The present invention relates to a method and to a device for controlling the position of a throttle for regulating the quantity of air admitted into an internal combustion engine propelling a motor vehicle and, more particularly, to such an electronic control device ensuring high angular resolution, a large dynamic range and a high speed of response, whilst offering good immunity to noise and a contained manufacturing cost.

Motor vehicles are usually equipped with an accelerator pedal which is mechanically linked to a disc mounted tiltably in the air inlet duct of the internal combustion engine propelling the vehicle. Through an action on this pedal, the driver controls the angular position of the disc in the duct, which position regu¬ lates through loss of load the quantity of air entering the engine. Such a mechanical link is difficult to instal, lacks precision and does not allow intervention, in the regulating of the air admitted, by electronic devices such as a calculator of petrol injection time, a wheel anti-slip or anti-lock device, etc. An electronic device for controlling the position of the air throttle which does not have- these limitations is known. Such a device has been represented diagrammatically in Figure 1, and essentially comprises a sensor 1 of the position of an accelerator pedal 2, this sensor delivering a signal representing the request by the driver, a control unit 3 for an electric motor 5 actuating a throttle 4 for driving the position of the throttle to a setpoint value requested by the driver or by an on-board electronic device, and a sensor 6 of position of the throttle. On the basis of a setpoint value of the posi¬ tion of the throttle 4 set by the driver or by one of the abovementioned electronic devices, and of the measurement

of the current position of the throttle effected by the sensor 6 and provided to the control unit 3, a calculator contained in the unit 3 formulates a command transmitted to the electric motor 5 which then ensures a suitable displacement of the throttle 4 in order to drive the letter's position to the requested setpoint.

To be satisfactory, such a device must have a very good angular resolution (of the order of 0.1° for a throttle diameter of 50 mm), a large dynamic range corresponding to a 90° rotation of a throttle between its open position and its closed position, and a high speed of response, the throttle having to pass from an open position into a closed position or vice versa in 100 ms, for example. These characteristics are indispensable to ensure good stability at idle speed (± 20 revs/min) whereas the speed of the engine is particularly sensitive to small variations in the angular position of the throttle. These characteristics also ensure, still at idle speed, better compensation for the speed variations when switching on an energy consumer such as air condi¬ tioning.

At idle speed, the said characteristics ensure the driver good driving comfort, that is to say good progressiveness and good accuracy of control of the throttle as well as a high speed of response so that no time lag between the request and the response can be perceived by the driver.

To obtain the desired resolution, the use can be envisaged of a reduction gear at the output of the electric motor. With the throttle then not being coupled directly to the motor, this solution is not secure from the mechanical point of view. Furthermore, in order to displace the throttle over 90°, several revolutions of the motor are required and a sufficiently rapid response cannot then be obtained.

The use can also be envisaged of a digital control. The dynamic range of 0 to 900 then required

(90°/0.1°) given the requested resolution (0.1°) then entails using signals digitised with 10 bits (dynamic range of 0 to 1023), this creating a certain number of problems. Firstly, it is then necessary to use an accurate and therefore expensive sensor of throttle angular position, certain types of sensors (Hall effect for example) not then being suitable through lack of accuracy. Furthermore, the signal provided by the sensor is delivered to an analog input of digital control means (microprocessor-based for example) which must then be equipped with an analog/digital converter operating on 10 bits, necessarily more expensive than a converter operating on 8 bits, for example. Furthermore, the use of an analog input digitised with 10 bits reduces the immunity to noise of the device whereas the latter, controlling an electric motor, generates a great deal of interference in the electrical supplies required. The sensor of angular position of the throttle being connec¬ ted to the control means by a cable, there is furthermore a risk of the signal provided by this sensor being disturbed by electromagnetically induced interference.

If it is observed that high accuracy is required only for small openings of the throttle, while the voltage signal provided by the sensor 6 is weak, the costs can be reduced by using an 8-bit analog/digital converter and by changing the reference voltage of the converter in such a way as to then enhance the effective resolution. However, the accuracy of the various reference voltages used must then be guaranteed and it must be observed that there is risk of the successive analog/digital conversion, before and after a switch of scale, providing different values since these conversions will not have involved the same elements of the converter.

SUBSTITUTE SHEET

The present invention therefore aims to provide a method and a device for controlling the position of a throttle for regulating the quantity of air admitted into an internal combustion engine propelling a motor vehicle, which do not have the abovementioned disadvantages.

