A novel boundary defect recognition method based on adaptive regularization-improved artificial gorilla troops optimizer

In the field of heat transfer, it is an important task to identify fault boundaries within samples or devices. In this paper, an Adaptive Regularization-Improved Artificial Gorilla Troop Optimizer (AR-IGTO) is proposed, which combines adaptive regularization with the Gorilla Troop Optimizer (GTO) algorithm to improve the stability and efficiency of Inverse Heat Conduction Problem (IHCP) solutions. The AR-IGTO is applied to defect recognition problems with unknown boundaries, based on temperature information, using a two-dimensional (2D) unsteady heat transfer model and a natural gas pipeline model as benchmarks. The simulation results show that AR-IGTO can reduce the influence of measurement errors on the recognition accuracy and converge faster. It is significantly better than the traditional particle swarm optimization (PSO) and GTO methods, proving its effectiveness and applicability in identifying defects in natural gas pipelines. It is also of reference significance for solving other IHCP.