Artificial Intelligence for Intelligent Control
The system includes modules with two main loops :
the control loop, which steers the process to the desired behavior
the Knowledge Learning loop, which is used to retrain the predictive model if the operating conditions changed that the existing model does not cover.
Our proprietary AML – patent-pending machine learning algorithm is designed to integrate with the process automation systems.
In reality, we have two types of process parameters: (1) external parameters which cannot be controlled but need to be managed, such as the weather, and (2) internal parameters which depend on the process itself and can be manipulated.
The objective of intelligent control is to minimize the process sensitivity to external forces so that it can operate optimally by manipulating its internal parameters that compensate for any external disturbance.
An intelligent model can be used for two purposes:
Running forward to predict the operations' behavior;
Running backward to produce operations' planning and scheduling
Our approach to the use of AI modules for creating intelligent models of processes is used to optimize the operation and minimize the cost of production. The purpose of intelligent models is to run in parallel to the process, predict its behavior and compare it to the actual performance.