V.V. Kvashennikov1, O.O. Kozeeva2

1,2 Kaluga Branch of the Bauman MSTU (Kaluga, Russia)

Mathematical model of the artificial intelligence in smart home systems and the processes of data transfer and processing in the terms of error-correcting coding and with systematic approach is described in the article. AI of smart home system has signals from sensors and other embedded devices as an input data. These signals are encoded with sequences of characters and are divided into allowable and forbidden code sequences. Allowable sequences correspond to images; forbidden ones separate images from each other. The large redundancy of character sequences allows for reliable pattern detection even if there are errors in the input sequences. Error correction increases code efficiency of code recovery. While receiving information, stochastic codes are decoded by comparing the code of the input data with its information part. So, not only the correction of errors and erasures of the input sequence is implemented, but also an optimal image encoding is obtained, that allows to minimize the number of symbols of the image code, so image processing is simplified as well. Images can be divided into simple and complex: complex images are designed on the basis of simple images or other complex images so the input data can be represented as a large number of simple images forming complex images. Therefore, more efficient encoding of complex images can be achieved.

Functional transformations in neural networks are applied for image processing. Neural network provides a task of a universal approximator of functions of many variables forming a nonlinear function for constructing images. High performance is ensured by parallel processing of a large number of input images in neural networks. The processing of single images is carried out by automatic control systems, the processing of complex images is implemented using AI technologies, which may include various devices, so «smart home» system includes three main AI subsystems and a diagram of it is presented in the article.

AI in the smart home is an intelligent assistant that controls multiple engineering systems in a home. The importance of developments in the field of smart home is that they directly affect human life. The study of the principles of coding and construction of images, as well as decision-making are important tasks in the development of AI technologies in smart home systems.

Kvashennikov V.V., Kozeeva O.O. Artificial intelligence in smart home systems. Neurocomputers. 2021. V. 23. № 6. P. 48−54. DOI: https://doi.org/10.18127/j19998554-202106-05 (in Russian)

1. Sleigl Dzh. Iskusstvennyi intellekt. M.: Mir. 2017. 320 c. (in Russian)
2. KHaikin S. Neironnye seti: polnyi kurs. Per. s angl. M.: Izdatelskii dom «Vilyams». 2006. 1104 s. (in Russian)
3. Nikolenko S., Kadurin A., Arkhangelskaya E. Glubokoe obuchenie. SPb.: Piter. 2018. 480 s. (in Russian)
4. Osmolovskii S.A. Stokhasticheskie metody zashchity informatsii. M.: Radio i svyaz. 2003. 320 s. (in Russian)
5. Piterson U., Ueldon E. Kody, ispravlyayushchie oshibki: Per. s angl.. Pod red. R.L. Dobrushina i S.I. Samoilenko. M: Mir. 1976. 594 s. (in Russian)
6. Hornik K., Stinchcombe M. and White H. Multilayer feedforward networks are universal approximators. Neural Networks. 1989. V. 2. № 5. P. 359−366.
7. Galushkin A.I. Neironnye seti: osnovy teorii. M.: Goryachaya liniya – Telekom. 2017. 496 s. (in Russian)
8. Gallager R. Teoriya informatsii i nadezhnaya svyaz. SShA. 1968. Per. s angl.. Pod red. M.S. Pinskera i B.S. Tsibakova. M.: Sov. radio. 1974. 720 s. (in Russian)
9. Kolesnik V.D., Mironchikov E.T. Dekodirovanie tsiklicheskikh kodov. M.: Svyaz. 1968. 251 s. (in Russian)
10. Uinston P. Iskusstvennyi intellekt. M.: Mir. 2017. 520 c. (in Russian)
11. Dementev A.D. Umnyi dom 21 veka. Izdatelskie resheniya. 2016. 100 s. (in Russian)
12. Ignatenko S.O., Islamova R.R. Iskusstvennyi intellekt v sisteme «umnyi dom». Materialy Mezhdunar. nauchno-praktich. konf. Belgorod. 2018. Chast II. S. 53−56. (in Russian)
13. Rassel S., Norvig P. Iskusstvennyi intellekt: sovremennyi podkhod. Per. s angl. i red. K.A. Ptitsyna. Izd. 2-e. M.: Izdatelskii dom «Vilyams». 2006. 1408 s. (in Russian)