The present invention also aims to provide such a method and such a device offering, apart from resolu¬ tion, the required speed of response and dynamic range, good immunity to noise and a moderate construction cost. The present invention further aims to provide such a method and such a device ensuring good progres- siveness of control at small openings of the throttle, which is required to obtain satisfactory stability of operation of the engine at idle speed. These aims of the invention, as well as others which will emerge on reading the present description, are achieved with a method of controlling the position of a throttle for regulating the quantity of air admitted into an internal combustion engine propelling a motor vehicle, according to which the angle of position of the throttle is measured, the measurement obtained is digitised with n bits, and the digitised measurement is used to control a stepper motor for actuating this throttle in such a way as to drive this angular position to a setpoint value. According to the invention, the stepper motor is control¬ led in closed loop on the basis of an error signal digitised with n bits and in open loop on the basis of an error signal digitised with (n + m) bits.

As will be seen below, it is thereby possible to achieve the resolution corresponding to the use of signals digitised with (n + m) bits, with sensors and analog/digital converters of moderate cost, the latter ensuring digitisation with n bits alone.

According to an essential feature of the method according to the invention, a microstep control of the stepper motor is used to ensure the driving of the position of the throttle to a setpoint value digitised

with (n + m) bits. This known technique of control in microsteps (see the USA Patent No. 4 855 660 in the name Of WRIGHT and granted to SIEMENS-BENDIX AUTOMOTIVE ELECTRONICS LP) reduces the risks of oscillations of the throttle about a position, and therefore of loss of step. It furthermore simplifies the control method according to the invention, as will be seen below.

Following another feature of the method according to the invention, an error signal of position of the closed-loop control is formulated from a setpoint value digitised with (n + m) bits and from a measurement of the position of the throttle digitised with n bits and formatted as (n + m) bits by addition of lower-order bits in the "zero" state. According to yet another feature of the method according to the invention, for each of cycle of calcula¬ tion of the error signal of the open-loop control, the initial position of the throttle is identified with the setpoint value of this position used in the preceding cycle.

According to a variant of the method according to the invention, the setpoint value of the position of the throttle is further acted upon on the basis of engine operating parameters, in such a way as to duplicate the closed loop for controlling the position of the throttle with a second closed loop, outside the first.

The invention also provides a device for imple¬ menting the method according to the invention, comprising a stepper motor for actuating the throttle, electrical supply means for the motor and means responsive to a signal delivered by a sensor of angular position of the throttle and to a signal representing a setpoint position of this throttle in order to control this supply. According to the invention, the device comprises means for digitising the angular position signal with n bits, the control means being supplied with digitised angular- position and setpoint signals in order to drive the

position of the stepper motor and of the throttle in closed loop with the aid of an error signal digitised with n bits, and in open loop with the aid of an error signal digitised with (n + m) bits. Means are provided for controlling the engine with microsteps according to an angular resolution corresponding to the resolution of the error signal with (n + m) bits used in the open-loop servo control.

According to a variant of the device according to the invention, the latter furthermore comprises sensors of operating parameters of the engine and means responsive to these parameters for acting upon the position setpoint of the throttle according to a second closed-loop control loop, outside the first. This second loop makes it possible to improve the positional control of the throttle and therefore to ensure, in particular at idle speed, good stability of speed.

Other features and advantages of the method and of the device according to the invention will emerge on reading the following description and on examining the attached drawing in which:

- Figure 1 is a functional diagram, described in the preamble of the present description, of a device for controlling the position of a throttle of the prior art,

- Figure 2 is a functional diagram of a device for controlling the position of a throttle for regulating the quantity of air admitted into an internal combustion engine, according to a first embodiment of the invention, - Figure 3 is a functional diagram of a variant of the device of Figure 2,

- Figure 4 is a flow diagram useful in describing the control method according to the invention, and - Figure 5 represents graphs of the rate of displacement of the throttle of the device according to

the invention, as a function of the error of instantaneous position and as a function of time, respectively.

