摘要
探讨了输油管道的动态压力信号存在混沌特性的可能性.以动态压力变送器的管道运行信号为研究对象,对一维的动态压力时间序列进行相空间重构,求得最大Lyapunov指数,验证了输油管道的动态压力信号确实存在混沌的特性.提出了一种利用混沌理论中的关联维数对管道泄漏进行识别的方法,根据关联维数所属数值区间对不同类型的动态压力信号进行区分.在兰成渝成品油管道的现场实验结果表明,该方法能够有效地识别管道泄漏事件,其总体识别正确率达到93.3%,具有较强的工况识别能力.
The chaos analysis theory was used to discuss whether the dynamic pressure signal of oil pipeline has chaotic characteristics. With the normal running pipeline signal of dynamic pressure transmitter as research object, the one-dimensional dynamic pressure time series were reconstructed in the high-dimensional phase space, and then the largest Lyapunov exponent was calculated to validate the existence of chaotic characteristics of the dynamic pressure signal. A recognition method for oil pipeline leak was proposed based on the correlation dimension of chaos theory to distinguish different types of dynamic pressure signals by the value range to which the correlation dimension belongs. The experimental data obtained at Lanchengyu products pipeline are used to evaluate the method and the overall recognition accuracy rate achieves 93.3% , which proves its high accuracy in the recognition of leak events and condition.
出处
《纳米技术与精密工程》
EI
CAS
CSCD
2009年第4期337-341,共5页
Nanotechnology and Precision Engineering
基金
国家自然科学基金重点资助项目(60534050)
天津市自然科学基金资助项目(06YFJMJC02000)
关键词
管道
泄漏检测
动态压力信号
混沌
关联维数
pipeline
leak detection
dynamic pressure signal
chaos
correlation dimension
