Gearboxes are widely used in various types of mechanical equipment and have become an essential part of connecting various components in the machine and transmitting power. A fault in the gearbox will cause the entire mechanical equipment to stop working, causing economic losses and safety hazards. Previous fault diagnosis models for gearbox used the vibration signals of the gearbox as input features. They designed a variety of feature extraction modules to learn the information contained in the vibration signals. However, the latent features learned by previous models have poor interpretability and robustness, thus affecting model performance. In addition, the prediction results of previous fault diagnosis models used for gearbox using single-scale features are often less robust. Moreover, the fault diagnosis models used for gearbox are usually end-to-end black box models, which cannot provide interpretive information about the predicted values, and it is difficult to handle the uncertain information in the vibration signals. Therefore, we propose a multi-scale fuzzy variational autoencoder (MFVAE) using a fuzzy neural network for Big Data-based fault diagnosis in gearbox. The Big Data technology can automatically collect vibration signals from sensors and provide high-quality training samples for fault diagnosis models. The MFVAE model first uses two variational autoencoders of different sizes to extract the latent features of the gearbox vibration signal. Subsequently, the MFVAE model fuses low-level latent features and high-level latent features in proportion to form the final head features, which are used to diagnose gearbox faults. Finally, the MFVAE model inputs the head features into the prediction layer to diagnose the fault of the gearbox. The prediction layer comprises a fully connected network and a fuzzy neural network. Experimental results on the real gearbox dataset verify the outstanding performance of the MFVAE model.

Accurate vessel trajectory prediction is vital for forecasting navigation trends, reducing potential risks, and ensuring maritime traffic safety. Using automatic identification system (AIS) data to improve the comprehensive performance of vessel future trajectory prediction is an urgent problem in intelligent transportation systems. In this paper, an improved Pelican optimization algorithm (IPOA) optimized bidirectional gated recurrent unit-attention mechanism (BiGRU-AM) vessel trajectory prediction framework based on AIS data is proposed. This method integrates the optimized hyperparameters of IPOA into the BiGRU-AM backbone network. Utilizing the BiGRU-AM network with tuned parameters to achieve bidirectional information enhancement, and establish an efficient IPOA-BiGRU-AM hybrid prediction model. Furthermore, we propose a shared fusion training mechanism (SFTM) for information exchange between modules. A comprehensive evaluation system covering global error, local error and spatial similarity is developed to address the limitations of single indicator evaluation. The trajectory prediction effectiveness of this method is verified on three datasets. Compared with the current methods, this approach improves accuracy by up to 79.30% across all indicators and reduces the average prediction error by 76.31%. In terms of model complexity, our model’s running speed has improved by 49.73% on average, and has a 1.5 times computational advantage compared to baseline models. The experimental results proved the superiority of the proposed method.

The extreme environment refers to the abnormal temperature, pressure, or vibration in the environment within a certain period of time, which will cause the fault of bearing equipment. Bearing fault diagnosis model can accurately identify the health status of bearing equipment, which can deal with the influence of extreme environments on the normal operation of bearings in a timely manner. However, current bearing fault diagnosis models have the following challenge: the sample size of faulty data is too small, which makes the parameters in the bearing fault diagnosis model unable to be effectively learned. Therefore, in order to solve the above issue in the field of bearing fault diagnosis, we draw on the siamese network and convolutional autoencoder, and propose a real-time bearing fault diagnosis model based on siamese convolutional autoencoder (RBFDSCA) in this work. First, we use an Industrial Internet of Things (IIoT) platform to collect, store and analyze bearing data. Second, to cope with the challenge of the small sample size of faulty data, RBFDSCA model constructs a siamese convolutional autoencoder. The siamese convolutional autoencoder contains a positive feature extraction network, a negative feature extraction network, and a prediction network. The four evaluation metrics of RBFDSCA model on the real bearing data set are 0.9638, 0.9640, 0.9641, and 0.9639, respectively, which verifies its excellent performance.

