AFM 1.0 is a simulation tool for fuzzy logic, which can automatically generate a database containing inference rules and function description parameters from given input-output pairs. The generated database is the optimal approximation of the input process/function. The tool can translate project files to FUZZYSTUDIO project files, MATLAB and C code, so that this environment can be used as a support for simulation and control.
The document describes the SGS-Thomson Microelectronics AFM 1.0 (Adaptive Fuzzy Modeller), a tool designed to automatically generate fuzzy logic knowledge bases from input/output data patterns. It supports both unsupervised rule learning and supervised membership function tuning using back-propagation algorithms.
| Input Variables | Up to 8 |
|---|---|
| Output Variables | Up to 4 |
| RAM (Recommended) | 16 Mbytes |
| Hard Disk Space Required | at least 1MBytes free space |
| Operating Temperature | 0- 70°C |
The system supports W.A.R.P. 1.1, W.A.R.P. 2.0, MATLAB, and ANSI C flow.
The system requires a 486 or Pentium compatible processor chip.