Currently, energy consumption in Indonesia has increased so that the utilization of renewable energy is more developed to supply projections for future energy needs. One of the renewable energy sources that is being developed is biogas, especially for household-scale biogas. There are several types of biogas implementation at the household scale, one of which is the use of biogas as generator fuel to produce electricity. Fuel generators can use biogas in full or mix gasoline with biogas fuel. Electric generator sets with dual gasoline-biogas fuel can save the use of gasoline as fuel and can also increase the performance of generators. The gasoline-biogas mixture ratio affects engine performance, one of which is the rotational speed. However, at present the ratio of gasoline to biogas is still manually regulated on household scale biogas usage. Based on these conditions, the artificial neural networks(ANN) method was developed in this study which aims to find the optimal ratio in order to get the generator set rotational speed characterization with the best engine performance value. A total of 300 variations of data were processed using 75% for training with the number of hidden nodes 100 net.trainParam.goal value = 0.0001, net.trainParam.lr = 0.01, and net.trainParam.epochs = 1000, and 25% for the test. This study produced a RMSE training value of 10.4812 at node 55 and a test RMSE value of 5.8301 with a rotational speed of 3445.87, and obtained the best ratio of 0.012 L / min gasoline and 5 L / min biogas.
Copyrights © 2020