Dynamic System Stability Modeling Approach with Sparrow-Inspired Meta-Heuristic Optimization Algorithm.

Area of Science:

• Aerospace Engineering
• Mechanical Engineering
• Computational Fluid Dynamics

Background:

• Compressor instability poses significant risks to operational safety and efficiency.
• Accurate prediction of stall inception is crucial for preventing catastrophic failures.
• Existing methods often lack the precision required for early detection in complex dynamic systems.

Purpose of the Study:

• To develop a dynamic system stability model for accurate compressor instability prediction.
• To enhance the early warning capabilities of dynamic systems in compressors.
• To propose a novel approach for system modeling and performance prediction.

Main Methods:

• Spatial mode for flow field feature extraction and modeling.
• Fuzzy entropy for nonlinear characteristic identification and instability determination.
• Sparrow Search Algorithm (SSA) optimized Radial Basis Function Neural Network (RBFNN) for performance prediction.
• Logistic SSA and Tent SSA (TSSA) for optimizing RBFNN parameters and improving model robustness.

Main Results:

• The RBFNN-LSSA hybrid model detected stall inception approximately 35.8 revolutions in advance.
• The RBFNN-TSSA model achieved a superior multi-step prediction accuracy, providing 42 revolutions of advance warning.
• The TSSA-based optimization demonstrated robust adaptation for complex nonlinear dynamic modeling.

Conclusions:

• The proposed RBFNN-TSSA model offers a novel and promising approach for compressor system modeling and instability prediction.
• The advanced warning capabilities significantly enhance the abnormal warning capacity of dynamic compressor systems.
• This methodology contributes to improved operational safety and reliability in compressor applications.