Aircraft Landing Gear Point Cloud Registration Method Based on Improved Teaching‐Learning‐Based Optimization

ABSTRACT

Point cloud registration of aircraft landing gear can be regarded as a six‐degree‐of‐freedom optimization problem, where the optimization performance directly affects registration accuracy. Although the Teaching–Learning‐Based Optimization (TLBO) algorithm has a simple structure and requires few control parameters, it is prone to premature convergence when solving high‐dimensional optimization problems. In this work, an improved TLBO algorithm, referred to as HAT‐TLBO, is proposed by combining Halton sequence initialization, adaptive learning weights and a tutoring phase. A mathematical optimization model is established by explicitly defining the decision variables, objective function, and optimization constraints. The proposed algorithm is evaluated on the CEC2017 test functions through parameter sensitivity analysis and ablation experiments. Benchmark results indicate that HAT‐TLBO achieves the best performance on 17 and 25 test functions in the 10‐dimensional and 50‐dimensional test suites, respectively. Registration experiments on public datasets and real aircraft landing gear point clouds show that HAT‐TLBO achieves higher registration accuracy and better convergence stability than the standard TLBO and PSO algorithms. The proposed method provides an effective solution for high‐precision point cloud registration of aircraft landing gear.