Federated Learning (FL) has been recognized as a transformative approach in vehicular networks, enabling collaborative training between vehicles and preserving data privacy. However, the high mobility and dynamic topology changes inherent in vehicular environments pose significant challenges, primarily due to the increased communication overhead associated with exchanging model parameters. To mitigate these issues, a novel framework for decentralized wireless FL with soft clustering and 1-bit compressed sensing (SC1BCS-WFL) is proposed in this article. The framework considers vehicle position, vehicle attributes, vehicle speed, and model cosine similarity when grouping vehicles. It utilizes an adaptive threshold mechanism based on 1-bit compression to reduce uplink transmission load while maintaining FL performance. In addition, an early stopping strategy is incorporated into the proposed framework to avoid unnecessary waste of computational and communication resources. Simulation results show that the SC1BCS-WFL framework can enhance the efficiency of federated learning in vehicular settings, particularly with non-independent and identically distributed data. Simulation results also validate the framework’s ability to reduce communication overhead while achieving high model accuracy, indicating its suitability for distributed Internet of Vehicles (IoV) scenarios and contributing to the development of smarter and more efficient IoV applications.

Traffic flow prediction is the key to transportation safety and efficiency. The advance in machine learning and deep learning has promoted the development of intelligent transportation systems. For example, the emergence of Deep Learning as a Service (DLaaS) has benefitted researchers a lot in dealing with large scale dataset and complex deep learning algorithms. In traffic forecasting, despite the success of deep learning-based models, there are still shortcomings, such as inadequate use of temporal and spatial traffic information, and indirect modeling of dependencies in traffic data. To address these challenges, we learn the transportation network in the form of a graph, and use graph wavelet as a key component to extract well-positioned features from the graph based on the transportation network. Compared with graph convolution, graph wavelets are very flexible and do not need to specify adjacent regions in the topological graph structure for feature extraction. At the same time, we propose to combine the multi-information fusion traffic control and guidance collaborative neural network and the results obtained are better than the benchmark algorithms. The results by comparison with several baseline methods show that our proposed method can outperform all the baseline methods.

Using traffic data to accurately predict the traffic flow at a certain time in the future can alleviate problems such as traffic congestion, which plays an important role in the healthy transportation and economic development of cities. However, current traffic flow prediction models rely on human experience and only consider the advantages of single machine learning model. Therefore, in this work, we propose a multi-layer model based on transformer and deep learning for traffic flow prediction (MTDLTFP). The MTDLTFP model first draws on the idea of transformer model, which uses multiple encoders and decoders to perform feature extraction on the initial traffic data without human experience. In addition, in the prediction stage, the MTDLTFP model using deep learning technology, which input the hidden features into the convolutional neural network (CNN) and multi-layer feedforward neural network (MFNN) to obtain the prediction score respectively. The CNN model can captures the correlation information between the hidden features, and the MFNN can captures the nonlinear relationship between the features. Finally, we use a linear model to combine the two prediction scores, which can make the final prediction value take into account the common advantages of both models. Multiple experimental results on two real datasets demonstrate the effectiveness of the MTDLTFP model. The experimental results on the$WorkDay$dataset are as follows, with the RMSE value of 0.191, MAE value of 0.165. The experimental results on the$HoliDay$dataset are as follows, with RMSE value of 0.227, MAE value of 0.192.

