Answers are grounded in this paper and cite their sources.
Void fraction in gas-liquid flow has been estimated using the artificial neural network (ANN) with electrical current response (ECR) from a voltage-current (VC) system, which measures electrical currents through the combination with eight electrodes attached around a pipe for gas-liquid flow. The ANN model for ECR with the number of neurons N = 3136 has the lowest relative error and absolute error, respectively, in estimating void fraction of gas-liquid flow, among five different numbers of neurons (N = 392, 784, 1568, 3136, and 6272) on a hidden layer of the ANN model for ECR. With regard to the change in total loss during the training process, the ANN model for ECR with N = 3136 shows good convergence as 28 × 10-4for void fraction estimation in gas-liquid flow compared to other ANN models with different numbers of neurons N. In this paper, electrical currents measured by the VC system in the static experiment with eight void fraction conditions are used to optimize the ANN model for ECR in order to estimate void fraction of gas-liquid flow. In the gas-liquid flow with seven void fraction conditions, the optimized ANN model for ECR is applied for void fraction estimation in order that the estimation performance according to the number of neurons on the hidden layer is evaluated.
No takes yet. Share an insight, caveat, or question.
Jeon et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: