GB2586864A

GB2586864A - EGR Valve Controller

Info

Publication number: GB2586864A
Application number: GB1912861.0A
Authority: GB
Inventors: Reza Kianifar Mohammed
Original assignee: Perkins Engines Co Ltd
Current assignee: Perkins Engines Co Ltd
Priority date: 2019-09-06
Filing date: 2019-09-06
Publication date: 2021-03-10
Anticipated expiration: 2039-09-06
Also published as: GB2586864B; GB201912861D0

Abstract

An exhaust gas recirculation (EGR) valve controller for an internal combustion engine is provided. The EGR valve controller 18 is configured to control an EGR actuator 16 for an EGR valve 14. The EGR valve controller comprises an EGR learn module 30 configured to update a reference value for the EGR actuator which locates the EGR valve in a reference position, the EGR learn module configured to: determine a candidate value for the EGR actuator; calculate an upper learn value for the EGR actuator and a lower learn value for the EGR actuator, and wherein the EGR learn module is configured such that, in an event that the candidate value falls between the upper learn value and the lower learn value, the reference value for the EGR actuator is updated based on the candidate value. The upper and the lower learn values are calculated based on an average of a plurality of previous updates to the reference value. By applying a limit to the range of EGR position values used in the learning module any incorrect readings can be filtered out of the calculations used when determining the EGR position and the EGR open and/or closed positions can be accurately determined despite fluctuations of these positions with age and wearing of the valve.

Description

EGR Valve Controller

Technical Field

The present disclosure relates to the control of a component of an internal combustion engine. In particular, the present disclosure relates to the control of an Exhaust Gas Recirculation (EGR) valve of an internal combustion engine.

Background

An internal combustion engine may be provided with an exhaust gas recirculation (EGR) system. An EGR system aims to reduce Nitrogen Oxide (N0x) emissions from an internal combustion engine. An EGR system is typically configured to recirculate a portion of an internal combustion engine's exhaust gas back to the engine cylinders. Recirculating exhaust gas dilutes the 02 in the incoming air stream and provides gases inert to combustion to act as absorbents of combustion heat, thereby reducing peak in-cylinder temperatures, which in turn may reduce NOx emissions.

Typically, an EGR system is configured to control the proportion of the exhaust gas which is recirculated to the internal combustion engine cylinders. Accordingly, an EGR system may comprise an EGR valve which can be controlled to vary the proportion of the exhaust gas recirculated by varying the degree of opening of the valve.

Typically, the opening of the valve is actioned by an EGR actuator, which may be an electronic actuator. The EGR actuator may be controlled by an EGR controller, for example an engine control unit (ECU) or other controller of the internal combustion engine. In order to accurately control the EGR valve, it is important that the controller is able to accurately define a position of the EGR valve (e.g. an EGR valve opening position) in order to control the EGR valve. Due to variations in manufacturing tolerances of the EGR system, the EGR actuator, and the controller, there may be some uncertainty in the precise electronic signal levels required to control the EGR actuator to locate the EGR valve in the EGR valve opening position. Furthermore, the EGR system may be subjected to wear over the duration of its lifetime thereby causing further changes in the operating characteristics of the EGR system. These changes introduce further uncertainty into the control characteristics of the EGR system.

In order to reduce the effects of uncertainty on the control of the EGR system, an EGR controller may be configured to determine, and update a reference position of an EGR valve. For example, US 9,458,785 discloses an EGR controller for an internal combustion engine which is able to learn a fully-closed position of an EGR valve. The EGR valve is driven in a direction where an opening degree of the EGR valve is increased and in a direction in which the opening degree of the EGR valve is decreased with respect to a full-close position of the EGR valve.

It is an object of the present disclosure to provide an improved EGR controller which tackles at least one of the problems associated with prior art EGR controllers or, at least, provide a commercially useful alternative thereto.

Summary

According to a first aspect of the disclosure an exhaust gas recirculation (EGR) valve controller configured to control an EGR actuator for an EGR valve is provided. The EGR valve controller comprises an EGR learn module. The EGR learn module is configured to update a reference value for the EGR actuator which locates the EGR valve in a reference positon, the EGR learn module configured to: determine a candidate value for the EGR actuator, calculate an upper learn value for the EGR actuator and a lower learn value for the EGR actuator, and wherein the EGR learn module is configured such that, in an event that the candidate value falls between the upper learn value and the lower learn value, the reference value for the EGR actuator is updated based on the candidate value. The upper learn value and the lower learn value are calculated based on an average of a plurality of previous updates to the reference value.

