Control Method and Device For Engine

• US 20100168989A1
• Filed: 04/09/2008
• Published: 07/01/2010
• Est. Priority Date: 04/09/2007
• Status: Active Grant

• Current Assignee: Shan Dong Shenpu Automobile Control Technology Co., Ltd.
• Original Assignee: Hau Zhao
• Inventors: Zhao, Hua, Gong, Chunyong, Gao, Xiaoqun
• Granted Patent: US 8,452,522 B2
• Time in Patent Office: Days
• US Class Current: 701/110
• CPC Class Codes: F02D 41/1401 characterised by the control or regulation method F02D41/1473, F02D41/1477 take precedence
• CPC Class Codes: F02D 41/2422 Selective use of one or more tables
• CPC Class Codes: F02D 41/2429 Methods of calibrating or learning
• CPC Class Codes: F02D 41/2432 Methods of calibration
• CPC Class Codes: F02D 41/2454 Learning of the air-fuel ratio control
• CPC Class Codes: F02D 41/2477 characterised by the method used for learning
• CPC Class Codes: F02D 41/26 using computer, e.g. microprocessor
• CPC Class Codes: G05B 13/0245 not using a perturbation signal
• CPC Class Codes: G05B 2219/2623 Combustion motor
• CPC Class Codes: Y02T 10/40 Engine management systems

Abstract

A control method and a control device of an engine are introduced. During the process of the engine controlling, the control unit does the adaptive learning for the actual target value of the feedback of different aims, and following the dynamic spectrogram generation strategy optimizing compares the adaptively learning parameter of the same working status and the same time with the basic spectrogram parameter. If the compared result doesn'"'"'t meet the condition, then keep the basic spectrogram parameter. And if it meets the condition, then the engine generates the dynamic spectrogram parameter. Based on the dynamic spectrogram combination strategy, the engine combines the basic spectrogram parameter and the dynamic spectrogram parameter generated to the combined spectrogram parameter instead of the basic spectrogram parameter.

Claims

1. A control method for an engine, which is executed by a control unit in the engine, comprising the following steps:

determining current working status according to basic operation conditions and feature signal values of a plural of sensors,

generating expected target values for different targets of the engine under the current working status according to basic parameter maps,

correcting the basic parameter maps,

and controlling the different targets of the engine;

which is characterized in that, during performing the control method for the engine, the control unit also executes the following steps;

step S1;

adaptively learning actual target values fed back by the different targets of the engine and performing comparison and judgment for optimization between the adaptively learned actual target values and the basic parameter maps under the same working status and the same basic operation conditions;

wherein, if a condition is not satisfied, the basic parameter maps are kept; and

if the condition is satisfied, dynamic parameter maps are generated, and

step S2;

forming combined parameter maps from the basic parameter maps and the dynamic parameter maps, and replacing the basic parameter maps with the combined parameter maps.

2. The control method for the engine according to claim 1, which is characterized in that:

the current working status comprises a medium and low load working status, a heavy load working status, a starting working status, an acceleration-deceleration working status and an idle working status.

3. The control method for the engine according to claim 1, which is characterized in that:

the combined parameter maps comprise combined ignition parameter maps of the engine, combined injection parameter maps of the engine, combined intake parameter maps of the engine, combined idle controlled parameter maps of the engine, combined EGR rate parameter maps, and combined ignition dwell angle parameter maps of the engine.

4. The control method for the engine according to claim 1, which is characterized in that:

the basic parameter maps comprise first parameter maps calibrated by an engine testing bench and second parameter maps calibrated by the engine testing bench and path parameter optimization.

