Metaheuristic Algorithm-Based Vibration Response Model for a Gas Microturbine.

Area of Science:

• Mechanical Engineering
• Condition Monitoring
• Vibration Analysis

Background:

• Rotating machinery fault diagnosis relies on dynamic analysis of components like rotors.
• Intelligent algorithms are used for optimization, but the Jaya algorithm is underutilized.
• Vibration analysis is crucial for identifying operational anomalies.

Purpose of the Study:

• To apply the Jaya algorithm for optimizing fault diagnosis in rotating machinery.
• To analyze vibration data from a gas microturbine under varying conditions.
• To determine optimal operating parameters and the impact of temperature on vibrations.

Main Methods:

• Collected radial vibration amplitude data from a gas microturbine at different rotational frequencies and temperatures.
• Developed response surface models using polynomial fitting tuned by the Jaya metaheuristic algorithm.
• Evaluated model performance using Mean Squared Error (MSE) and tested different polynomial orders.

Main Results:

• A fifth-order polynomial model demonstrated superior performance with lower MSE.
• Optimal operating conditions identified at low/high rotational frequencies and 300°C.
• High temperatures increased measurement variability and shifted minimum vibration amplitude points.

Conclusions:

• The Jaya algorithm effectively optimizes vibration analysis for fault diagnosis in rotating machinery.
• Response surface models, particularly using averaged data, can identify optimal operating conditions.
• The proposed method offers a computationally efficient approach for vibration analysis.