Influence Of Controlled Magnetic Field On Fouling Correlation Parameters And Prediction Model Construction Of Fouling Resistance

Heat exchanger is one of the basic thermal equipment widely used in industrial production.In the process of heat transfer,the fouling accumulated on the heat exchanger surface seriously restricts the heat transfer performance.Magnetic field is an effective and physical method for fouling control.The researches show that the fouling formation has significant nonlinear characteristics,and the fouling resistance,as an indicator to evaluate the accumulation state of fouling on heat exchange surface,is also a complex nonlinear function of multiple characteristic variables.Therefore,it is quite difficult to accurately predict and estimate the fouling resistance under the action of magnetic field based on the mechanism of fouling formation.Aiming at the prediction problem of fouling resistance,the selection of experimental conditions,the analysis of parameter characteristics,selection of input variables and prediction model construction of fouling resistance under the action of alternating magnetic field were studied in this paper.Specific research work is as follows:（1）Aiming at the selection of working conditions,an experimental system was designed and built independently.Taking the asymptotic value of fouling resistance as the evaluation index,the synergistic effect of four factors,namely magnetic induction intensity,magnetic field frequency,medium flow rate and inlet temperature,on the fouling formation of Ca CO₃ on heat exchange surface was studied.The results showed that magnetic induction intensity and magnetic field frequency,as external controllable factors,had certain influence on Ca CO₃fouling.Medium flow rate and inlet temperature were operating factors of the system.Increasing medium flow rate and decreasing inlet temperature can also effectively reduce fouling,but the setting of two operating parameters was limited by the operating state of the system.Therefore,in the follow-up study,when the medium flow rate and inlet temperature reached the best treatment effect,the alternating magnetic field will be treated,and the prediction of fouling resistance will be studied on the basis of further improving the fouling inhibition effect.（2）Aiming at the selection of variables related to fouling resistance,three parameters associated with Ca CO₃ fouling are selected in the process of the experiment:conductivity,p H and outlet temperature of the experimental fluid.The variation characteristics and mechanism of fouling resistance,conductivity,p H and outlet temperature under different alternating magnetic field parameters were analyzed.The results showed that the fouling resistance increased first and then became stable.The fouling resistances in the alternating magnetic field treatment group were lower than that in the non-alternating magnetic field treatment group,and there were extreme characteristics.The conductivity and p H of the alternating magnetic field treatment group decreased compared with that of the non-alternating magnetic field treatment group,and the decrease was more obvious when the fouling resistance was smaller.The change trend of outlet temperature was opposite to the fouling resistance,and the reduction of fouling resistance was equivalent to strengthen the heat transfer.According to the variation characteristics and mechanism,conductivity,p H and outlet temperature were related to the fouling resistance.However,whether they can be used as the input variables of the prediction model needs to be analyzed for correlation characteristics.（3）Aiming at the selection of input variables of prediction model,scatterplot matrix,spearman correlation coefficient and grey correlational analysis of degree of balance and approach analysis were used to analyze and quantify the correlation between fouling resistance and three related variables.The results showed that the absolute values of correlation coefficients between fouling resistance and its related variables were about 0.8,and the degrees of balance and approach were above 0.7.It shows that the fouling resistance had strong correlation with three related variables.Conductivity,p H and outlet temperature can be used as input variables to construct the fouling resistance prediction model.（4）Aiming at the prediction model construction of fouling resistance,a fouling resistance prediction model based on support vector regression method was proposed,taking conductivity,p H and outlet temperature as input variables and fouling resistance as output variables.The model took SVR method as the main body.The grey wolf optimization（GWO）algorithm with adaptive differential evolution（ADE）mechanism was used to optimize the penalty factor and kernel function parameter.The prediction results of two prediction models,SVR and ADEGWO-SVR,were compared.Experimental data verified that in the two types of fouling resistance prediction,the ADEGWO-SVR prediction model showed higher prediction accuracy and stronger generalization ability.It can closely track the variation trend of fouling resistance in the growth stage,gradually approximate the asymptotic value of fouling resistance in the stable stage,and has strong nonlinear processing ability and robustness.The accurate prediction of fouling resistance under alternating magnetic field was realized.