A Data-Driven Method for Determining the Loading Spectrum in Fatigue Tests for Working Device of Engineering Machinery.

The working device of a wheel loader is a critical component of construction machinery and is subjected to long-term random loads during operation, making it prone to fatigue damage. To evaluate whether it satisfies the design life requirements, load spectra are typically extracted and fatigue life analysis is performed. However, the determination of the load spectrum remains a key challenge, as no standardized method is currently available in the field of construction machinery. To this end, this paper proposes an inverse VIM based on a data-driven model and a genetic algorithm to invert the input load spectrum and apply it to the working device of a loader, so as to realize the fatigue life assessment of the working device of a construction machinery in order to guide its structural design. The initial step involves applying random loads to the four actuators of the fatigue testing platform and measuring the corresponding output loads at various measurement points. By utilizing data identification techniques and the NARX model, a numerical model for the working device of the loader is established. Subsequently, employing genetic algorithms and the measured target output loads, the input loads are inversely determined. Finally, the obtained input loads are applied to the fatigue testing platform, and the output loads at different measurement points are collected and compared with the target output loads to validate the effectiveness and reliability of this method in terms of damage assessment. The results indicate that the inverse virtual iteration technique based on the NARX model and genetic algorithm provides a useful reference for determining the loading spectrum of construction machinery.