ProOpt performs process quality and efficiency optimisation using nonlinear model predictive control. The system learns the dynamics of the production process, after which is can predict the process output quality metrics given the current values for the control parameters. The user can also make simulated predictions about the key quality indicators with alternative control parameter values, so as to seek for better control. The system can also find optimal control values given the desired output quality and optimisation constraints, and can thus be used as an advisory tool for the operator or as an autonomous closed-loop controller.
- Predicts the quality indicators of the process given the historical measurements and chosen values for the process control parameters. - Can make simulated predictions about quality indicators with alternative control parameters. -Automatically finds optimal control parameter values in order to reach the desired product quality. Can also be used to find the most cost-efficient control parameter values (e.g. minimum amount of chemicals needed in order to reach desired output quality).
- Introduction - Process optimisation concept and approach - Example customer use case - Technical architecture in RAMP - Demonstration