Method for controlling an amount of material delivered during a material transfer

• US 6,441,322 B1
• Filed: 06/01/2000
• Issued: 08/27/2002
• Est. Priority Date: 11/24/1999
• Status: Expired due to Term

• Current Assignee: The Procter & Gamble Company
• Original Assignee: The Procter & Gamble Company
• Inventors: Ash, Raymond Houston, Chappell, David Arthur
• Primary Examiner(s): Gibson, Randy W.
• Application Number:
  US09/584,883
• Time in Patent Office: 817 Days
• Field of Search: 177/1, 177/116, 177/60, 177/119, 222/601, 222/77, 222/55, 222/56, 141/1, 141/192, 141/86, 141/311.A, 141/83
• US Class Current: 177/116
• CPC Class Codes: G01G 13/12 Arrangements for compensating for material suspended at cut-off, i.e. for material which is still falling from the feeder when the weigher stops the feeder
• CPC Class Codes: G01G 13/2857 Arrangements for compensating for material suspended at cut-off, i.e. for material which is still falling from the feeder when the weigher stops the feeder
• CPC Class Codes: G01G 23/3728 with wireless means
• CPC Class Codes: G01G 23/3735 using a digital network

Abstract

A method for controlling an amount of material delivered during a material transfer. This method includes the steps of inputting a target amount for a material to be delivered from a source location to a target location, updating the target amount during the delivery of the material using a predictive model-based algorithm, and updating the predictive model-based algorithm based on process performance data using a recursive least-squares procedure. An adaptive selection algorithm which is used to determine whether at least one previous predictive model-based algorithm parameter should be replaced with at least one new predictive model-based algorithm parameter in order to adapt the predictive model-based algorithm in response to a process change. A method for overlapping a delivery of at least one independently measured material with a delivery of a main material during a material transfer process.

Claims

1. A method for controlling an amount of material delivered during a material transfer, comprising the steps of:

inputting a target amount for a material to be delivered from a source location to a target location;

updating the target amount during the delivery of the material using a predictive model-based algorithm; and

updating said predictive model-based algorithm based on process performance data using a recursive least-squares procedure, wherein said predictive model-based algorithm updates the target amount to equal a dispensed amount plus a predicted spill amount, wherein said predicted spill amount is defined by the following predictive model;

2. The method of claim 1, where initial downward velocity of the material is equal to zero or independent of flow rate, such that:

3. The method of claim 1, where initial downward velocity of the material is proportional to flow rate, such that:

4. The method of claim 1, wherein said K₁ and said K₂ are updated from process performance data using said recursive least-squares procedure.

5. The method of claim 1, wherein an adaptive selection algorithm is used to determine whether at least one previous predictive model-based algorithm parameter should be replaced with at least one new predictive model-based algorithm parameter in order to adapt the predictive model-based algorithm in response to a process change, said determination is based on process performance data.

12. The method of claim 1, wherein said method is self-starting in that initial values of said predictive model-based algorithm parameters are automatically set from data of a first material delivery.

13. The method of claim 1, wherein said method is self-correcting in that said predictive model-based algorithm parameters are reset when changed process or material characteristics meeting specified criteria are detected.

14. The method of claim 1, wherein said Q is equal to the maximum measured or calculated flow rate of the material.

15. The method of claim 1, wherein said steps are contained within a computer-readable medium.

16. The method of claim 1, wherein said steps are contained within a digital control device.

17. The method of claim 1, wherein said steps are contained within a computer data signal embedded in a carrier wave.

6. A method for controlling an amount of material delivered during a material transfer, comprising the steps of:

inputting a target amount for a material to be delivered from a source location to a target location;

updating the target amount during the delivery of the material using a predictive model-based algorithm;

updating said predictive model-based algorithmn based on process performance data using a recursive least-squares procedure, wherein an adaptive selection algorithm is used to determine whether at least one previous predictive model-based algorithm parameter should be replaced with at least one new predictive model-based algorithm parameter in order to adapt the predictive model-based algorithm in response to a process change, said determination is based on process performance data;

defining a coordinate system to display values of a dependent variable versus values of at least one independent variable, said dependent variable being related to said independent variable by a mathematical function defining a predictive model;

defining a closed reference box on said coordinate system, said closed reference box having a fixed center point located at values of said dependent and independent variables based on predetermined acceptable material delivery data values, wherein dimensions of said closed reference box are initially set based on predetermined acceptable material delivery data values, wherein said dimensions and said center point being subject to a rule-based adjustment using process performance data;

defining a finite number of successively smaller closed selection boxes on said coordinate system, said closed selection boxes having each a center point located at values of said dependent and independent variables based on current averaged values of said dependent and independent variables as calculated from process performance data, wherein dimensions of said closed selection boxes are defined by successively smaller fractional multiples of said closed reference box dimensions; and

defining at least one decision rule by which a specific new data point having values of said dependent and independent variables derived from process performance data can be used to select a specific rule by which said predictive model will be modified, said selection depending on which of said boxes contain said data point.

7. The method of claim 6, further comprises the steps of:

defining two said closed selection boxes and four said decision rules as follows;

if said new data point is contained within both of the smaller of said selection boxes and said closed reference box, then updating said predictive model parameters,

else if said new data point is contained within both the larger of said selection boxes and said closed reference box but not within said smaller selection box, then reset predictive model parameters,

else if said new data point is contained within said closed reference box but not within either of said selection boxes, then on first occurrence within said region make no change to said predictive model parameters, else reset said predictive model parameters,

else if new data point is outside said closed reference box, then make no change to said predictive model parameters.

8. The method of claim 6, wherein said dependent variable is an actual spill amount.

9. The method of claim 6, wherein the number of said independent variables is equal to one and said independent variable is a cut-off flow rate of said material.

10. The method of claim 6, wherein said number of closed selection boxes is 2.

11. The method of claim 6, wherein said number of closed selection boxes is 3.

18. A method for controlling an amount of material delivered during a material transfer, comprising the steps of:

inputting a target amount for a material to be delivered from a source location to a target location;

delivering the material to the target location, wherein the rate of said delivery need not be controlled;

updating the target amount during the delivery of the material using a predictive model-based algorithm; and

updating said predictive model-based algorithm based on process performance data using a recursive least-squares procedure.