Road Surface Profile Estimation via Half-Car Parameter Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing road surface profile estimation methods face significant estimation errors due to difficulties in measuring inertial moment and maintaining consistent experimental conditions, leading to inaccuracies in road surface profile estimation.
Innovation Solution
A road surface profile estimating device that dynamically updates variance-covariance matrices and uses a half car model to estimate profiles based on vertical displacement and velocity of both front and rear tires, ensuring parameter determination and profile matching, thereby reducing estimation errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If parameters are estimated from actual measurement or hump climbing response, then road surface profile can be estimated, but estimation accuracy deteriorates due to difficulty in measuring inertial moment and maintaining consistent experimental conditions
Solution Approach 1:
The system uses the vehicle itself as the measurement platform, utilizing its own sensors (accelerometers, gyroscopes) and existing motion data to estimate road surface profiles. This eliminates the need for separate experimental setups and external measurement equipment, allowing parameter estimation to be performed during normal vehicle operation without additional experimental complexity
Solution Approach 2:
The system estimates simulation model parameters (such as inertial moment, damping coefficients, spring constants) by optimizing the match between simulated and actual vehicle responses during normal operation. By treating these parameters as variable quantities to be determined through data assimilation rather than fixed values from separate measurements, the system achieves accurate road surface profile estimation without requiring controlled experimental conditions
2Reliability
If simulation model parameters are determined through separate experiments, then model accuracy improves, but the process becomes complex and time-consuming
Solution Approach 1:
The system performs preliminary estimation of simulation model parameters using available vehicle operation data before actual road surface profile estimation begins. By pre-determining parameters such as inertial moment and damping coefficients through optimization against measured vehicle responses, the system prepares accurate simulation models in advance, enabling reliable real-time road surface profile estimation during normal vehicle operation
Solution Approach 2:
The system continuously refines simulation model parameters by comparing simulated vehicle responses with actual sensor measurements and adjusting parameters to minimize differences. This feedback-based parameter optimization ensures the simulation model remains accurate without requiring separate experimental calibration processes, as parameters are automatically updated using ongoing vehicle operation data
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided is a road surface profile estimating device and the like that determine a plurality of parameters included in a simulation model together with estimation of a road surface profile. A road surface profile estimating device 10 includes an acquisition unit 11 configured to acquire physical quantities representing a motion of a vehicle 30 that is traveling on a road surface, a prediction unit 12 configured to predict time evolution of state variables of the vehicle 30 on the basis of a simulation model M1 including a plurality of parameters P, a calculation unit 13 configured to calculate an observation value of the physical quantities from the state variables predicted by the prediction unit 12 on the basis of an observation model M2, an update unit 14 configured to update the state variables through data assimilation, an estimation unit 15 configured to estimate respective profiles of the road surface under a first condition and a second condition, and a determination unit 16 configured to determine the plurality of parameters such that an absolute value of an evaluation function for evaluating a difference between the profile of the road surface estimated under the first condition and the profile of the road surface estimated under the second condition is small.