5. The control method for the engine according to any of claim 1, which is characterized in that, the basic parameter maps are corrected according to a correction strategy, comprising the following steps:

correcting the basic parameter maps using the feature signal values of a plural of associated sensors which reflect the current working status of the engine;

correcting the basic parameter maps using soft measuring feature signal values which are obtained by performing a soft measuring method with a new feature signal processing way of the associated sensors on the variables that can'"'"'t be directly measured; and

correcting the basic parameter maps using the feature signal values of the associated sensors which are adaptively learned by performing weighted value matching during the correction step between the expected target values and the actual target values;

wherein, the feature signals of the associated sensors comprise the signals of the associated sensors of the engine and the feature signals deduced using soft measuring way, including a crankshaft position and rotational speed signal of the engine, a top dead center signal, a torque signal, an injection pulse width signal, a throttle valve position signal, an oxygen sensor signal, a fuel temperature signal, a machine oil temperature signal, an environment pressure signal, a power supply loop voltage signal, a water temperature sensor signal, an exhaust temperature signal, an intake pressure signal, an air-fuel ratio signal, a knock signal, a flame conduction angle signal, frequency selection probability distribution of knock signals in several cycles, and probability of a flame transmitting angle signal exceeding time threshold.

6. The control method for the engine according to claim 1, which is characterized in that, the step S1 further comprises:

generating a set of adaptive parameters according to learning the change of the working status conditions and operation conditions and the change of self factors of the engine, and storing the adaptive parameters temporarily by experiential clustering distinguished from the working status and operation conditions;

during the control process of different targets of the engine, continuously performing comparison and judgment according to optimization conditions between the basic parameter maps and the temporarily stored adaptively learned parameters under the same working status and the same operation conditions;

wherein, when the temporarily stored adaptively learned parameters match a dynamic map generation strategy, the dynamic parameter maps under the working status and the operation conditions are generated, which will be continuously learned in the later control and be continuously updated by performing repeatedly the above process.

7. The control method for the engine according to claim 6, which is characterized in that, the dynamic parameter maps are generated according to the dynamic map generation strategy, comprising the following steps:

a. determining a dynamic map generation region;

the basic parameter maps of a certain control target under the same working status and the same operation conditions and the various associated feature signal values that represent the working status and the operation conditions at the moment are taken as a data node;

the corrected basic parameter map y of this node is taken as the central value; and

the dynamic map generation region (y−

Δ

y, y+Δ

y) is generated with the difference between the expected control target value and the actual control target value as the basic reference radius;

b. determining a dynamic map optimization region in the generation region;

in the spatial region of the same dimension, the dynamic map generation trend is judged based on the magnitude of the changing rates of the various associated feature signal values representing this working status in this data node to determine whether the smaller region is on the (y−

Δ

y) side or on the (y+Δ

y) side;

after the determination, the median of the (y−

Δ

y) or (y+Δ

y) region is taken as the new node after target approaching, which is taken as the center to determine the new approached dynamic map generation region;

this process repeats and continuously approaches the target until the minimum region min(y−

Δ

y, y+Δ

y) appears as the optimization region;

c. generating a dynamic map;

when the various associated feature signal values representing this working status tend to become a constant ε

near zero, and the probability distribution of the associated feature signals after performing probability statistic processing is in the allowable range, the median point y_m in min(y−

Δ

y, y+Δ

y) is determined as the generated dynamic parameter map;

d. determining the dynamic map;

the above processes a-c are repeated and experiential clustering is performed;

when the changing rates ε

of the various associated feature signal values and the probability distributions of the associated feature signals are stable in an allowable changing range, this dynamic parameter map is determined and stored in a ferroelectric memory;

at the moment, the various associated feature signal values corresponding to the determined dynamic parameter map are a set of data nodes, and the set of data nodes are working status condition signals to determine the dynamic parameter map;

e. updating the dynamic map;

due to the characteristic change of the engine and the usage environment change during the control process, the control targets of the generated dynamic map also change;

during the a-d process taken on the composed data nodes, when the changing rates ε

of these various associated feature signal values and the probability distributions of the associated feature signals are determined to be out of the allowable changing range, the new dynamic parameter map is regenerated, determined by experiential clustering, and then used to update the unit of the original data nodes addresses.

