基于时延的高精度泄漏点超声定向检测方法  被引量:1

High accuracy method of ultrasonic gas leak direction detection based on time delay estimation

在线阅读下载全文

作  者:廖平平[1] 蔡茂林[1] 

机构地区:[1]北京航空航天大学自动化科学与电气工程学院,北京100191

出  处:《北京航空航天大学学报》2013年第3期391-395,共5页Journal of Beijing University of Aeronautics and Astronautics

基  金:国家科技支撑计划资助项目(2011BAF05B04)

摘  要:针对当前气体泄漏点检测定向精度低的情况,提出一种基于时延的高精度超声检测方法.该方法利用3个呈等边三角形分布的超声传感器接收由泄漏点产生的超声波,根据3路信号的相对时延值确定泄漏点的方向.为克服采样间隔对时延估计精度的限制,采用基于三次样条插值的时延估计算法估计信号时延值,并对不同核窗长度下的各100组实验数据的时延值进行误差统计,得到其均方差并与其Cramér-Rao下界比较,发现二者的变化趋势具有良好的一致性.在此基础上研究了时延值均方差,超声传感器间距和泄漏点距离对定向精度的影响.结果表明:定向误差随时延值均方差增大而增大,随超声传感器间距增大而减小,随泄漏点距离增大而增大;定向精度比单超声传感器检测方法提高7~10倍.In order to solve the low accuracy problem in current gas leak detection, a new ultrasonic leak detection method based on time delay estimation (TDE) was proposed. Three ultrasonic sensors arranged in an equilateral triangle and received the ultrasound generated by a gas leak and the leak direction can be deter- mined according to time delays between the outputs of every two sensors. A TDE algorithm based on cubic spline interpolation was adopted to overcome the accuracy limit caused by sampling interval. For each kernel window length, delay estimates of 100 sets of experimental data were obtained and their mean squared errors (MSE) were calculated. Comparison between MSE of experimental data and Cram^r-Rao lower bound showed that their changing tendency was accordant. Influences of MSE, the space between sensors and the distance be- tween leak and sensor on direction accuracy were analyzed. Results show that the direction error increases with MSE, decreases with the space between sensors, and increases with distance between leak and sensor. The di- rection detection accuracy is improved by 7 - 10 times compared with detection method with single ultrasonic sensor.

关 键 词:泄漏检测 超声 时延 三次样条插值 定向精度 

分 类 号:TH412[机械工程—机械制造及自动化] TH865

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象