EP4273392A1

EP4273392A1 - Method and device for controlling a throttle valve of an internal combustion engine

Info

Publication number: EP4273392A1
Application number: EP23168997.7A
Authority: EP
Inventors: Fabio Ramundo; Tommaso CANGEMI; Raul Rodriguez
Original assignee: FPT Industrial SpA
Current assignee: FPT Industrial SpA
Priority date: 2022-04-26
Filing date: 2023-04-20
Publication date: 2023-11-08

Abstract

Method of controlling a throttle valve of an internal combustion engine, the valve comprising an electric actuator for controlling a position of a shutter and a position sensor for detecting a position of the shutter, the method comprising a step of controlling said actuator with a control signal (Volt_act) proportional to a signal representing a position error (e_pos) when an absolute value of the position error is less than or equal to a predetermined threshold (e_s) and proportional to a signal representing a speed error (e_vel) when the absolute value of the position error is greater than said predetermined threshold (e_s).

Description

Field of the invention

The present invention relates to the field of methods and devices for controlling throttle valves of an internal combustion engine.

State of the art

The position control of the butterfly valves consisting of a flat shutter fixed on a rotating shaft, which lies substantially in the plane of the plug. The rotation of the valve is actuated by a DC motor, sometimes by a gear train. Preloaded springs are associated with the shaft, for the return of the shutter to a pre-established position. Additionally, a position sensor is associated with the shaft to determine shaft position and control valve position.
Two possible variants are commercially available:

one referred to as "smart", in which the valve manufacturer supplies a device equipped not only with the electromechanical part consisting of the valve and its actuator, but also with the electronic module and the relative software/firmware for controlling the position of the shutter, and
one indicated as "full power", in which the manufacturer supplies only the electromechanical portion.

Therefore, in the first case the valve, in addition to the necessary electrical power supply, has a data input for acquiring a value of a target position and a data output which returns a value of a current position measured by the aforementioned sensor. Therefore, the "smart" valves are generally connected to the vehicle data network, typically CAN, and to the vehicle power supply, typically in direct current.
In the second case the valve is not equipped with an electronic module and therefore the reading of the valve position is made in analog domain and the electronic module with the relative software/firmware is placed in another device, typically the engine control unit, universally indicated as ECU (Engine Control Unit).
The preloaded springs, in the absence of power to the electric actuator, return the valve to a position called "Limp home", generally close to, but not coinciding with, the position of complete closure of the valve. Typical values of the Limp home position are in the order of 7 - 12 degrees with respect to the fully closed position of the valve.
The springs introduce a significant non-linearity in the model associated with the valve.
In the automotive field, the correct functioning of these devices, both in terms of precision, i.e. of error at steady state, and in terms of response promptness, is fundamental for the control of the torque of spark ignition heat engines.
Similar types of valves can also be used to control the flow of exhausted gas to be recirculated at the intake of the heat engine. Generally, these systems are referred to as EGR and mainly implemented in the context of Diesel cycle engines.
Various solutions for valve control are known.
A "single closed loop" feedback scheme is known, in which the measurement of the shutter position is retro-activated and controllers with "gain scheduling" are used. In other words, these are controllers, in which the gains Kp and Ki are not constant but are selected by means of lookup tables as a function of two or more inputs, which in turn are a function of predetermined operating conditions.
These controllers are generally proportional and integral PI controllers or proportional, integral and derivative PID controllers. They can be configured to compensate for the torque contribution given by the preloaded springs and/or static friction.
They are configured to return a current signal to be converted into voltage or, more commonly, to directly return a voltage value.
This solution is widely implemented as it allows the controllers to be calibrated on several points by means of gain scheduling in order to take into account the intrinsic non-linearities of the system being controlled, but at the same time they involve a huge consumption of time.
The position control of the butterfly valves (flap valve) in the automotive field has the peculiarity that the sole available sensor is the position sensor.
Among the many control techniques that can be used for electric position drives, the "classic" one based on the three "closed loops" in cascade should certainly be mentioned:

a first, outer loop with position feedback via a proportional controller,
an intermediate loop with speed feedback via a proportional and integral controller,
a third, innermost loop with current feedback via a proportional and integral controller, the so-called torque or current loop, which returns a voltage value.

This scheme can be modified and simplified by eliminating the innermost loop, i.e. the current loop. Therefore, they remain

an outer loop with position feedback via proportional controller
an internal loop with speed feedback via a proportional-integral controller.