Referring to Figure 2 of the attached drawing in which are found in .the device represented, according to the invention, the inlet throttle 4, the motor 5 for controlling the position of this throttle and the sensor 6 of the position occupied by the said throttle. The device further comprises digital calculation means 7 consisting for example of a microprocessor, the latter delivering control signals to a power control unit 8 which regulates the electrical supply of phases l, Φ2 of the motor 5 which, according to the invention, is a stepper motor. The control method according to the invention which will be described below, is implemented with the aid of appropriate programming of the micro¬ processor 7, and with the aid of a controller 9 associa¬ ted for example with a device for controlling fuel injection into the internal combustion engine equipped with the device according to the invention. As represen- ted in Figure 2, the controller 9 delivers to the micro¬ processor 7 a setpoint signal of position of the throttle, calculated as a function of one or more control strategies each adapted to one particular phase of operation of the engine, such as a phase of operation at idle speed, for example. Of course, this controller is associated with means (not shown) allowing account •to be taken of a request by the driver of the vehicle, indi¬ cated with the aid of an accelerator pedal, as was seen above. According to an essential feature of the control method according to the invention, the step motor is controlled in closed loop on the basis of an error signal digitised with n bits and in open loop on the basis of an error signal digitised with (n + m) bits. As was seen above, to achieve the necessary resolution of 0.1°, calculations must be performed on values digitised with 10 bits. Hence, the controller 9

delivers to the microprocessor 7 a setpoint signal of position of the throttle digitised with 10 bits. As was seen further above, considerations of cost and of immunity to noise induce the use, in the calculations, of a signal of actual position of the throttle digitised with 8 bits only. Hence, the analog signal delivered by the sensor 6 is digitised in an analog/digital converter 10, which may be outside or inside the microprocessor 7, in such a way as to deliver a position signal digitised with 8 bits. According to another essential feature of the method according to the invention, in order to achieve the resolution of 0.1° required in the position control of the throttle 4, the stepper motor 5 can be controlled in microsteps according to the technique described in the United-States Patent No. 4 855 660 mentioned above. By way of non-limiting example, use can be made in the invention of a stepper motor of the hybrid type with 200 steps per revolution (namely 50 steps over 90°), and whose microstep control technique allows the position to be controlled with a resolution of l/32nd of a step, namely an angular resolution of 1.8/32 = 0.056°, much greater than the minimum resolution of 0.1° required for the positional control of the throttle, in particular at idle speed. Reference may be made to the abovementioned American patent, incorporated through reference to the present application, for further details regarding the microstep control technique. In the present application, the dynamic range obtained through control in l/32nd of a step is from 0 to 1600 (50 steps x 32 microsteps = 1600), this dynamic range being greater than that (900) which is required for a throttle which can move over 90°.

From the above, it follows that the device according to the invention represented in Figure 2 comprises a microprocessor 7 which ensures acquisition of the angular position setpoint of the throttle and of the current position of the throttle, in order to control, via the power control unit 8, a stepper electric motor 5.

The latter directly drives the inlet throttle 4 which is itself coupled to the position sensor 6. The signal delivered by this sensor, after digitisation with 8 bits only, enables the microprocessor to drive the position of the throttle to the desired setpoint, as will be under¬ stood on reading the following description of the control method according to the invention, given in relation to the flow diagram of Figure 4.

As was seen above, in the method according to the invention, the throttle 4 is positionally controlled in closed loop with n bits (n = 8 in the example described) with a dynamic range from 0 to 255, and an open-loop control is used to achieve a finer resolution by virtue of the use of the microstep control technique, the open-loop control being effected with the aid of signals digitised with (n + m) bits (n + m = 10 in the example described, namely m = 2). The positional control thus obtained makes it possible to achieve excellent progres- siveness of control (l/32nd of a step, namely 0.056°) and good accuracy (8 bits, namely 90°/256 = 0.35°). As was seen above, reference can be made to the abovementioned American Patent for a complete description of the microstep control technique. In brief, the latter makes it possible to obtain a half-step when the two phases Φl and Φ2 (commonly called "cosine" and "sine") are simul¬ taneously supplied by currents of like amplitude. By modulating the currents flowing in the two coils Φl and Φ2, progressive rotations are l/32nd of a step can be obtained, this making it possible to achieve the above- mentioned angular resolution of 0.056°.

The microstep control, by virtue of an improved damping of the inertia of the motor, allows the rates of rotation of the latter to be increased. To obtain good positional control of the throttle (suppression of the oscillations of the throttle about the desired position), the rate of rotation of the motor and of the throttle will depend on the positioning error and will follow a

S BSTITUTE SHEET

law of proportionality as a function of time, with saturation, as represented in the graph of Figure 5A. Over time, this rate of displacement will be able to follow the law illustrated in Figure 5B, which shows a linear increase with time, then a saturation, and a final linear decrease.

Thus, in practice, the position error allows determination of an optimal rate of displacement, accord¬ ing to the law illustrated by Figure 5A. The law of variation over time defined by the graph of Figure 5B may however provoke the choice of a rate of displacement intermediate between those defined by the two laws, if the law of Figure 5A does not make it possible to comply with the change over time defined by Figure 5B. With each calculation cycle, the opportunity to use the optimal rate of displacement is thus reexamined.

