Vehicle Speed Profile Prediction Using Networked Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting a vehicle's speed profile are limited by the range of onboard sensors and the availability of collected driving profiles, which can result in unreliable forecasts, especially in areas with limited network coverage or for vehicle manufacturers with a small market share.
Innovation Solution
A method that generates a prediction model using input data including geocoordinates, digital map information, average traffic flow data, and speed profiles from networked vehicles, with a situation analysis selecting the most suitable data based on criteria such as data availability, cost, and road class.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If networked vehicle data is used for speed profile prediction, then prediction accuracy is improved, but data availability deteriorates in areas with limited network coverage
Solution Approach 1:
The system performs preliminary actions by storing speed profile data locally in the vehicle before reaching areas with limited network coverage. The prediction model is pre-loaded with relevant data, and the system switches to using locally stored data when network connection is unavailable, ensuring continuous prediction capability throughout the journey.
Solution Approach 2:
The patent introduces an intermediary mechanism that combines multiple data sources (digital map attributes, onboard sensor data, and locally stored historical data) to bridge the gap when networked vehicle data is unavailable. This intermediary approach allows the system to maintain prediction accuracy by substituting networked data with alternative data sources when necessary.
2Loss of information
If digital map attributes are used for speed profile prediction, then prediction can be made without network coverage, but prediction accuracy deteriorates because maximum speed cannot be reached due to braking operations
Solution Approach 1:
The system merges multiple data sources including digital map attributes, onboard sensor data, and locally stored historical speed profiles to create a comprehensive prediction model. This combination allows the system to overcome the limitations of individual data sources, particularly the inability of digital map attributes alone to capture actual driving behavior and speed variations.
Solution Approach 2:
The prediction system dynamically adapts by switching between different data sources based on availability and relevance. When networked data is available, it supplements digital map attributes with real-time vehicle data. When networked data is unavailable, it relies on locally stored historical data that captures actual driving patterns, making the system dynamic rather than static.
3Adaptability or versatility
If crowdsourcing approach is used, then speed profile prediction is possible, but reliability deteriorates due to lack of guarantee across all route sections
Solution Approach 1:
The system applies local quality by storing and using location-specific speed profile data that is relevant to particular route sections. Each vehicle stores historical data for routes it has previously traversed, and the prediction model selects and applies the most appropriate local data based on the current route section, ensuring reliable predictions even in areas where networked data may be unavailable.
Data Source
AI summary
Various embodiments include a method for operating a vehicle based on a future speed curve along a predetermined travel route up to a specific preview horizon. The method may include: generating a prediction model; supplying data to the model; using an algorithm to generate a predicted speed profile; selecting a datum for the second input data group depending on a situation analysis using predetermined criteria; and using a controller to: implement operating strategies for a vehicle, control exhaust aftertreatment systems, increase an accuracy of navigation algorithms, and/or predict operating states depending on the speed forecast. The input data include a first input data group containing geocoordinates of the travel route and a second input data group containing at least one datum selected from the group consisting of: location information for a digital map, average traffic flow data along the travel route, and speed profiles of networked vehicles.