Traffic Prediction via Waterfall Data Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing traffic prediction systems face difficulties in accurately separating individual vehicle tracks from waterfall data when traffic is dense, leading to challenges in predicting vehicle numbers and speeds.
Innovation Solution
A traffic prediction apparatus that acquires and preprocesses waterfall data, trains a model using ground truth traffic properties, and predicts traffic properties within a predetermined time and position range, utilizing a distributed vibration sensor and optical fiber to detect vibrations along a road.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If basic signal processing principles are used to identify vehicle tracks, then individual vehicle tracks can be identified when traffic volume is low, but when traffic volume is high and vehicles are densely present, the vehicle tracks overlap and mix making it difficult to separate individual tracks
Solution Approach 1:
The patent transforms the waterfall data through parameter changes including normalization of vibration amplitudes and conversion to time-distance representation. This transformation changes the parameter space in which vehicle tracks are analyzed, enabling better separation and identification of individual tracks even in dense traffic conditions where original waterfall data shows overlapping trajectories.
2Measurement precision
If a trained model is used to predict traffic properties, then accurate prediction of vehicle numbers and speeds can be achieved, but the system complexity increases due to the need for training data processing and model generation
Solution Approach 1:
The patent performs preliminary actions by pre-processing the waterfall data including normalization and transformation before feeding it to the training model. This preliminary processing prepares the data in an optimized format that reduces the complexity of the subsequent model training and improves prediction accuracy, thereby managing system complexity more effectively.
3Loss of information
If waterfall data is acquired and processed to predict traffic properties, then traffic flow analysis is improved, but the processing time and computational resources increase
Solution Approach 1:
The patent extracts only the essential features from the waterfall data through normalization and transformation processes. By taking out and focusing on the most relevant characteristics (vibration amplitudes, time-distance relationships) rather than processing the entire raw dataset, the system reduces computational burden and processing time while maintaining accurate traffic property prediction.
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
Enables accurate prediction of vehicle numbers and speeds by transforming and processing waterfall data, effectively handling dense traffic conditions and varying sensitivity, thereby improving traffic flow analysis.
Implementation Method 1
A distributed vibration sensor that detects a vibration generated by a vehicle (moving object) running on a road by way of an optical fiber provided along the road
Data Source
AI summary
The present disclosure provides a traffic prediction apparatus, system, method and program capable of predicting the number of vehicles and the speed of vehicles in a predetermined time and in a predetermined range based on waterfall data of vehicles. The traffic prediction apparatus comprises acquisition means for acquiring waterfall data comprising a generation position of a vibration, a generation time of the vibration and an amplitude of the vibration generated by a vehicle traveling on a road, pre-processing means for transforming the acquired waterfall data, generation means for training a portion of the plurality of processed waterfall data and at least one corresponding ground truth traffic property used as plurality of labels to generate a trained model, wherein the ground truth may be obtained from a secondary acquisition means, and prediction means for predicting at least one traffic property for a processed waterfall data.