Traffic Flow Prediction Model Using CGAN for New Roads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to accurately predict traffic flow on new roads due to the lack of historical traffic data, especially when considering various contextual factors like weather, day types, and seasons.
Innovation Solution
The use of a conditional generative adversarial network (CGAN)-based prediction model that is trained using traffic data from existing roads and input traffic data corresponding to a default context of the new road, allowing for the prediction of traffic flow across various contexts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a prediction model is trained using historical traffic data from existing roads, then prediction accuracy for various contexts is improved, but the model cannot be applied to new roads without historical data
Solution Approach 1:
The patent creates a virtual copy of historical traffic data from existing roads and adapts it to new road contexts using context information. The CGAN model generates synthetic traffic flow data that replicates the patterns of existing roads while incorporating new road characteristics, allowing the model to be applied to new roads without requiring actual historical data from those specific roads.
Solution Approach 2:
The patent changes the contextual parameters (weather conditions, day types, seasons) associated with the historical traffic data to match the new road's context. By transforming the context parameters rather than the raw traffic data itself, the model maintains the statistical patterns learned from existing roads while adapting to the specific conditions of new roads.
2Measurement precision
If traffic data corresponding to various contexts is collected for new roads, then prediction accuracy for different contexts is improved, but it takes a lot of time to collect sufficient data
Solution Approach 1:
The patent performs preliminary action by pre-training the prediction model using historical traffic data from existing roads covering various contexts (different weather conditions, day types, seasons). This pre-training establishes the model's ability to recognize traffic patterns across different contexts before deployment, eliminating the need to collect extensive contextual data for each new road.
Solution Approach 2:
The model copies the contextual relationships learned from existing roads and applies them to new roads. By replicating the pattern associations (e.g., how traffic flow changes with weather or time of day) rather than requiring actual collected data for each context, the system achieves contextual prediction capability without time-consuming data collection.
3Adaptability or versatility
If a prediction model is trained without historical traffic data for new roads, then the model can be immediately applied to new roads, but prediction accuracy deteriorates
Solution Approach 1:
The patent introduces an intermediary approach by using context information as a bridge between existing road data and new road predictions. The model uses contextual parameters (weather, time, season) as intermediaries to transfer knowledge from existing roads to new roads, enabling immediate applicability while maintaining accuracy through context-aware pattern matching rather than direct data transplantation.
Data Source
AI summary
An embodiment apparatus for predicting traffic flow on a new road includes a memory, an input device, and a controller configured to input traffic data corresponding to a default context of the new road received by the input device to a prediction model stored in the memory, training of which is completed, and to predict a traffic flow corresponding to various contexts of the new road based on the prediction model.