Modern computer vision techniques for forest fire detection face a trade-off between computational efficiency and detection accuracy in complex forest environments. To address this, we propose a lightweight YOLOv11n-based framework optimized for edge deployment. The backbone network integrates a novel C3k2MBNV2 (Cross Stage Partial Bottleneck with 3 convolutions and kernel size 2 MobileNetV2) block to enable efficient fire feature extraction via a compact architecture. We further introduce the SCDown (Spatial-Channel Decoupled Downsampling) block in both the backbone and neck to preserve critical information during downsampling. The neck further incorporates the C3k2WTDC (Cross Stage Partial Bottleneck with 3 convolutions and kernel size 2, combined with Wavelet Transform Depthwise Convolution) block, enhancing contextual understanding with reduced computational overhead. Experiments on a forest fire dataset demonstrate that our model achieves a 53.2% reduction in parameters and 28.6% fewer FLOPs compared to YOLOv11n (You Only Look Once version eleven), along with a 3.3% improvement in mean average precision. These advancements establish an optimal balance between efficiency and accuracy, enabling the proposed framework to attain real-time detection capabilities on resource-constrained edge devices in forest environments. This work provides a practical solution for deploying reliable forest fire detection systems in scenarios demanding low latency and minimal computational resources.

Grain security is a fundamental guarantee for social stability and sustainable development. Accurate monitoring and prediction of overall granary temperature are essential for reducing storage losses and improving warehouse management efficiency. As an integrated system, the temperature evolution of the grain pile is deeply affected by its inherent physical structure and heat transfer pathways. Therefore, a multi-level warehouse–surface–line–point (WSLP) structural modeling approach driven by the physical properties of the grain pile is proposed to extract the joint environmental and spatial characteristics. Building upon the WSLP framework, a dual-channel time-series prediction architecture integrating both long short-term memory (LSTM) and iTransformer through a mutual verification fusion mechanism is developed to enable synchronized temperature forecasting across different regions of the grain piles. Experiments are conducted using real granary data from Shandong, China. The results demonstrate that the proposed model achieves more than 30% improvement over baseline methods in terms of MAE and RMSE. Moreover, the WSLP-LSTM–iTransformer framework significantly improves prediction accuracy in complex warehouse environments and enhances the interpretability and applicability of deep learning models for grain condition forecasting by incorporating real environmental characteristics.

The surface temperature of grain piles is sensitive to environmental fluctuations and exhibits nonlinear, multi-scale temporal patterns, making accurate prediction crucial for grain storage risk early warning. This paper proposes a decomposition–reconstruction prediction method integrating Sample Entropy (SampEn), variational mode decomposition (VMD), and a variant Long Short-Term Memory network (vLSTM). SampEn determines the optimal decomposition parameters, VMD extracts intrinsic mode functions (IMFs), and vLSTM, with peephole connections and coupled gates, conducts synchronous multi-IMF prediction. To explicitly account for environmental influences, a support vector regression (SVR) model driven by dew point temperature and vapor pressure deficit is employed to estimate the surface temperature variation $\mathrm{\Delta }T$. This component enhances the adaptability of the framework to dynamic storage conditions. The environment-derived $\mathrm{\Delta }T$ is then integrated with the VMD-SampEn-vLSTM output to obtain the final forecast. Experiments on real-granary data from Liaoning, China demonstrate that the proposed method reduces mean absolute error (MAE) and root mean square error (RMSE) by 25% and 14%, respectively, compared with baseline models, thus achieving a significant improvement in prediction performance. This integration of data-driven prediction with environmental adjustment significantly improves forecasting accuracy and robustness.