GB2460397A

GB2460397A - Controlling the operation of an engine

Info

Publication number: GB2460397A
Application number: GB0809050A
Authority: GB
Inventors: Khizer Tufail
Original assignee: Ford Global Technologies LLC
Current assignee: Ford Global Technologies LLC
Priority date: 2008-05-19
Filing date: 2008-05-19
Publication date: 2009-12-02
Anticipated expiration: 2028-05-19
Also published as: CN101586500B; GB2460397B; DE102009021887A1; CN101586500A; GB0809050D0

Abstract

A method for controlling operation of a turbocharged diesel engine 10 with exhaust gas recirculation in which deviations of NOx and particulate matter emissions from target output values are used to prioritise control of an exhaust gas recirculation valve 34 and a turbocharger inlet flow control device 44 so as to improve emission performance and reduce fuel use. According to another aspect of the invention virtual NOx and particulate matter sensors are formed by data stored in look up tables saved in a memory of an electronic controller 24.

Description

A Method and System for Controlling the Operation of an Engine This invention relates to turbocharged internal combustion engines and more particularly to a turbocharged internal combustion engine having an exhaust gas recirculation (EGR) system.

As is known in the art, high performance, high speed engines are often equipped with turbochargers to increase power density over a wider engine operating range, and EGR systems to reduce the production of NOx emissions.

More particularly, turbochargers use a portion of the exhaust gas energy to increase the mass of the air charge (i.e., boost) delivered to the engine combustion chambers.

The larger mass of air can be burned with a larger quantity of fuel, thereby resulting in increased power, torque and fuel efficiency as compared to naturally aspirated engines.

As is also known in the art, EGR systems are used to reduce NOx emissions by increasing the dilution fraction in the intake manifold. EGR is typically accomplished with an EGR valve that connects the intake manifold and the exhaust manifold. In the cylinders, the recirculated burned exhaust gas acts as an inert gas, thus lowering the flame and in-cylinder gas temperature and, hence, decreasing the formation of NOx. On the other hand, the recirculated burned exhaust gas displaces fresh air and reduces the air-to-fuel ratio of the in-cylinder mixture.

A typical turbocharger includes a compressor and turbine coupled by a common shaft. The exhaust gas drives the turbine, which drives the compressor, which in turn, compresses ambient air and directs it into the intake manifold. A continuously variable geometry turbocharger (VGT) allows the intake airflow to be optimized continuously over a range of engine speeds. In diesel engines, this is accomplished by changing the angle of the inlet guide vanes on the turbine stator and an optimal position for the inlet guide vanes is determined from a combination of desired torque response, fuel economy and emissions requirements.

It is an object of this invention to provide an improved method for controlling operation of an engine system.

According to a first aspect of the invention there is provided a method for controlling a diesel engine in which NOx emissions are controlled by varying exhaust gas recirculation flow through an exhaust gas recirculation system and particulate matter emissions are controlled by varying the boost from a turbocharger wherein the method comprises prioritising control of the exhaust gas flow and the turbocharger boost based upon which of the two emissions is currently the most significant.

If NOx emissions are currently the most significant, then priority may be given to increasing exhaust gas circulation flow.

Whereas, if particulate matter emissions are currently the most significant, then priority may be given to increasing turbocharger boost.

The diesel engine may have an exhaust manifold and an inlet manifold, the turbocharger may have a turbine arranged so as to be driven by the exhaust gases from the engine, a compressor for selectively increasing the pressure in the inlet manifold of the engine and an electronically controlled inlet flow control device to regulate the exhaust gas entering the turbine and the exhaust gas recirculation system may selectively recirculate gas from a position in the exhaust manifold located upstream from the turbocharger to the inlet manifold of the engine and may include an electronically controlled exhaust gas recirculation control valve to control the flow of exhaust gas passing from the exhaust manifold to the inlet manifold, and the method may further comprise prioritising control of the exhaust gas recirculation control valve or the inlet flow control device based upon which is the larger of the differences between (i) a target NOx output value and a current NOx output value, and (ii) a target particulate matter output value and a current particulate matter output value.

The method may further comprise obtaining the differences by subtracting the target values from the current values and using the larger positive difference as the larger overproduction difference.

The method may further comprise obtaining the differences by subtracting the current values from the target values and using the larger negative difference as the larger overproduction difference.

