Adaptive data-driven modeling, prediction and optimal control for loaded transmission error of helicopter zerol spiral bevel gear transmission system

Abstract

In complex multi-stage transmission system especially for the helicopter zerol spiral bevel gears, the loaded transmission error (LTE) is a primary source of noise and vibration, and the misalignment is always an important noise factor affecting geometric and physical performances. An innovative adaptive data-driven LTE modeling, prediction and optimal control are proposed by modifying misalignment evaluations. Firstly, misalignment evaluations on helicopter transmission system are used as the data-driven design variables for tooth contact analysis with error (eTCA). Then, numerical loaded tooth contact analysis (NLTCA) considering flank elastic contact deformation, loaded contact pattern and load distribution is proposed for data-driven LTE modeling. To distinguish with the conventional minimizing LTE, an adaptive data-driven prediction and control system is provided. By prescribing the design target relating to the actual requirements, an adaptive data-driven functional relationship with respect to misalignment evaluations is established. For the input signal namely initial misalignment and its predicted LTE, by predicting its gap from the prescribed design target, this result is used to establish data-driven control objective function. Finally, after giving misalignment optimal operation, adaptive control strategy and solution are proposed to get a robust control result. The numerical instance is provided to verify the proposed methodology.