基于粒子群-拟牛顿混合算法的管道机器人定位  

Pipeline robot localization system based on PSO-BFGS hybrid algorithm

作　　者：魏明生[1,2] 童敏明[2] 訾斌[3] 唐守锋[2] 

机构地区：[1]江苏师范大学物理与电子工程学院,徐州221116 [2]中国矿业大学信息与电气工程学院,徐州221008 [3]中国矿业大学机电工程学院,徐州221008

出　　处：《仪器仪表学报》2012年第11期2594-2600,共7页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(50905179;51275515)资助项目

摘　　要：针对管道机器人定位问题,通过磁偶极子理论建立了以极低频电磁脉冲为检测手段的定位模型,推导出了定位模型的非线性方程组。在此基础上设计了基于粒子群优化-拟牛顿混合算法用于定位方程组的求解,提高了管道机器人的定位精度,避免了粒子群优化算法局部精细搜索能力差和拟牛顿算法对初值敏感的不足。数值仿真与实验结果验证了管道机器人定位模型和数值算法的有效性和可行性,实验定位平均误差满足管道机器人工作中定位的需要,采用改进的混合算法后,平均定位误差在x、y、z三个方向上分别降低了4.19 cm、3.81 cm、4.65 cm,提高了定位精度。In order to locate robot in pipeline, a localization model was established using extremely low frequency electromagnetic pulse as the detection means based on magnetic dipole theory; and the nonlinear equation set of the localization model was derived. Based on above, a particle swarm optimization （PSO）-quasi-Newton （BFGS） hybrid algorithm was designed to solve the nonlinear equation set, which could improve the positioning precision of pipeline robot and avoid the deficiencies of the poor local fine searching ability of PSO algorithm and the low sensitivity to initial value of quasi-Newton algorithm. Numerical simulation and experimental results verify the effectiveness and feasibility of the pipeline robot positioning model and the hybrid algorithm. The experimental average position errors can meet the requirements of non-contact, online, high precision and rapid localization for pipeline robot. Using the improved hybrid algorithm, the average position errors decrease by 4.19 cm, 3.81 cm and 4.65 cm in x, y and z directions, respectively; and the positioning precision is improved.

关 键 词：管道机器人 磁偶极子模型 粒子群优化-拟牛顿混合算法 定位 

分 类 号：TP242.3[自动化与计算机技术—检测技术与自动化装置] TM937[自动化与计算机技术—控制科学与工程]