According to a second aspect of the disclosure an exhaust gas recirculation (EGR) apparatus is provided. The EGR apparatus comprises an EGR conduit, an EGR valve disposed in the EGR conduit, an EGR actuator configured to displace the EGR valve in order to control the flow rate of exhaust gas through the EGR conduit, and an EGR valve controller according the first aspect of the disclosure.

According to a third aspect of the disclosure, a method of controlling an exhaust gas recirculation (EGR) actuator for an EGR valve is provided. According to the method of the third aspect a reference value for the EGR actuator which locates the EGR valve in a reference positon is updated. The method comprises: i) determining a candidate value for the EGR actuator; ii) calculating an upper learn value for the EGR actuator and a lower learn value for the EGR actuator; and iii) if the candidate EGR value falls between the upper learn value and the lower learn value, updating the reference value for the EGR actuator based on the candidate value, wherein the upper learn value and the lower learn value are calculated based on an average of a plurality of previous updates to the reference value.

Brief Description of the Drawings

The disclosure will now be described in relation to the following non-limiting figures. Further important features of the disclosure are apparent by reference to the detailed description when considered in conjunction with the figures in which: - Figure 1 shows a diagram of an internal combustion engine connected to an exhaust gas recirculation apparatus according to an embodiment of the disclosure; - Figure 2 shows a flow chart of a method of calculating a reference value for an EGR actuator according to an embodiment of the disclosure; Figure 3 shows a graph of candidate values and reference values calculated over

time according to an embodiment of the disclosure.

Detailed description

Figure 1 shows a block diagram of an internal combustion engine 1 connected to an exhaust gas recirculation apparatus 10 according to an embodiment of the disclosure. As shown in Fig. 1 the EGR apparatus 10 is connected between an exhaust gas outlet 3 and an air intake 5 of the internal combustion engine 1.

The EGR apparatus 10 according comprises an EGR conduit 12, an EGR valve 14, an EGR actuator 16, and an EGR valve controller 18.

As shown in Fig. 1 the EGR valve is disposed in the EGR conduit 12 between the air intake and the exhaust gas outlet 3. The EGR valve 14 is configured to control the flow of exhaust gas from the exhaust gas outlet 3 to the air intake 5. Such EGR valves 14 for 4 -recirculating exhaust gas are well known in the art. Various different types of EGR valve 14 are known to the skilled person. As is known in the art, the design of the EGR valve 14 may be adapted to the EGR conduit 12 in which the EGR valve 14 is to be located.

In order to control the flow of exhaust gas through the EGR conduit 12, the EGR valve 14 may be controlled between a fully closed position and an open position. In some embodiments, a degree of opening of the EGR valve 14 may be set in order to control the rate of flow of exhaust gas through the EGR conduit 12. In order to control the degree of opening of the EGR valve 14, an EGR actuator 16 is provided. The EGR actuator 16 is configured to displace the EGR valve 14 in the EGR conduit 12 in order to control the flow rate of exhaust gas through the EGR conduit. As such, the EGR actuator 16 is configured to control the degree of opening of the EGR valve 14. Various types of EGR actuator 16 are known to the skilled person. For example, in some embodiments the EGR actuator 16 may be an electromechanical actuator such as a solenoid or a stepper motor.

In order to provide the EGR apparatus 10 with a desired performance level, the EGR actuator 16 is controlled by an EGR valve controller 18. The EGR valve controller 18 is configured to control the EGR valve actuator 16 in order to locate the EGR valve 14 in a desired position. For example, the EGR valve controller 18 may output a signal to the EGR actuator 16 which causes the EGR valve 14 to be fully closed.

The EGR valve controller 18 may be configured to control the EGR valve 14 in order to provide a desired flow rate of exhaust gas through the EGR conduit 12 (a desired EGR mass flow rate). The desired EGR mass flow rate may be determined by the EGR valve controller 18 or a desired EGR mass flow rate may be provided to the EGR valve controller 18 from an engine control module (not shown). The desired EGR mass flow rate may be determined in order to provide a desired EGR performance (i.e. based on emissions requirements, operating state of the internal combustion engine etc.).

