JPH01138335A - Engine fuel system dynamic characteristic analysis method - Google Patents

Abstract

PURPOSE:To simulate accurately the model of fuel flow properties within inlet pipes in various operating ranges by computing parameters for a fuel system model based on both the quantity of fuel injection converted into mass flow per unit time and time series data for the measured value of an air-fuel ratio of exhaust gas. CONSTITUTION:In an engine with a single point fuel injection system, fuel is injected by a single point fuel injection valve 24 into all cylinders of the engine. In this case, detected signals from an air capacity sensor 23, a throttle angle sensor 25, an internal pressure sensor 26, a crank angle sensor 27, a water temperature sensor 28 and an air-fuel ratio sensor 29 are inputted into a computer 21 respectively via a A/D converter 22. And parameters for the fuel system model representing fuel flow properties within inlet pipes are computed based on both the commanding value for the quantity of fuel injection which represents the quantity of fuel injection converted into mass flow per unit time and time series data for the measured value of an air-fuel ratio of exhaust gas. This constitution thereby allows the model of fuel flow properties within inlet pipes to be accurately simulated.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention is an engine fuel system dynamic characteristic analysis method suitable for formulating fuel flow characteristics in an intake pipe.

[Traditional techniques, techniques]

As a conventional method for formulating fuel flow characteristics in an intake pipe, a method described in Society of Automotive Engineers of Japan, Academic Lecture Preprint Collection 842049 is known. In this method, engine state quantities such as engine speed and intake pipe internal pressure are kept constant, and the response of the exhaust gas air-fuel ratio is measured with an air-fuel ratio sensor when the fuel injection amount is changed in steps. The parameters in the model were determined to best match the results of a simulation under the same conditions using a flow model, and these were formulated as a map of engine state quantities.

[Problem that the invention seeks to solve]

However, with conventional methods, it is difficult to calculate optimal parameters because the agreement between experimental results and simulation results is visually determined, and a trial-and-error method must be used to calculate parameters, which requires a lot of man-hours. There was a problem that it was necessary.

Furthermore, when formulating the calculated parameters as a map of engine state quantities, no consideration was given to the selection of state variables such as the number of revolutions and intake pipe internal pressure, resulting in the problem that accurate modeling was not possible. For this reason, when controlling the engine using a control algorithm derived from a conventional model, the expected control performance cannot be obtained in various driving regions, leading to problems such as deterioration in exhaust gas purification performance, power performance, fuel efficiency, etc. .

Further, in the conventional method, consideration is not clearly given to various differences such as a response delay of an air-fuel ratio sensor, a flow delay of exhaust gas in an exhaust pipe, etc., and as a result, there is a problem that the accuracy of modeling is reduced.

An object of the present invention is to provide a fuel system dynamic characteristic analysis method that eliminates the above-mentioned problems.

[Means for solving problems]

The above purpose is to eliminate the delay between fuel injection and its effect appearing in the output of the air-fuel ratio sensor, such as response delay of the air-fuel ratio sensor to detect the exhaust gas air-fuel ratio, and flow delay of exhaust gas in the exhaust pipe. A step for formulating a delay that cannot be expressed by a fuel system model that expresses fuel flow characteristics in the pipe, and input obtained from the delay model formulated in the step and the fuel system model including unknown parameters is a fuel injection amount, A step of uniquely calculating the parameters of the fuel system model based on experimental data so that the equation error of the difference equation of the exhaust gas air-fuel ratio is minimized; This is achieved by providing a step of formulating the characteristics as a map or function of engine state quantities such as .

[Effect]

The sensor response delay, exhaust gas flow delay, etc. are formulated as follows: An injection valve for gaseous fuel such as LP gas that does not adhere to the intake pipe wall is installed at the fuel injection position of the engine. The response of the exhaust gas air-fuel ratio when the gas injection amount is changed in steps is measured using an air-fuel ratio sensor, and based on this measurement value, the delay is determined using a transfer function or an impulse response model to determine the unknown parameters of the formula fuel system model.
Since the error equation of the difference equation obtained from the fuel system model and the delay model formulated using the above method is linear with respect to the unknown parameters, the optimal parameters in the sense that the root mean square of the equation error is the minimum can be easily calculated. is uniquely determined.

Furthermore, regarding parameter characteristic formulation, rational modeling can be performed by considering the physical meaning of the parameters and using them as maps or functions of engine state quantities that are considered to have a large influence.

〔Example〕

An embodiment of the present invention will be described below with reference to FIGS. 1 to 4.

The following equation is used as a model to express the fuel flow characteristics in the intake pipe, which is the object of modeling.

G*e=(L X)・Gz+−・Mi −(
1) dt τ where Gi: The fuel injected by the injector is the fuel flow rate (g
/s) (hereinafter referred to as fuel injection) Mi: Amount of fuel adhering to the intake pipe wall (g) (hereinafter referred to as liquid film) Gze Fuel flowing into the cylinder is expressed as fuel flow rate (g/s)
X: The rate at which the injected fuel adheres to the intake pipe wall (hereinafter referred to as the adhesion rate) m: The rate at which the liquid film evaporates per unit time (hereinafter, τ is the evaporation rate and τ is the evaporation time constant) ) τ.