The speed is estimated, as there is no speed sensor.
Figs. 1 and 2 show two generic feedback control schemes.
In figure 1, the feedback control scheme of a device D provides a single feedback loop, which returns the value of a variable under control, which value is compared with the value of an input signal, which represents the target value that it is desired to force the variable object of control, and the comparison, ie the error, is used to correct the control signal "output" so that the error tends to zero.
In Fig. 2, the control scheme is a cascade of two feedback controls, completely similar to that of Fig. 1. Here the external controller C1 generates a signal homogeneous with the feedback signal of the internal loop, so that the internal comparison has a physical meaning. Each of the two blocks "C1" and "C2" can include a controller for example of the PID, proportional, integral, derivative type or a subset, for example PI or PD. The controller C2, being the last block of the control scheme, is the one in charge of generating the actuation signal for the actuation of the device D.
They can be configured to compensate for the torque contribution given by the preloaded springs and/or static friction.
Unless specifically excluded in the detailed description that follows, what is described in this chapter is to be considered as an integral part of the detailed description.

Summary of the invention

The object of the present invention is to present a control scheme for a valve that is reliable and simple to implement, i.e. one which does not require a long controller calibration process.
The basic idea of the present invention is to implement a cascade scheme with only two feedback loops, in which an outer loop feedback a signal representative of the valve position and an inner loop feedback a signal representative of an actuation speed of the valve, calculated by means of an estimator, and in which when a position error is less than or equal, in absolute value, to a predetermined position error threshold then the actuation control signal is a function of a signal representative of a position error whereas, when the position error exceeds, in absolute value, said threshold, then the actuation control signal is a function of a signal representative of a speed error. Thus the cascade scheme, having only two feedback loops, is arranged to switch to a single loop scheme when the position error is less than or equal to the above threshold.
Advantageously, the fact of estimating the speed using an estimator entails decoupling the dynamics of the position from those of the speed, allowing the use of scalar and/or vectorial parameters for the controllers, avoiding the implementation of matrix gain scheduling, with an undoubted simplification of the calibration procedures.
This decoupling is particularly useful in consideration of the non-linearities introduced by the springs.
The present solution allows to maintain the linear control scheme in the two domains identified by the aforementioned threshold unlike other control schemes which, to compensate for the non-linearity of the controlled system, use gains, which are in turn extremely non-linear.
The error signal, be it the position error or the speed error, is addressed to a proportional or proportional-integral or proportional-integral-derivative controller, equipped with scalar or vectorial gains.
In particular, it is preferred that the gain of the proportional contribution of the loop controller relative to the position error and the gain of the integral contribution of the loop controller relative to the speed error are scalar, while the gain of the proportional contribution of the of the speed error loop either scalar or vectorial.
The present preferred solution of the invention effectively solves the problem of static friction compensation, eliminating the limit cycles, which are triggered by small position errors. Preferably, the gain of the proportional contribution of the controller relative to the internal loop is vectorial and equal to the value of an inversely proportional function, of the type 1/x and relative powers, of the absolute value of the position error.
Therefore, the vector gain calibration of the proportional controller is greatly simplified.
This solution also allows the internal loop to be kept linear since the proportional portion of the relative controller does not depend on the speed error but on the position error which is external to this loop.
The dependent claims describe preferred variants of the invention, forming an integral part of the present description.

Brief description of the figures

Further objects and advantages of the present invention will become clear from the detailed description that follows of an embodiment of the same (and of its variants) and from the annexed drawings given for purely explanatory and non-limiting purposes, in which:

Figs. 1 and 2 show feedback control schemes according to the prior art;
Fig. 3 shows an example of a control diagram of a butterfly valve according to the present invention;
Fig. 4 shows a second example of a control scheme according to the present invention.

The same reference numbers and letters in the figures identify the same elements or components or functions.
It should also be noted that the terms "first", "second", "third", "superior", "inferior" and the like may be used herein to distinguish various elements. These terms do not imply a spatial, sequential, or hierarchical order for the modified items unless specifically indicated or inferred from the text.
The elements and characteristics illustrated in the various preferred embodiments, including the drawings, can be combined with each other without however departing from the scope of protection of the present application as described below.

