Monitor and Control of Drug Preparation using Deep Q Neural Network

In the pharmaceutical industry, an evaporator is used to vaporize the solvent from a solution and prepare the drug for use. Because of its significance, evaporation is often treated as a separate process. If it is being evaporated, the remaining material will dry out and become solid. Evaporation may seem like it would dry up a solution, but in reality, it concentrates it. Keeping the solution's concentration constant is a huge task. Maintaining the proper concentration of the solution is a primary focus of this study. As a first step, we construct the mathematical model of the evaporator. Second, controllers like Model Predictive Controllers (MPC) and Deep Q Network (DQN) controllers are designed. Third, the controllers are evaluated using error measures like Integral Square Error (ISE), Integral Absolute Error (IAE), and Integral Time Square Error (ITSE) by giving the required set-point to the system. The evaluation result shows that the DQN produce minimal error of ISE, IAE, and ITSE and their values are 14.26, 9.45, and 671.82. Additionally, the DQN controller reach the set-point with minimal settling time, and no overshoot and undershoot characteristics.