Study On The Etching Process Of High Voltage Anode Foil Based On Neural Network

With the development of science and technology, electronic devices are tend to be smaller and smaller, the volume of capacitor is an important factor to restrict the miniaturization of electronic devices. The effect of etching technology of aluminum foil is the key factor for the development of electrolytic capacitor with small volume high volume aluminum. Through the corrosion of aluminum foil have many inpacts and the process is more complicated, the increase of the capacitance is directly restricted. In this paper, the aluminum foil corrosion expansion process was researched, the process parameters of the corrosion were optimized, the uniformity and density of corrosion holes was improved and got high voltage anode foil which is needed for small volume of aluminum electrolytic capacitor.In this study, in order to obtain the best corrosion process parameters, orthogonal experiments, neural network technology, scanning electron microscopy were used to analyze the influence to the aluminum foil corrosion, which includes corrosion temperature, etching solution composition, current density, corrosion time, etc. At the same time, the influence of the concentration of sulfuric acid, hydrochloric acid and the concentration ratio on pitting of the aluminum foil was analyzed. Specific content as follows:First of all, on the basis of previous experimental data, the orthogonal test of aluminum foil etching process was designed to determine the scope of the initial selection parameters. Preliminary range of the parameters to be optimized for the process of the pitting: the corrosion temperature was 73~79?, the current time was 50~70 s, the current density was 0.25~0.35 A/cm2, the concentration of sulfuric acid was 2.5~3.5 mol/l, the concentration of hydrogen chloride was 0.8~1.2 mol/l; Preliminary range of the parameters to be optimized for the process of the etching: the corrosion temperature was 74~78?,the corrosion time was 11~13 min, the concentration of nitric acid was 1.8~2.2 mol/l.Secondly, in order to establish the prediction model of the process, the influence of pitting process was choosed as variable and the effect of the corrosion expansion is used as the objective of optimization. The corrosion temperature, corrosion time, corrosion current density, concentration of sulfuric acid, hydrochloric acid concentration as the decision variables of the neural network, corrosion foil bending strength and volume do for neural network optimization objectives, and the mind evolutionary algorithm of neural network weights optimization in order to improve the prediction model prediction accuracy and training speed. The experiment shows that the model can fit data accurately and predict the complex corrosion process parameters effectively.Finally, the optimal corrosion conditions were chosen through neural network simulating volume and bending strength of each factor under different levels that the corrosion temperature was 75?, the corrosion time was 63 s, the current density was 0.34 A/cm2. the concentration of hydrogen chloride was 2.9 mol/l, the concentration of sulfuric acid was 0.8 mol/l. The interaction in HCl-H2SO4 mixed acid corrosion system were analyzed and the optimum proportion of high specific quality etched foil was obtained.