. A., Komarchuk, D. S., Okhrymenko, P. H., Chyrchenko, D. V. (2015). Intelektual'na systema pidtrymky ta pryynyattya rishen' shchodo orhanizatsiyi vyroshchuvannya, zboru ta peretvorennya enerhetychnykh kul'tur u biometan [Intelektualna system pіdtrimki that of acceptance rіshen schodo organіzatsії viroschuvannya, Assembly that peretvorennya energetichnih cultures have bіometan]. Scientific Bulletin of National University of Life and Environmental Sciences of Ukraine, Tekhnika ta enerhetyka APK, 209 (2), 140–147.
Chirchenko, D., Shvorov, S. (2014). Intellectual systems of decision making support in the management of the collection and processing of organic raw materials. Annals of Warsaw University of Life Sciences, 68, 154.
Shvorov, S., Reshetiuk, V., Bolbot, I., Shtepa, V., Chirchenko, D. (2012). Theoretical issues construction and operation of agricultural mission robotic system . Annals of Warsaw University of Life Sciences – SGGW, 60, 177.
Shvorov, S., Komarchuk, D., Ohrimenko, P., Chyrchenko, D. (2015) Methodical Framework of the Support and Decision-making System for the Collection and Disposal of Organic Raw Materials. Energy Engineering and Control Systems, 1 (1), 29–34.
Downloads
Published
Issue
Section
License
Relationship between right holders and users shall be governed by the terms of the license Creative Commons Attribution – non-commercial – Distribution On Same Conditions 4.0 international (CC BY-NC-SA 4.0):https://creativecommons.org/licenses/by-nc-sa/4.0/deed.uk
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