The method may further comprise, if the larger difference is the NOx difference, adjusting the position of the exhaust gas recirculation valve so as to drive the NOx output from the engine towards the target NOx output value.

The method may further comprise adjusting the position of the exhaust gas recirculation valve so as to drive the NOx output from the engine towards the target output value based upon the magnitude of the NOx difference.

The method may further comprise, if the larger difference is the particulate matter difference, adjusting the position of the inlet flow control device so as to drive the particulate matter output from the engine towards the target particulate matter output value.

The method may further comprise adjusting the position of the inlet flow control device so as to drive the particulate matter output from the engine towards the target output value based upon the magnitude of the particulate matter difference.

The method may further comprise summing the NOx and particulate matter differences and, if the sum of the NOx and the particulate matter differences is greater than a predetermined value, reducing the fuel supplied to the engine.

The method may further comprise summing the NOx and particulate matter differences and, if the sum of the NOx and the particulate matter differences is greater than a predetermined value, providing a warning to a user of the engine.

The predetermined value may vary based upon a function of NOx output magnitude and particulate matter output magnitude.

According to a second aspect of the invention there is provided a control system for controlling a diesel engine in which NOx emissions are controlled by varying exhaust gas recirculation flow through an exhaust gas recirculation system and particulate matter emissions are controlled by varying the boost from a turbocharger wherein the system includes an electronic controller arranged to prioritise control of the exhaust gas flow and the turbocharger boost based upon which of the two emissions is currently the most significant.

If NOx emissions are currently the most significant, then the electronic controller may be operable to give priority to increasing exhaust gas circulation flow.

Whereas, if particulate matter emissions are currently the most significant, then the electronic controller may be arranged to give priority to increasing turbocharger boost.

The diesel engine may have an exhaust manifold and an inlet manifold, the turbocharger may have a turbine arranged so as to be driven by the exhaust gases from the engine and a compressor for selectively increasing the pressure in the inlet manifold of the engine and the exhaust gas recirculation system may selectively recirculate gas from a position in the exhaust manifold located upstream from the turbocharger to the inlet manifold of the engine and the system may further comprise an inlet flow control device controlled by the electronic controller to regulate the exhaust gas entering the turbine of the turbocharger, an exhaust gas recirculation control valve controlled by the electronic controller to control the flow of exhaust gas passing from the exhaust manifold to the inlet manifold though the exhaust gas recirculation system, the electronic controller being operable to prioritise control of the exhaust gas recirculation control valve or the inlet flow control device based upon which is the larger of the differences between (i) a target NOx output value and a current NOx output value, and (ii) a target particulate matter output value and a current particulate matter output value.

The differences may be obtained by subtracting the target values from the current values and using the larger positive difference as the larger overproduction difference.

The differences may be obtained by subtracting the current values from the target values and using the larger negative difference as the larger overproduction difference.

The electronic controller may be further operable if the larger difference is the NOx difference to adjust the position of the exhaust gas recirculation valve so as to drive the NOx output from the engine towards the target NOx output value.

The electronic controller may be further operable to adjust the position of the exhaust gas recirculation valve so as to drive the NOx output from the engine towards the target output value based upon the magnitude of the NOx difference.

The electronic controller may be further operable if the larger difference is the particulate matter difference to adjust the position of the inlet flow control device so as to drive the particulate matter output from the engine towards the target particulate matter output value.

The electronic controller may be further operable to adjust the position of the inlet flow control device so as to drive the particulate matter output from the engine towards the target output value based upon the magnitude of the particulate matter difference.

The electronic controller may be further operable to sum the NOx and particulate matter differences and, if the sum of the NOx and the particulate matter differences is greater than a predetermined value, reduce the fuel supplied to the engine.

The electronic controller may be further operable to sum the NOx and particulate matter differences and, if the sum of the NOx and the particulate matter differences is greater than a predetermined value, provide a warning to a user of the engine.

The predetermined value may vary based upon a function of NOx magnitude and particulate matter magnitude.

Determining the current NOx output level may comprise measuring the current NOx output level using a NOx sensor located in the exhaust manifold.

Determining the current NOx output level may comprise using a virtual NOx sensor generated using a number of look up tables stored in the electronic controller to provide a value for current NOx.

Determining the current particulate matter output level may comprise measuring the current output level of particulate matter using a particulate matter sensor located in the exhaust manifold.

Determining the current NOx output level may comprise using a virtual particulate matter sensor generated using a number of look up tables stored in the electronic controller to provide a value for current NOx.