The microprocessor 7 receives a signal digitised with 8 bits and defining the current position of the throttle and a signal digitised with 10 bits representing a setpoint value of the position of this throttle. The microprocessor then calculates the distance over 8 bits of these two values in order to extract therefrom an error signal with 8 bits. The microprocessor operates through successive cycles of calculation which are renewed, for example, every 5 ms. At the start of a calculation cycle, the microprocessor first captures the measurement of the current position of the throttle (with 8 bits) then acquires the position setpoint of the throttle (the latter measurement is effected on the basis of an input in frequency and does not therefore require any analog input with 10 bits). The microprocessor then carries out a first calculation of position error with 8 bits. For this purpose, the throttle position setpoint is reduced to 8 bits, that is to say the two least signifi- cant bits of the setpoint with 10 bits are deleted.

If the position error thus calculated with 8 bits is not zero, the current position setpoint of the

throttle, reduced to 8 bits and formatted with 10 bits so as to be acceptable to the microprocessor, that is to say by setting to 0 the two least significant bits, is saved for the following calculation cycle, as initial position value of the throttle. According to the invention, it is in fact assumed that during the 5 ms for which a cycle lasts, the stepper motor has time to meet the setpoint fixed for it during this cycle. A closed-loop control of the position of the throttle is thereby set up on the basis of calculations, the accuracy of which is 8 bits.

Next, the number of steps or microsteps which the stepped motor must make in order to achieve the new throttle position setpoint (see Figure 4) is determined. The sign of the position error indicates the direction of rotation of the motor 5. The number of microsteps to be made enables the microprocessor to determine the rate of rotation of the motor in accordance with the laws shown in the graphs of Figure 5, placed in memory in the micro¬ processor. It will be noted that the rate of rotation is limited in so as not to entail losses of step, for example during sudden starting or stopping of the motor.

Once the calculation of position error with 8 bits gives a zero result, the microprocessor executes a second calculation of position error with 10 bits. For this purpose it uses as value of position of the throttle (with 10 bits) the value of the setpoint used in the earlier cycle, assumed (as indicated above), to represent the current position of the throttle with 10 bits. The current setpoint is saved for use in a possible subsequent calculation cycle, as a measurement of initial position in the cycle, as was seen above. Finally, on the basis of the amplitude of the error and of its sign, the amplitude, the direction and the rate of rotation of the motor are regulated, as described above. Thus, according to the invention, an open-loop control with 10 bits intervenes when the position servo control, with 8 bits (closed loop) is carried out. The

position error is then small and now only requires small rates of displacement of the throttle. This reduces the risks of positioning error resulting for example from losses of step due to the inertia of the rotor of the stepped motor, to errors in position measurement, etc. Furthermore, the errors in position being small, it is possible to guarantee that the motor will have time to achieve the new setpoint within 5 ms, this allowing the use, according to the invention, of the value of the old setpoint as the current position of the throttle at the start of a new calculation cycle.

In accordance with the teachings of the above- mentioned American patent relating to the technique of microstep control of the motor, the microprocessor triggers cyclically the commands required for the posi¬ tional control of the stepper motor. On the basis of the determination of the number of microsteps to be made and of the direction of rotation of the motor, the previously calculated position error is decremented and an index is incremented in a "map" placed in memory in order to determine the amplitude of the currents to be sent in the phases Φl and Φ2 (sine and cosine) of the stepped motor.

It is now apparent that the control of the motor which proceeds on the basis of an error calculation with 10 bits corresponds, as stated above, to an open-loop control, this control phase being preceded by a phase of closed-loop control with 8 bits.

Incidentally, it will be noted that the control of a stepped motor does not, in principle, require the use of a sensor to check the position which it occupies. In fact, knowing the start position of the motor, it suffices to count the steps taken during the displacements in order to know the new position of the motor. In practice, such counting operations may be disturbed by losses of step when the motor is started or stopped, as a result of the latter's inertia. In order not to be too limited in terms of rate of rotation, that is to say in

terms of speed of response, a position sensor is then used to check the displacements effected.