8. The control method for the engine according to claim 1, which is characterized in that, the combined parameter maps are formed according to a dynamic map combination strategy in the step S2, comprising:

a. comparing the basic parameter maps and the generated dynamic parameter maps under the same working status and the same operation condition or under the same working status and very close operation conditions;

b. setting wholly or partly the dynamic parameter maps as the combined parameter maps of the engine according to the changing rates of the various associated feature signal values when the various associated feature signal values are same while the target parameters are different in the elements composing the data nodes, and shielding the corresponding basic parameter maps;

wherein, the step of the dynamic parameter maps being wholly or partly set as the combined parameter maps of the engine according to the changing rates of the various associated feature signal values comprises;

abandoning the dynamic parameter maps when the target is controlled by the determined dynamic parameter maps and the changing rates of the various associated feature signal values can'"'"'t be stable in the allowable region, setting the basic parameter maps under the working status and the operation conditions as the combined parameter maps of the engine, and regenerating the dynamic parameter maps;

c. calculating respectively the changing rates of the different feature signal values according to the value of the previous cycle and the value of the current cycle value when the various associated feature signal values are not all the same while the target parameters are same, comparing the changing rates to take the minimum to judge the optimization, and determining the combined parameter maps which include whole or partial dynamic parameter maps.

9. A control device for an engine according to the control methods in claim 1, comprises:

a microprocessor,

a power detecting and voltage stabilizing circuit, a communication interface CAN, LIN and external diagnosis circuit, a high-power drive circuit, a switching quantity drive circuit and a drive circuit, which are individually connected to the microprocessor; and

,

an input conditioning circuit, an input conditioning buffer circuit, analog signal channels and digital signal channels;

wherein a portion of analog signals from a plural of sensors are transported to the microprocessor via the input conditioning circuit and the analog signal channels, while the other portion of analog signals from the sensors are transported to the microprocessor via the input conditioning circuit and the digital signal channels;

wherein digital signals from the sensors are transported to the microprocessor via the input conditioning buffer circuit and the digital signal channels; and

wherein the microprocessor determines a current working status according to basic operation conditions, the analog signals and the digital signals;

generates expected target values for different targets of the engine under the current working status according to basic parameter maps; and

controls the different targets of the engine after the expected target values are corrected;

which is characterized in that, the control device for the engine further comprises;

a cerebellum model arithmetic controller CMAC which is connected to the microprocessor, and

a ferroelectric memory which is connected to the cerebellum model arithmetic controller CMAC;

wherein the cerebellum model arithmetic controller CMAC is used to learn adaptively actual target values fed back by different targets of the engine during performing the control method, performs comparison and judgment for optimization between the adaptively learned actual target values and the basic parameter maps, and when a condition is satisfied, generates the new dynamic parameter maps and stores into the ferroelectric memory; and

wherein the microprocessor forms combined parameter maps from the basic parameter maps and the dynamic parameter maps stored in the ferroelectric memory, and replaces the basic parameter maps with the combined parameter maps.

10. The control device for the engine according to claim 9, which is characterized in that:

the analog signals comprise an intake pressure or intake flowrate signal, a throttle valve position signal, an atmospheric pressure signal, an intake temperature signal, a cooling water temperature signal, an oxygen sensor signal, an environment temperature signal, an accelerator pedal signal and a system voltage variation signal;

the digital signals comprise a crankshaft position signal, an injection pulse width signal, a vehicle speed signal, a knock signal, an air-condition requesting signal, a power-assisted steering requesting signal, a neutral gear signal and a headlamp switch signal.

11. The control method for the engine according to claim 3, which is characterized in that, the combined parameter maps are formed according to a dynamic map combination strategy in the step S2, comprising:

a. comparing the basic parameter maps and the generated dynamic parameter maps under the same working status and the same operation condition or under the same working status and very close operation conditions;

b. setting wholly or partly the dynamic parameter maps as the combined parameter maps of the engine according to the changing rates of the various associated feature signal values when the various associated feature signal values are same while the target parameters are different in the elements composing the data nodes, and shielding the corresponding basic parameter maps;

wherein, the step of the dynamic parameter maps being wholly or partly set as the combined parameter maps of the engine according to the changing rates of the various associated feature signal values comprises;

abandoning the dynamic parameter maps when the target is controlled by the determined dynamic parameter maps and the changing rates of the various associated feature signal values can'"'"'t be stable in the allowable region, setting the basic parameter maps under the working status and the operation conditions as the combined parameter maps of the engine, and regenerating the dynamic parameter maps; and

c. calculating respectively the changing rates of the different feature signal values according to the value of the previous cycle and the value of the current cycle value when the various associated feature signal values are not all the same while the target parameters are same, comparing the changing rates to take the minimum to judge the optimization, and determining the combined parameter maps which include whole or partial dynamic parameter maps.

