Abstract
A method to find dynamic parameters of a system using Neuro-Fuzzy techniques is proposed in this paper. In many model-based and adaptive control strategies, the precise or estimated knowledge of these parameters is required. In order to determine these parameters, we first begin by building an input/output mapping using ANFIS (Adaptive Neuro-Fuzzy Interference Systems) architecture based on input/output data pairs collected from experiment or simulation. Then, we use principles based on the properties of equation which described the dynamic model of robots in order to derive the dynamic parameters. Simulations on the systems where the mathematical dynamic models are well known demonstrate that proposed method is quite effective.
| Original language | English |
|---|---|
| Pages (from-to) | 15-32 |
| Number of pages | 18 |
| Journal | UPB Scientific Bulletin, Series D: Mechanical Engineering |
| Volume | 70 |
| Issue number | 2 |
| Publication status | Published - 2008 |
Keywords
- Dynamic parameters
- Neuro-fuzzy
- Pantograph mechanism
- Walking robot