Dual model-based temperature controller

Patent metadata

Field	Value
Publication number	US-12572164-B2
Application number	US-202318344543-A
Country	US
Kind code	B2
Filing date	Jun 29, 2023
Priority date	Jun 30, 2022
Publication date	Mar 10, 2026
Grant date	Mar 10, 2026

Abstract

Temperature control systems and methods for a thermal reactor having a process chamber, the control system comprising a first control loop comprising a first Model-Based Predictive Controller (MBPC) and a second control loop comprising a second MBPC, wherein the first and second MBPC are provided with predictive models representing the behavior of the thermal reactor.

First claim

The invention claimed is:

1 . A temperature control system for a thermal reactor having a process chamber, the temperature control system comprising:
a first control loop comprising a first Model-Based Predictive Controller (MBPC), which uses a spike temperature sensor signal from a spike temperature sensor as input, and which provides a power output signal that controls power to a heating element of the thermal reactor, the spike temperature sensor being located in proximity to the heating element and spaced from the process chamber, wherein the first MBPC is provided with a first predictive model representing behavior of the thermal reactor, the first MBPC being configured to calculate a first output value based on first calculations over a first predictive time horizon, using the first predictive model, said first output value controlling the power output signal; and a second control loop comprising a second MBPC, which uses a paddle temperature sensor signal from a paddle temperature sensor and the spike temperature sensor signal as inputs, and which provides as a spike temperature control setpoint that is used as input for the first MBPC, the paddle temperature sensor being spaced from the heating element and located inside or in proximity to the process chamber, wherein the second MBPC is provided with a second predictive model representing the behavior of the thermal reactor, the second MBPC being configured to calculate a second output value based on second calculations over a second predictive time horizon, using the second predictive model, the second output value controlling the spike temperature control setpoint.

2 . The temperature control system according to claim 1 , wherein the first or the second MBPC is provided with a generic linear dynamic model that characterizes thermal response of the thermal reactor.

3 . The temperature control system according to claim 2 , wherein the first and the second MBPC is provided with the generic linear dynamic model that characterizes the thermal response of the thermal reactor.

4 . The temperature control system according to claim 3 , wherein a steady state gain factor for the first and the second MBPC is different.

5 . The temperature control system according to claim 2 , wherein the first or the second MBPC adds model mismatch correction factors to a predictive calculation and an output optimization calculation to adapt the generic linear dynamic model.

6 . The temperature control system according to claim 1 , wherein the first or second MBPC comprises a trajectory planner which automatically reduces a specified ramp rate when approaching a constant temperature control setpoint.

7 . A control system comprising:
a first control loop comprising a first Model-Based Predictive Controller (MBPC) for controlling a plant with an output signal that controls power to a heating element in a process chamber of said plant, wherein said first MBPC is configured to receive first sensor data from at least one spike temperature sensor located in proximity to the heating element and spaced from the process chamber, and said output signal is based at least in part on first calculations in said first MBPC over a first predictive time horizon; and a second control loop comprising a second MBPC, wherein said second MBPC is configured to provide a control setpoint to said first MBPC, said control setpoint is based on second calculations in said second MBPC over a second predictive time horizon, and said second MBPC is further configured to receive second sensor data from: (1) at least one paddle temperature sensor located inside or in proximity to the process chamber and spaced from the heating element; and (2) the at least one spike temperature sensor.

8 . The control system according to claim 7 , wherein the first or the second MBPC is provided with a generic linear dynamic model that characterizes thermal response of the process chamber.

9 . The control system according to claim 8 , wherein the first and the second MBPC is provided with the generic linear dynamic model that characterizes the thermal response of the process chamber.

10 . The control system according to claim 9 , wherein a steady state gain factor for the first and the second MBPC is different.

11 . The control system according to claim 8 , wherein the first or the second MBPC adds model mismatch correction factors to a model predictive calculation and an output optimization calculation to adapt the generic linear dynamic model.

12 . A method for controlling a plant having a process chamber, comprising:
providing control inputs to said plant from a first control loop, said first control loop comprising a first Model-Based Predictive Controller (MBPC) configured to receive first sensor data from at least one spike temperature sensor located in proximity to a heating element and spaced from the process chamber, and said control inputs being based at least in part on first calculations in said first MBPC over a first predictive time horizon; and providing a control setpoint to said first control loop, said control setpoint being computed by a second control loop comprising a second MBPC configured to receive second sensor data from: (1) at least one paddle temperature sensor located inside or in proximity to the process chamber and spaced from the heating element; and (2) the at least one spike temperature sensor; said second MBPC further being configured to receive a control process sequence for said plant, and said second MBPC being configured to calculate said control setpoint based on second calculations in said second MBPC over a second predictive time horizon.

13 . The method according to claim 12 , wherein the first or the second MBPC is provided with a generic linear dynamic model that characterizes thermal response of the process chamber.

14 . The method according to claim 13 , wherein the first and the second MBPC is provided with the generic linear dynamic model that characterizes the thermal response of the process chamber and wherein a steady state gain factor for the first and the second MBPC is different.

15 . The method according to claim 13 , further comprising:
adapting, with the first or the second MBPC, the generic linear dynamic model by adding model mismatch correction factors to a model predictive calculation and an output optimization calculation.

Assignees

• Asm Ip Holding Bv

Inventors

• Zhenduo Liu
• Oosterlaken Theodorus G M

Classifications

• H05B1/0247

  For chemical processes · CPC title

• H10P72/0602

  Temperature monitoring · CPC title

• H10P72/0431

  Apparatus for thermal treatment · CPC title

• G05B13/042

  in which a parameter or coefficient is automatically adjusted to optimise the performance · CPC title

• G05B13/048 Primary

  using a predictor · CPC title