In some embodiments, the EGR valve controller 18 may incorporate some form of feedback from the EGR valve 14. For example, as shown in Fig. 1 a sensor 20 may be provided as part of the EGR apparatus 10. The sensor 20 may be configured to sense a position (or degree of opening) of the EGR valve 14 and provide a corresponding signal to the EGR controller 18 representative of the position of the EGR valve 14. In some embodiments, a differential pressure sensor (not shown) may also be provided across the

-

EGR valve 14 to sense the pressure across the EGR valve 14 (EGR valve differential pressure). The EGR valve may be configured to predict EGR mass flow rate through a statistical prediction model based on an operating condition of the EGR valve 14. For example, in some embodiments, the EGR valve controller 18 may be provided with data representative of operating state of the internal combustion engine, including the fuel mass flow rate and the engine speed, and data representative of the EGR valve 14 operating condition, including data representative of the EGR valve position and EGR valve differential pressure sensor, to minimize the difference between the desired EGR mass flow (e.g. provided from the Engine Control Unit (ECU)) and the predicted EGR mass flow by a statistical model. Of course, in other embodiments, other combinations of variables may be used by the EGR valve controller 18 in order to determine predicted EGR mass flow rate. Data representative of the desired operating condition of the EGR valve position may be the signal sent to EGR actuator 16, or data derived from one or more sensors connected to the EGR apparatus 10.

The EGR valve controller 18 may minimise the difference between the desired EGR mass flow rate and the predicted EGR mass flow rate by adjusting (controlling) the position of the EGR valve 14 via the EGR actuator 16. Due to variations in manufacturing tolerances of the EGR apparatus 10, the EGR actuator 16, the EGR valve 14, or the EGR valve controller 18 there may be some uncertainty in the real position of the EGR valve position (the position sensed by sensor 20 and/or the signal set by the EGR valve controller 18 for the EGR actuator 16) and accordingly the actual EGR mass flow rate that results from such a position. Furthermore, the EGR valve 14 may be subjected to wear over the duration of its lifetime, or a build-up of particulate matter in the exhaust gas conduit 12, thereby causing further changes in the operating characteristics of the EGR apparatus 10. For example, a build-up of particulate matter around the EGR valve 14 may introduce uncertainty into the operating characteristics of the EGR valve 14. Accordingly, the above noted sources of uncertainty may cause the predicted EGR mass flow rate determined by the EGR valve controller 18 to deviate from the desired EGR mass flow rate intended for operation of the internal combustion engine. This in turn may result in the actual performance of the EGR apparatus 10 differing from the performance expected by the EGR valve controller 18.

In order to reduce the effects of uncertainty associated with the EGR apparatus 10, the EGR valve controller 18 includes an EGR learn module 30. The EGR lean module 30 is 6 -configured to determine a reference value for the EGR actuator 16 which locates the EGR valve in a reference positon. The EGR learn module 30 updates this reference value over time to account for changes in the EGR apparatus 10. That is to say, the reference value for the EGR actuator 16 which locates the EGR valve 14 in a reference position may change over time (for example due to soot build up or actuator wear and the like).

In some embodiments, the reference position of the EGR valve 14 may be a fully closed position of the EGR valve 14. In other embodiments, the reference position may be another position of the EGR valve 14 such as a fully open position or a position corresponding to a partially throttled flow rate of exhaust gas. As such, the reference position may be a position of the EGR valve 14 which provides a predetermined EGR mass flow rate.

In some embodiments, a range of EGR actuator values may result in the EGR valve 14 being fully closed. For example, this may be due to particulate matter build up, or due to the type of EGR actuator 16 used in the EGR apparatus. Accordingly, in some embodiments, the reference value for the EGR actuator 16 may be the value for the EGR actuator 16 which locates the EGR valve 14 at the point of opening (i.e. the lift point for the EGR valve 14). That is to say, at the point of opening the actual EGR mass flow rate is 0 kg/s, and any further increase in the opening of the EGR valve 14 would result in an increase in the actual EGR mass flow rate (if the EGR apparatus 10 is in operation).

