Railroad Crossing Prediction Using Velocity Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Railroad crossing prediction systems face challenges in providing accurate warning times due to noisy position signals, leading to either short or long warning times, which can be misleading for pedestrians and vehicles, as they struggle to differentiate between noise and actual motion on the track.
Innovation Solution
A method that determines vehicle position and velocity by filtering sensed data, with a greater emphasis on filtering velocity above a threshold frequency, to improve the accuracy of time-to-crossing predictions, using a controller with an algorithm that calculates warning times based on averaged velocity rather than position, thereby reducing the impact of noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If position signal is used directly for crossing prediction, then the system responds quickly to train motion, but the warning time becomes inaccurate due to noise
Solution Approach 1:
The system performs preliminary filtering of the position signal to extract velocity information before calculating warning time. By deriving velocity from position changes and filtering the velocity signal, the system prepares cleaned data in advance, which is then used for more accurate time-to-crossing predictions without losing responsiveness.
Solution Approach 2:
Velocity serves as an intermediary parameter between position and warning time calculation. Instead of using position directly, the system introduces velocity as a intermediate step that filters out noise while preserving motion information, thereby improving the accuracy of the final warning time prediction.
2Measurement precision
If velocity signal is filtered to reduce noise, then warning time accuracy improves, but the system may delay detection of actual motion changes
Solution Approach 1:
The filtering approach is made dynamic by applying frequency-dependent filtering to the velocity signal. The system adapts the filtering strength based on the frequency content of the signal, allowing it to preserve quick motion changes while still filtering out noise. This dynamic approach balances noise reduction with timely motion detection.
3Measurement precision
If position signal has high noise, then it is difficult to differentiate between noise and actual train motion, but filtering position reduces responsiveness to real motion
Solution Approach 1:
The system replaces direct position-based motion detection with a velocity-based detection mechanism. By calculating velocity from position changes and using this derived velocity signal for motion detection, the system substitutes a more noise-resistant measurement approach that maintains responsiveness to actual train motion while filtering out position noise.
Data Source
AI summary
Various methods and systems are provided for predicting crossing times. In one embodiment, a method comprises determining a vehicle position of a vehicle from sensed data, filtering the vehicle position, determining vehicle velocity based on the filtered vehicle position, filtering the vehicle velocity to a greater extent than vehicle position is filtered, at least above a threshold frequency, and determining a time-to-crossing based on the filtered vehicle position and filtered velocity.