12. The control device for the engine according to claim 9, which is characterized in that:

the current working status comprises a medium and low load working status, a heavy load working status, a starting working status, an acceleration-deceleration working status and an idle working status.

13. The control device for the engine according to claim 9, which is characterized in that:

the combined parameter maps comprise combined ignition parameter maps of the engine, combined injection parameter maps of the engine, combined intake parameter maps of the engine, combined idle controlled parameter maps of the engine, combined EGR rate parameter maps, and combined ignition dwell angle parameter maps of the engine.

14. The control device for the engine according to claim 9, which is characterized in that:

the basic parameter maps comprise first parameter maps calibrated by an engine testing bench and second parameter maps calibrated by the engine testing bench and path parameter optimization.

15. The control device for the engine according to claim 9, which is characterized in that, the microprocessor corrects the basic parameter maps according to a correction strategy, comprising:

correcting the basic parameter maps using the feature signal values of a plural of associated sensors which reflect the current working status of the engine;

correcting the basic parameter maps using soft measuring feature signal values which are obtained by performing a soft measuring method with a new feature signal processing way of the associated sensors on the variables that can'"'"'t be directly measured; and

correcting the basic parameter maps using the feature signal values of the associated sensors which are adaptively learned by performing weighted value matching during the correction step between the expected target values and the actual target values;

wherein, the feature signals of the associated sensors comprise the signals of the associated sensors of the engine and the feature signals deduced using soft measuring way, including a crankshaft position and rotational speed signal of the engine, a top dead center signal, a torque signal, an injection pulse width signal, a throttle valve position signal, an oxygen sensor signal, a fuel temperature signal, a machine oil temperature signal, an environment pressure signal, a power supply loop voltage signal, a water temperature sensor signal, an exhaust temperature signal, an intake pressure signal, an air-fuel ratio signal, a knock signal, a flame conduction angle signal, frequency selection probability distribution of knock signals in several cycles, and probability of a flame transmitting angle signal exceeding time threshold.

16. The control device for the engine according to claim 9, which is characterized in that, the cerebellum model arithmetic controller CMAC executes:

generating a set of adaptive parameters according to learning the change of the working status conditions and operation conditions and the change of self factors of the engine, and storing the adaptive parameters temporarily by experiential clustering distinguished from the working status and operation conditions;

during the control process of different targets of the engine, continuously performing comparison and judgment according to optimization conditions between the basic parameter maps and the temporarily stored adaptively learned parameters under the same working status and the same operation conditions;

wherein, when the temporarily stored adaptively learned parameters match a dynamic map generation strategy, the dynamic parameter maps under the working status and the operation conditions are generated, which will be continuously learned in the later control and be continuously updated by performing repeatedly the above process.

17. The control device for the engine according to claim 16, which is characterized in that, the cerebellum model arithmetic controller CMAC generates the dynamic parameter maps according to the dynamic map generation strategy, comprising:

a. determining a dynamic map generation region;

the basic parameter maps of a certain control target under the same working status and the same operation conditions and the various associated feature signal values that represent the working status and the operation conditions at the moment are taken as a data node;

the corrected basic parameter map y of this node is taken as the central value; and

the dynamic map generation region (y−

Δ

y, y+Δ

y) is generated with the difference between the expected control target value and the actual control target value as the basic reference radius;

b. determining a dynamic map optimization region in the generation region;

in the spatial region of the same dimension, the dynamic map generation trend is judged based on the magnitude of the changing rates of the various associated feature signal values representing this working status in this data node to determine whether the smaller region is on the (y−