In order to update the reference value of the EGR actuator 16, the EGR learn module 30 is configured to: (i) determine a candidate value for the EGR actuator 16; (ii) calculate an upper learn value for the EGR actuator 16 and a lower learn value for the EGR actuator 16; and if the candidate EGR value falls between the upper learn value and the lower learn value, the reference value for the EGR actuator is updated based on the candidate value.

The upper learn value and the lower learn value are calculated by the EGR learn module based on an average of a plurality of previous updates to the reference value.

An example of a method carried out by the EGR learn module 30 is shown as a flow chart in Fig. 2 of this disclosure. In accordance with step (i) a candidate value (LN) for the EGR 7 -actuator 16 is determined by the EGR learn module 30. A candidate value may be determined by the EGR learn module 30 based on a number of different techniques.

In some embodiments a candidate value LN may be determined by the EGR lean module 30 based on an open loop calculation of the reference value based on one or more parameters of the Engine Control Unit (ECU). In some embodiments, the ECU (or EGR valve controller 18) may be configured to perform a test routine to determine a candidate value LN for the reference value. For example, a test routing to determine a candidate value LN for the EGR actuator 16 may comprise an open-loop duty cycle ramp of the signal set by the EGR valve controller 18 for the EGR actuator 16 in order to detect the opening point of the EGR valve 14. The opening point of the EGR valve may be based on a voltage response of the EGR actuator 16 and/or values output from one or more sensors 20 (if present) of the EGR apparatus 10. It will be appreciated that the method of determining candidate values for the reference value may depend on the type of EGR valve 14 and the data available to the EGR learn module 30. Accordingly, any method of determining a candidate value for the reference value known to the skilled person may be applicable to methods of the present disclosure.

Once a candidate reference value LN for the EGR actuator 16 has been determined, the EGR learn module 30 may perform a series of calculations to determine whether or not the candidate value LN is a suitable value for the reference value or whether the candidate value is unduly influenced by stochastic noise. In order to determine whether or not the candidate value is suitable the EGR learn module 30 calculates an upper learn value (Lu) for the EGR actuator and a lower learn value (LL) for the EGR actuator. The upper learn value and the lower learn value are calculated based on an average of a plurality of previous updates to the reference value. For example, in some embodiments, at least the previous three updates to the reference value may be stored by the EGR learn module 30. In some embodiments, at least the previous 5, 7, 10 or 15 updates to the reference value may be stored by the EGR learn module 30. Accordingly, an average of the stored previous updates to the reference values may be calculated from which upper and lower learn values may be determined.

In some embodiments, the upper learn value and the lower learn value may be calculated based on the mean (m) or median (M) of the (stored) previous updates to the reference value, and an offset value. For example, the offset value may be a fixed proportion of the 8 -average of the previous updates. For example, the offset value (k) may be at least: 5 %, 10%, 15% or 20% of the average. As such, the upper learn value (Lu) and the lower learn value (LL) may be calculated based on the mean average as: Lu = m +km = m -km In other embodiments, the median average M may be used in place of the mean average.

In the embodiment of Fig. 2, the upper learn value and the lower learn value are calculated based on the mean of a plurality of previous updates to the reference value and a standard deviation (S) of the previous updates to the reference value. In the embodiment of Fig. 2, the previous ten updates to the reference value are stored by the EGR module 30. The mean (m) and standard deviation (S) of the stored previous updates to the reference value are then calculated. As shown in Fig. 2 the lower learn value is then calculated as the mean less three standard deviations. The upper learn value is calculated as the mean plus three standard deviations. Of course, in other embodiments, the upper and lower learn values may be calculated using at least one, two, three, four, or five standard deviations away from the mean. It will be appreciated that increasing the number of standard deviations thereby increasing the range between the upper and lower learn values will reduce the selectivity of the EGR learn module 30.

In some embodiments, the upper learn value and the lower learn value may be calculated based on the median of the (stored) previous updates to the reference value. The upper learn value and the lower learn value may also take into account the standard deviation of the previous updates.