Below, the fuel system model equations expressed by equations (1) and (2) are defined.

formulation, i.e., the formulation τ of the model parameters X,− Explain the method of conversion.

Parameter formulation is performed through three steps shown in FIG.

First, as a first step, we will formulate the delay between fuel injection and its effect appearing in the output of the air-fuel ratio sensor installed in the exhaust pipe, which cannot be expressed by the fuel system model, as shown in Figure 2. Perform experimentally with equipment.

In the experimental equipment, an injection valve for gaseous fuel such as LP gas is installed instead of an injector in the original fuel injection valve installation position, and the engine status such as intake air amount and internal pressure is monitored by sensors and A.
The data is imported into the computer through a /D converter. In such experimental equipment, intake air volume, intake pipe internal pressure, and water temperature.

The throttle opening and rotation speed are kept constant, the engine is operated steadily, and the response of the exhaust gas air-fuel ratio is measured with an air-fuel ratio sensor installed in the exhaust pipe when the gas injection amount is changed in steps. The sampling data of the exhaust gas air-fuel ratio from the time when the sampling period of the exhaust gas air-fuel ratio is changed by step of Δt gas injection amount is 5(k) (k=o, 1.2...)
shall be.

Here, in FIG. 2, 21 is a calculator, 22 is an AD converter, 23 is an air amount sensor, 24 is a gas fuel injection valve, 2
5 is a throttle angle sensor, 26 is an internal pressure sensor, 27 is a crank angle sensor, 28 is a water temperature sensor, and 29 is an air-fuel ratio sensor.

At this time, the delay that cannot be expressed by the above fuel system model can be solved using the relational expression between the injection amount Gto (g/s) and the measured value A/F of the exhaust gas air-fuel ratio when the gas injection amount is converted to a value equivalent to the fuel flow rate of gasoline. It is expressed as the following formula.

...(3) S (ko) S (0) (i=o,..., kol) Here, Qa: Intake air amount (g/5) ko: Time at which the air-fuel ratio response settles down It is.

Here, the reason why the delay is formulated as a moving average of the fuel-air ratio is to make the mathematical model of two variables, the fuel injection amount and the exhaust gas air-fuel ratio, which will be described later, linear τ with respect to the parameters X and -.

Next, in the second step, experimental data is collected under various operating conditions, and is adjusted to correspond to the operating conditions.
3) Formulate a delay model in the form of equation (4). A difference equation between two variables, the command value Gi of the fuel injection amount and the measured value A/F of the non-gas air-fuel ratio, is derived, and the parameter X is set so that the equation error is minimized.

− is determined. Considering the fuel flow delay in the intake pipe τ included in the model of equation (3) as dead time, the above difference equation sets GzO equal to Gze, and the difference equation of equations (1) and (2) and (3) It is obtained by combining the equations and becomes the following equation.

In equation (5), the error equation is the parameter X.

Since it is linear with respect to -, the optimal parameter in the sense that the square of the equation error is minimum is calculated by the following equation.

(Σa(k)2)−(Σb(k)”)−(Σa(k)b
(k))”...(6) (Σa(k)2)・(Σb(k)2)−(Σa(k)−
bOc))”...(7) ...(8) ...(9) ・(Gn(i)-Gz(i 1))
...(10) In order to calculate X, - using equations (6) to (10), τ time series data of fuel injection amount Gz(k), time series data of measured value of exhaust gas air-fuel ratio A/ F(k) is required. These data were obtained under the same engine operating conditions as when the lag model was derived in the first step, i.e., the intake air amount, water temperature, etc. were kept constant, the engine was operated steadily, and the fuel injection amount was randomly varied. It is obtained by measuring the response of the exhaust gas air-fuel ratio at

In order to calculate The calculated parameters are made to correspond to the engine operating conditions during the experiment.

Finally, in the third step, the parameters X, - determined corresponding to the engine operating state are formulated as a function or map of at least one variable expressing the operating state τ. Selecting variables that are considered to have a particularly large influence on the adhesion phenomenon include throttle opening, intake air amount,
Intake pipe internal pressureWater temperature, variables that are considered to have a particularly large influence on the liquid film evaporation phenomenon are water temperature, intake air amount, and intake pipe internal pressure. The parameters X, - are characteristically formulated as a relationship between each selected variable or as a τ map. Note that, after formulating using the above method, if a variable that has a small influence on the parameters is found, that variable is excluded from the characteristic expression.

This allows the memory in the ROM to be saved when control is performed using the characteristics of X, -.

