Model predictive control for HPI closed-loop fuel pressure control system

• US 5,609,136 A
• Filed: 06/28/1994
• Issued: 03/11/1997
• Est. Priority Date: 06/28/1994
• Status: Expired due to Fees

• Current Assignee: Cummins Engine IP Inc (Cummins Incorporated)
• Original Assignee: Cummins Engine Company Incorporated (Cummins Incorporated)
• Inventors: Tuken, Taner
• Primary Examiner(s): Moulis, Thomas N.
• Application Number:
  US08/267,606
• Time in Patent Office: 987 Days
• Field of Search: 123/478, 123/480, 123/458, 123/674, 123/357, 123/339.2, 364/149, 364/150, 364/151, 364/164, 364/165
• US Class Current: 123/357
• CPC Class Codes: F02D 2041/1409 using at least a proportional, integral or derivative controller
• CPC Class Codes: F02D 2041/141 using a feed-forward control element
• CPC Class Codes: F02D 2041/1412 using a predictive controller
• CPC Class Codes: F02D 2041/1415 using a state feedback or a state space representation
• CPC Class Codes: F02D 2041/1418 Several control loops, either as alternatives or simultaneous
• CPC Class Codes: F02D 2041/1431 the system including an input-output delay
• CPC Class Codes: F02D 2041/1432 the system including a filter, e.g. a low pass or high pass filter
• CPC Class Codes: F02D 2041/1433 using a model or simulation of the system
• CPC Class Codes: F02D 2041/1434 Inverse model
• CPC Class Codes: F02D 2200/0602 Fuel pressure
• CPC Class Codes: F02D 2250/31 Control of the fuel pressure
• CPC Class Codes: F02D 41/14 Introducing closed-loop corrections
• CPC Class Codes: F02D 41/1401 characterised by the control or regulation method F02D41/1473, F02D41/1477 take precedence
• CPC Class Codes: F02D 41/32 of the low pressure type F02D41/3082 takes precedence
• CPC Class Codes: F02D 41/3809 Common rail control systems common rail apparatus F02M55/025, F02M63/0225
• CPC Class Codes: G05B 13/048 using a predictor

Abstract

A closed-loop control system having two feedback paths 106, 108. A first feedback path contains an expected system output value generated by a system model 102. The second feedback path provides insensitivity to noise by heavily filtering the error between the model output 104 and the actual measured system output 74. Because of the delay associated with the filtering operation, the dual feedback paths result in an initial fast (but approximated) system response, followed by a delayed correction to that response.

Claims

1. A method of providing feedback in a closed-loop control system which controls a first system, comprising the steps of:

(a) predicting an expected output of the first system using a model of the first system;

(b) feeding back the expected output as a first feedback input to the control system;

(c) computing an error between an actual output of the first system and the expected output;

(d) filtering the error in order to attenuate unwanted frequency components;

(e) creating a second feedback input based on the filtered error;

(f) feeding back the second feedback input to the control system, wherein there is a delay between feedback of the first and second feedback inputs;

(g) producing a control system output based upon the first and second feedback inputs; and

(h) applying the control system output to the first system in order to control the first system.

2. The method claim 1 wherein step (d) comprises filtering the error.

3. The method of claim 1 wherein the error includes noise associated with measuring the actual output of the first system.

4. The method of claim 1 wherein tile first system is a fuel supply system for an internal combustion engine.

5. A method of providing fast response from a first system controlled by a closed-loop control system, comprising the steps of:

(a) inputting a desired first system output signal to the control system;

(b) generating a first system input signal from the control system based on the desired first system output signal;

(c) generating an expected first system output signal based on the first system input signal;

(d) feeding back the expected first system output signal as a first feedback input to the control system;

(e) applying the first system input signal to the first system;

(f) generating an actual first system output signal from the first system;

(g) generating an error input signal equal to a difference between the actual first system output signal and the expected first system output signal;

(h) generating an error output signal based on the error input signal; and

(i) feeding back the error output signal as a second feedback input to the control system, wherein there is a delay between feedback of the first and second feedback inputs;

(j) producing a control system output based upon the first and second feedback inputs; and

(k) applying the control system output to the first system in order to control the first system.

