Speed Profile Prediction Using Segmented Route Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems struggle to predict vehicle speed profiles accurately, especially in real-world conditions, due to idealized assumptions and limited representation of traffic disruptions, leading to imprecise arrival time predictions and energy consumption calculations.
Innovation Solution
A method that defines a dataset of route points with input and output speeds, using measured values to create a detailed speed profile by dividing routes into processes and assigning driving styles, allowing for non-linear speed changes and situational contexts, thereby reflecting real driving behaviors and conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If idealized assumptions are used for speed profile calculation (constant speeds per road type, abrupt changes), then the calculation is simple and fast, but the prediction accuracy of arrival time and energy consumption deteriorates
Solution Approach 1:
The route is divided into multiple route sections based on road type changes, intersections, and other significant points. Each section is assigned a specific speed value, allowing for detailed speed profile calculation while maintaining manageable computation complexity through structured segmentation of the route.
Solution Approach 2:
Speed profiles are pre-calculated and stored for different road types and route sections based on historical data and statistical analysis. During actual navigation, the system retrieves and combines these pre-calculated profiles rather than computing everything in real-time, improving both accuracy and calculation speed.
2Measurement precision
If detailed measured values from multiple routes are collected and processed, then the speed profile realism and prediction accuracy improve, but the data processing complexity and computational requirements increase
Solution Approach 1:
The system automatically collects speed measurement data from multiple routes, processes it through statistical analysis, and generates updated speed profiles without requiring manual intervention. The navigation system serves itself by continuously improving its own data set through actual usage, reducing the need for external data processing infrastructure.
Solution Approach 2:
The system transforms raw speed measurement data into statistical parameters (average speeds, standard deviations, confidence intervals) for different route sections. This parameter transformation simplifies the data structure while preserving the essential information needed for accurate speed profile prediction, reducing computational complexity.
3Measurement precision
If acceleration and deceleration phases are included in speed profile calculation, then the energy consumption calculation accuracy improves, but the computational time and processing requirements increase
Solution Approach 1:
The system applies acceleration and deceleration calculations selectively at critical points (intersections, road type changes, elevation changes) rather than continuously along the entire route. This partial application of complex calculations maintains energy consumption accuracy where it matters most while minimizing overall computational time.
Solution Approach 2:
Different calculation methods are applied to different route sections based on their characteristics. High-precision acceleration/deceleration modeling is applied only to sections where it significantly impacts energy consumption (e.g., steep gradients, frequent stops), while simpler models are used for sections with minimal speed variations, optimizing the balance between accuracy and computation time.
Data Source
Figure 1~4
Figure 2~3
Figure 5~6
AI summary
Operations specified in a data set are defined, preferably a plurality of such operations each related to different driver types. Each operation is characterised by an input speed and an output speed. A provisional speed profile can be divided into route sections, to each of which one operation out of the defined operations corresponds. The assigned speed profiles may be very complex, and in particular also comprise non-linear sections between the input speed and the output speed; thus a constant travel may be simulated for example by an oscillatory change in the speed. Shutdown phases with a specific frequency and duration can also just as well be contained in an operation. The situational context given by other operations can be taken into consideration when establishing the speed profile.