A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Prediction of Robusta green bean coffee moisture content based on bioelectric properties with artificial neural network method
An artificial neural network (ANN) is presented for predicting the moisture content of Robusta green-bean coffee. Moisture content is measured based on bioelectric properties using a capacitance sensor, where coffee beans are considered capacitors. This research aimed to develop predictive models of the moisture content of Robusta green bean coffee using bioelectrical properties with the ANN method. Moisture content was affected by the bioelectrical properties, and the bioelectric model of green bean coffee moisture content became a resistor-inductance-capacitor (R-L-C) series. Moisture content is observed for 37.5 hours, with data collection time intervals every 2.5 hours. This research obtains 4800 data with eight samples at a frequency of 100 Hz, 1 kHz, and 10 kHz. The best ANN structure to predict moisture content based on the bioelectrical properties is 9-30-30-1. The selected ANN topology results in an R training correlation coefficient of 0.99123, an R validation correlation coefficient of 0.90343, a training MSE of 0.0099, and a validation MSE of 0.1047. ANN models based on the bioelectrical properties have been proposed to develop an accurate, simple, and reliable technique as a sensor for the detection of the moisture content of green bean coffee during the drying process.
Prediction of Robusta green bean coffee moisture content based on bioelectric properties with artificial neural network method
An artificial neural network (ANN) is presented for predicting the moisture content of Robusta green-bean coffee. Moisture content is measured based on bioelectric properties using a capacitance sensor, where coffee beans are considered capacitors. This research aimed to develop predictive models of the moisture content of Robusta green bean coffee using bioelectrical properties with the ANN method. Moisture content was affected by the bioelectrical properties, and the bioelectric model of green bean coffee moisture content became a resistor-inductance-capacitor (R-L-C) series. Moisture content is observed for 37.5 hours, with data collection time intervals every 2.5 hours. This research obtains 4800 data with eight samples at a frequency of 100 Hz, 1 kHz, and 10 kHz. The best ANN structure to predict moisture content based on the bioelectrical properties is 9-30-30-1. The selected ANN topology results in an R training correlation coefficient of 0.99123, an R validation correlation coefficient of 0.90343, a training MSE of 0.0099, and a validation MSE of 0.1047. ANN models based on the bioelectrical properties have been proposed to develop an accurate, simple, and reliable technique as a sensor for the detection of the moisture content of green bean coffee during the drying process.
Prediction of Robusta green bean coffee moisture content based on bioelectric properties with artificial neural network method
Damayanti, Retno (author) / Ristianingrum, Wahyu Dwi (author) / Ubaidillah, Nazhif (author) / Al Riza, Dimas Firmanda (author)
2023-01-31
Advances in Food Science, Sustainable Agriculture and Agroindustrial Engineering (AFSSAAE); 6th International Conference on Green Agro-industry and Bioeconomy (ICGAB) July 2022 - Special Issue; 41-50 ; 2622-5921
Article (Journal)
Electronic Resource
English