According to a third aspect of the invention there is provided a motor vehicle having a control system constructed in accordance with said second aspect of the invention.

According to a fourth aspect of the invention there is provided method of producing a virtual sensor for an engine having a pair of interrelated variables wherein the method comprises selecting a predetermined number of speed and load combinations, for each speed and load combination, setting one variable to a predetermined operating level and controlling other variable so as to sweep through its normal operating range while capturing data forming the subject of the sensor, repeating the process for predetermined operating levels throughout the normal operating range of the one variable and storing the data as a series of look up

tables.

The method may further comprise, setting the other variable to a predetermined operating level and controllinq said one variable so as to sweep through its normal operating range while capturing data forming the subject of the sensor, repeating the process for predetermined operating levels throughout the normal operating range of the other variable and storing the data as a series of look

up tables.

The method may further comprise producing a first virtual sensor for the subject of the sensor using a method as claimed in claim 22 or in claim 23 producing a second virtual sensor for the same subject using a method as claimed in claim 22 or in claim 23 with different variables to those used to produce the first virtual sensor and combining the values from the first and second virtual sensors to produce a compensated sensor value for the subject matter of the sensor.

One pair of variables may be turbocharger boost position and exhaust gas recirculation flow valve position.

One pair of variables may be fuel rail pressure and fuel injection timing.

The virtual sensor may be a virtual NOx sensor, the subject may be NOx and the data stored may be values of NOx produced.

The virtual sensor may be a virtual particulate matter sensor, the subject may be particulate matter and the data stored may be values of particulate matter produced.

The invention will now be described by way of example with reference to the accompanying drawing of which: -10 -FIG.1 is a schematic view of an engine system having an EGR system and a variable geometry turbocharger (VGI) according to the invention; FIG.2 is a high level flow chart showing a prior art method for controlling the EGR system and VGI shown in Fig. 1; Fig.3 is a high level flow chart showing a method for controlling the EGR system and VGT shown in Fig.1 according to this invention; Fig.4 is a graph showing the relationship between particulate matter emissions and NOx for a typical diesel engine; Fig.5 is a table showing various differences for points indicated on the graph of Fig.4; Fig.6A is a matrix showing various combinations of engine speed and load that are used as reference points for determining NOx values; Fig.6B is a matrix showing various combinations of engine speed and load that are used as reference points for determining particulate matter values; Fig.7A is a matrix showing various combinations of EGR valve position and turbocharger inlet flow control device position that are mapped as dynamic NOx values; Fig.7B is a matrix showing various combinations of EGR valve position and turbocharger inlet flow control -11 -device position that are mapped as dynamic particulate matter values; Fig.8A is a matrix showing various combinations of fuel rail fuel pressure and fuel injection timing that are mapped as dynamic NOx values; Fig.8B is a matrix showing various combinations of fuel rail fuel pressure and fuel injection timing that are mapped as dynamic particulate matter values; Fig.9 is a flow chart of a method according to this invention.

Referring now to FIG.1, a diesel engine system 10 is shown. The engine system 10 includes an exhaust gas recirculation (EGR) system 12 and a variable geometry turbocharger 14.

The turbocharger 14 has a compressor 36 and a turbine 38 coupled by a common shaft 40 and an inlet flow control device which is in this case is a set of moveable turbine vanes 44 moved by an actuator. Closing the vanes 44 increases the inlet flow speed and therefore will increase boost and vice-versa.

The turbocharger 14 uses exhaust gas energy to increase the mass of the air charge (i.e., boost) delivered to the engine combustion chambers 18 resulting in more torque and power as compared to a naturally aspirated, non-turbocharged engine. The exhaust gas 30 drives the turbine 38 which drives the compressor 36, which in turn, compresses ambient -12 -air 42 and directs it in the direction of arrow 43 into the intake manifold 26.

The VGT 14 is adjusted as a function of engine operating conditions such as, for example, engine speed during engine operation by varying the turbine flow area between either a relatively open position and a relatively closed position. This is accomplished by changing the angle of the inlet guide vanes 44 on the turbine 38.

The relatively open or closed position of the guide vanes 44 is determined from the desired engine operating characteristics at various engine speeds. It is to be noted that at a given operating condition, when in the relatively open position, the boost indicated by arrow 43 is relatively low whereas when in the relatively closed position the boost is relatively high. Further, when in the closed position, the pressure in the exhaust manifold, and hence at the input to the EGR valve 34 is relatively high while in the open position the pressure is relatively low.

