An expert system based modelling scheme for tactile data interpretation
| Autor: | Hugh C. Wood, C. S. Vaidyanathan |
|---|---|
| Rok vydání: | 1991 |
| Předmět: |
Scheme (programming language)
Computer science business.industry General Mathematics Model selection Control engineering computer.software_genre Expert system Computer Science Applications Task (project management) Image (mathematics) Control and Systems Engineering Embedding Limit (mathematics) Artificial intelligence business computer Software computer.programming_language |
| Zdroj: | Robotics and Autonomous Systems. 8:213-237 |
| ISSN: | 0921-8890 |
| DOI: | 10.1016/0921-8890(91)90006-7 |
| Popis: | Models are used extensively to help understand the behaviour of a system. In order to interpret dynamic forces acquired from a tactile sensing robotic gripper during a task, it is useful to model and calibrate the lactile sensing system while keeping the modelling errors within an acceptable limit. The error limit is governed by uncertainties contributed by the system components. The development of a modelling scheme based on an expert system, which enables the user to select the best model consistent with the system hardware, has been described. The project consists of the development of a mathematical model, estimation of system uncertainty, development of an expert system for model selection, and embedding of the expert system in the processing software to form an integrated user-friendly application program. The case study describes the implementation and testing procedure used for validating this scheme for a two-fingered tactile sensing robotic gripper. The mean value of modelling error and a suitably defined confidence factor associated with it are provided for on-line decision making. This technique was used to extract an accurate force image, which could be subsequently interpreted during the performance of a task. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|