机械与电子2025，Vol.43Issue(12)：24-30,7.
北斗和GRU神经网络算法的机械作业风险估测
Risk Estimation of Mechanical Operations Using Beidou and GRU Neural Network Algorithms
摘要
Abstract
There are various potential dangers in mechanical operations.Therefore,a mechanical oper-ation risk estimation method based on Beidou and GRU neural network algorithms is proposed.The coor-dinates of the mechanical operation estimation points in the geodetic right-angle coordinate system are converted to the station-centered coordinate system of the Beidou positioning system,and the mechanical operation parameters are obtained by using Beidou positioning technology.The risk features of mechanical operations are extracted by using lightweight deep learning.The extracted mechanical operation risk fea-tures are taken as the input of the GRU neural network.Through the gating mechanism,the evolution law of the mechanical operation risk features over time is learned and classified.Based on this,a mapping func-tion for the risk characteristics existing in the current data is established,and risk estimation is achieved ac-cording to the mapping function.The experimental results show that the risk estimation accuracy of the research method is high,and the positioning residual of the Beidou technology is always less than±3 cm.关键词
北斗定位技术/GRU神经网络算法/机械作业/风险估测/轻量级深度学习Key words
Beidou positioning technology/GRU neural network algorithm/mechanical operations/measuring risks/lightweight deep learning分类
信息技术与安全科学引用本文复制引用
舒晴川,罗阳,赵必舜..北斗和GRU神经网络算法的机械作业风险估测[J].机械与电子,2025,43(12):24-30,7.基金项目
2023年南方电网佛山供电局科技项目(030600KC23070020) （030600KC23070020）
