Evaluating the performance of the adaptive neuro-fuzzy inference system (ANFIS) in predicting reheat gas turbine cycle output using machine learning techniques

Gas turbines serve as a backbone in numerous industrial applications such as electric power generation, aerospace propulsion, oil and gas processing, and the food industry. Evaluating the behaviour and performance of the entire gas turbine system—both prior to deployment and during operation—is crucial for enhancing efficiency, reducing downtime, and achieving cost-effectiveness due to its complex thermodynamic nature and high operational costs. To address these challenges, advanced modelling and optimisation techniques are progressively being adopted. Among them, the Adaptive Neuro-Fuzzy Inference System (ANFIS) has emerged as a robust hybrid intelligence approach that combines the learning ability of neural networks with the decision-making logic of fuzzy systems. This study focuses on the modelling, analysis, and optimisation of a gas turbine operating on a reheat cycle, which is known to enhance power output and thermal efficiency. The primary objective is to determine optimal operating conditions that either maximise thermal efficiency or maximise specific work output, depending on system requirements. These optimisation objectives are framed as a neuro-fuzzy problem and addressed using ANFIS implemented in MATLAB, which is particularly suitable for solving constrained, nonlinear optimisation problems involving multiple performance parameters and system limitations. The ANFIS model learns from a set of input and output data, enabling it to predict turbine performance accurately under varying operating conditions. Comparative analysis is also conducted with results obtained from the Lagrange Multiplier Method, serving as a benchmark to validate the reliability and precision of the ANFIS-based approach. The outcomes reveal that ANFIS offers a high level of adaptability, speed, and accuracy, making it a promising tool for real-time optimisation and intelligent control of gas turbine systems.