General Parameters

Configure these parameters when configuring the controller with the MPC Editor.

Note:

In the table below, "I" and "O" in the "I/O" column refer to input and output for the MPC block respectively.

A parameter can be classified as:

● I (input only)
● O (output only)
● IO (input and output)
● N/A (neither, set in offline configuration)

Parameter	Type	Description	Effect if increased	Effect if decreased	Notes	I/O
Name	String	The name of the controller.	N/A	N/A	None.	N/A
ControllerState	Integer	Sets execution mode of controller. 0 - offline/stopped (controller does not execute), 1 - online/running (controller executes, writes new MVs), 2 - warm (controller executes, echoes current MVs)	N/A	N/A	None.	IO
QPMaxIterations	Integer	The maximum number of iterations allowed for the quadratic programming solver per execution.	Increases maximum number of allowable iterations.	Decreases maximum number of allowable iterations.	None.	IO
QPTolerance	Double	After each iteration, a complimentary gap is calculated, and a tolerance on this value must be specified. After the tolerance is achieved, the output flag will be set to one, indicating that a feasible solution has been found, and the algorithm will then terminate.	Controller execution converges in fewer iterations, could potentially lead to poor results.	Controller execution takes longer to converge, potentially exceeding QPMaxIterations.	None.	IO
Sampling Interval	Integer	Execution time of controller.	Faster controller.	Slower controller.	Should be approximately smallest time constant divided by 3.	O
Prediction Horizon	Integer	Number of execution intervals the controller predicts into the future.	Above a certain threshold, has no real effect on controller performance, but causes longer execution time.	Must at least capture all dynamics, allow all models to settle at steady state. Must be less than or equal to the Model Horizon.	None.	O
Control Horizon	Integer	Number of execution intervals the controller can use to satisfy all constraints.	Slower controller, more moves are available to control to setpoint. Must be less than the Prediction Horizon.	More aggressive controller, fewer moves are available to reach control objectives.	None.	O
Model Horizon	Integer	Number of execution intervals for which the controller models are specified for to reach steady state.	Slower controller, more moves are available to control to setpoint. Must be less than the Prediction Horizon.	More aggressive controller, fewer moves are available to reach setpoint.	None.	O
Duality Gap	Double	Used to test for convergence.	N/A	N/A	None.	O
Iterations Used	Integer	Number of iterations used to solve QP problem.	N/A	N/A	None.	O
Converged	Integer	0 - QP has not converged to solution 1 - QP has converged to solution	N/A	N/A	None.	O
Execution Time Ms	Double	Time taken to perform an execution step (ms)	N/A	N/A	None.	O