变电站内SF_(6)分解物检测追踪系统设计

Design of SF_(6)decomposition gas plume detection and tracking system in substation

六氟化硫(SF_(6))气体分解与泄漏现象日趋严重,给电气设备的绝缘性能带来了严重的负面影响,甚至对电气设备在运行过程中的安全性产生了极大的威胁。为快速有效地对SF_(6)气体的分解物(二氧化硫SO_(2)、硫化氢H_(2)S等)进行检测并定位泄露源,该研究基于高级精简指令集计算机(advanced RISC machine,ARM)控制器设计了可以检测以上两种主要有毒气体的电子鼻系统。接着以模拟退火算法为基础,提出气体烟羽追踪方法,旨在解决气体泄露源定位问题。气体浓度检测实验表明,相比MX6商业电子鼻,该文自制的电子鼻系统的气体浓度检测相对误差在0.7×10^(-6)以内,可以有效地检测以上两种有毒气体。此外,考虑到以上气体具有毒性,以乙醇气体为气味源进行了气体烟羽追踪实验。结果表明,该文所提出算法的有效性和实时性较好,平均定位误差只有0.63 m。以上实验证明该设计的电子鼻及模拟退火算法能够满足移动机器人进行SF_(6)分解气体泄漏检测的应用需求。

The decomposition and leakage of sulfur hexafluoride(SF_(6))gas is becoming more and more serious in recent years,which will reduce the insulation performance of electrical equipment and endanger its safe operation.To quickly and effectively detect the decomposition products(sulfur dioxide SO_(2),hydrogen sulfide H_(2)S)of SF_(6)gas and locate the leak source,this research designed an electronic nose system(e-nose)based on the Advanced RISC Machine(ARM)controller that can detect these two toxic gases.Subsequently,aiming at the problem of locating the source of gas leak,this research designed a gas plume tracking method based on simulated annealing algorithm.Gas concentration detection experiments show that,compared with the MX6 commercial e-nose,the relative error of the gas concentration detection of the self-made e-nose in this paper is within 0.7×10^(-6),which can effectively detect the above two toxic gases.In addition,considering that the above gases are toxic,this paper uses ethanol gas as the odor source to conduct gas plume tracking experiments.The proposed method has high effectiveness and timeliness,shown by results,with an average positioning error of 0.63 m.The above experiments demonstrate that the designed e-nose and simulated annealing algorithms can meet the application requirements of mobile robots for SF_(6)decomposition gas leak detection.