An enhanced fault diagnosis approach for rolling bearings with composite faults using an optimized Squeeze and Excitation ResNet (SE-ResNet) model is proposed. This method integrates grid search (GS), support vector regression (SVR), ensemble empirical mode decomposition (EEMD), and low-rank multimodal fusion (LMF) to effectively handle the signals of acoustic–vibration fusion. By combining these techniques, the aim is to improve the accuracy and reliability of rolling bearing fault diagnosis. Firstly, improved EEMD combined with GS-SVR and a window function is used for rolling bearing vibration signal decomposition. Singular value methods are used to filter and reconstruct the results. Secondly, Markov transition fields (MTFs) are used to encode vibration signals into 2D images. LMF is used for the fusion of vibration and sound signals. An improved Squeeze and Excitation ResNet50 network is proposed for feature identification and classification of rolling bearing composite fault data. Finally, the method undergoes rigorous testing and evaluation using rolling bearing data. The experimental outcomes demonstrate that, in comparison to traditional neural networks, the enhanced SE-ResNet, integrated with GS-SVR-EEMD and LMF, attains superior diagnostic accuracy. Additionally, the proposed approach can be effectively utilized for diagnosing rolling bearing composite faults.

Due to the large size of the heavy duty machine tool-foundation systems, space temperature difference is high related to thermal error, which affects to system’s accuracy greatly. The recent highly focused deep learning technology could be an alternative in thermal error prediction. In this paper, a thermal prediction model based on a self-organizing deep neural network (DNN) is developed to facilitate accurate-based training for thermal error modeling of heavy-duty machine tool-foundation systems. The proposed model is improved in two ways. Firstly, a dropout self-organizing mechanism for unsupervised training is developed to prevent co-adaptation of the feature detectors. In addition, a regularization enhanced transfer function is proposed to further reduce the less important weights of the process and improve the network feature extraction capability and generalization ability. Furthermore, temperature sensors are used to acquire temperature data from the heavy-duty machine tool and concrete foundation. In this way, sample data of thermal error predictive model are repeatedly collected from the same locations at different times. Finally, accuracy of the thermal error prediction model was validated by thermal error experiments, thus laying the foundation for subsequent studies on thermal error compensation.

Bolted joints are one of the most common fastening methods in engineering applications. To meet the requirements of structural parts, the torque method is often used for controlling the bolted joint performance. However, only a few investigations have been carried out on the conversion efficiency of bolt torque to the tensile force, leading to uncertainty and potential safety hazards during the bolt tightening. In order to study the input torque distribution and overcome problems caused by the Motosh method and experimental investigations, a new energy-based torque distribution model is established in the present study. In the proposed model, numerous affecting parameters, including the connector compression work, effective bearing radius, effective thread contact radius, and spiral angle are considered. Then a parameterized thread mesh model using finite element technology is proposed to analyze the influence of different bolt friction coefficients on the bolt tightening process. Based on 16 types of tightening analyses, it is concluded that as bolt friction coefficient increases, the corresponding torque conversion rate decreases from 14.45% to 7.89%. Compared with the Motosh method, the torque conversion rate obtained by the proposed method is relatively large, which makes the actual pre-tightening force larger than the design value. However, there is still a possibility of bolt failure.

Under environmental excitation and based on observability, an online model to predict the state of heavy-duty machine tool–foundation systems is proposed aimed to address the difficulties of directly measuring machine tool displacement states. The aim of the model is to address the difficulties associated with directly measuring machine tool displacements in real time. In this paper, to accurately obtain contact parameters of the joint surface, three states—elasticity, plasticity, and fracture—of concrete micro-bumps were studied. To obtain the equivalent elastic modulus of the secondary pouring material and reinforced concrete material, a composite foundation constitutive model is proposed to determine the equivalent elastic modulus of the concrete foundation. Surface topography features were reconstructed by truncating the peaks of curves, force balance relationships were defined at the joint surfaces, and a metal–concrete joint contact model based on fractal theory was deduced. Based on the joint contact model, a dynamic model of the heavy-duty machine tool–foundation system was established. The dynamic parameters were detected in real time and used to reconstruct the above dynamic model based on observation theory. Further, an estimation model was established to describe the state of a heavy-duty machine tool–foundation system, and online estimation of the machine tool displacement was realized. Finally, the estimation model was validated using an experimental setup of the heavy-duty machine tool–foundation system that considers joint surface factors. In conclusion, the model provides a theoretical basis for stable online control of heavy-duty machine tools.

Because of the characteristics of heavy-duty machine tools such as large self-weight and heavy load, the working precision and service life of their lathe beds, columns, and other large structural parts are all directly influenced by the foundation. In view of the considerable influence of joint surfaces on system characteristics, this study involved obtaining joint surface parameter values from a microscopic perspective, deriving a static joint surface parameter model from Reynolds equation, adopting fractal theory to develop a bolted joint surface parameter model, and thus completing the embedding of joint surface parameters under uneven loads; a simulation model for a heavy-duty machine tool-foundation system was also devised considering the influence of joint surfaces. To identify the structural micro-deformation status of heavy-duty numerical control machine tool-foundation systems, the authors constructed a fiber grating technology-based experimental platform for detecting the deformation of structural parts, verified the correctness of the above simulation model via experiments, and proposed a method for detecting the deformation of heavy-duty machine tool-foundation systems using fiber grating technology. Based on the above simulation model, the influence of reinforced layer position, foundation outline specifications and soil properties around the foundation on the bearing deformation of heavy-duty machine tool-foundation systems were studied, and some guidelines on the construction of concrete foundations were formulated. This model and the related detection method laid a theoretical foundation for guiding the design and optimization of heavy-duty machine tool structure and foundation.