The EGR learn module 30 determines if the candidate value falls between the upper learn value and the lower learn value. When the candidate value falls between the upper and lower learn values, the reference value for the EGR actuator is updated based on the candidate value. If the candidate value does not fall between the upper learn value and the lower learn value, the EGR learn module 30 is configured to increment a flag counter value (FO). The flag counter value may be used by the EGR learn module 30 or the EGR controller 18 to monitor the performance of the EGR learn module and/or the EGR apparatus 10. For example, monitoring the number of increments to the flag counter value over a period of time may provide a useful performance indicator of the performance of the EGR learn module 30. In some embodiments, a number of consecutive updates to the flag counter value may be indicative of a mechanical fault with the EGR apparatus 10. That is to say, if a number of consecutive candidate values determined by the EGR learn module 30 result in increments to the flag counter value, this may be indicative of a fault in the EGR apparatus 10. For example, the EGR learn module 30 and/or the EGR controller 18 may monitor the flag counter value and may determine a potential mechanical fault has occurred if at least five, seven, or ten consecutive increments to the flag counter value occur.

In some embodiments, if the candidate value falls within the range defined by the upper and lower learn values, the candidate value may be set as the new reference value for the EGR actuator 16. In the embodiment of Fig. 2, the candidate learn value may be further processed by the EGR learn module 30 to try to smooth a transition between a previous reference value and a new reference value for the EGR actuator. Accordingly, the candidate value may be scaled by a scaling equation which is based on the previously calculated reference values in order to smooth a transition to a new reference value. For example, in the embodiment of Fig. 2, the candidate value is processed by the following equation in order to calculate the updated reference value (LE): LF = M + (LN -M)/K In the above equation M is the median of the previous stored updates to the reference value, and K is a scaling parameter. In the embodiment of Fig. 2, K is shown as being equal to 3. In some embodiments, K may be a scaling parameter which limits the maximum change in subsequent processed EGR reference values. In some embodiments, K may range from at least 1 up to about 10.

In some embodiments, the calculation of the upper learn value and the lower learn may each be subject to a minimum range limit. In the event that the upper learn value and/or the lower learn value calculated falls outside the respective minimum range limit, the upper learn value and/or the lower learn value may be calculated based on the minimum range limit.

-10 -As such, the upper learn value and the lower learn value may be calculated as being at above and below the average of the previous updates respectively by at least a minimum range limit (X). For example, when the upper learn value and the lower learn value are calculated based on the mean, the upper learn value and the lower learn value may be subject to the following limits: If: Lu m + X; then Lu = m + X If: LL m -X; then LL = m -X In other embodiments, the median M may be used in place of the mean average.

In some embodiments, the minimum range limit X may be a fixed scalar value. The fixed scalar value may be chosen based on the expected range of normal operating values for the EGR reference value. In other embodiments, the minimum range limit may be determined based on a proportion of the average of the previous updates. For example, the minimum range limit X may be at least: 5%, 10 %, 15%, or 20 % of the average of the previous values. In the example above, the minimum range limit X is the same for the upper learn value and the lower learn value. In other embodiments, the minimum range limit for the upper learn value (Xu) may be different to the minimum range limit for the lower learn value (XL).

In some embodiments, the calculation of the upper learn value and the lower learn may be subject to a maximum range limit. In the event that the upper learn value and/or the lower learn value calculated falls outside the respective maximum range limit, the upper learn value and/or the lower learn value may be calculated based on the maximum range limit.

As such, the upper learn value and the lower learn value may be calculated as being at above and below the mean respectively by no greater than a maximum range limit (Y). For example, when the upper learn value and the lower learn value are calculated based on the mean average, the upper learn value and the lower learn value may be subject to the following limits: If: Lu m +Y; then Lu = m + Y If: LL m -Y; then LL = m -Y By specifying a maximum range limit Y, the EGR learn module may prevent the range between the upper and lower learn values from increasing past a fixed limit during a period of time in which the reference value is changing. Such a limit may also provide further protection in the event of a fault with the EGR apparatus 10 by limiting the range over which the reference value may be updated.