6. The method of claim 5 wherein the first system is a fuel supply system for an internal combustion engine.

7. The method of claim 6 wherein the desired first system output signal is a desired pressure.

8. The method of claim 6 wherein the first system input signal is a commanded duty cycle.

9. The method of claim 6 wherein the first system output signal is a fuel rail pressure.

10. The method of claim 5 wherein step (h) comprises filtering the error input.

11. The method of claim 5 wherein the error input signal includes noise associated with measuring the actual first system output.

12. A fuel pressure system employing model predictive closed-loop control, comprising;

a reservoir adapted to hold a quantity of fuel;

a fuel pump operable to supply the fuel to a fuel injection system at a rate;

a pressure control device operatively coupled to the fuel injection system and operable to establish a fuel pressure in the fuel injection system, wherein the fuel is delivered to at least one fuel injector at the fuel pressure;

a pressure transducer operatively coupled to the fuel injection system and operable to produce a measured fuel pressure signal proportional to the fuel pressure; and

a control system operable to receive an input from the pressure transducer and provide a pressure control output to the pressure control device, wherein the control system generates the pressure control output using a model predictive first feedback signal and a second feedback signal based on the measured fuel pressure signal.

13. The fuel pressure system of claim 12, wherein the pressure control device comprises a valve in the fuel line operable to allow a selected quantity of tile fuel to flow past the valve and drain back to the reservoir, wherein selection of the selected quantity determines the fuel pressure.

14. The fuel pressure system of claim 12, wherein the control system comprises an electronic control unit laving at least a throttle position input and an engine speed input, the electronic control unit comprising:

means for producing a desired fuel pressure signal based on at least the throttle position input and the engine speed input;

means for producing a commanded, fuel pressure signal based on the desired fuel pressure signal, the first feedback signal and the second feedback signal;

means for predicting an expected measured fuel pressure signal by applying the commanded fuel pressure signal to a model of the fuel pressure system;

means for feeding back the expected measured fuel pressure signal as the first feedback signal;

means for producing an error correction signal based on a difference between the measured fuel pressure signal and the expected measured fuel pressure signal; and

means for feeding back the error correction signal as the second feedback signal.

15. The fuel pressure system of claim 14, wherein there is a delay between feeding back the first feedback signal and the second feedback signal.

16. The fuel pressure system of claim 14, wherein the error correction signal is produced by filtering the difference between the measured fuel pressure signal and the expected measured fuel pressure signal.

17. A closed-loop control system employing model predictive control to control a first system, comprising:

means for predicting an expected output of the first system using a model of the first system;

means for feeding back the expected output as a first feedback input to the control system;

means for computing an error between an actual output of the first system and the expected output;

means for filtering the error in order to attenuate unwanted frequency components;

means for creating a second feedback input based on the filtered error;

means for feeding back the second feedback input to the control system, whereby there is a delay between feedback of the first and second feedback inputs;

means for producing a control system output based upon the first and second feedback inputs; and

means for applying the control system output to the first system in order to control the first system.

18. The closed-loop control system of claim 17, wherein the means for creating a second feedback input comprises a filter.

19. The closed-loop control system of claim 17, wherein the error includes noise associated with measuring the actual output of the first system.

20. The closed-loop control system of claim 17, wherein the first system is a fuel supply system for an internal combustion engine.

21. A method of providing feedback in a closed-loop control system which controls a first system, comprising the steps of:

(a) calculating a first portion of a control algorithm in a feedforward path;

(b) predicting an expected output of the first system using a model of the first system;

(c) using the expected output to calculate a second portion of the control algorithm, thereby creating a first feedback input;

(d) feeding back the first feedback input to the first portion;

(e) computing an error between an actual output of the first system and the expected output;

(f) filtering the error in order to attenuate unwanted frequency components;

(g) creating a second feedback input based on the filtered error;

(h) feeding back the second feedback input to the first portion, wherein both pole and zero locations of the control algorithm may be assigned arbitrarily;

(i) producing a control system output based upon both the first and second feedback inputs; and

(j) applying the control system output to the first system in order to control the first system.

22. The method of claim 21 wherein step (a) comprises calculating a PI algorithm.

23. The method of claim 21 wherein step (c) comprises using a compensator for the second portion of the control.