In modern manufacturing processes, requirements for automatic fault diagnosis has been growing increasingly as it plays a vital important role for the reliability and safety of industry facilities. Rolling bearing systems represent a critical part in most of the industrial applications. In view of the strong environmental noise in the working environment of rolling bearing, its vibration signals have non-stationary and nonlinear characteristics and those features are difficult to be extracted. In this paper, we proposed a new intelligent fault diagnosis method for rolling bearing with unlabelled data by using the Convolutional Neural Network (CNN) and Fuzzy C-Means (FCM) clustering algorithm. A CNN is first utilized to automatically extract features from rolling bearing vibration signals. Then the principal component analysis (PCA) technique is used to reduce the dimension of the extracted features, and the first two principal components are selected as the fault feature vectors. Finally, the FCM algorithm is introduced to cluster those rolling bearing data in the derived feature space and identify the different fault types of rolling bearing. The results indicate that the newly proposed fault diagnosis method can achieve higher accuracy than other existing results in literature.

This paper deals with contour tracking control of a networked multi-axis motion control system (NMCS) with repetitive tasks. To improve the control performance of the system under the overall effects of aperiodic timevarying delays and exogenous disturbances, a networkinduced delay is first modeled to be a disturbance. Next, an improved equivalent-input-disturbance (IEID) approach is devised to suppress the overall effect. Then, an iterative learning controller is designed in combination with the IEID estimator to perform individual-axis tracking control. The stability of the system is analyzed using the separation theorem. After that, a cross-coupled controller is used to further improve contour-tracking performance. Finally, simulations and experiments are carried out to verify the feasibility and effectiveness of the new approach.

Low-cost inertial measurement units (IMUs) suffer from low sensitivity and high random walk noise, which makes it challenging to use them directly for dead reckoning. Regular model-based inertial navigation methods require accurate modeling of IMU noise to get better results, while learning-based methods need plentiful datasets. In this article, a novel low-cost IMU dead-reckoning algorithm for wheeled mobile robot is introduced by integrating model-based and learning-based approaches, which inherits the merits of both methods. It achieves the dead reckoning by using invariant extended Kalman filter (InEKF) and IMU error model, and computes the noise parameters of the model with the aid of a deep-learning-based method. Our deep-learning-based strategy is designed to obtain noise-reduced inertial information of robots from low-cost IMU data such that the InEKF can converge. The experimental results show that the proposed method can accurately estimate the attitude, velocity, and position of the wheeled mobile robot, and can compete with vision algorithms. In addition, our proposed method consumes few computational resources to satisfy the needs of low-cost applications.

In object goal navigation tasks, the robot’s understanding of semantic relationships in the environment is a key factor in its ability to localize target objects. Previously, learning-based methods trained robots using 3D scene datasets to learn semantic relationships. However, these approaches perform poorly in new environments with unfamiliar semantic contexts. In this paper, we propose ChatNat which leverages the powerful knowledge summarizing and reasoning capabilities of a Large Language Model (LLM) for zero-shot inference of explicit semantic relationships. These relationships are further integrated into the navigation system for efficient localization of target objects. ChatNav employs a spatial object clustering algorithm to collect semantic clues and designs common-sense-based prompts for interacting with LLM. It then uses a gravity-repulsion model to convert inference results into heuristic factors for robust navigation decision-making. Our approach requires no additional training and can consistently obtain accurate semantic relationships from LLM, making it well-suited for navigating unknown environments. Experimental results demonstrate the outstanding navigation performance of our proposed method on the Gibson and HM3D datasets, surpassing the current state-of-the-art object goal navigation methods.

Recent studies revealed that reaction-diffusion (RD) dynamics can be significantly influenced by the structure of the underlying network. In this paper, a framework is established to study a closely related problem, i.e., to control the proportion of active particles in an RD process by adjusting the structure of the underlying diffusion network. Both distributed and centralized rewiring and reweighting control schemes are proposed for unweighted and weighted networks, respectively. Simulations show that the proportion of active particles can indeed be controlled to a certain extent even when the distributed control mechanism is totally random, while quite high precision can be achieved by centralized control schemes. More interestingly, it is found that the reactants in heterogeneous networks have wider controllable ranges than those in homogeneous networks with similar numbers of nodes and links, if only the weights of links are changed with a fixed bound. Therefore, it is believed that heterogeneous networks fit the changeable environment better, which provides another explanation for some common observations on many heterogeneous real-world networks.

This paper investigates the disturbance rejection control problem of a magnetic levitation system (MLS) in the presence of non-differentiable uncertainties and measurement noise. In practical applications, disturbances often include abrupt or discontinuous changes, making them non-differentiable. This paper models such uncertainties as a total disturbance, further divided into two components: a bounded non-differentiable signal and a differentiable signal. By leveraging this decomposition, we derive an extended state-space model for the MLS, where the differentiable component is treated as an extended state. To estimate both the extended state and the system state, we propose a high-order low-pass filter-based extended state observer (LPFESO). The high-order filter enhances the accuracy of state estimation by filtering the system output. Additionally, an output feedback controller with disturbance compensation is designed based on the internal model principle. The stability of the proposed LPFESO is rigorously demonstrated. Comparative experimental results are presented to validate the superiority and effectiveness of the proposed approach in enhancing the noisy system’s disturbance rejection capability.