Search

For intelligent condition monitoring (ICM) tasks of machine tools (MTs), physics-based and data-driven models typically suffer from two major challenges that constraint their applicability: first, the complex machining parameters set up along with the incompleteness of physics-based models and second, the limited representation ability of small-scale dataset for data-driven models. Considering that it is impractical for the cases of MTs to obtain sufficient scale and well-balanced dataset due to unaffordable specimen cost and strict manufacturing schedule. Accordingly, this paper proposes a new physics-informed scaling evolutionary transformer network, abbreviated as PIS-ETN, to incorporate prior knowledge into the ICM model. Specifically, it mainly includes three parts. First, a texture digital twin (TDT) model is designed to exploit prior knowledge from machining parameters and semi-observable sensor information. Secondly, a texture knowledge embedding module is designed to enhance representation capability. Thirdly, the Pareto-optimal solution space is adopted for further architecture optimization. The experiments indicate that the designed TDT model can effectively provide rich prior empirical knowledge for the designed scaling lightweight model and accelerate model convergence. The proposed lightweight architecture with its Pareto optimal training strategy shows promising fine-grained texture representation ability.

Condition monitoring of assets is significant to the efficiency and reliability of industrial automation systems. However, the accuracy of condition monitoring results is easily impaired by variational environments and volatile operations, especially for complex automation systems. In this article, an environmentally adaptive and contrastive representation learning method is proposed to address the problem. To suppress the unexpected effects of environmental variations on operating data, a regression model between the operational and environmental variables is developed. The variable regression adjustment is achieved by solving a penalized optimization problem based on spline functions, and the solution is explicitly derived. Then, negative samples and pseudo labels are generated based on the designed pattern of data augmentation, and valid data representations for asset condition monitoring can be obtained by contrastive learning. Moreover, the reference statue of healthy assets is established by kernel density estimation, and control charts are employed for online monitoring with alarm thresholds. Taking wind turbine blades as examples, the remarkable performance of the developed method is demonstrated with real-world measurements from wind farms. Furthermore, comparative analysis with benchmark approaches and ablation study are conducted to reveal the superiority and effectiveness of the proposed method.