测量异常下基于信任融合滤波的煤矿机器人定位方法  

Coal Mine Robot Positioning Method Based on Trust Fusion Filtering under Measurement Anomalies

作　　者：王坦 朱洪波 Wang Tan;Zhu Hongbo(College of Electrical and Information Engineering,Anhui University of Science and Technology,Huainan 232001,China)

机构地区：[1]安徽理工大学电气与信息工程学院,安徽淮南232001

出　　处：《煤矿机械》2024年第11期197-200,共4页Coal Mine Machinery

基　　金：国家自然科学基金项目(62003001);安徽高校自然科学研究项目重大项目(2023AH040157)。

摘　　要：针对矿井恶劣环境中无线传感网络的部分传感器节点出现测量异常时的煤矿机器人定位问题,构建了一种基于信任机制的分布式无迹卡尔曼滤波(DUKF)融合算法,旨在提高测量异常下对煤矿机器人定位的精确性。首先,各传感器节点根据采集的局部多源测量信息实时更新,获得煤矿机器人状态的局部后验估计。其次,基于K-means降维两簇聚类的信任机制将获得的估计信息分为信任估计信息和非信任估计信息,忽略非信任估计信息而保留信任估计信息。最后,通过融合信任估计信息进行下一时刻的预测更新。仿真结果表明,该算法在面对传感器测量异常时可以实现对煤矿机器人的精准定位。Aiming at the coal mine robot positioning problem when some sensor nodes of the wirelesssensor networks in the harsh environment of the mine have measurement anomalies,a distributed unscented Kalman filter(DUKF)fusion algorithm based on the trust mechanism was constructed,so as to improve the positioning accuracyof the coal mine robot under measurement anomalies.First,each sensor node updates in real time based on the collected local multi-source measurement information to obtain local posterior estimations of the coal mine robot states.Secondly,the trust mechanism based on K-means dimensionality reduction two-cluster clustering divides the obtained estimations information into trust estimation information and non-trust estimation information,ignores the non-trust estimations information and retains the trust estimations information.Finally,the prediction is updated at the next moment by fusing the trust estimations information.The simulation results show that this algorithm can achieve precise positioning of the coal mine robot when facing sensor measurement anomalies.

关 键 词：煤矿机器人 DUKF 聚类融合 测量异常 精准定位 状态估计 

分 类 号：TD655[矿业工程—矿山机电]