R&D of an adaptive and multivariable neuro-fuzzy controller for use in industrial processes

• Grant number: 15/50599-1
• Support Opportunities: Research Grants - Innovative Research in Small Business - PIPE
• Start date: October 01, 2016
• End date: August 31, 2018
• Field of knowledge: Engineering - Electrical Engineering - Industrial Electronics, Electronic Systems and Controls
• Agreement: FINEP - PIPE/PAPPE Grant
• Principal Investigator: Leonardo Oliveira de Freitas e Souza
• Grantee: Leonardo Oliveira de Freitas e Souza
• Company: Sensycal Instrumentos e Sistemas Ltda
• City: São Paulo

Abstract

This research project aims to develop an "plug & play" advanced controller for industrial processes, allowing application directly to the Ethernet network in industries and plants. Also this controller can model your control model automatically, using historical data from the process under control. Based on the research and prototype developed, the controller will have neuro-fuzzy, adaptive (neural networks) and predictive control characteristics (combination of neuro-fuzzy with neural networks). There is a clear advantage of the product to be developed in relation to existing solutions because it will allow the definition and simulation of the result of advanced control to be implemented in a given industrial process without interfering with the actual process. (AU)