TBC-IG Random Variable-Dimension Algorithm for Aero-Engine Gas Path Sensor Optimization

doi:10.3390/aerospace12110970

Return

TBC-IG Random Variable-Dimension Algorithm for Aero-Engine Gas Path Sensor Optimization

2025-10-30

DOI：10.3390/aerospace12110970

Original

Original request for help

Save

Abstract

En 中文

The complex configuration of the internal flow field in aero-engines leads to limitations on sensor installation positions, and how to accurately identify the disturbances of the installation influence parameters under this constraint has long been a significant challenge. To address this issue, this study proposes an optimization algorithm to identify the optimal sensor layout. This is achieved by employing mutually distinct integer encoding, which ensures the uniqueness of each sensor position and prevents duplication. More importantly, an objective evaluation system incorporating tracking error, sensor comprehensiveness, and spatial coverage is integrated into the fitness function design, thereby overcoming the one-sidedness and limitations of single-indicator evaluation. Building upon this foundation, a sensor optimization scheme is proposed for identifying installation influence parameters. This scheme integrates the rapid search capability of the Tabu Bee Colony Random Variation Dimension Algorithm with the global optimization capability of an Improved Genetic Random Variation Dimension Algorithm, resulting in a Tabu Bee Colony-Improved Genetic Random Variation Dimension Optimization Algorithm (TBC-IG-RVDOA). For each installation influence parameter, different perturbation conditions were established, and the selected optimal sensor combination was then validated using the Extended Kalman Filter (EKF). Experimental studies show that, under all perturbation scenarios, the TBC-IG-RVDOA demonstrates strong convergence, high computational efficiency, and fitness function values consistently exceeding 0.92, thereby accurately capturing the changes in each installation influence parameter.

Keywords:

aero-engine

sensor optimization

tabu search

bee colony

improved genetic

random variable-dimension