The safe operation of marine diesel engines (MDEs) is an important safeguard for ships and engine crews at sea. In this paper, a combined neural network prediction model (PCA-CNN-BiLSTM) is proposed for the problem of multi-parameter prediction and fault warning for MDEs. PCA is able to reduce the data dimensions and diminish the redundant information in the data, which helps to improve the training efficiency and generalization ability of the model. CNN can effectively extract spatial features from data, assisting in capturing local patterns and regularities in signals. BiLSTM works to process time series data and capture the temporal dependence in the data, enabling prediction of the failure conditions of MDE, condition monitoring, and prediction of a wide range of thermal parameters with more accuracy. We propose a standardized Euclidean distance-based diesel engine fault warning threshold setting method for ships combined with the standard deviation index threshold to set the diesel engine fault warning threshold. Combined with experimental verification, the method can achieve real-time monitoring of diesel engine operating condition and abnormal condition warning and realize diesel engine health condition assessment and rapid fault detection function.

Domain adaptation techniques have effectively tackled fault diagnosis under varying operational conditions. Many existing studies presume that machine health states remain consistent between training and testing data. However, in real-world scenarios, fault modes during testing are often unpredictable, introducing unknown faults that challenge the effectiveness of domain adaptation-based fault diagnosis methods. To address these challenges, this paper proposes a Deep Open Set Domain Adaptation Network (DODAN). Firstly, a feature extraction module based on multi-scale depthwise separable convolutions is constructed for discriminative feature extraction. To improve the model’s adaptability, an adversarial training strategy is implemented to learn generalized features that are resilient to unknown domain shifts. Additionally, an outlier detection module is employed to determine the optimal decision boundaries for each class representation space, enabling the classification of known fault modes and the identification of unknown ones. Extensive diagnostic experiments on two marine machinery datasets validate the effectiveness of the proposed method. Furthermore, ablation studies verify the efficacy of the proposed modules and strategies, highlighting significant potential for practical applications.

Magnetoresistance, the change in resistance with applied magnetic fields, is crucial to the magnetic sensor technology. Linear magnetoresistance has been intensively studied in semimetals and semiconductors. However, the air-stable oxides with a large linear magnetoresistance are highly desirable but remain to be fully explored. In this paper, we report the direct observation of linear magnetoresistance in polycrystalline MoO₂ without any sign of saturation up to 7 T under 50 K. Interestingly, the linear magnetoresistance reaches as large as 1500% under 7 T at 2 K. The linear field dependence is in great contrast to the parabolic behavior observed in single-crystal MoO₂, probably due to phonon scattering near the grain boundaries. Our results pave the way to comprehending magneto-transport behavior in oxides and their potential applications in magnetic sensors.