Detailed description

Fig. 3 shows a first example of a control scheme according to the present invention.
The input pos_des represents the target position signal. This signal is subtracted from the signal of the measured position pos_mis by the position sensor associated with the valve shutter shaft. The result of the comparison is the position error signal e_pos. The position error signal is input to a first P-type proportional controller C1. This first controller is also referred to as position loop or outer loop controller. The output signal of the first controller is referred to as target speed vel_des. A signal of the estimated speed vel_stim is subtracted from the output signal of the first controller. The result of the comparison is the speed error signal e_vel which enters a switch SW.
The switch SW has as control signal the result of the comparison between the absolute value of the position error signal |e_pos| and a threshold e_s. When the absolute value of the position error signal is less than or equal to the threshold e_s then (Yes) a first input is selected, while in the negative case (No) a second input is selected. The first input receives the position error signal e_pos, while the second input receives the speed error signal e_vel, therefore depending on the above comparison, the switch output is represented by the position error or by the speed error.
The threshold value is approximately three times the resolution value of the analog-to-digital conversion. For example, if sampling is done at 12bit, then the resolution is approximately 0.025% of the entire shutter aperture range from 0 to 100% open. Therefore, the threshold is 0.075.
The switch output is routed to the actuator, indicated as "Valve" in the diagram, preferably via a second controller C2. This second controller C2 generates a current signal to be converted into a voltage signal or, more preferably, directly generates a voltage signal Volt_act.
The presence of the second controller is completely optional and depends exclusively on the type of signal that it is desired to generate to control the actuator. For example, the actuator may include a conversion block external to the present control scheme.
Conversely, when the second controller is present, this means that the actuator is made in such a way as to receive a current or better a voltage signal as input.
The representative signal of the speed estimation Vel_stim is generated by an estimator, "Estimator", for example implementing a Kalman filter or a Fuzzy logic, which according to the control signal of the actuator Volt_act and the signal of the measured position Pos_mis, generates the signal representative of the estimated speed vel_stim. Evidently the estimator implements an electro-mechanical model of the valve which can be found in the literature ["Variable-Structure Control of Electronic Throttle Valve",Yaodong Pan; Umit Ozguner; Oguz Hasan Dagci. IEEE Transactions on Industrial Electronics (Volume: 55, Issue: 11, Nov. 2008). Page(s): 3899 - 3907].
The second controller C2 is of the proportional-integral type and does not include matrix "gain scheduling", on the contrary, the gains are scalar or at most vectorial. Preferably, the gain of the proportional contribution of the first controller C1 is scalar.
Fig. 4 shows a further control scheme which differs from that of Fig. 3 exclusively for the second controller C2.
In this case, the gain of the integral contribution is a scalar, while the gain of the proportional contribution is a vectorial function, as it provides multiple output values as a function of the single input value represented by the absolute value of the position error.
The gain of the proportional contribution of the second controller C2, in particular, is represented by an inversely proportional function of the absolute value of the position error, of the type 1/x and relative powers, 1/x², 1/X³, etc.
The substantial advantage of using a proportional contribution of the second controller C2 as described in Fig. 4 is that for small variations in position, the error is small and therefore the gain is relatively high, allowing static friction to be overcome quickly and guaranteeing speed and accuracy in correcting the error because it ensures that no limit cycles are triggered. Conversely, for a high error, the gain is relatively low, and together with the integral contribution of the controller, it provides for a rapid correction of the error.
Thanks to the present variant, regardless of the starting position of the valve shutter, excellent control behaviour is obtained, both in terms of speed, i.e. response time, and precision, i.e. error at zero steady state.
Preferably, the gain of the integral contribution is greater than the smallest gain value of the proportional contribution of the same second controller, and preferably between one and two orders of magnitude greater.
In other words G_I ≈ (10 ÷ 100) [G_P]_Min, where G_I is the scalar gain of the integral contribution and [G_P]_Min is the smallest gain expressed by the vectorial function of the proportional contribution of the same controller C2.
This ensures that the torque of the preloaded springs is quickly compensated for, which is strongly non-linear around the Limp Home position precisely because they are preloaded.
It is preferable that the gains of the proportional contribution of the first controller C1 and of the integral contribution of the second controller C2 are scalar since this choice is sufficient to ensure the desired performance for valve control, however if one or both of them is a vector/matrix, this would not exceed the scope of protection of the present patent application.
Furthermore, the second controller C2 can also include a derivative contribution by defining a PID and also the gain of the derivative contribution is a scalar or at most a vector function.
According to a further preferred variant of the invention, a saturator filter can be provided with the relative "anti-windup" arranged between the second controller C2 and the "Valve" actuator.
The control scheme, according to any of the proposed variants, can be implemented in an engine control processing unit ECU through a suitable software/firmware module or it can be implemented through a dedicated control unit associated with the valve according to the "smart" solution indicated above.
The present invention can advantageously be implemented through a computer program comprising coding means for carrying out one or more steps of the method, when this program is executed on a computer. Thus, it is understood that the scope of protection extends to said computer program and also to computer readable means comprising a recorded message, said computer readable means comprising program coding means for carrying out one or more steps of the method, when said program is run on a computer. Variants of the non-limiting example described are possible, without however departing from the scope of protection of the present invention, including all equivalent embodiments for a person skilled in the art, to the contents of the claims.
From the description given above, the person skilled in the art is capable of realizing the object of the invention without introducing further constructive details.

