The transmission spectrum of apples is affected by the fruit’s size, which leads to poor prediction performance of the soluble solids content (SSC) models built for their different apple sizes. In this paper, three sets of near infrared (NIR) spectra of apples with various apple diameters were collected by applying NIR spectroscopy detection equipment to compare the spectra differences among various apple diameter groups. The NIR spectra of apples were corrected by studying the extinction rates within different apples. The corrected spectra were used to develop a partial least squares prediction model for their soluble solids content. Compared with the prediction model of the soluble solids content of apples without size correction, the R_p of PLSR improved from 0.769 to 0.869 and RMSEP declined from 0.990 to 0.721 in the small fruit diameter group; the R_p of PLSR improved from 0.787 to 0.932 and RMSEP declined from 0.878 to 0.531 in the large fruit diameter group. The proposed apple spectra correction method is effective and can be used to reduce the influence of sample diameter on NIR spectra.

Distracted driving is one of the major factors leading drivers to ignore potential road hazards. In response to the challenges of high computational complexity, limited generalization capacity, and suboptimal detection accuracy in existing deep learning-based detection algorithms, this paper introduces a novel approach called StarDL-YOLO (StarNet-detectlscd-yolo), which leverages an enhanced version of YOLOv8n. Initially, the StarNet integrated into the backbone of YOLOv8n significantly improves the feature extraction capability of the model with remarkable reduction in computational complexity. Subsequently, the Star Block is incorporated into the neck network, forming a C2f-Star module that offers lower computational cost. Additionally, shared convolution is introduced in the detection head to further reduce computational burden and parameter size. Finally, the Wise-Focaler-MPDIoU loss function is proposed to strengthen detection accuracy. The experimental results demonstrate that StarDL-YOLO significantly improves the efficiency of the distracted driving behavior detection, achieving an accuracy of 99.6% on the StateFarm dataset. Moreover, the parameter count of the model is minimized by 56.4%, and its computational load is decreased by 45.1%. Additionally, generalization experiments are performed on the 100-Driver dataset, revealing that the proposed scheme enhances generalization effectiveness compared to YOLOv8n. Therefore, this algorithm significantly reduces computational load while maintaining high reliability and generalization capability.

Wave energy is a promising source of sustainable clean energy, yet its inherent intermittency and irregularity pose challenges for stable grid integration. Accurate forecasting of wave energy power is crucial for reliable grid management. This paper introduces a novel approach that utilizes a Bidirectional Gated Recurrent Unit (BiGRU) network to fit the power matrix, effectively modeling the relationship between wave characteristics and energy output. Leveraging this fitted power matrix, the wave energy converter (WEC) output power is predicted using a model that incorporates a Convolutional Neural Network (CNN), a Bidirectional Long Short-Term Memory (BiLSTM) network, and deformable efficient local attention (DELA), thereby improving the accuracy and robustness of wave energy power prediction. The proposed method employs BiGRU to transform wave parameters into power outputs for various devices, which are subsequently processed by the CNN-BiLSTM-DELA model to forecast future generation. The results indicate that the CNN-BiLSTM-DELA model outperforms BiLSTM, CNN, BP, LSTM, CNN-BiLSTM, and GRU models, achieving the lowest mean squared error (0.0396 W) and mean absolute percentage error (3.7361%), alongside the highest $R^2$ (98.69%), underscoring its exceptional forecasting accuracy. By enhancing power forecasting, the method facilitates effective power generation dispatch, thereby mitigating the adverse effects of randomness on the power grid.

Pesticide residues in crops are harmful to people’s health and have become a focal point that continues to be addressed. In this study, Silver Nanowires (AgNWs) were used as the surface-enhanced Raman spectroscopy (SERS) substrate to detect phosmet and dimethoate residues in oranges and achieved better results. The acquired spectral data were subsequently modeled based on partial least squares methods. For standard solutions of phosmet with concentrations ranging from 1 to 35 mg/l, the modeling effect was better after multiple scattering correction (MSC) preprocessing, where the correlation coefficient (Rp) was 0.967; for standard solutions of dimethoate, the Rp was 0.956 after smoothing and first derivatives preprocessing. For phosmet residues on the skin of navel oranges, which had a concentration of 5 to 35 mg/l, the modeling effect was better after second derivative preprocessing, where the Rp was 0.930 and the root mean square error of prediction (RMSEP) was 3.169 mg/l; for dimethoate residues on the skin of navel oranges, the Rp was 0.957 and the RMSEP was 2.969 mg/l after MSC preprocessing. To determine the accuracy of the modeling, the spectral data at the locations of the characteristic peaks were modeled. After intercepting the spectra of the phosmet-dependent characteristic peak, the modeling effect was better after MSC preprocessing, where the Rp was 0.911 and the RMSEP was 4.317 mg/l. After preprocessing the spectra of dimethoate through the second derivatives, the model had an Rp of 0.853 and an RMSEP of 5.159 mg/l. Analytical modeling results show that phosmet residues on oranges could be detected using SERS coupled with AgNWs, and the high sensitivity and rapidity of this method make it promising for food safety applications.