An engine block 16 is shown having four combustion chambers 18, each of which includes a direct-injection fuel injector 20. The duty cycle of the fuel injectors 20 is determined by an electronic controller which in this case is in the form of an engine control unit (ECU) 24 and transmitted along signal line 22.

Air enters the combustion chambers 18 through an intake manifold 26 and combustion gases are exhausted through an exhaust manifold 28 in the direction of arrow 30.

-13 -To reduce the level of NOx emissions, the engine is equipped with the EGR system 12 which comprises a conduit 32 connecting the exhaust manifold 28 to the intake manifold 26. This allows a selective portion of the exhaust gases to be circulated from the exhaust manifold 28 to the intake manifold 26 in the direction of arrow 31. An EGR valve 34 regulates the amount of exhaust gas recirculated from the exhaust manifold 28 and in the combustion chambers, the recirculated burned exhaust gas acts as an inert gas, thus lowering the flame and in cylinder gas temperature and decreasing the formation of NOx. It will be appreciated by those skilled in the art that the flow of exhaust gas through the EGR valve 34 is a function of the pressure across the valve 34 and the valve position demanded by the electrical signal provided to the EGR valve 34 on line 46 from the ECU 24. That is to say there is not a linear relationship between EGR flow rate and EGR valve 34 position.

The electrical signal on line 46 is produced by the ECU 24 from relationships stored in the ECU 24 in accordance with a computer program stored in the ECU 24.

All of the engine systems, including the EGR valve 34, VGT 14 and the fuel injectors 20 are controlled by the ECU 24. For example, the signal 46 from the ECU 24 regulates the EGR valve position, a signal on line 48 regulates the position of the VGT vanes 44 and a signal on the line 47 controls a throttle valve 49.

In the ECU 24, the command signals 46, 48 to the EGR 34 and VGT 14 actuators are calculated from measured or estimated variables and engine operating parameters by means -14 -of a control algorithm. Sensors and calibratable lookup tables residing in ECU memory provide the ECU 24 with engine operating information.

An intake manifold pressure (MAP) sensor 50 provides a signal 52 to the ECU indicative of the pressure in the intake manifold 26, an air charge temperature sensor 58 provides a signal via line 60 to the ECU 24 indicative of the temperature of the intake air charge and a mass air flow sensor (MAF) provides a signal via line 66 of air flow entering the compressor portion 36. Additional sensor inputs are also received by the ECU 24 along signal line 62 such as engine coolant temperature, fuel rail pressure, fuel injector timing, engine speed and throttle position.

Operator inputs 68 are received along signal line 70 such as the accelerator pedal position.

Based on the sensor inputs, data stored in memory such as, for example, engine mapping data and various algorithms, the ECU 24 controls the EGR 34 to regulate the EGR flow fraction and the position of the vanes 44 in order to provide emission reduction by the EGR and high fuel economy provided by the VGT boost.

Referring now to Fig.2 a typical prior art control

strategy. The method starts at blocks 100A and 100B where predicted values for MAF and MAP are generated normally from steady state look up tables generated during pre production testing of the engine 10.

At blocks 120A, 120B the predicted values are compared with actual sensed values obtained from the MAP and MAF sensors 50, 64. These comparisons generate error signals -15 -that are sent to the EGR valve 34 at block 160A and to the actuator control for the vanes 44 at block 160B.

At blocks 160A and 160B the system operates so as to try and eliminate these errors and produce new NOx and particulate matter levels as indicated by blocks 180A and 180B.

There are however a few drawbacks with this approach, firstly the variables that are important to control are NOx and particulate matter (primarily soot but also other solid matter in the exhaust stream) but these are not the inputs used to produce control of the engine. Secondly, there is an interaction between exhaust gas flow and boost pressure because the exhaust gas flow used to rotate the turbine 38 is the same flow of exhaust gas used for recirculation.

Therefore if more exhaust gas is recirculated there is less exhaust gas to power the turbocharger.

Therefore in some circumstances the control signals sent to the actuator for the vanes 44 and to the EGR valve 34 cause the two systems to fight with one another thereby failing to achieve the desired reductions in NOx and particulate matter emissions.

Referring now to Fig.3 there is shown the high level strategy used in this invention to avoid this fight and produce improved emissions and lower fuel use. A more detailed description of the method follows with respect to Fig.9.