FIG. 3 shows a simplified embodiment of the procedure τ for formulating the parameters X and - described above. It is easy to formulate the actual formulation based on this flow.111 First, in step 31, what values should be maintained for the throttle opening, intake air amount, and intake pipe internal pressure 2 rpm at which the engine should be operated? , to set engine operating conditions.

Next, in step 32, based on the operating conditions set in step 31, the engine is operated steadily, and the gas injection amount of P gas is changed step by step, and the exhaust gas air-fuel ratio response 5 ( k). 22, k=
o is the time at which the fuel is changed in steps.

In step 33, the response A/F(k) of the exhaust gas air-fuel ratio when the fuel injection amount Gi(k) is randomly changed by the injector under the same engine operating conditions is measured.

In step 34, the model gain G(k) is calculated from the time series data 5(k) collected in step 32 using equation (4).
Calculate.

In step 35, the nine time series data at(k), A/F(k) collected in step 33, and the gain Gz(lc) of the lag model calculated in step 34.
,as well as.

From the intake air amount Qa set in step 31, (6) to (
10) Calculate the adhesion rate X and evaporation rate - based on the formula. The above steps are repeated under various operating conditions (step 36).Next, in step 37, X and - determined corresponding to various engine operating conditions are adjusted to throttle opening, intake air amount, intake pipe internal pressure, and Formulated as a function or map of water temperature, intake pipe internal pressure, intake air amount, and water temperature.

Finally, in step 38, after the characteristic formulation, X.

If a variable is found that seems to have a small influence on −, then τ If so, perform characteristic expression excluding that variable.

This completes the formulation of the fuel system model.

In addition, if the evaporation rate - characteristic τ, which is formulated assuming a constant water temperature, changes rapidly on the line where the intake pipe internal pressure is constant, as shown in Figure 4, then it will change in the region to the right of the dotted line in the figure. It is assumed that a phenomenon in which the liquid film flows through the intake pipe wall surface]2 and directly flows into the cylinder is occurring.In this region, a newly established variable, the liquid film cylinder inflow rate α (α: Mj per unit time Let us express the evaporation rate τ calculated by equation (7) as the sum of the amount of liquid film flowing into the cylindane (a variable) and the evaporation rate τ, that is, (α+−). This makes it possible to model fuel flow characteristics including the liquid film cylinder inflow phenomenon using the original τ method.

As described above, according to the present invention, it is possible to uniquely determine the optimal parameters X, - in terms of the minimum root mean square error, and reduce the number of man-hours for parameter calculation. Furthermore, since all delays other than the fuel flow delay in the intake pipe are taken into account, the parameters X and - can be calculated with high accuracy.

τ Also, since the parameter characteristics are formulated rationally, highly accurate modeling of fuel flow characteristics can be performed in various operating regions.

〔Effect of the invention〕

As described above, according to the present invention, fuel flow characteristics in the intake pipe can be accurately modeled in various operating regions.

Therefore, if the model formulated by the method of the present invention is used for engine control such as air-fuel ratio control, the expected control performance can be obtained in various driving ranges, and the exhaust gas purification performance and
This has the effect of improving power performance, fuel efficiency, etc.

[Brief explanation of the drawing]

Claims (1)

[Claims] 1. In an engine with a single point fuel injection system in which fuel is supplied to all cylinders by a single injection valve, at least the intake air amount, intake pipe internal pressure, water temperature, throttle opening, and rotation speed are An air-fuel ratio sensor installed in the exhaust pipe measures the response of the exhaust gas air-fuel ratio when the fuel injection amount is changed under specified conditions under a constant engine operating state. An engine characterized in that parameters of a fuel system model expressing fuel flow characteristics in an intake pipe are calculated from a command value of a fuel injection amount, which is an amount converted to a mass flow rate, and time series data of a measured value of an exhaust gas air-fuel ratio. fuel system dynamic characteristics analysis method. 2. A gaseous fuel injection valve is provided at the fuel injection position, and the air-fuel ratio sensor measures the response of the exhaust gas air-fuel ratio when the gas injection amount is changed under predetermined conditions under the engine operating state; From the time-series data of the command value of gas injection amount and the measured value of exhaust gas air-fuel ratio, we can calculate the delay that cannot be expressed by the fuel system model among the delays from fuel injection until the effect appears in the output of the air-fuel ratio sensor. 2. The engine fuel system dynamic characteristic analysis method according to item 1, characterized in that mathematical modeling is performed. 3. From the model obtained in the second term and the fuel system model, derive a difference equation whose input is the command value of the fuel injection amount and the output is the measured value of the exhaust gas air-fuel ratio, so that the equation error of the error component equation is minimized. 2. The engine fuel system dynamic characteristic analysis method according to item 1 or 2, characterized in that parameters of the fuel system model are calculated. 4. Parameter 1, characterized in that the parameters of the fuel system model corresponding to various engine operating states are calculated, and the parameter characteristics are formulated as a function or a map of at least one variable expressing the engine operating state. 3. A fuel system dynamic characteristic analysis method for an engine according to item 3.