688 / 2019-04-30 12:42:06

Application of BP neural network based on genetic algorithm optimization in thermal conductivity of nanofluids

BP neural network, optimization, genetic algorithm(GA), Thermal Conductivity

全体主题 > 纳米传热、纳米流体流动和纳米材料

In this paper, a hybrid model (BP-GA) including a back propagation network and a genetic algorithm is used to estimate the thermal conductivity of the nanofluid. The genetic algorithm is used to optimize the initial weight and threshold of BP neural network, so that the optimized BP neural network can better predict the function output. The purpose of genetic algorithms optimization BP neural network is to obtain better initial weights and thresholds of the network through genetic algorithm. The basic opinion is to use BP neural network prediction error to represent the initial weight and threshold of the network and initialize the individual value as the fitness value of the individual, and find the optimal individual through the selection, crossover and mutation operations, that is, the optimal initial BP neural network weight. In this study, CuO-ZnO mixed nanofluid with the mass fraction of 0%, 1%, 2%, and 3% were studied at temperatures of 25, 30, 35, 40, 45, 50, 55, and 60 °C, respectively. The thermal conductivity of the mixed nanofluid base liquid is 20:80, 40:60, 50:50, 60:40, 80:20, and the mass ratio of CuO-ZnO is 50:50. The mass fraction of CuO-ZnO, the temperature and the mixing ratio of different base liquids were used as input parameters, and the thermal conductivity was the output parameter, forming a 3-input ANN neural network. The predicted thermal conductivity results of BP neural network and genetic algorithm optimized BP neural network (GA-BP) was compared with experimental data. The results illustrate that the genetic algorithm-optimized prediction model has a good effect on the accuracy and stability of the prediction results.