Referring now to Figure 3 of the attached drawing, in which has been represented a variant of the device of Figure 2, a variant which it proves particu¬ larly advantageous to employ to regulate the rate of rotation of the internal combustion engine during idling operation of the latter, which entails very small openings of the throttle. In Figure 3 are again found all the members of the device of Figure 2, which are associated with an engine 11 comprising a flywheel 12 on its output shaft, the rate of rotation of this flywheel being detected by a sensor 13, with variable reluctance for example, this sensor providing a corresponding signal to a throttle controller 9 incorporated in a device 15 provided in order to control, for example, the injection of fuel and/or the ignition of the air/fuel mixture introduced into the engine. The capture by the device 15 and the controller 9, of the speed (N) of the engine, allows a second servo control loop to be set up, parallel to be first, this second servo control allowing the resolution of the positional control of the throttle to be improved, thereby raising the stability of the idle speed. So as to further improve this stability, there can be provision for an inlet pressure sensor 14 for provid¬ ing the device 15 and the controller 9 with an additional parameter such as the inlet pressure (P), enabling the operation of the engine to be observed, in particular at idle speed. The second servo control loop may be activa- ted only at idle speed.

The method and the device described above for control according to the invention have numerous advanta¬ ges. Apart from that which is gained from the use of a hybrid stepper motor (large number of steps per revolution, high mass torque, significant internal damping, "magnetic memory" for position), the use of a throttle position signal digitised with only 8 bits

enables a position sensor and an analog/digital converter of moderate price to be incorporated in the device. It will be possible to further contain the price of the device by the use of a microprocessor with integrated 8- bit analog/digital converter which is currently found on the market.

Moreover, control of the motor in microsteps reduces the risks of oscillations about a position, and this avoids losses of step. It will furthermore be noted that the positional marking of the throttle performed by the position sensor can be duplicated with a counting of the steps, and this may offer a significant advantage when demanding a high level of safety and therefore back¬ up measurements of the position of the throttle. This results in an improvement in the performance of the control in terms of stoppage or starting of the stepped motor, that is to say an improvement in response time. Finally, this control technique enables the implementa¬ tion of the method according to the invention to be simplified. In fact, it will be possible to use a conven¬ tional stepped control within the closed-loop control phase, and then a microstep control in the second, open- loop control phase.

Owing to a digitising with 8 bits of the measurements made by the throttle position sensor, the immunity of the device to noise will be improved relative to that which would be observed if digitising with 10 bits. This may make it possible to dispense with the shielding of the cable which connects the sensor to the microprocessor.

If the accuracy of the controller obtained is determined by the analog/digital converter used (namely 8 bits or 0.35°), the resolution of the control (l/32nd of a step, namely 0.056°) ensures excellent progressiveness at small throttle openings. If a more accurate control is required, it will be possible to perform a servo control by counting microsteps, at the cost of a lower rate of

displacement of the motor, in order to avoid losses of microsteps in the counting operations.

If it is desired to use a large throttle on a low-power engine, the resolution of the control of this throttle must of course be increased. The invention makes it possible to obtain a high resolution and, to this end, allows a certain standardisation of the throttle on engines of various powers.

During the idling regulation phases, the second servo control provided by the invention (see Figure 3) by virtue of observation of the change in the engine speed and/or in the inlet air pressure of the engine, makes it possible to improve the accuracy of the positional control of the throttle and therefore to improve the stability of the idle speed. It will be observed that the outer servo control loop set up by virtue of the sensor 13 and/or 14 is slow but accurate, since it currently operates with an accuracy better than 10 bits in respect of the engine speed. The "manifold pressure" information provided by the sensor 14 may, it is true, be of only 8 bit accuracy. However, it must then be noted that the function connecting the position of the throttle to the pressure of the manifold is not linear. In particular, in the field of small throttle openings into which the invention most particularly applies, small variations in the position of the throttle bring about large variations in the pressure of the manifold, the signal provided by the pressure sensor then contributing greatly to the increase in accuracy of the regulation of the air flow rate. It is therefore by virtue of the accuracy of the outer closed loop that the accuracy of the open loop used in the invention remains sufficiently good for the sought-after application.

Of course, the invention is not limited to the embodiments described and represented, which were given merely by way of example. Thus, under the assumption that the accuracy of a motor controlled solely step by step

would be judged sufficient, it would be possible to abandon the use of a control in microsteps of this motor. The invention would however retain its value in that it would enable the cost of the device to be lowered through the use which it makes of measurements digitised with 8 bits only instead of 10 for example.

Similarly, although the invention advantageous¬ ly calls upon a stepper motor of the hybrid type to achieve the requested resolution, the scope of the invention would not be exceeded by using stepper motors of other types, with variable reluctance or with permanent magnet for example.

The invention extends finally to control methods combining a closed-loop servo control and an open- loop control employed on quantities digitised with numbers n or (n + m) of bits which differ from those given above by way of example.