Multisensor System Based on Porous Silicon with Multiagent Data Processing

Authors S.E. Pritchin , I.V. Shevchenko , A.P. Oksanich , M.G. Kogdas , Y.A. Rastoropov, О.S. Prytchyn, V.A. Palagin
Affiliations

Kremenchuk Mykhailo Ostrogradsky National University, 39600 Kremenchuk, Ukraine

Е-mail pritchinse@gmail.com
Issue Volume 17, Year 2025, Number 6
Dates Received 24 September 2025; revised manuscript received 16 December 2025; published online 19 December 2025
Citation S.E. Pritchin, I.V. Shevchenko, A.P. Oksanich, et al., J. Nano- Electron. Phys. 17 No 6, 06002 (2025)
DOI https://doi.org/10.21272/jnep.17(6).06002
PACS Number(s) 07.07.Df ,61.43.Gt
Keywords Porous silicon (3) , Gas concentration sensors, Multisensor system, Multiagent system, Extended feature space.
Annotation

The article proposes a method for improving the selectivity of determining gas concentrations in gas mixtures using multisensor systems based on porous silicon with the application of a multiagent information processing system. A general formal description of the multi-agent system is provided, which is improved by the introduction of a set of fuzzy cognitive maps designed to recognize situations, forming a spectrum of representations, focusing attention, and evaluating the success of agents' actions, which allows for the construction of various multi-agent systems using a single methodology, in particular those used to improve the selectivity of sensors. The task description includes a set of classes to be recognized, a set of features, a measurement procedure, a set of associative models for transforming a set of features into a set of classes, which is a set of fuzzy cognitive maps, as well as procedures for automatic adjustment and correction of gas concentration measurement results. The proposed method is based on determining the composition of a gas mixture obtained by a matrix of gas sensor agents based on porous silicon with different porosity and at different temperatures. Using a multi-agent system, an extended feature space is formed, and the distances between measurement points in this space are calculated. The set of distances is converted into a distribution of gas concentrations, which simplifies the concentration measurement algorithm while maintaining the required measurement accuracy for determining the composition of the gas mixture and the concentration of its individual elements. The experiments performed and the analysis of the results show that the proposed system is capable of determining the composition of gas mixtures with sufficient accuracy.

List of References