The use of an electronic tongue is proposed for the first time in the supervision of the nutrient solution composition produced by a fertigation system. This new approach in chemical analysis consists of an array of non-specific sensors coupled with a multivariate calibration tool. For the studied case, the proposed system was formed by a sensor array of eight potentiometric sensors based on polymeric (PVC) membranes. The subsequent cross-response processing was based on a multilayer artificial neural network (ANN) model. This model predicted the concentration values of ammonium, potassium and nitrate, and the undesired saline sodium and chloride compounds in nutrient solution, drainage solution and tap water, departing from the information supplied by the sensor array plus temperature. Results showed good prediction ability for the different chemical species except for chloride, where the strong interfering effects of nitrate could not be balanced. The approach appears as a feasible method for the in-line assessment of nutrients and undesired compounds in fertigation solutions. © 2007 Elsevier B.V. All rights reserved.
|Journal||Computers and Electronics in Agriculture|
|Publication status||Published - 1 May 2007|
- Artificial neural networks
- Electronic tongue
- Potentiometric sensors
- Soilless culture