Claims

Method of controlling a throttle valve of an internal combustion engine, the valve comprising an electric actuator for controlling a position of a shutter and a position sensor for detecting a position of the shutter, the method comprising a step of controlling said actuator with a control signal (Volt_act) which is alternately
- function of a signal representative of a shutter position error (e_pos), when an absolute value of the shutter position error is less than or equal to/a predetermined threshold (e_s) and

- function of a signal representing a shutter speed error (e_vel), when the absolute value of the shutter position error is greater than said predetermined threshold (e_s), wherein said threshold is a value strictly greater than zero.
Method according to claim 1, wherein said speed error is calculated as the difference between
- a signal representing a target speed (vel_des), proportional to said signal representing the position error and

- a signal representing an estimated speed (vel_stim) calculated by an observer (Estimator).
Method according to claim 2, wherein said signal representative of the position error (e_pos) is filtered by a first controller (C1) comprising a proportional controller (P), so as to generate said signal representative of the target speed.
Method according to any one of claims 1 - 3, wherein said control signal (Volt_act) is a function of said signal representative of the speed or position error, filtered by a second controller (C2) comprising
- a proportional contribution (P) and

- an integral contribution (I).
Method according to claims 3 or 4, wherein said observer (Estimator) implements an electro-mechanical model of the valve and comprises two inputs: said control signal (Volt_act) and a position signal (pos_mis) generated by said sensor position.
Method according to claim 4 or 5, wherein a gain of said proportional contribution (P) of said first controller (C1) and/or a gain of said integral contribution (I) of said second controller (C2) is a scalar or a vector, but not a matrix.
Method according to any one of claims 4 - 6, wherein a gain of said proportional contribution (P) of said second controller (C2) is a scalar or a vector.
Method according to claim 7, wherein said vector is representative of an inversely proportional function of the absolute value of the position error, of the type 1/x and relative powers.
Method according to claim 8, wherein the gain of said integral contribution of said second controller (C2) is between one and two orders of magnitude greater than the smallest value of the vector defining the gain of the proportional contribution (P) of said second controller (C2) .
The method according to any one of the preceding claims, implementing a cascade control scheme consisting of a first external feedback control loop and a second internal feedback control loop, wherein said signal representative of the position error is generated by the first outer loop, and said speed error representative signal is generated by the second inner loop.
A computer program comprising program coding means adapted to carry out all steps of any one of claims 1 to 10, when said program is run on a processing unit operatively connected with a position sensor and an electric actuator of a butterfly valve and to a signal source representative of a target position (pos_des) of the butterfly valve plug.
A computer readable means comprising a recorded program, said computer readable means comprising a program coding means adapted to carry out all steps of any one of claims 1 to 11, when said program is run on a processing unit operatively connected with a position sensor and an electric actuator of a butterfly valve and a signal source representative of a target position (pos_des) of the butterfly valve plug.
A control system for a throttle valve comprising an internal combustion engine, the valve comprising an electric actuator for controlling a position of a shutter and a position sensor for detecting a position of the shutter, the system comprising processing means configured to control said actuator with a control signal (Volt_act)
- proportional to a signal representing a position error (e_pos) when an absolute value of the position error is less than or equal to/a predetermined threshold (e_s) and

- proportional to a signal representing a speed error (e_vel) when the absolute value of the position error is greater than said predetermined threshold (e_s).
Vehicle equipped with an internal combustion engine and a throttle valve disposed on a related intake or exhaust manifold of the internal combustion engine and equipped with the throttle valve control system according to claim 13.