Bearings are the core component of mechanical equipment. The health status of bearings is the key to the stable operation of the system. Bearing fault diagnosis model can discover damaged bearings in time, which has a large economic value for enterprises. The previous bearings fault diagnosis model suffers from problems such as small fault data and unrepresentative features, which leads to poor model generalization performance. Therefore, in this work, we propose a masked one-dimensional convolutional autoencoder (MOCAE) for bearing fault diagnosis based on digital twin enabled industrial internet of things (IIoT). The model monitors the bearing data using a set of IIoT platforms. The digital twin technology is used to build a digital twin model of the bearing device, and the parameters of the digital twin model are trained by the fault data obtained from the IIoT platform. The trained digital twin model can then simulate whether the bearing is faulty. In this digital twin model, MOCAE model is proposed for diagnosing faulty bearing signals. The MOCAE model first extracts the features from the time series signal of the bearing using a one-dimensional convolutional autoencoder, which can enhance the reconstruction ability of hidden features to make them more representative. Next, the MOCAE model automatically extracts the feature information contained in the time series signal data by self-training in order to reduce the dependence on the labeled data. The comprehensive experimental results on real bearing datasets show the superiority of the MOCAE model.

Deep learning methods have shown significant performance in medical image analysis tasks. However, they generally act like ”black box” without explanations in both feature extraction and decision processes, leading to lack of clinical insights and high risk assessments. To aid deep learning in envisioning diseases with visual clues, we propose Representation Group-Disentangling Network (RGD-Net), which can completely disentangle feature space of input X-ray images into several independent feature groups, each corresponding to a specific disease. Taking several semantically related and labeled X-ray images as input, RGD-Net firstly extracts completely group-disentangled representations of diseases through Group-Disentangle Module, which applies group-swap and linking operations to construct latent space by enforcing semantic consistency of attributes. To prevent learning degenerate representations defined as shortcut problem, we further introduce adversarial constricts on mapping from features to diseases, thus avoiding model collapse with former free-form disentanglement. Experiments on chestxray-14 and ChestXpert datasets demonstrate that RGD-Net are effective in predicting diseases with remarkable advantages, which leverage potential factors contributing to different diseases, thus enhancing interpretability in working patterns of deep learning methods.

Intelligent transportation system (ITS) plays an important role in solving today’s transportation problems, and short-term traffic flow prediction is at its core. Deep learning can extract and capture abstract high-order features, and introducing attention mechanism to improve the performance of deep learning algorithm has been verified in many fields. Due to the complexity and randomness of traffic flow, accurate traffic flow prediction is not a simple task. Reasonable use of deep learning to predict traffic flow is of great significance to the whole transportation system. In this paper, the reason of choosing recurrent neural network (RNN) as the basic network for traffic flow prediction is explained. Aiming at the problem of gradient disappearance in practical application, the long short-term memory network (LSTM) is introduced to improve the model, and the model framework, algorithm and training process are described in detail. Attention mechanism is introduced into LSTM-RNN to build a short-term traffic flow prediction model. Applying the proposed model to observed traffic flow data, we found that the proposed model has higher prediction accuracy and model efficiency.

With the advance of artificial intelligence and communication technology in the smart city, various location-based data of users can be collected via location-based social networks (LBSNs). How to make full use of these data for accurate point-of-interest (POI) recommendation is challenging because POI selection is influenced by various factors. In this article, we propose a network representation learning-enhanced multisource information (MSI) fusion model for POI recommendation in the context of LBSNs. The proposed model jointly considers various factors, including user preference, geographical influence, and social influence for a recommendation. Specifically, the social influence is modeled by performing network representation learning methods on the constructed co-visiting user networks so that the hidden complex social relationships among users can be measured automatically. Moreover, considering the significance of user preference and geographical influence, a fusion model is designed to jointly consider user preference, social influence, and geographical influence for POI recommendation. Our method is evaluated based on two publicly available data sets and extensive experimental results demonstrate that the proposed MSI fusion model outperforms several state-of-the-art algorithms for POI recommendation in terms of precision, recall, and F1.

