Neural Network based Throttle Controller for Compressor

The centrifugal Compressor is used in various industries, including oil, aerospace, and gas facilities, to increase pressure and oil output. If the Compressors are failed, then the entire system's production will be halted.The primary objective is to model the centrifugal Compressor with Anti-surge control (ASC) mechanism using neural networks to avoid surge issues and improve performance metrics. Different controllers like PID controller, followed by Fuzzy logic controller (FLC), Neuro-fuzzy controller (NFC) are incorporated in Anti surge control(ASC) mechanism for Compressor recycle system to realize the performance comparison. The modelling of the CRS with ASC mechanism is carried out using MATLAB Simulink environment. The simulation results like different flow rates, suction and plenum pressures, speed, and compressor maps are realized in detail for all the controller designs. The performance metrics error responses are analyzed in detail for CRS with ASC mechanism using different controllers. Based on the NNPC output response in the ASC mechanism of the CRS system, the OP position is adjacent to the Surge control line (SCL).

Autorentext

La Dra. Divya M N, doctora en electrónica aeroespacial por la VTU de Belgaum, cuenta con 22 años de docencia y 1 año de experiencia en la industria. Ha publicado más de 22 trabajos de investigación en revistas y conferencias internacionales. Sus áreas de especialización son los sistemas de control, el aprendizaje automático y el procesamiento de señales.