Fluid Network Sensor Placement via Graph Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fluid networks, such as water and gas distribution systems, face challenges in detecting leaks, corrosion, and contamination issues in a timely manner, leading to significant impacts on infrastructure and health, as seen in events like the Elk River Chemical Spill.
Innovation Solution
A method and system for optimally placing sensors in fluid networks by creating a model with directionally connected nodes and analyzing data using a processor to predict anomalies, allowing for early or real-time detection and preemptive actions, involving the use of sensors to collect data and a processor to analyze it based on a model of the network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensors are placed throughout the fluid network to enable early detection of problems, then detection capability is improved, but device complexity and cost increase
Solution Approach 1:
The fluid network is divided into discrete nodes and segments that can be independently modeled and analyzed. This segmentation allows the system to focus monitoring efforts on specific high-risk segments rather than uniformly deploying sensors throughout the entire network, reducing complexity while maintaining detection capability.
Solution Approach 2:
The patent transitions from physical sensor deployment to a mathematical modeling dimension by representing the fluid network as a graph with nodes and edges. This dimensional shift enables virtual monitoring through data analysis and prediction algorithms, eliminating the need for extensive physical sensor infrastructure while achieving comprehensive detection coverage.
2Reliability
If comprehensive monitoring is implemented to detect leaks and contamination, then reliability is improved, but loss of time for problem discovery decreases (detection is faster)
Solution Approach 1:
The system performs preliminary actions by establishing a comprehensive mathematical model of the fluid network beforehand, incorporating all nodes, edges, and potential failure points. This pre-modeling enables rapid detection and prediction of anomalies without requiring physical sensors to be deployed throughout the network, achieving both high reliability and fast detection time.
Solution Approach 2:
The patent replaces the mechanical sensor-based monitoring system with a computational model-based system. Instead of using physical sensors to detect leaks and contamination, the system uses mathematical models and data analysis to predict and identify anomalies, eliminating the time required for physical sensor deployment while maintaining or improving detection speed.
3Ease of operation
If a mathematical model is created to represent the fluid network, then optimal sensor placement is achieved, but device complexity increases due to modeling requirements
Solution Approach 1:
The patent creates a virtual copy or mathematical representation of the fluid network through graph modeling. This copy includes all the essential characteristics of the physical network (nodes, edges, flow directions) but exists in a simplified mathematical form that is easier to analyze and optimize for sensor placement without the complexity of the physical system.
Data Source
AI summary
Systems and methods are provided for optimally determining sensor or infrastructure placement in a fluid network, for determining an anomaly of interest in the fluid network, and for determining sensor coverage in a fluid network, which are based on a model of the fluid network represented by a directed graph.