Vehicle Speed Estimation Using Transverse Position Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting vehicle driving maneuvers, such as the Intelligent Driver Model, are prone to errors due to short-term inaccuracies in detecting vehicles in front, leading to unreliable estimates of future speed and distance, especially when vehicles change lanes or are incorrectly recognized as 'ghost vehicles'.
Innovation Solution
A method that filters the recorded distance to the vehicle in front using low-pass filtering or smoothing, while also considering the transverse position of the vehicle to differentiate between valid and invalid detection segments, allowing for robust estimation of future speed and distance by switching between filtered and unfiltered distances based on lane changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the Intelligent Driver Model is used to estimate future speed based on detected distance to front vehicle, then the estimation adapts to actual traffic conditions, but short-term detection errors (ghost vehicles, missed detections) cause unreliable speed estimates
Solution Approach 1:
The system performs preliminary actions by detecting the transverse position of the front vehicle before using the distance data for speed estimation. By提前 checking whether the front vehicle is on the current path (using transverse position threshold), the system prevents detection errors from affecting the speed estimate. This preliminary validation step ensures that only reliable distance measurements (from vehicles actually on the path) are used in the Intelligent Driver Model.
Solution Approach 2:
The transverse position detection acts as an intermediary mechanism between the raw distance detection and the speed estimation process. This intermediary step filters out invalid distance measurements caused by ghost vehicles or missed detections by checking whether the front vehicle's transverse position indicates it is on the current path. Only when the intermediary check passes is the distance data transmitted to the speed estimation algorithm.
2Reliability
If filtering is applied to the distance course to smooth detection errors, then estimation robustness improves, but response to actual lane changes or vehicle departures is delayed
Solution Approach 1:
The system dynamically adjusts the filtering behavior based on the transverse position of the front vehicle. When the transverse position indicates the vehicle has left the path, the system automatically switches from filtered distance values to unfiltered measured distance values. This dynamic adaptation allows the system to maintain robustness during normal operation while quickly responding to lane changes or vehicle departures without artificial smoothing delays.
Solution Approach 2:
The transverse position detection provides continuous feedback about the front vehicle's lateral status. This feedback mechanism allows the system to detect when a vehicle leaves the current path and adjust the distance filtering accordingly. The feedback loop ensures that filtering is applied only when appropriate (vehicle on path) and discontinued when the vehicle departs (transverse position exceeds threshold), maintaining both robustness and responsiveness.
3Measurement precision
If transverse position detection is added to differentiate valid from invalid detection segments, then detection accuracy improves, but system complexity increases
Solution Approach 1:
The system achieves multi-functionality by using the transverse position detection for dual purposes: (1) identifying when the front vehicle is on the current path to validate distance measurements, and (2) detecting lane changes or vehicle departures to trigger filtering mode changes. This single transverse position measurement serves multiple functions, improving precision without proportionally increasing complexity.
Solution Approach 2:
The system changes the parameter being monitored from just longitudinal distance to include transverse position. By adding this dimensional parameter, the system gains the ability to distinguish valid from invalid detection segments. The transverse position parameter provides additional information that enables more precise distance detection without requiring complex new hardware, simply by utilizing an additional dimension of the existing sensor data.
Data Source
Figure 1a
Figure 1b
Figure 1c
AI summary
The invention relates to a method for estimating a future speed of a vehicle and/or distance of a vehicle from a reference point, comprising: detecting the speed of the vehicle at a first distance of the vehicle along a path from a reference point; the vehicle being located at the first distance from the reference point at a first time; detecting the distance and the course of the distance of the vehicle from a preceding vehicle driving in front of the vehicle between the first time and a second time; during the detection, the preceding vehicle being located on a path on which the vehicle will also located under the assumption of a first driving maneuver from the first time to the second time; creating at least one first estimate of the speed of the vehicle for the second time and/or of the distance of the vehicle from the reference point for the second time; the first estimate being based on the speed of the vehicle at the first distance, the first distance from the reference point, and the second time; the first estimate of the speed and/or of the distance furthermore being based on a first course of the desired speed of the vehicle determined in advance for the assumption of the first driving maneuver in dependence on the distance of the vehicle from the reference point; characterized by: detection of the lateral deviation and the course of the lateral deviation of the preceding vehicle between the first time and the second time; the lateral deviation being the perpendicular distance of the vehicle center axis, in particular of the center or of the rear end of the vehicle center axis, from the path; the first estimate also taking into account the course of the distance from the preceding vehicle and a filtering, in particular low-pass filtering or smoothing, of the course of the detected distance in dependence on the course of the lateral deviation.