Neural network based performance estimator for high energy electron beam irradiation process

摘要

High energy electron beam irradiation process has potential applications in water, wastewater, ans industrial waste treatment. Electron beam irradiation treatment of contaminated aqueous solutions have been experimentally demonstrated for a number of organic compounds under different operational conditions. At an acceleration voltage of about one million volts, electrons can travel less than a centimeter in water and produce excited states of compounds present in water, positive ions and electrons. Limited understanding of the process chemistry, fast reaction kinetics, complicated process chemistry, and limitations to adjust the process variables have been major challenges for the modeling efforts for the electron beam technology. A neural network based performance estimation program was developed to predict the effectiveness of electron beam technology for treatment of contaminated aqueous solutions. The program was used to estimate the process output in terms of destruction efficiency in relation to process variables which include pH, concentration, and irradiation dosage. The model was demonstrated for the case of phenol contaminated aqueous solutions.