The method starts at blocks 100C and 100D where target values for NOx and Particulate matter are retrieved from -16 -look up tables generated during pre production testing of the engine 10 corresponding to the current engine load and speed.

At blocks 120C, 120D the target values are compared with actual sensed values obtained from NOx sensor 115C and Particulate matter sensor 115D. As shown, the NOx sensor comprises of two sensors hOC and 112C positioned at different locations in the exhaust manifold 28 and the NOx sensor 115C output is a composite value obtained by combining the outputs from the two NOx sensors hOC and 112C. This provides redundancy should one of the sensors hOC, 112C fail and increases the consistency of the measurement.

Similarly, the particulate matter sensor 115D comprises of two sensors hOD and 112D positioned at different locations in the exhaust manifold 28 and the particulate matter sensor 115D output is a composite value obtained by combining the outputs from the two particulate matter sensors hOD and 112D. As before, this provides redundancy should one of the sensors hOC, 112C fail and increases the accuracy of the measurement.

Redundancy is important because the environment in the exhaust manifold is a severe one and it is possible for contamination or particulate matter deposition to interfere with the correct operation of sensors located there.

However, it will be appreciated that the invention is not limited to the use of two NOx and two particulate matter sensors and that the benefits and advantages of the invention are obtainable if only one sensor is used to -17 -measure NOx and one sensor is used to measure particulate matter output.

These comparisons generate error signals or differences from the target values (steady state optimised demand values) that are forwarded to the EGR valve 34 at block 160C and to the actuator control for the vanes 44 at block 160D.

At blocks 160C and 160D the system operates so as to eliminate the apparent errors so as to produce new NOx and particulate matter levels as indicated by blocks 180C and 180D but in this case the system is operable to determine which of the two variables NOx and particulate matter is most in need of reduction and to bias or prioritise control of the system so as to reduce the respective error towards zero. It will be appreciated that because the system is repetitively carrying out the comparison at blocks 120C and 120D the prioritisation will change over time. For example the initial need might be to reduce NOx and so priority will be given to maximising EGR and then as NOx falls particulate matter may constitute a more significant problem and so priority will switch to reducing particulate matter by increasing boost pressure.

Referring now to Figs.4, 5 and 9 there is shown in more detail the method described above in outline.

The method starts at block 200 which is a key-on event, then, at block 210, target values are determined for NOx and particulate matter. These may be obtained from look up tables stored in the memory of the ECU 24 for the current load and speed conditions of the engine 10.

-18 -Then at block 220 dynamic NOx and particulate matter values are obtained from, for example, the sensors 115C, 115D then at block 230 error or difference values for NOx and particulate matter are obtained using the equations:-Aparticulate matter (PM) = Dynamic (PM) -Target (PM) (1) ANOx = Dynamic NOx -Target NOx (2) Dynamic (PM) = Current particulate emissions Dynamic NOx = Current NOx emissions Then at block 240 it is determined whether the sum of the two differences is greater than a predetermined limit.

This test is done in order to determine whether there has been a failure in the engine 10 or its associated control system that will prevent ongoing safe running of the engine within prescribed limits. That is to say the test at block 240 acts as an OBD check.

Referring now briefly to Figs.4 and 5 examples for several situations are shown.

The reference point R' corresponds to the location of the target values of NOx and particulate matter and the line Q' upon which R' lies represents one relationship between NOx and particulate matter for the engine 10. Note that the point R' is a position where both of the variables NOx and particulate matter are low it is not the lowest NOx point nor the lowest particulate matter point but an optimised or compromise position therebetween.

-19 -It will be noted that between the points A' and B' large changes in particulate matter emissions occur with no significant change in NOx and similarly at the other extreme between the points E' and F' large changes in NOx emissions occur with no significant change in particulate matter. Therefore if the boost pressure is increased while the engine is operating in the A' to B' region it will have no significant effect on NOx but a dramatic reducing effect on particulate matter output and similarly if EGR flow is greatly increased while the engine 10 is operating in the E' to F' region it will have no significant effect on particulate matter but a dramatic reducing effect on NOx output. In fact, whenever the particulate matter emission error is positive (to the left of point R') increasing the turbocharger boost to reduce particulate matter emissions will not have a serious effect on NOx because in this operating region NOx is always lower than the optimised value R' and similarly whenever the NOx emission error is positive (to the right of point R') increasing the EGR flow will not have a serious effect on particulate matter emissions because in this operating region particulate matter emissions are always lower than the optimised value R' and it this realisation that forms the foundation for this invention.

