This study presents an advanced conveyor belt inspection system that integrates laser scanning technology with deep learning to achieve high-precision micro-damage detection. The system is designed to overcome two major challenges in industrial inspection: ensuring reliable operation under harsh environmental conditions such as dust, vibration, and low illumination, and enabling the early identification of subtle belt defects that are often overlooked by conventional approaches. To this end, an innovative hardware platform was developed, combining laser-based illumination with high-speed imaging to enhance defect visibility. On the algorithmic side, an improved You Only Look Once (YOLO)v7 model was proposed, incorporating four enhancements-funnel rectified linear unit (F-ReLU) activation, spatial pyramid pooling fast cross stage partial convolution (SPPFCSPC) module, efficient intersection over union (EIoU) loss function, and squeeze-and-excitation (SE-Net) attention mechanism. A comprehensive dataset was constructed from both laboratory test benches and field-collected samples in a coal coking plant, ensuring robustness across diverse operating conditions. Experimental results demonstrate that the improved YOLOv7 achieves a mean average precision (mAP@0.5) of 96.6%, significantly surpassing the baseline YOLOv7 (90.7%) and outperforming recent detectors such as DETR and RT-DETR in both accuracy and efficiency. Moreover, long-term reliability tests, including 72-hour continuous operation and low-light industrial deployment, validated the system’s stability and adaptability. These contributions highlight not only the technical novelty of combining laser-enhanced imaging with deep learning, but also the practical value for predictive maintenance, safe production, and sustainable operation. This work offers a robust and scalable inspection framework, advancing the digitalization and intelligent automation of conveyor systems in line with Industry 4.0 principles.

Industrial integrated multi-energy systems must coordinate electricity, heat, gas, storage and flexible demand while limiting carbon emissions. Renewable uncertainty complicates this dispatch because wind and photovoltaic profiles are stochastic and correlated. This study develops a scenario-informed dispatch framework in which a multicollinearity-reduced conditional variational autoencoder (MCLRCVAE) generates joint wind–photovoltaic scenarios, a probabilistic-distance method retains representative scenarios with probabilities, and the reduced set is embedded directly in a low-carbon dispatch model. The multicollinearity-reduction step removes redundant linear dependence before latent-variable learning, while the conditional variational structure produces scenario distributions rather than a single forecast. To ensure that every numerical result is independently reproducible and to address concerns about parameter provenance, the numerical study is built entirely on transparent, fully public data sources with documented assumptions. On this benchmark, MCLRCVAE attains the lowest Wasserstein distance among the tested generators, indicating the closest match to the empirical joint distribution, while different baselines lead on individual point-error metrics. Scenario-based dispatch reduces total cost and expected CO₂ emissions relative to deterministic dispatch under the benchmark assumptions. The results support multicollinearity-aware scenario generation for low-carbon dispatch.