In some embodiments, the maximum range limit Y may be a fixed scalar value. The fixed scalar value may be chosen based on the expected range of normal operating values for the EGR reference value. In other embodiments, the maximum range limit may be determined based on a proportion of the average of the previous updates. For example, the maximum range limit Y may be no greater than: 50%, 40 %, 35 %, or 30 % of the average of the previous values. In the example above, the maximum range limit Y is the same for the upper learn value and the lower learn value. In other embodiments, the maximum range limit for the upper learn value (Yu) may be different to the maximum range limit for the lower learn value (YL).

In some embodiments, both the maximum range limit Y and the minimum range limit X may be applied when calculating the upper learn value and the lower learn value. As such, in some embodiments the upper learn value and the lower learn value may be subject to the following limits: m +X Lu m +Y -Y LL m -X In the embodiment of Fig. 2, the maximum range limit Y and the minimum range limit X are implemented as a filter on the range of values for the standard deviation (S). That is to say, in embodiments which use the standard deviation (S) to determine the upper and lower learn values, the standard deviation may be subject to a minimum standard deviation value (Sram) and maximum standard deviation value (Smax). For example, in the embodiment of Fig. 2 where the upper limit Lu and the lower limit LL are calculated based on three standard deviations, Smia and S. may be calculated as scaled versions of the minimum range limit X and maximum range limit Y described above (i.e. Sa,m = X/3 and Smax = Y/3).

Fig. 3 shows a graph of the calculated updates to the reference value calculated by an EGR learn module 30 according to an embodiment of the disclosure. Fig. 3 also shows the -12 -candidate values calculated by EGR controller 18 for reference. Accordingly, it will be appreciated that the EGR learn module 30 filters out large variations in the reference value which may be the result of stochastic uncertainty. As shown in Fig. 3, the EGR learn module 30 provides for a smoother change in reference value over time, compared to the noise associated with the candidate values.

Industrial applicability

The present disclosure provides an EGR valve controller, an EGR apparatus, an engine control module and a method of controlling an EGR actuator for an EGR valve which may improve the operation of an EGR valve. In particular, the present disclosure may provide improvements in the calculation of a reference value for an EGR actuator. Importantly, the EGR valve controller according to this disclosure is configured to reduce or prevent undesirable updates to the reference value of the EGR actuator.

For example, methods of determining reference values for EGR actuators may be subject to a degree of stochastic noise, due to the noise in the signal or random hardware issues such as EGR valve stickiness. Filtering out updates to the reference value which have been unduly influenced by stochastic noise is challenging, as the EGR system must still take into account variations in manufacturing tolerances of the EGR system and lifetime wear of the EGR system.

Accordingly, the EGR valve controller of this disclosure checks if a candidate value for updating the reference value falls within a range of values (upper learn value -lower learn value) based on an average of previous updates to the reference value. Thus, the EGR valve controller may detect if a candidate value represents a significant departure from the previous reference value (e.g. due to stochastic noise) and therefore may be discarded.

Importantly, the EGR valve controller bases the range of allowable values based on the previous values of the reference value. Thus, the EGR valve controller may take into account changes in the EGR apparatus over the lifetime of the EGR apparatus. Furthermore, as the EGR valve controller of this disclosure may determine the updates to the reference value in real time, the range between the upper learn value and lower learn value may not need to take into account the same range of uncertainty as prior art systems. That is to say, the EGR valve controller may use real-time data to identify that a -13 -candidate value which may fall within an allowable operating range for the EGR actuator represents a significant departure from the previously calculated reference values and may discard that candidate value accordingly. As such, the EGR valve controller of the disclosure may be more selective when updating the reference value for the EGR actuator

than prior art systems.

In some embodiments, the EGR valve controller may be provided as part of an engine control module. In some embodiments, the EGR controller may be provided as part of an EGR apparatus for an internal combustion engine.

It will be appreciated that the above embodiments of the disclosure are provided by way of example only. Various modifications to, and combinations of, one or more of the above described embodiments of the invention will be apparent to the skilled person without departing from the scope of this disclosure.