Point A' represents a situation where there is a very high level of particulate matter emissions and the difference calculated using equation (2) is 10-3 giving a difference or error of +7 (e.g. an overproduction of particulate matter) but the NOx level calculated using equation (1) is 1-3 giving a difference or error of -2 is lower than the target level R' (e.g. an underproduction or -20 -better than expected output of NOx) . Fig.5 shows that the sum of these values is +5.

Point B' represents a situation where there is a high level of particulate matter emissions (error +5) but the NOx level (error-2) is lower than the target level R' . Fig.5 shows that the sum of these values is +3.

Point C' represents a situation where there is a slightly high level of particulate matter emissions (error +2) but the NOx level (error -1.5) is lower than the target level R' . Fig.5 shows that the sum of these values is +0.5.

Point F' represents a situation where there is a very high level of NOx emissions (error +8) but the particulate matter level (error -2) is lower than the target level R' Fig.5 shows that the sum of these values is +6.

Point E' represents a situation where there is a high level of NOx (error +5) but the particulate matter level (error-2) is lower than the target level R' . Fig.5 shows that the sum of these values is +3.

Point D' represents a situation where there is a slightly high level of NOx emissions (error +1.5) but the particulate matter level (error -1.0) is lower than the target level R' . Fig.5 shows that the sum of these values is +0.5.

Point G' represents a failure situation indicated by the fact that the point G' lies well off the line Q' and there is a high level of NOx emissions (error +4) and the -21 -particulate matter level (error + 4) is also higher than the target level R' . Fig.5 shows that the sum of these values is +8. The line L' represents a predetermined limit above and to the right of which an OBD failure is considered to have occurred and for which adjustment of the boost or EGR flow will have no significant effect.

It will be appreciated that minor deviations from the optimal relationship line Q' may occur and that these must not be the source of false OBD failure indications or unnecessary fuel supply restrictions.

It will be appreciated that these figures are based on subtracting the target values from the dynamic values in equations 1 and 2 so that overproduction of NOx or particulate matter produces positive differences which need to be reduced and this is seen as the easiest to understand and use approach. It will however be appreciated by those skilled in the art that equations 1 and 2 could be replaced by equations in which the dynamic values are subtracted from the target values and in this case it would be the largest negative difference that would be used to control the process and not the largest positive difference because a large negative difference would then represent an overproduction of NOx or particulate matter.

Returning now to block 240 on Fig.9, it will be appreciated that the test used is merely one of many that could be used and that the invention is not limited to such a test. For example, for every NOx output level (magnitude of output) there could be a particulate matter output above which a fault will automatically be indicated or for every particulate matter output level (magnitude of output) there could be a NOx output above which a fault will automatically be indicated. What is important is that minor deviations from the line Q' must not result in a fault being indicated.

If the test at block 240 is passed indicating a system failure then at step 244 the fuel supply to the engine 10 is reduced and at block 246 an warning is provided to a user of the engine 10 by, for example, lighting a warning lamp.

After block 246 the method advances via block 248 to block 500 to determine whether the key-on state is still on, if not the method ends at block 900 otherwise the method returns to block 210. Reducing the fuel supply may comprise limiting the maximum amount of fuel that can be supplied to the engine 10 irrespective of accelerator pedal position.

If, the test at block 240 is failed, then at block 250 it is determined which of the two errors is the largest positive error. Note that this is not just the largest error it is the largest positive error because positive errors need to be reduced whereas negative errors indicate that the engine 10 is operating with that particular variable (NOx or particulate matter) below the optimised value R' on Fig.4.

Then at block 300 the control of the engine 10 is prioritised based on the NOx error if it is the larger and the ECU 24 operates to reduce NOx by increasing EGR flow by opening the EGR valve 34.

If the NOx error is the larger positive error then at block 310 the control of the EGR valve 34 takes place based upon the actual magnitude of the error. That is to say, if -23 -the error is very large the change in valve position to produce increased EGR flow is large and rapid but if the error is small the change in EGR valve position is smaller and takes place at a slower rate thereby reducing the risk of system instability due to the system responding too slowly. After block 310 the method advances to block 500 to determine whether the key-on state is still on, if not the method ends at block 900 otherwise the method returns to block 210.

