Model generating method, model generating program, and simulation apparatus

• US 6,862,514 B2
• Filed: 11/10/2003
• Issued: 03/01/2005
• Est. Priority Date: 11/27/2002
• Status: Expired due to Fees

• Current Assignee: Toyota Jidosha Kabushiki Kaisha (Toyota Motor Corporation)
• Original Assignee: Toyota Jidosha Kabushiki Kaisha (Toyota Motor Corporation)
• Inventors: Ehara, Masato
• Primary Examiner(s): Vo, Hieu T.
• Application Number:
  US10/703,550
• Publication Number:
  US 20040102891A1
• Time in Patent Office: 477 Days
• Field of Search: 701/102, 731/173, 703/14, 703/19, 716/6
• US Class Current: 701/102
• CPC Class Codes: G05B 17/02 electric

Abstract

The present invention is directed to reduce the number of measuring processes of an actual apparatus. A final model for estimating the behavior of an engine 3 is generated by the following procedure. First, a simulation model in which physical properties of the engine 3 are considered is set. Next, measurement is roughly performed by an actual apparatus. Based on the difference between the simulation data and the measurement data, a correction function is Err, 130 derived. By combining the physical model and the correction function Err, 130, a final model is obtained. By using the final model, a final adaptive value is determined.

Claims

1. A simulation apparatus for generating a model to estimate a state of an object under a predetermined control condition, comprising:

an error data generating device for calculating error data indicative of a difference between actual measurement data obtained by measuring a state of said object corresponding to input data and simulation data obtained by inputting said input data to a physical model in which physical properties of said object are considered;

a correction function generating device for generating a correction function to correct said simulation data so as to come close to said actual measurement data based on a change in said error data corresponding to said input data; and

a final model generating device for generating a final model of said object by combining said physical model and said correction function.

2. The simulation apparatus according to claim 1, wherein said correction function generating device comprises:

a presenting device for presenting a change in said error data corresponding to said input data to a user and prompting the user to enter a form of said correction function;

an input device for outputting an output signal according to an operation of the user; and

a parameter determining device for specifying the form of said correction function including at least one parameter based on said output signal and determining a value of said parameter so that correction data obtained by inputting said simulation data to said correction function comes close to said actual measurement data.

3. The simulation apparatus according to claim 1, wherein said correction function generating device stores a form of said correction function which is predetermined and includes at least one parameter, and determines a value of said parameter so that correction data obtained by inputting said simulation data to said correction function comes close to said actual measurement data.

4. The simulation apparatus according to claim 1, wherein said correction function generating device determines a form of said correction function including at least one parameter by a predetermined computation based on a change in said error data corresponding to said input data, and determines a value of said parameter so that correction data obtained by inputting said simulation data to said correction function comes close to said actual measurement data.

5. The simulation apparatus according to claims 1, further comprising an adaptation data generating device for inputting a control condition of said object to said final model and generating adaptation data indicative of the state of said object.

6. The simulation apparatus according to claim 5, wherein the number of pieces of said adaptation data generated by said adaptation data generating device is larger than the number of measurement points at which said actual measurement data is measured.

7. The simulation apparatus according to claim 5, wherein said object is an internal combustion engine for burning a fuel, and

said control condition is an operating condition of said internal combustion engine.

8. A model generating program for generating a model to estimate a state of an object under a predetermined control condition, which allows a computer to execute the steps of:

calculating error data indicative of a difference between actual measurement data obtained by measuring a state of said object corresponding to input data and simulation data obtained by inputting said input data to a physical model in which physical properties of said object are considered;

deriving a correction function to correct said simulation data so as to come close to said actual measurement data based on a change in said error data corresponding to said input data; and

generating a final model of said object by combining said physical model and said correction function.

9. A model generating method of generating a model to estimate a state of an object under a predetermined control condition, comprising the steps of:

calculating error data indicative of a difference between actual measurement data obtained by measuring a state of said object corresponding to input data and simulation data obtained by inputting said input data to a physical model in which physical properties of said object are considered;

deriving a correction function for correcting said simulation data so as to be close to said actual measurement data based on a change in said error data corresponding to said input data; and

generating a final model of said object by combining said physical model and said correction function.

10. The model generating method according to claim 9, wherein the step of deriving said correction function includes the steps of:

determining a form of said correction function including at least one parameter based on a change in said error data corresponding to respect to said input data; and

determining a value of said parameter so that correction data obtained by inputting said simulation data to said correction function comes close to said actual measurement data.