Claims

-14 -CLAIMS: 1. An exhaust gas recirculation (EGR) valve controller configured to control an EGR actuator for an EGR valve comprising: an EGR learn module configured to update a reference value for the EGR actuator which locates the EGR valve in a reference positon, the EGR learn module configured to: determine a candidate value for the EGR actuator; calculate an upper learn value for the EGR actuator and a lower learn value for the EGR actuator, and wherein the EGR learn module is configured such that, in an event that the candidate value falls between the upper learn value and the lower learn value, the reference value for the EGR actuator is updated based on the candidate value; wherein the upper learn value and the lower learn value are calculated based on an average of a plurality of previous updates to the reference value.
2. An EGR valve controller according to claim 1 wherein if the candidate value does not fall between the upper learn value and the lower learn value, the EGR learn module is configured to increment a flag counter value.
3. An EGR valve controller according to claim 2 wherein the flag counter value is monitored by the EGR controller in order to determine if a mechanical fault has occurred.
4. An EGR valve controller according to any preceding claim, wherein: the upper learn value and the lower learn value are calculated based an average of at least the 5 previous updates to the reference value.
5. An EGR valve controller according to any preceding claim, wherein: the upper learn value and the lower learn value are calculated based on a median of the plurality of previous updates to the reference value and a standard deviation of the plurality of previous updates to the reference value.
6. An EGR valve controller according to claim 5, wherein: the upper learn value and the lower learn value are calculated as being at least three standard deviations above and below the median respectively.
7. An EGR valve controller according to claim 6, wherein the upper learn value and the lower learn value are further calculated as being at above and below the mean respectively by at least a minimum range limit; and/or the upper learn value and the lower learn value are further calculated as being at above and below the mean respectively by no greater than a maximum range limit.
8. An EGR valve controller according to any preceding claim, wherein the EGR learn module is configured to update the reference value (LE) for the EGR actuator based on the candidate value (LN) by: determining a median (M) of the plurality of previous updates to the reference valve, and the reference value (LE) is updated based on the median (M), the candidate value (LN) and a scaling factor (K) according to the equation: LF = M + ([N-M)/K.
9. An EGR valve controller according to any preceding claim, wherein the reference position of the EGR valve is a position of the EGR valve in which the EGR valve is at the point of opening.
10. An exhaust gas recirculation (EGR) apparatus comprising: an EGR conduit; an EGR valve disposed in the EGR conduit: an EGR actuator configured to displace the EGR valve in order to control the flow rate of exhaust gas through the EGR conduit: an EGR valve controller according to any one of claims 1 to 7.
11. An engine control module comprising the EGR valve controller of any one of claims 1 to 9.
12. A method of controlling an exhaust gas recirculation (EGR) actuator for an EGR valve in which a reference value for the EGR actuator which locates the EGR valve in a reference positon is updated comprising: determining a candidate value for the EGR actuator; -16 -calculating an upper learn value for the EGR actuator and a lower learn value for the EGR actuator, and if the candidate EGR value falls between the upper learn value and the lower learn value, updating the reference value for the EGR actuator based on the candidate value; wherein the upper learn value and the lower learn value are calculated based on an average of a plurality of previous updates to the reference value.
13. A method according to claim 12, wherein if the candidate value does not fall between the upper learn value and the lower learn value, the EGR learn module increments a flag counter value.
14. A method according to claim 13, wherein the flag counter value is monitored by the EGR controller in order to determine if a mechanical fault has occurred.
15. A method according to any of claims 12 to 14, wherein: the upper learn value and the lower learn value are calculated based an average of at least the 5 previous updates to the reference value.
16. A method according to any of claims 12 to 15, wherein the upper learn value and the lower learn value are calculated based on a median of the plurality of previous updates to the reference value and a standard deviation of the plurality of previous updates to the reference value.
17. A method according to any of claims 12 to 16 wherein the upper learn value and the lower learn value are calculated as being at least three standard deviations above and below the median respectively.
18. A method according to any of claims 12 to 17, wherein the upper learn value and the lower learn value are further calculated as being at above and below the mean respectively by at least a minimum range limit; and/or the upper learn value and the lower learn value are further calculated as being at above and below the mean respectively by no greater than a maximum range limit.
19. A method according to any of claims 12 to 18, -17 -wherein the EGR learn module updates the reference value (LF) for the EGR actuator based on the candidate value (LN) by: determining a median (M) of the plurality of previous updates to the reference valve, and the reference value (LF) is updated based on the median (M), the candidate value (LN) and a scaling factor (K) according to the equation: LF = M (LN-M)/K