Multi-Task Learning for Vehicle Stay Behavior Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for analyzing and predicting vehicle stay behaviors fail to capture the temporal-spatial correlation and task dependence between different prediction tasks, such as stay location and stay duration, leading to reduced prediction accuracy.
Innovation Solution
A method and system utilizing multi-task learning with deep recurrent neural networks to extract spatial-temporal characteristics from vehicle GPS and OBD data, processing data through first- and second-level recurrent neural networks to input into a multi-task learning and predicting network, which correlates stay location and stay duration prediction tasks based on historical behaviors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If single-task learning methods are used to predict traffic data, then the model is simpler and easier to implement, but the prediction accuracy is reduced due to ignoring task dependence between different prediction tasks
Solution Approach 1:
The patent combines multiple single-task learning models into a unified multi-task learning framework that simultaneously predicts traffic flow, speed, and travel time. The model shares common layers for extracting spatial-temporal features from trajectory data, while maintaining task-specific output layers. This merging approach allows the model to capture task dependencies and correlations between different traffic predictions, improving overall prediction accuracy while maintaining reasonable model complexity through parameter sharing.
Solution Approach 2:
The patent creates a universal multi-task learning model that can perform multiple prediction functions simultaneously. The shared neural network architecture learns general spatial-temporal patterns from trajectory data that are applicable across different traffic prediction tasks. This multi-functional model replaces multiple separate single-task models, improving prediction accuracy by leveraging correlations between tasks while avoiding the redundancy of training separate models for each traffic parameter.
2Measurement precision
If deep recurrent neural networks with multiple levels are used to extract spatial-temporal characteristics, then the extraction accuracy is improved, but the computational complexity and processing time increase
Solution Approach 1:
The patent segments the complex task of spatial-temporal feature extraction into multiple hierarchical levels using deep recurrent neural networks. The first-level RNNs process trajectory data at a fine-grained level to capture local temporal patterns, while the second-level RNNs aggregate these features to capture broader spatial-temporal correlations. This segmentation into hierarchical levels improves extraction accuracy by progressively building complex representations from simpler components, while managing computational complexity through modular architecture design.
Data Source
AI summary
The present application discloses a method and a system for analyzing and predicting a vehicle stay behavior based on multi-task learning, and the method includes the following steps: acquiring vehicle GPS and OBD data including a vehicle ID, a travel start time, a start longitude, a start latitude, an end time, an end longitude, and an end latitude after desensitization; preprocessing vehicle GPS and OBD data to obtain vehicle stay behavior data including stay location and stay duration; extract a spatial-temporal characteristic of the preprocessed vehicle stay behavior data by a deep recurrent neural network; inputting the spatial-temporal characteristic into a multi-task learning and predicting network, and obtaining the correlation between a stay location prediction task and the stay duration prediction task based on the historical stay behavior of the vehicle through the multi-task learning and predicting network to predict the stay location and stay duration.