Autonomous Vehicle Load Estimation Using Driving Statistics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion planning and control systems for autonomous driving vehicles do not accurately account for vehicle load variations, leading to inconsistent performance across different driving conditions and vehicle types.
Innovation Solution
A computer-implemented method that measures and collects driving statistics to select a load calibration table matching the current vehicle conditions, using a set of predetermined tables created based on various load and environmental factors, allowing for real-time load determination and adjustment of control commands.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional motion planning and control systems apply the same control to all vehicles without load consideration, then the system complexity is low, but the control accuracy and driving performance deteriorate
Solution Approach 1:
The system performs preliminary load calibration by collecting driving statistics under known load conditions and storing them in calibration tables before actual operation. This pre-established data structure enables rapid load determination without real-time complex calculations, resolving the contradiction between accurate load estimation and system complexity
Solution Approach 2:
Instead of directly measuring physical load parameters, the system creates a virtual representation by matching current driving statistics against pre-collected calibration data. This copying approach allows accurate load determination through statistical comparison rather than direct physical measurement, maintaining simplicity while improving accuracy
2Measurement precision
If the system uses predetermined load calibration tables with driving statistics, then the load determination accuracy improves, but the data processing and table selection complexity increases
Solution Approach 1:
The system automatically selects the appropriate calibration table by comparing current driving statistics with stored calibration data and identifying the best match. This self-service mechanism eliminates manual table selection and ensures accurate load determination without requiring complex external intervention or management overhead
3Measurement precision
If the system collects and processes driving statistics in real-time, then the load estimation accuracy improves, but the computational time and processing load increase
Solution Approach 1:
The system pre-collects and organizes driving statistics under various known load conditions into calibration tables before operation. This preliminary preparation stores processed reference data that enables rapid comparison and load determination during actual driving, avoiding real-time complex computations while maintaining high accuracy
Solution Approach 2:
The system replaces direct physical load measurement with a computational approach using statistical pattern matching. By substituting physical sensors and direct measurement with algorithmic comparison of driving statistics against calibration data, the system achieves accurate load estimation through information processing rather than mechanical measurement
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A first set of driving statistics is measured and collected from an autonomous driving vehicle (ADV) at different points in time in response to various control commands while the ADV is driving in various driving environments. Based on the first set of driving statistics, a search is conducted in each of the load calibration tables to find a load calibration table having similar driving statistics. One of the load calibration tables is selected, which contains a second set of driving statistics that are most similar to the first set of driving statistics. A current load of the ADV is determined based on the selected load calibration table, for example, by designating the load associated with the selected load calibration table as the current load of the ADV. The load of the ADV can be utilized as a factor for generating subsequent control commands for the ADV.