2417 results about "Tubular network" patented technology

The invention discloses an underground pipe network leakage detection method fusing big data analysis and machine learning. The underground pipe network leakage detection method comprises the step of deploying an acoustic sensor, a pressure sensor and a flow sensor at preset positions of an underground pipe network. Performing noise reduction and standardization processing on the acquired operation data, extracting frequency domain, energy and time sequence characteristics of acoustic signals to generate acoustic characteristic vectors, and calculating initial space coordinates of a leakage position based on time difference of arrival of the acoustic signals and sensor arrangement; and generating a pressure feature vector, and analyzing the spatial-temporal correlation of the flow data to generate a flow feature vector. And fusing acoustic, pressure and flow feature vectors to form a multi-dimensional feature vector, inputting the multi-dimensional feature vector and the initial space coordinates into a pre-trained cascade deep learning model, and outputting the leakage probability and the leakage position of underground pipe network leakage. And when the leakage probability exceeds a preset threshold value, triggering an alarm mechanism. According to the invention, efficient detection and accurate positioning of underground pipe network leakage are realized, and high real-time performance and positioning accuracy are achieved.

The invention provides a digital twinborn-based park pipe network monitoring and early warning method and system, and relates to the technical field of pipe network monitoring and early warning, and the method comprises the steps: constructing a park pipe network domain knowledge graph, collecting real-time monitoring data and historical operation and maintenance data, and forming a knowledge enhanced pipe network data set; extracting pipe network performance attenuation characteristics by using a bidirectional gating recurrent neural network, and deeply fusing the pipe network performance attenuation characteristics with expert empirical rules in the domain knowledge graph to construct a performance attenuation analysis model; a performance degradation curve and remaining useful life are calculated based on the model, potential fault risks are predicted, and maintenance suggestions are generated. According to the invention, precise evaluation of pipe network performance degradation, early warning of fault risks and intelligent generation of maintenance suggestions are realized, and the intelligent level and reliability of operation and maintenance of the park pipe network are improved.