Multivariable model predictive controller

1. A system, comprising:

a multistep plant; and

a model predictive controller (MPC) comprising a model of the multistep plant, wherein the MPC is operable to linearize at least two steps of the multistep plant into cycle steps based on the model, to process an output signal from the multistep plant, and to independently control the cycle steps based on the output signal to optimize an output of the multistep plant.

2. The system of claim 1 , wherein:

the cycle steps comprise durations, and the MPC is further operable to independently vary the durations of the cycle steps.

3. The system of claim 1 , wherein:

the MPC comprises a Kalman filter operable to predict processing changes in the multistep plant based on operating conditions of the multistep plant, and to estimate initial conditions of the cycle steps.

4. The system of claim 3 , wherein:

the MPC further comprises an optimization module operable to perform a convex quadratic operation on the initial conditions of the cycle steps to generate control signals that independently control the cycle steps.

5. The system of claim 1 , wherein:

at least a portion of the MPC comprises firmware implemented in an embedded controller.

6. The system of claim 1 , wherein:

the multistep plant comprises one or more of a hydrogen separation plant, an air pre-purification plant, a carbon dioxide capture processing plant, an air separation processing plant, and a cardiac device.

7. The system of claim 1 , wherein:

the multistep plant comprises a medical oxygen concentrator (MOC) operable to produce oxygen from ambient air.

8. The system of claim 7 , wherein:

the MOC uses a cyclic multistep Rapid Pressure Swing Adsorption (RPSA) process to produce the oxygen.

9. The system of claim 8 , wherein:

the RPSA process comprises one or more of a Pressure Swing Adsorption (PSA), a Temperature Swing Adsorption (TSA), a Vacuum Swing Adsorption (VSA), and a Vacuum Pressure Swing Adsorption (VPSA).

10. The system of claim 7 , wherein:

the MPC is operable to optimize a product composition and a tank pressure of the MOC with each cycle step.

11. The system of claim 1 , wherein:

the MPC is further operable to perform a system identification of the multistep plant using a plurality of Pseudo-Random Binary Sequences to stimulate the at least two steps of the multistep plant.

12. The system of claim 1 , wherein:

the MPC is operable to process a plurality of output signals from the multistep plant to independently control the cycle steps of the multistep plant.

13. A method operable with a multistep plant, comprising:

generating a model of the multistep plant; and

via a model predictive controller (MPC):

linearizing at least two steps of the multistep plant into cycle steps based on the model;

processing an output signal from the multistep plant; and

independently controlling the cycle steps based on the output signal to optimize an output of the multistep plant.

14. The method of claim 13 , wherein:

the cycle steps comprise durations, and the method further includes, via the MPC independently varying the durations of the cycle steps.

15. The method of claim 13 , further comprising:

predicting processing changes in the multistep plant based on operating conditions of the multistep plant via a Kalman filter; and

estimating initial conditions of the cycle steps.

16. The method of claim 15 , further comprising:

performing a convex quadratic operation on the initial conditions of the cycle steps to generate control signals that independently control the cycle steps via an optimization module.

17. The method of claim 13 , wherein:

at least a portion of the method is implemented as firmware in an embedded controller.

18. The method of claim 13 , wherein:

the multistep plant comprises one or more of a hydrogen separation plant, an air pre-purification plant, a carbon dioxide capture processing plant, an air separation processing plant, and a cardiac device.

19. The method of claim 13 , wherein:

the multistep plant comprises a medical oxygen concentrator (MOC) operable to produce oxygen from ambient air.

20. The method of claim 19 , wherein:

the MOC uses a cyclic multistep Rapid Pressure Swing Adsorption (RPSA) process to produce the oxygen.

21. The method of claim 20 , wherein:

the RPSA process comprises one or more of a Pressure Swing Adsorption (PSA), a Temperature Swing Adsorption (TSA), a Vacuum Swing Adsorption (VSA), and a Vacuum Pressure Swing Adsorption (VPSA).

22. The method of claim 19 , further comprising:

optimizing a product composition and a tank pressure of the MOC with each cycle step.

23. The method of claim 13 , further comprising:

performing a system identification of the multistep plant using a plurality of Pseudo-Random Binary Sequences to stimulate the at least two steps of the multistep plant.

24. The method of claim 13 , further comprising:

processing a plurality of output signals from the multistep plant to independently control the cycle steps of the multistep plant.

25. A non-transitory computer readable medium, comprising instructions that, when executed by a model predictive controller (MPC), direct the MPC to:

generate a model of a multistep plant; and

linearize at least two steps of the multistep plant into cycle steps based on the model;

process an output signal from the multistep plant; and

independently control the cycle steps based on the output signal to optimize an output of the multistep plant.

26. The non-transitory computer readable medium of claim 25 , wherein:

the cycle steps comprise durations, and the non-transitory computer readable medium further comprises instructions that direct the MPC to independently vary the durations of the cycle steps.

27. The non-transitory computer readable medium of claim 25 , further comprising instructions that direct the MPC to:

predict processing changes in the multistep plant based on operating conditions of the multistep plant via a Kalman filter; and

estimate initial conditions of the cycle steps.

28. The non-transitory computer readable medium of claim 27 , further comprising instructions that direct the MPC to:

perform a convex quadratic operation on the initial conditions of the cycle steps to generate control signals that independently control the cycle steps via an optimization module.

29. The non-transitory computer readable medium of claim 25 , wherein:

at least a portion of the instructions are implemented as firmware in an embedded controller.

30. The non-transitory computer readable medium of claim 25 , wherein:

the multistep plant comprises one or more of a hydrogen separation plant, an air pre-purification plant, a carbon dioxide capture processing plant, an air separation processing plant, and a cardiac device.

31. The non-transitory computer readable medium of claim 25 , wherein:

the multistep plant comprises a medical oxygen concentrator (MOC) operable to produce oxygen from ambient air.

32. The non-transitory computer readable medium of claim 31 , wherein:

the MOC uses a cyclic multistep Rapid Pressure Swing Adsorption (RPSA) process to produce the oxygen.

33. The non-transitory computer readable medium of claim 32 , wherein:

the RPSA process comprises one or more of a Pressure Swing Adsorption (PSA), a Temperature Swing Adsorption (TSA), a Vacuum Swing Adsorption (VSA), and a Vacuum Pressure Swing Adsorption (VPSA).

34. The non-transitory computer readable medium of claim 31 , further comprising instructions that direct the MPC to:

optimize a product composition and a tank pressure of the MOC with each cycle step.

35. The non-transitory computer readable medium of claim 25 , further comprising instructions that direct the MPC to:

perform a system identification of the multistep plant using a plurality of Pseudo-Random Binary Sequences to stimulate the at least two steps of the multistep plant.

36. The non-transitory computer readable medium of claim 25 , further comprising instructions that direct the MPC to:

a plurality of output signals from the multistep plant to independently control the cycle steps of the multistep plant.