Predictive control systems and methods with hydrocracker conversion optimization

A method for operating a hydrocracking unit in an oil refining system, the method comprising:

• obtaining an objective function that quantifies a value of operating the hydrocracking unit as a function of at least one of a yield or a product quality of one or more output oil products provided as outputs of the hydrocracking unit;
• using a neural network model to generate a target control device setpoint predicted to optimize the objective function when the hydrocracking unit operates at the target control device setpoint;
• and
• operating a control device using the target control device setpoint to modulate at least one of the yield or the product quality of the one or more output oil products.

A control system for automatic operation of a hydrocracking unit in an oil refining system, the control system comprising:

• a control device operable to affect at least one of a yield or a product quality of one or more output oil products provided as outputs of the hydrocracking unit;
• and
• a controller comprising a processing circuit configured to:
• obtain an objective function that quantifies a value of operating the hydrocracking unit as a function of at least one of the yield or the product quality of the one or more output oil products;
• use a predictive model to generate a target control device setpoint predicted to optimize the objective function when the hydrocracking unit operates at the target control device setpoint;
• and
• operate the control device using the target control device setpoint to modulate at least one of the yield or the product quality of the one or more output oil products.