At present, medical images have played a more and more important role in clinical treatment. Lung images provide an important reference for doctors to make a diagnosis. Especially for surgical patients, a tumor can be accurately removed based on the full cognition about its size, position, and quantity. Therefore, computer-aided diagnosis for the analysis and treatment of a lot of lung tumor images is very important. Aiming at complexity and self-adaption of image segmentation in lung tumors, this article proposed a parallel deep learning algorithm with hybrid attention mechanism for image segmentation. First, lung parenchyma was extracted via preprocessing images. Then, images were input into hybrid attention mechanism and densely connected convolutional networks (DenseNet) module, respectively, where hybrid attention mechanism consisted of a spatial attention mechanism and a channel attention mechanism. Finally, four feasible solutions were proposed for the verification through changing the convolution quantity of dense block in DenseNet. The network structure with the better performance was achieved. The experimental results prove the parallel deep learning algorithm with hybrid attention mechanism performed well in image segmentation of lung tumors, and its accuracy can reach 94.61%.

Smart hospitals are important components of smart cities. An intelligent medical system for brain tumor segmentation is required to construct smart hospitals. To achieve intelligent brain tumor segmentation, morphological variety and serious category imbalance must be managed effectively. Conventional deep neural networks have difficulty in predicting high-accuracy segmentation images due to these issues. To solve these problems, we propose using multimodal brain tumor images combined with the UNET and LSTM models to construct a new network structure with a mixed loss function to solve sample imbalance and describe an intelligent segmentation process to identify brain tumors. To verify the practicability of this algorithm, we used the open source Brain Tumor Segmentation Challenge dataset to train and verify the proposed network. We obtained DSCs of 0.91, 0.82, and 0.80; sensitivities of 0.93, 0.85, and 0.82; and specificities of 0.99, 0.99, and 0.98 in three tumor regions, including the whole tumor ( WT ), tumor core ( TC ), and enhanced tumor ( ET ). We also compared the results of the proposed network with those of other brain tumor segmentation methods, and the results showed that the proposed algorithm could segment different tumor lesions more accurately, highlighting its potential application value in the clinical diagnosis of brain tumors.

With the advancement of Internet of Things technology, we are able to collect massive people's trajectory data from various GPS services. These large amounts of trajectory records enable us to better understand human mobility patterns. Meanwhile, we are able to extract social relationships based on these digital records to provide personalized recommendation services, such as points of interests (POI) recommendation and friend recommendation. In this paper, we propose to recommend friends for taxi drivers based on vehicular trajectory records. For this purpose, we propose to construct a vehicular ad hoc network based on co-occurrence phenomenon. Furthermore, we take advantages of the network representation learning technique on the vehicular ad hoc network for learning driver vectors. Finally, potential friends are recommended based on the similarity of driver vectors. Extensive experimental results on two real-world datasets demonstrate that our proposed method has the best performance on friend recommendation compared with several state-of-the-art methods. To the best of our knowledge, this is the first attempt to recommend friends for taxi drivers based on vehicular ad hoc network representation learning.

Bogies are an important component of high-speed trains. The level of mechanical performance of bogies has a major influence on the safety and reliability of high-speed train. Therefore, conducting fault diagnoses on bogies with big data is very important. Fault mechanisms of bogies are very complex, and feature signals are nonobvious. For these reasons, fault information of bogies cannot be effectively extracted using the traditional signal processing method. Therefore, this paper adopted the deep neural network to recognize faults in bogies. The deep neural network offers numerous benefits in this context. Using deep neural networks, fault information in a signal spectrum can be extracted in a selfadaptive method. This technique is free of dependence on extensive signal processing knowledge and diagnostic experience. Compared with the traditional intelligent diagnosis method, the deep neural network can obtain a higher diagnostic accuracy. Additionally, the deep neural network does not depend on the sample size, and it can obtain high diagnostic accuracy even when the sample size is relatively small. It also achieves very high diagnostic accuracy applied to high-speed trains with different speeds and different faults, which shows that the method is extensively applicable. Furthermore, the recognition accuracy rate of the deep neural network under normal conditions can reach 100%. This method provides a new paradigm for fault diagnosis of the high-speed train with big data and plays an important role in this field.