Traffic Flow Model Corrects DAS Missed Detections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing traffic monitoring systems face challenges in accurately monitoring traffic flow on multi-lane roads due to masking effects and limited resolution, particularly with distributed acoustic sensing (DAS) systems, which struggle to distinctly detect vehicles further away from the sensing fiber, leading to missed detections and inaccurate flow measurements.
Innovation Solution
A method and apparatus using a traffic flow model that accounts for the likelihood of vehicle masking and detection probabilities to correct for missed detections, integrating with DAS systems to provide an estimate of true traffic flow by comparing detected and predicted flow properties, potentially with additional monitoring systems like video-based systems for calibration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If distributed acoustic sensing is used to monitor traffic flow, then the system provides continuous monitoring capability along the road, but vehicles further away from the sensing fiber are acoustically masked by vehicles in closer lanes, leading to missed detections and reduced measurement accuracy
Solution Approach 1:
The patent introduces a traffic flow model as an intermediary that processes the detected vehicle data and compensates for missed detections. The model uses detection probability calculations and traffic flow relationships to estimate the true traffic flow, effectively mediating between the imperfect sensor data and the desired accurate measurement.
Solution Approach 2:
The system implements feedback by using the traffic flow model to continuously adjust and correct the traffic flow estimates based on the detected vehicles and expected detection probabilities. The model feeds back corrections to account for masked vehicles, improving the overall measurement accuracy over time.
2Measurement precision
If multiple sensors are deployed across different lanes to improve detection accuracy, then vehicle detection accuracy improves, but the system complexity and installation cost increase significantly
Solution Approach 1:
The patent makes the single optical fiber sensor multi-functional by using it to detect vehicles across multiple lanes simultaneously. Instead of requiring separate sensors for each lane, the distributed acoustic sensing system uses one fiber to monitor traffic flow in all lanes, with the traffic flow model compensating for the masking effect.
Solution Approach 2:
The patent replaces the mechanical approach of deploying multiple physical sensors across different lanes with a computational approach. The traffic flow model and detection probability algorithms substitute for the physical presence of multiple sensors, achieving accurate multi-lane monitoring through data processing rather than hardware multiplication.
3Difficulty of detecting and measuring
If pressure or strain based sensors are embedded in the road to detect vehicle weight and type, then vehicle classification capability improves, but the sensors are subject to severe wear and tear requiring robust construction and regular maintenance
Solution Approach 1:
The patent replaces mechanical pressure and strain sensors with distributed acoustic sensing using optical fiber. Instead of physical sensors that contact the road and suffer from wear, the system uses acoustic wave detection through the ground, eliminating mechanical wear and improving reliability while maintaining vehicle classification capability through acoustic signature analysis.
4Measurement precision
If induction loops are embedded within the carriageway to monitor traffic volume and flow, then the sensors provide reliable detection, but installation requires cutting into the road and maintenance is much harder
Solution Approach 1:
The patent replaces embedded induction loops with distributed acoustic sensing. Instead of cutting roads to install electrical loops, the system uses optical fiber that can be laid alongside or on the road surface, detecting traffic through acoustic emissions. This eliminates the need for road cutting and simplifies both installation and maintenance.
5Speed
If radar or lidar systems are mounted on overhead gantries to transmit pulses and detect vehicle speed, then the systems provide accurate speed measurement, but they require overhead structures which may not be available in various parts of the network
Solution Approach 1:
The patent replaces overhead radar and lidar systems with distributed acoustic sensing that does not require overhead structures. Instead of active electromagnetic pulse transmission from gantries, the system uses passive acoustic detection through the ground, enabling deployment in locations without overhead infrastructure while maintaining speed measurement capability through Doppler shift analysis of acoustic signals.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances the accuracy of traffic flow monitoring by compensating for missed detections and improving the performance of DAS systems, providing a more reliable estimate of traffic flow and enabling better traffic management decisions.
Implementation Method 1
performing distributed acoustic sensing to provide a measurement signal from each of a plurality of sensing portions of a first length of a sensing optical fibre
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This application relates to apparatus and methods for traffic monitoring, especially for monitoring of road traffic flow using fibre optic distributed acoustic sensing. The method involves performing distributed acoustic sensing (DAS) using a suitable interrogator unit (103) to provide a measurement signal from each of a plurality of sensing portions of a first length of a sensing optical fibre (102), where the first length of the sensing optical fibre runs alongside a road (101) having a plurality of lanes. The measurement signals from the sensing portions are processed to detect vehicles (104) travelling on the road and to determine at least one detected traffic flow property. The method further involves using a traffic flow model to relate the detected flow property to a modelled flow property, where the traffic flow model is configured to model vehicle detections that will be missed by the distributed acoustic sensing i.e. an indication of vehicles present on the road but not detected by the DAS sensor.