Abstract
Diagnostic expert system (DES) for NFT-hydroponic culture is designed and examined based on a computer integrated cultivation system (CICS) .
DES is composed of four frames of knowledge referred to as“searched out mechanical and electric error functions on sensors, ”“searched out error functions on controllers, ”“continuous control”and“searched out error functions caused on crops, ”and these frames consist of production rules. Measurement data are compared with simulated data from system identification on autoregressive (AR) model. In case of these differences increasing, DES will search out error functions and continuously control.
In this paper, DES diagnosed trouble caused by an ion-sensor. DES searched out an error function on the ion-sensor, and continuously controlled ion-concentrations.
DES uses not only ion concentration control but also other applications. For automation of cultivation and control in NFT-hydroponic culture, CICS as proposed in this paper is the most suitable system at the moment.