基于可视化近超声的狭长管道人员定位系统

Narrow and long pipeline personnel positioning system based on visual near ultrasound

针对工业系统室内管道空间所具有的狭长和弯折等复杂特征,以及现有室内定位技术精度不够、实用性不强的问题,提出一种基于可视化近超声的狭长管道人员定位系统。采用逐帧归一化互相关方法和联合小波导数寻峰的方法确定到达时间。将改进变异率和选择策略的差分进化算法(EA)与极大似然的TDOA算法结合进一步增强定位准确度,将MDS分解与基站定位融合实现基站自标定功能。仿真和实验结果表明,该方法可以很好克服狭长弯折空间以及环境噪声干扰带来的影响,满足狭长管道空间人员定位系统的高精度定位需要。

Aiming at the complex features of the narrow and long pipe space in the industrial system,as well as the insufficient accuracy and practicability of the existing indoor positioning technology,a visual near ultrasound-based narrow and long pipeline personnel positioning system was proposed.The arrival time was determined using the method of frame-by-frame normalized cross-correlation combined with small wave guide number peak finding.The improved EA algorithm and the maximum likelihood TDOA algorithm were combined to further enhance the positioning accuracy,and the MDS decomposition and base station positioning were integrated to realize the base station self-calibration function.The results in simulation and experiment show that the proposed method can overcome non-line-of-sight and environmental noise interference,and meet the high-precision positioning needs of the narrow and long pipeline space personnel management and control system.