21 July 2020 in Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture
In the process of high strength steel turning, tool wear will reduce the surface quality of the workpiece and increase cutting force and cutting temperature. To obtain the fine surface quality and avoid unnecessary loss, it is necessary to monitor the state of tool wear in the dry turning. In this article, the cutting force, vibration signal and surface texture of the machined surface were collected by tool condition monitoring system and signal processing techniques are being used for extracting the time-domain, frequency-domain and time-frequency features of cutting force and vibration. The gray level processing technique is used to extract the features of the gray co-occurrence matrix of the surface texture and found that these features changed simultaneously when the cutting tool broke. After this, an intelligent prediction model of tool wear was built using the support vector regression (SVR) whose kernel function parameters were optimized by the grid search algorithm (GS), the genetic algorithm (GA) and the particle swarm optimization algorithm respectively. The features extracted from the signals and surface texture are used to train the prediction model in MATLAB. It was found that after the surface texture features were fused using the intelligent prediction model on the basis of the features of cutting force and vibration, prediction accuracy of the proposed method is found as 97.32% and 96.72% respectively under the two prediction models of GA-SVR and GS-SVR. Moreover, the intelligent prediction model can not only predict the tool wear under different cutting conditions, but also the different wear stages in a single wear cycle and the absolute error between the predicted value and the actual value is less than 10 μm, the confidence coefficient of prediction curve is around 0.99.
22 July 2026 in Measurement Science and Technology
As the end effector of the intelligent machining system, the accurate monitoring of cutting tool wear condition is directly correlated with machining quality, production efficiency and manufacturing cost. Domain generalization model-driven indirect tool wear monitoring (TWM) does not require access to cutting signal data under new operating conditions, demonstrating broad application prospects. Existing domain generalization models do not consider the raw input sample transferability and the decision-making process of classifier on generalization capability. To overcome these limitations, a novel rational invariance constrained domain generalization model with weighted adversarial training is proposed for TWM under unknown working conditions. First, a novel feature extractor (FE) based on channel-wise and temporal-wise self-attention mechanism is designed to learn features with strong discriminative capability. Then a weighted domain discriminator is proposed by integrating an entropy-based transferability metric to boost the learning capability of the FE. Furthermore, the rational invariance regularization term is employed to constrain the classifier, enabling the classifier to make more rational decisions. The effectiveness and universality of the domain generalization model are enhanced from both the FE and classifier levels. Moreover, the SHapley (SHAP) method is employed to conduct importance analysis on raw multi-channel cutting signals, enhancing the model’s interpretability. Two types of tool wear datasets are executed to validate the domain generalization model’s performance, achieving an average monitoring accuracy of 87.29% across eight domain generalization tasks. On the other hand, comparison and ablation experiments further exhibit the superior applicability of the proposed domain generalization model under unknown machining conditions.
25 February 2022 in Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture
Micro eccentric shaft has important application in many high-tech fields because of its small specific gravity, material, and energy saving. The machined surface texture has an indispensable influence on the surface integrity and the final functional capability. However, due to the micro scale and weak rigidity, it is difficult to characterize the surface texture and evaluate the functionality by traditional quantitative parameters. In order to comprehensively realize the surface texture characterization and functional analysis, a mathematical model is established to simulate the surface texture machined with different cutting tools. Then the surface microscopic profile and functional performance of the surface texture are analyzed by amplitude distribution function (ADF) and bearing area curve (BAC), and the surface texture is also evaluated by fractal dimension, which can avoid the negative effects of scale and resolution. Furthermore, power spectrum density (PSD) is utilized to analyze the relationship between the process dynamic state and geometrical specification of the surface texture. The validity of experimental results shown that the microscopic height distribution of surface machined by flat end milling cutter tends to be more random and there are more microscopic geometric features than that of the ball end milling cutter. The machined surface obtained by the flat end milling cutter has better load bearing, wear resistance, and liquid retention capability.