Implementing Artificial Neural Intelligence on Cyclic Voltammetry Method for Electrochemical Process toward Pyroprocessing Safeguards Application

A novel data simulation using an artificial neural intelligence (ANI) method has been applied on massive cyclic voltammetry (CV) electrochemical data sets (350,000 points) to predict elemental trends in an absence of experimental data with a high accuracy of prediction. Here, CV data sets reported by Hoover (2014) for 5, 7.5, and 10 wt% of UCl3 in LiCl-KCl molten salt under different scan rates at 773 K have been considered. All codes were written in Matlab with an ANI implementation through iterations and interrelationships among system variables such as currents, potentials, concentrations, scan rates, processing time, and weight percent. Possible structures giving a minimum average percent error were used and the simulated CV results were compared with the actual experimental data. This work illustrates ANI capability as a rapid concentration detection and measurement for CV graphs by predicting the trend of species in a blind and unseen situation with a high accuracy. ANI possibly can be used as an alternative method for signal detection towards safeguards application in the electrochemical process.