计算机科学
滤波器(信号处理)
卷积(计算机科学)
判别式
噪音(视频)
人工智能
核(代数)
财产(哲学)
联轴节(管道)
计算机视觉
瞬态(计算机编程)
振动
融合
鉴定(生物学)
模式识别(心理学)
卷积神经网络
干扰(通信)
编码器
棱锥(几何)
自动化
桥(图论)
接头(建筑物)
锐化
交通标志识别
人工神经网络
传感器融合
声学
双线性插值
混叠
作者
Bingrui Wang,Wang Sheng yi,Haiyan Jiang,Zhongye Du,Qinghui Song,Qingjun Song
标识
DOI:10.1088/1361-6501/ae739e
摘要
Abstract Accurate and real-time coal-gangue identification is quintessential for advancing the automation of top-coal caving (TCC). To tackle the persistent challenges of intense underground background noise and the non-stationary nature of impact signals, this paper proposes a cross-modal spatiotemporal gated fusion method. Specifically, a mixture-of-expert-based Spectrum-Aware Dynamic Vibration Encoder is developed to accommodate the transient impact characteristics of vibration signals, achieving sample-level adaptive kernel aggregation via variational mode decomposition-decomposed multi-channel inputs. Concurrently, to resolve the physical property coupling in acoustic features, a Time–Frequency Decoupled Three-Stream Audio Pyramid Network is designed to reconstruct coal-gangue physical attributes across three orthogonal dimensions, significantly enhancing fine-grained discrimination. To bridge the semantic gap between heterogeneous modalities, we further introduce a bidirectional symmetric multi-head cross-modal attention module and a dual-dimensional spatiotemporal gating mechanism, which adaptively filter high-signal-to-noise ratio features through a dynamic competition strategy. Experimental results on a simulated TCC platform demonstrate that the proposed method achieves a recognition accuracy of 96.94%. This research provides a low-cost, high-reliability for online coal-gangue monitoring.
科研通智能强力驱动
Strongly Powered by AbleSci AI