| In this paper,the scaling and corrosion of tube-in-tube heat transfer are studied,and three kinds of intelligent algorithm prediction models are established and compared.In order to reduce the corrosion rate,the model with the highest prediction accuracy is selected and the related variables are optimized.First of all,four types variables affecting corrosion are chosen.They are:equipment basic data,real-time data,laboratory data and calculation data.Eight main variables are selected by Grey Relational Analysis(GRA)and qualitative analysis.The selected variables are input variables for the later models.Secondly,three corrosion rate prediction models are established,which are Grey prediction model,Elman model and GA-Elman model.GM(1,1)is selected as the Grey prediction model.The number of prediction samples is 10,20,30,40 and 50 respectively,The average relative error under different number of prediction samples is calculated according to the real value and predicted value of corrosion rate,and the predicted values of corrosion rate in the next 15 periods are calculated.However,the Grey prediction model does not consider the relevant factors affecting the scaling and corrosion of heat exchanger so the universality is poor and the Grey prediction model only suits for short-term prediction.For Elman model,the number of neurons in the hidden layer of the network is set to 7 by the empirical function,and the learning rate of the model is 0.0025 by the cross test method.When the model is trained,the corrosion rate on the test set is predicted.Genetic algorithm is used to optimize the initial weights and thresholds of Elman neural network.The accuracy of the three models are evaluated by average absolute error MAD and average absolute percentage error MAPE.It is found that the accuracy of GA-Elman model is higher than the other two models.Finally,the inlet temperature and pressure drop of the operating variables in the pipeline are optimized by GA-Elman model to slow down the corrosion rate.The results show that when the inlet temperature of circulating cooling water decreases 0.1Centigrade,the predicted corrosion rate decreases about 0.1%.It is found that when the inlet temperature is lower than 22 ℃,the above relationship is vanished.And the corrosion rate is reduce by 0.18% on avarage.In the same way,The corrosion rate decreases as the pressure drop increases.Hence,theoretically,the corrosion rate can be reduced by decreasing the inlet temperature or increasing the press drop to prolong the life of the heat exchanger. |
