System Excitation Selection for Faster Parameter Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing control methods for autonomous systems, particularly in vehicle lateral guidance, fail to adequately address uncertainties such as external faults and parameter uncertainties, leading to time-consuming and costly adjustments in the application phase, and lack effective analytic methods for determining relevant system excitations for parameter identification.
Innovation Solution
A method for determining system excitations using a system model, parameter library, and grade measures to select optimal maneuvers for system identification, which can be implemented in a computer system, reducing the need for extensive empirical catalogs and enabling efficient parameter identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If extensive empirical maneuver catalogs are used for system identification, then comprehensive parameter identification is achieved, but time and cost increase significantly
Solution Approach 1:
The patent performs preliminary analysis of system excitations during the design phase by calculating grade measures that quantify the information content of different excitations for identifying specific parameters. This preliminary assessment allows selection of optimal excitations before actual system identification, avoiding the need for extensive empirical testing later.
Solution Approach 2:
The system uses automated computational methods to evaluate and select system excitations based on mathematical criteria (grade measures) rather than relying on extensive empirical catalogs developed through years of application. The system serves itself by automatically determining which excitations will provide the necessary information for parameter identification.
2Measurement precision
If comprehensive maneuver catalogs are used for system identification, then all parameters can be identified, but development cycles lengthen and market entry is delayed
Solution Approach 1:
The methodology performs preliminary calculation of grade measures during the design phase to identify which system excitations will provide the necessary information for parameter identification. This advance planning enables streamlined system identification later, reducing overall development time without compromising parameter identification completeness.
Solution Approach 2:
The patent transforms the approach from using fixed, extensive empirical catalogs to a dynamic method where system excitations are selected based on calculated grade measures that quantify their information content for specific parameters. This parameter-based selection optimizes the balance between identification completeness and development speed.
3Device complexity
If nominal models are used for control, then control implementation is simple, but uncertainties such as external faults and parameter uncertainties are not addressed
Solution Approach 1:
The system performs preliminary identification of parameters and uncertainties during the design phase using optimized system excitations. By identifying parameters early and accounting for uncertainties before deployment, the control system can be designed to be more robust without requiring overly complex adaptive control mechanisms during operation.
Data Source
AI summary
A method for determining system excitations for system identification is disclosed. The method includes (i) receiving a system model, a plurality of system excitations, and a parameter library that comprises at least one parameter to be identified, (ii) determining a plurality of output variables based on the plurality of system excitations and the at least one parameter to be identified using the system model, (iii) determining a plurality of grade measures of the determined plurality of output variables with respect to the at least one parameter to be identified, and (iv) selecting system excitations for system identification based on the determined plurality of grade measures.