Prefix-Based Estimator for Intermittent Vehicle Sensor Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Control and estimation algorithms in automotive applications often face challenges due to intermittent sensor measurements, such as packet drops in communication networks, sensor glitches, and occlusions, which necessitate robustness against missing data.
Innovation Solution
A prefix-based estimator system that uses a pattern of previous measurements to calculate an estimated updated vehicle state, adapting filter gains based on observed missing data patterns to provide bounded error estimates, leveraging Q-parametrization for convex optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional estimation algorithms are used with intermittent sensor measurements, then the system can handle missing data, but the estimation accuracy and recovery levels deteriorate
Solution Approach 1:
The estimator dynamically adapts its parameters and structure based on the observed missing data patterns. The algorithm switches between different estimation modes depending on the current data availability, allowing it to maintain high accuracy during periods with missing measurements while recovering quickly when data becomes available again.
Solution Approach 2:
The system continuously monitors the pattern of missing data and uses this information to adjust its estimation strategy. By incorporating feedback about data availability patterns, the estimator can optimize its performance and maintain accuracy even when measurements are intermittently lost.
2Measurement precision
If complex estimation algorithms are developed to handle missing data patterns, then estimation accuracy improves, but computational complexity and synthesis difficulty increase
Solution Approach 1:
The estimation problem is segmented into distinct phases based on missing data patterns. Rather than using a single complex algorithm, the system divides the estimation process into multiple simpler sub-problems that can be solved sequentially, reducing overall computational complexity while maintaining accuracy.
Solution Approach 2:
The algorithm changes its parameters based on the observed missing data patterns. By adapting parameters dynamically rather than using fixed complex structures, the system achieves high estimation accuracy with computationally efficient implementations.
3Ease of operation
If traditional estimators are used without considering missing data patterns, then the system is simpler to implement, but recovery levels and error bounds worsen
Solution Approach 1:
The system performs preliminary analysis of missing data patterns to determine the appropriate estimation strategy. By proactively identifying patterns in data availability, the estimator can prepare and switch to optimal estimation modes in advance, improving recovery levels without adding significant implementation complexity.
Data Source
AI summary
A method of the disclosure implements a prefix-based estimator system. The method includes: receiving a plurality of vehicle states at a plurality of time points from at least one sensor coupled to a vehicle, wherein each of the plurality of vehicle states includes at least one parameter information and a discrete state; determining a most recent vehicle state of the plurality of vehicle states has at least one parameter information missing; identifying a prefix comprising a missing data pattern that matches a sequence of discrete states of a subset of time-ordered vehicle states including the most recent vehicle state, wherein the subset of time-ordered vehicle states correspond to the prefix; calculating an estimated updated vehicle state of the vehicle using an optimized prefix-based dynamic estimator based on the prefix and the subset of time ordered vehicle states; and providing the estimated updated vehicle state to a driving control system of the vehicle.