Vehicle Route Determination Using Bayesian Filters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining a vehicle's route through a transport network junction rely on track-side infrastructure, which is expensive to install and maintain, and not autonomous. Additionally, Global Navigation Satellite System (GNSS) sensors suffer from signal availability and continuity issues in environments like tunnels.
Innovation Solution
A computer-implemented method using Bayesian estimation filter algorithms to determine a vehicle's route through a transport network junction without relying on track-side infrastructure. This method involves obtaining track geometry data, generating Bayesian estimation filter algorithms for each route option, monitoring their output as the vehicle passes through the junction, and selecting the route with the highest probability based on the algorithm outputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If track-side infrastructure (balises or beacons) is used to detect vehicle position and determine route, then route detection accuracy is improved, but installation and maintenance costs increase significantly
Solution Approach 1:
The patent extracts the route detection function from the track-side infrastructure and relocates it to the vehicle's onboard system. The vehicle uses its own sensors (accelerometers, gyroscopes, magnetometers) and a database of track geometry data to independently determine its position and route through junctions, eliminating the need for balises or beacons at each location.
Solution Approach 2:
The vehicle performs self-positioning and self-route-determination using its onboard sensors and processing capabilities. The system continuously monitors sensor data, compares it with expected track geometry, and autonomously determines the vehicle's position and selected route without requiring external track-side detection infrastructure.
2Reliability
If track-side infrastructure is used to transmit position signals, then route information reliability is improved, but system complexity and communication dependency increase
Solution Approach 1:
The vehicle independently determines its own position and route using onboard sensors and local processing. The system compares sensor measurements with pre-stored track geometry data to autonomously identify the vehicle's location and selected route, eliminating dependency on communication links with track-side infrastructure or network signalling systems.
Solution Approach 2:
The patent replaces the mechanical/communication-based track-side infrastructure system with an onboard inertial and magnetic sensing system. Instead of relying on electrical or radio communication from track-side devices, the vehicle uses physical sensor measurements (acceleration, rotation, magnetic field) processed through algorithms to determine position and route.
3Extent of automation
If GNSS sensors are used to determine vehicle position, then autonomous positioning capability is improved, but signal availability and continuity deteriorate in environments like tunnels
Solution Approach 1:
The patent merges multiple sensing systems (accelerometers, gyroscopes, magnetometers) with local track geometry data to create a hybrid positioning system. This combination allows the vehicle to maintain autonomous positioning capability in environments where GNSS signals are unavailable, such as tunnels, by relying on inertial and magnetic field measurements compared against known track geometry.
Solution Approach 2:
The system changes the physical parameters used for positioning from satellite-based electromagnetic signals (GNSS) to local physical measurements (inertial acceleration, rotational motion, magnetic field strength and direction). This parameter change enables continuous autonomous positioning regardless of satellite signal availability.
Data Source
AI summary
A computer-implemented method of determining a position of a vehicle 10 within a transport network comprises: obtaining track geometry data indicating track geometry of at least a part of the transport network; determining, based upon the track geometry data, that the vehicle is approaching a junction; determining, based upon the track geometry data, a plurality of route options from the junction; generating a plurality of Bayesian estimation filter algorithms each associated with a respective one of the plurality of route options and configured to estimate a position of the vehicle based upon the track geometry data indicative of the associated route option, wherein the plurality of Bayesian estimation filter algorithms are configured to output data indicative of probabilities of the vehicle taking the associated route options; monitoring the output of the plurality of Bayesian estimation filter algorithms as the vehicle passes through the junction; and determining the route option taken by the vehicle by selecting one of the plurality of route options which presents the highest probability based upon the output of the plurality of Bayesian estimation filter algorithms.