A decision-level multimodal fusion deep learning strategy is proposed for the effective fault detection of rolling bearings based on long-term fault signals collected from multiple sensors. First, key features are extracted from the multimodal signal set using singular spectrum analysis (SSA), and these features are transformed into a composite dataset that combines short-time Fourier transform (STFT) images and time series data. Based on this, a recursive gated convolutional neural network (RGCNN) is designed to process the STFT image data, while a 1D convolutional neural network (1DCNN) is specifically optimized for training with time series data. Furthermore, decision-level multimodal feature fusion is achieved by applying a weighted average method to integrate the features from different deep learning models, aiming to obtain more comprehensive fault prediction results. The proposed method, multimodal fusion fault detection (MFFD), is validated on the Paderborn and Ottawa rolling bearing datasets, which include various typical faults. Experimental results demonstrate the effectiveness of the proposed approach. Compared to traditional single-modality deep learning models, the proposed method shows significant improvements in fault diagnosis accuracy and generalization capability.

To tackle the issue of the traditional intelligent diagnostic algorithm’s insufficient utilization of correlation characteristics within the time series of fault signals and to meet the challenges of accuracy and computational complexity in rotating machinery fault diagnosis, a novel approach based on a recurrence binary plot (RBP) and a lightweight, deep, separable, dilated convolutional neural network (DSD-CNN) is proposed. Firstly, a recursive encoding method is used to convert the fault vibration signals of rotating machinery into two-dimensional texture images, extracting feature information from the internal structure of the fault signals as the input for the model. Subsequently, leveraging the excellent feature extraction capabilities of a lightweight convolutional neural network embedded with attention modules, the fault diagnosis of rotating machinery is carried out. The experimental results using different datasets demonstrate that the proposed model achieves excellent diagnostic accuracy and computational efficiency. Additionally, compared with other representative fault diagnosis methods, this model shows better anti-noise performance under different noise test data, and it provides a reliable and efficient reference solution for rotating machinery fault-classification tasks.

The fault diagnosis in rotating machinery is crucial for ensuring the safe and dependable operation of intricate mechanical systems. Addressing the limitations inherent in traditional deep learning approaches concerning extended time sequence encoding and subpar generalization capability is paramount. The study utilizes the Gramian Angular Field (GAF) and Squeeze and Excitation (SE) attention mechanisms to alleviate these constraints. GAF enhances feature extraction by emphasizing the angular relationships among adjacent signal points to uncover latent fault characteristics. Simultaneously, through the integration of SE with DenseNet architecture, the network facilitates global information exchange and improves multi-scale fusion, thereby enhancing the precise identification of fault type and location within the signal. Experiments conducted on two datasets achieved accuracies of 100% and 99.85%, respectively, outperforming other methods and models, thereby validating the effectiveness of this study.

In order to fully harness obstacle information in path planning and improve the coordination between global and local path planning, a novel mobile robot path planning method is proposed. The novelty of the proposed path planning strategy lies in its integration of obstacle gap characteristics into both global and local planning processes. Specifically, this method addresses the issues of low search efficiency, excessive redundant points, and poor path quality in the traditional A* algorithm for global path planning by extracting gap grids in the global grid map and incorporating their influence into the heuristic function, thereby guiding the search more effectively. The generated global path is further optimized at gap points to remove redundant nodes. For local path planning, which employs the Dynamic Window Approach (DWA) and often exhibits weak compatibility with global planning and a lack of smoothness through obstacle gaps, this method calculates feasible steering angles based on the distance between the robot and obstacles as well as gap attributes. Additionally, the geometric relationship between global and local paths is established using the Bernstein equation, generating segmented guidance control points for DWA. Simulation experiments demonstrate that the proposed algorithm significantly enhances path efficiency and obstacle avoidance capability in tight space environments, reducing path length by approximately 4.79% and motion time by approximately 15.22% compared to conventional algorithms.