Δ

y) side or on the (y+Δ

y) side;

after the determination, the median of the (y−

Δ

y) or (y+Δ

y) region is taken as the new node after target approaching, which is taken as the center to determine the new approached dynamic map generation region;

this process repeats and continuously approaches the target until the minimum region min(y−

Δ

y, y+Δ

y) appears as the optimization region;

c. generating a dynamic map;

when the various associated feature signal values representing this working status tend to become a constant ε

near zero, and the probability distribution of the associated feature signals after performing probability statistic processing is in the allowable range, the median point y_m in min(y−

Δ

y, y+Δ

y) is determined as the generated dynamic parameter map;

d. determining the dynamic map;

the above processes a-c are repeated and experiential clustering is performed;

when the changing rates ε

of the various associated feature signal values and the probability distributions of the associated feature signals are stable in an allowable changing range, this dynamic parameter map is determined and stored in a ferroelectric memory;

at the moment, the various associated feature signal values corresponding to the determined dynamic parameter map are a set of data nodes, and the set of data nodes are working status condition signals to determine the dynamic parameter map;

e. updating the dynamic map;

due to the characteristic change of the engine and the usage environment change during the control process, the control targets of the generated dynamic map also change;

during the a-d process taken on the composed data nodes, when the changing rates ε

of these various associated feature signal values and the probability distributions of the associated feature signals are determined to be out of the allowable changing range, the new dynamic parameter map is regenerated, determined by experiential clustering, and then used to update the unit of the original data nodes addresses.

18. The control device for the engine according to claim 9, which is characterized in that, the microprocessor forms the combined parameter maps according to a dynamic map combination strategy, comprising:

a. comparing the basic parameter maps and the generated dynamic parameter maps under the same working status and the same operation condition or under the same working status and very close operation conditions;

b. setting wholly or partly the dynamic parameter maps as the combined parameter maps of the engine according to the changing rates of the various associated feature signal values when the various associated feature signal values are same while the target parameters are different in the elements composing the data nodes, and shielding the corresponding basic parameter maps;

wherein, the step of the dynamic parameter maps being wholly or partly set as the combined parameter maps of the engine according to the changing rates of the various associated feature signal values comprises;

abandoning the dynamic parameter maps when the target is controlled by the determined dynamic parameter maps and the changing rates of the various associated feature signal values can'"'"'t be stable in the allowable region, setting the basic parameter maps under the working status and the operation conditions as the combined parameter maps of the engine, and regenerating the dynamic parameter maps;

c. calculating respectively the changing rates of the different feature signal values according to the value of the previous cycle and the value of the current cycle value when the various associated feature signal values are not all the same while the target parameters are same, comparing the changing rates to take the minimum to judge the optimization, and determining the combined parameter maps which include whole or partial dynamic parameter maps.

19. The control device for the engine according to claim 13, which is characterized in that, the microprocessor forms the combined parameter maps according to a dynamic map combination strategy, comprising:

a. comparing the basic parameter maps and the generated dynamic parameter maps under the same working status and the same operation condition or under the same working status and very close operation conditions;

b. setting wholly or partly the dynamic parameter maps as the combined parameter maps of the engine according to the changing rates of the various associated feature signal values when the various associated feature signal values are same while the target parameters are different in the elements composing the data nodes, and shielding the corresponding basic parameter maps;

wherein, the step of the dynamic parameter maps being wholly or partly set as the combined parameter maps of the engine according to the changing rates of the various associated feature signal values comprises;

abandoning the dynamic parameter maps when the target is controlled by the determined dynamic parameter maps and the changing rates of the various associated feature signal values can'"'"'t be stable in the allowable region, setting the basic parameter maps under the working status and the operation conditions as the combined parameter maps of the engine, and regenerating the dynamic parameter maps;

c. calculating respectively the changing rates of the different feature signal values according to the value of the previous cycle and the value of the current cycle value when the various associated feature signal values are not all the same while the target parameters are same, comparing the changing rates to take the minimum to judge the optimization, and determining the combined parameter maps which include whole or partial dynamic parameter maps.