Affiliations: [a] IT4Innovations, VŠB Technical University of Ostrava, Ostrava, Czech Republic | [b] Department of Computer Science & Engineering, Jadavpur University, Kolkata, India | [c] Department of Computer & System Sciences, Visva-Bharati University, India | [d] CEGESS, Indian Institute of Engineering Science and Technology, Howrah, India
Corresponding author: Varun Kumar Ojha, IT4Innovations, VŠB Technical University of Ostrava, Ostrava, Czech Republic, E-mail:firstname.lastname@example.org
Abstract: In this article, we proposed a multi-agent concurrent neorosimulated annealing (CNSA) algorithm, which was used for the supervised training of the neural networks (NN). The proposed CNSA is a population based parallel version of the basic simulated annealing (SA) algorithm. In this work, CNSA was applied for designing an intelligent sensory system that detects proportion of component gases of manhole gas mixture. The proposed intelligent sensory system was modeled using NN, where, the training of NN was supplemented by the proposed parallel version of SA algorithm, that is, CNSA. Once the training of the NN was covered, the sensory system was used for sensing the accumulated toxic gas components of manholes. The manhole gas-mixture problem was treated as pattern recognition and noise reduction problem. This article offers a critical performance analysis of CNSA algorithm, where its performance was compared with backpropagation, conjugate gradient algorithm, particle swarm optimization, and genetic algorithm in both empirical and statistical sense. We found that the proposed CNSA performed significantly well in comparison to its counterparts as far as this case study was concerned.