Aiming at multiresonance phenomena excited by bearing compound fault and the masking effect of strong shocks on weak fault shocks, a novel multifault diagnosis method is proposed based on comprehensively characterizing the impulsivity and periodicity of multifault shocks in the shared resonance frequency band. First, the adaptive redundant lifting wavelet packet is presented to decompose the vibration signal into various narrow bands. Then, spike period volatility factor (SPVF) is designed to quantify the multiperiodic impulsive characteristics of narrowband signals. Consequently, the SPVF spectrogram is constructed to highlight the shared resonance bands of multifaults. Finally, the SPVF-cyclic frequency spectrum is developed to synchronously detect the multifault characteristics frequencies. Simulation and experimental analysis showed that the proposed method can simultaneously diagnose multiple bearing faults with good sensitivity to fault-related impulses and robustness to random interferences.

Fault diagnosis under specific working conditions has achieved remarkable success. However, due to variations in working conditions, the assumption that training and test samples are independent and identically distributed is often violated, which makes the diagnostic model brittle under unseen working conditions. To this end, a generic generalization strategy, namely, regularized domain adaptive weight optimization strategy (RDAWOs), is devised for fault diagnosis of rotating machinery. We first design the architecture of a 1-D convolutional neural network. Then, the hyperparameter regularization term and an adaptive pooling layer are designed to control the complexity and improve the adaptability of the overparameterized deep model, respectively. Finally, domain adaptive weight optimization is established to identify the working condition abundant in spurious label-related information and to mine the robust fault knowledge under various working conditions. Obtained results indicate the strong generalization ability for out-of-distribution samples, as well as relatively high diagnostic accuracy of the RDAWOs-based deep model under unseen working conditions.

Multi-scenario & multi-task learning has been widely applied to many recommendation systems in industrial applications, wherein an effective and practical approach is to carry out multi-scenario transfer learning on the basis of the Mixture-of-Expert (MoE) architecture. However, the MoE-based method, which aims to project all information in the same feature space, cannot effectively deal with the complex relationships inherent among various scenarios and tasks, resulting in unsatisfactory performance. To tackle the problem, we propose a Hierarchical information extraction Network (HiNet) for multi-scenario and multi-task recommendation, which achieves hierarchical extraction based on coarse-to-fine knowledge transfer scheme. The multiple extraction layers of the hierarchical network enable the model to enhance the capability of transferring valuable information across scenarios while preserving specific features of scenarios and tasks. Furthermore, a novel scenario-aware attentive network module is proposed to model correlations between scenarios explicitly. Comprehensive experiments conducted on real-world industrial datasets from Meituan Meishi platform demonstrate that HiNet achieves a new state-of-the-art performance and significantly outperforms existing solutions. HiNet is currently fully deployed in two scenarios and has achieved 2.87% and 1.75% order quantity gain respectively.

Damage detection is an inevitable part of composite laminate product verification during both manufacture and maintenance inspections. The article presents a damage characterization method to detect and quantize the fatigue damage in composite laminates. The cross recurrence plot (CRP) of Lamb signal is introduced to assess the dynamic difference between the composite laminates in undamaged and damaged states. The cross recursive quantification analysis features derived from the CRPs are optimally selected and fused into a damage value (DV) based on the support vector data description model. Then, the DVs of diagnostic paths are integrated into a unified damage index for quantizing the fatigue damage of composite laminate subjected to the cyclic loading. The feasibility of the proposed method is validated by the simulation data acquired from the finite element models with different severity degrees of microcracking and the fatigue damage progressing data from NASA prognostics data repository.

Customer response prediction is critical in many industrial applications such as online advertising and recommendations. In particular, the challenge is greater for ride-hailing platforms such as Uber and DiDi, because the response prediction models need to consider historical and real-time event information in the physical environment, such as surrounding traffic and supply and demand conditions. In this paper, we propose to use dynamically constructed heterogeneous graph for each ongoing event to encode the attributes of the event and its surroundings. In addition, we propose a multi-layer graph neural network model to learn the impact of historical actions and the surrounding environment on the current events, and generate an effective event representation to improve the accuracy of the response model. We investigate this framework to two practical applications on the DiDi platform. Offline and online experiments show that the framework can significantly improve prediction performance. The framework has been deployed in the online production environment and serves tens of millions of event prediction requests every day.

An unforeseen fault of the key bearing of production system due to different reasons has the potential to cause an interruption in the entire production line, resulting in economic and production losses. To improve the reliability of industry production, this article presents an intelligent diagnosis method for element rolling bearing based on the integrated interaction relationship of vibration nonlinear features. The nonlinear features of vibration signals are extracted using recurrence quantification analysis (RQA) and regrouped into different subsets of nonredundant features with the same level of discrimination ability through the technology of ReliefF-affinity propagation clustering. The weighted voting variable predictive model class discrimination (WV-VPMCD) is proposed to fully utilize the interaction of RQA features to do intelligent diagnostics for bearing faults. The experimental results have showed that the WV-VPMCD outperformes the conventional intelligent diagnosis methods in terms of accuracy, consistency, stability, and robustness, especially in the case of small number of samples.

The directional adaptive kernel distribution as a new time‐frequency analysis method is proposed to analyse the vibration signal of rotor‐bearing system. This method extends the adaptive kernel distribution to process the constructed complex‐valued signal by defining directional ambiguity function (DAF). Based on the cyclic autocorrelation analysis and the complex‐valued signal decomposition, the DAF is proposed, which is the product of the adaptive optimal kernel function and directional cyclic autocorrelation function. The kernel function taken as a two‐dimension filter is optimised and used to suppress the cross‐terms in the DAF. Then, the directional adaptive kernel distribution is obtained through the positive and inverse Fourier transform on the DAF. The new time‐frequency transform is used to analyse the lateral vibration signals of the rotor and the bearing pedestal operating at oil whirling and whipping speeds. The experimental results verified that the proposed method is effective in the characterisation of the fault instantaneous characteristic frequency, rub‐impact information, instantaneous planar motion and modulation information etc. in oil‐film instability states of rotor‐bearing system.