Sensor Delay Estimation for Autonomous Vehicle State Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sensor systems in autonomous driving face challenges in accurately estimating delay times, especially for sensors installed in infrastructure, leading to position errors due to unsynchronized clocks and communication delays, which affect the precision of vehicle state estimation.
Innovation Solution
A sensor delay time estimation device that includes a processor configured to acquire observables from multiple sensors, estimate delay times based on a fixed time interval, and compute vehicle states using Gaussian process regression and moving horizon estimation to interpolate discrete data into continuous values, thereby minimizing discretization errors and improving estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If sensor information is acquired at different timings using separate onboard computers with independent clocks, then each computer can process information independently, but the position estimation accuracy deteriorates due to time synchronization errors and unaccounted delay times
Solution Approach 1:
The patent introduces a time delay estimation unit as an intermediary component that calculates the time delay between sensor information acquisition timings. This mediator enables the system to account for synchronization errors without requiring complex hardware synchronization, thus maintaining independent processing while improving position estimation accuracy by compensating for timing differences mathematically.
Solution Approach 2:
The patent changes the parameter of time delay from an unknown variable to a calculated parameter. By estimating the time delay between sensor acquisitions and using this estimated value in subsequent position calculations, the system transforms the synchronization problem into a parameter compensation problem, thereby improving measurement precision while preserving operational independence.
2Device complexity
If time delay estimation is performed using discrete sensor data points, then the calculation is computationally simple, but discretization errors accumulate and reduce estimation accuracy
Solution Approach 1:
The patent applies preliminary action by performing time delay estimation using multiple discrete sensor data points collected over a predetermined time period before final position estimation. This preliminary estimation of delay characteristics allows the system to compensate for timing errors in subsequent calculations, reducing discretization errors while maintaining computational simplicity.
Solution Approach 2:
The patent continuously accumulates sensor information over a predetermined time period to estimate time delay, transforming discrete data points into a continuous estimation process. This continuous accumulation and averaging of delay measurements reduces the impact of individual discretization errors, improving overall estimation accuracy without significantly increasing computational complexity.
Data Source
AI summary
A sensor delay time estimation device that includes: a memory; and a processor coupled to the memory, wherein the processor is configured to: acquire an observable required to estimate a vehicle state from a plurality of sensors that detect and output the observable, each of the plurality of sensors being different, and based on the observable for a fixed time interval acquired from each of the plurality of sensors, estimate a time delay from detection to acquisition of the observable for each of the plurality of sensors, and estimate a vehicle state for the fixed time interval.