Dynamic Sampling for Hydraulic Transient Detection in Pipe Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current water supply systems face challenges in monitoring dynamic hydraulic conditions due to stochastic demand and operational changes, leading to increased pipe stress, fatigue failures, and hydraulic instabilities, which existing monitoring devices fail to capture effectively due to limited sampling frequency and energy constraints.
Innovation Solution
A liquid-flow monitoring device that samples data at high frequencies (64-1024 samples per second) and stores stratified streams of subsampled data, allowing for continuous monitoring and retrospective analysis of hydraulic conditions, including transient events, with a modular design for efficient energy use and data management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If monitoring devices sample at low frequency (15min intervals) to minimize power consumption, then battery life is extended and power consumption is reduced, but dynamic hydraulic conditions and transient events cannot be captured
Solution Approach 1:
The monitoring device dynamically adjusts its sampling frequency based on detected hydraulic conditions. During normal steady-state conditions, sampling occurs at low frequency (15min intervals) to conserve power. When transients or surge events are detected, the device automatically increases sampling frequency to capture the dynamic events, then returns to low-frequency sampling afterward. This dynamic adaptation resolves the contradiction between power consumption and measurement precision.
Solution Approach 2:
The device changes the sampling rate parameter from a fixed low value to a variable value that increases during transient events and decreases during normal operation. This parameter change allows the system to capture dynamic hydraulic conditions when necessary while maintaining low power consumption during stable periods, effectively resolving the contradiction between measurement precision and energy use.
2Measurement precision
If monitoring devices increase sampling frequency to capture transient events, then dynamic hydraulic conditions are captured, but power consumption increases and battery life decreases
Solution Approach 1:
The system employs dynamic sampling where the sampling frequency is adjusted in real-time based on hydraulic conditions. During transient events, high-frequency sampling captures the dynamics, but this is temporary. Between events, the device returns to low-frequency sampling, thereby capturing necessary data while minimizing overall power consumption and preserving battery life.
Solution Approach 2:
The monitoring device uses periodic low-frequency sampling as the baseline operation and temporarily switches to high-frequency sampling only when transients are detected. This periodic action pattern ensures that power-intensive high-frequency sampling occurs only when necessary, resolving the contradiction between capturing transient events and managing power consumption.
3Reliability
If surge monitoring uses gradient detectors and threshold triggers to capture extreme events, then sudden transients are detected, but low amplitude high-frequency oscillations and stochastic hydraulic conditions are omitted
Solution Approach 1:
The monitoring approach segments hydraulic events into different categories: extreme surges detected by gradient detectors, and low-amplitude oscillations captured by continuous high-frequency sampling. By segmenting the detection methodology, the system achieves both reliable extreme event detection and comprehensive capture of stochastic hydraulic conditions that would otherwise be missed.
Solution Approach 2:
The monitoring device performs multiple functions: it uses gradient detectors for extreme surge detection, continuous high-frequency sampling for low-amplitude oscillations, and statistical analysis for stochastic conditions. This multi-functionality allows the system to detect the full spectrum of hydraulic events, resolving the contradiction between reliable extreme event detection and adaptability to various hydraulic conditions.
4Measurement precision
If high-frequency sampling is implemented continuously to capture all hydraulic conditions, then complete data is acquired, but energy sources and data management requirements become excessively large
Solution Approach 1:
Instead of continuous high-frequency sampling, the device applies partial action by using high-frequency sampling only when and where needed (during transient events). The majority of the time, low-frequency sampling is sufficient. This partial application of high-frequency sampling captures the essential dynamic information while avoiding the excessive data volume and energy requirements of continuous high-frequency operation.
Solution Approach 2:
The system discards the approach of continuous high-frequency sampling and recovers by implementing event-triggered high-frequency sampling. By discarding the excessive data collection during normal steady-state conditions and recovering only when transients occur, the system achieves complete capture of hydraulic events while managing data volume and energy requirements sustainably.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A liquid-flow monitoring device (10) arranged to monitor the flow of liquid in a conduit. The device (10) comprises sampling means (20), processing means (30) and data storage means (40). The sampling means is arranged to receive data substantially continuously from a sensor (12) indicative of at least one variable sensed by the sensor (12) indicative of fluid flow in a conduit and to sample the data to produce a stream of sampled data. The processing means (30) is arranged to process the stream of sampled data to extract at least one sub-sampled stream therefrom, the sub-sampled stream comprising a plurality of data sets, each data set being a statistical subset of the stream of sampled data over a respective data-set period. The data storage means (40) is arranged to store the at least one sub-sampled stream.