Pemodelan dan Simulasi Produktivitas Perkebunan Kelapa Sawit Berdasarkan Kualitas Lahan dan Iklim Menggunakan Jaringan Syaraf Tiruan Modeling And Simulation Of Palm Oil Plantation Productivity Based On Land Quality And Climate Using ...

H. Hermantoro
Journal article Agromet • June 2009

Download full text
(Bahasa Indonesia, 7 pages)

Abstract

Crop growth and production on particular land and climate is strongly influenced by the interaction between plants, climate, soil, and management. Land quality and climate greatly affect the expected production of oil palm are: soil type, soil depth, altitude, soil pH, rainfall / year, average temperature, water deficit in mm / yr, air humidity, and solar radiation. Oil palm production as a function of land quality and climate can be predicted using various methods. Artificial Neural Network (ANN) is one recognized method for predict land productivity. In this study ANN Back propagation algorithm is used. The aim of this research is to develop ANN model and simulation of Oil Palm Plantation Productivity. Through the optimization procedure obtained the best ANN architecture is 11 neurons in input layer - 3 neurons in the hidden layer and - 1 neuron in the output layer, at 30,000 iterations of training step obtained the best model of oil palm productivity prediction with a value of R2: 0.98 and RMSE: 0:49, while from the test step obtains the value of R2: 0.94 and RMSE: 1.63. The results of simulation show that the simultaneous influence of several climatic changes that decrease the quantity of rainfall 100 mm / yr, 1 0C temperature rise, and increasing water deficit 50 mm / yr reduce the productivity of oil palm plantations for 2.15 tons / ha / year. From this research can be concluded that ANN can be used to predict the production of palm oil based on quality of land and local climate with very good results.

Metrics

  • 84 views
  • 179 downloads

Journal

Agromet

Agromet publishes original research articles and literature reviews in the fields of agricultural... see more