If the particulate matter error is the larger positive error then at block 400 the ECU 24 closes the vanes 44 SO as to increase boost thereby reducing particulate matter emissions then at 410 the control of the vanes 44 takes place based upon the actual magnitude of the error. That is to say, if the error is very large the change in vane position to produce increased boost is large and rapid but if the error is small the change in vane position is smaller and takes place at a slower rate. After block 410 the method advances to block 500 to determine whether the key-on state is still on, if not the method ends at block 900 otherwise the method returns to block 210.

It will be appreciated that the methods shown in Fig.3 and 9 are performed by the electronic controller or ECU 24 in real time.

Referring back now to Fig.3 although the NOx and particulate matter sensors 115D are described above as being physical sensors the inventors have realised that virtual sensors stored as a multiplicity of look up tables in the memory of the ECU 24 could be used. One advantage of such a virtual sensor is that it is not exposed to the harsh -24 -environment of the exhaust manifold 28 and so is unaffected by the temperature and conditions present there. A further advantage is that once produced a virtual sensor is very cheap to use as it merely has to be programmed or stored in the memory of the ECU 24 and so adds no real cost to the engine 10 nor complexity to the wiring loom.

With particular reference to Figs.6A to 8B there are shown some of the steps required to produce a virtual sensor for an engine having two controllable interrelated variables such as boost pressure and EGR flow rate, fuel rail pressure and injection timing or other variables that affect the parameter to be sensed e.g. NOx or particulate matter.

The method requires the use of an engine operated on a fully instrumented dynamometer rig that includes physical sensors for the parameters that need to be measured.

The first step of the method comprises selecting a predetermined number of speed and load combinations. As shown in Fig.6A a matrix of points XY to X5Y5 for the normal operating range of the engine 10 is produced for use in determining the transient NOx output from the engine 10 and similarly in Fig.6B a matrix of points XY to X5Y5 for the normal operating range of the engine 10 is produced for use in determining the transient particulate matter output from the engine 10. It will be appreciated that the same points may be used for both NOx and particulate matter or different points may be used.

Then, for each speed and load combination, one variable is set to a predetermined level and the other variable is swept through its normal operating range while capturing -25 -data forming the subject of the sensor. For example and with reference to Figs.7A, EGR valve position is held at level Ni while vane position P' is swept from P to P5 and the data (NOx in this case) is stored as a look up table.

The process is then repeated for predetermined levels N2 to N5 representing steps in the normal operating range of the EGR valve 34 and in each case the data is stored as a look

up table.

Although these look up tables are sufficient to meet the needs of the invention it is preferred if a second sweep as detailed below is performed in order to confirm the values obtained from the first sweep.

The other variable namely vane position is set to a predetermined level P and the EGR valve position is swept from N to N5 corresponding to its normal operating range while capturing data representing the NOx level which in this case forms the subject of the sensor, and the process is repeated for predetermined levels P2 to P5 of the normal operating range of the vanes 44 and the data is saved as a series of look up tables.

Then, with reference to Fig.7B a sensor for particulate matter is produced using the same technique. For each of the points XY to X5Y5 shown on Fig.6B the EGR valve position is held at level Ni while vane position P' is swept from P to P5 and the data (particulate matter in this case) is stored as a look up table.

The process is then repeated for predetermined levels N2 to N5 representing steps in the normal operating range of the -26 -EGR valve 34 and in each case the data is stored as a look

up table.

Although these look up tables are sufficient to meet the needs of the invention it is preferred if a second sweep as detailed below is performed in order to confirm the values obtained from the first sweep. The other variable namely vane position is set to a predetermined level P and the EGR valve position is swept from N to N5 corresponding to its normal operating range while capturing data representing the particulate matter level which in this case forms the subject of the sensor, and the process is repeated for predetermined levels P2 to P5 of the normal operating range of the vanes 44 and the data is saved as a series of

look up tables.

After this process is complete a series of look up tables providing NOx output for each of the points XY to X5Y5 on Fig.6A has been generated and a series of look up tables providing particulate matter output for each of the points XY to X5Y5 on Fig.6B has been generated.

These look up tables could then be stored in the memory of the ECU24 and be used to produce virtual sensor outputs corresponding to the outputs from 115C, 115D indicated on Fig.3 for comparison with the target values of NOx and particulate matter. That is to say single virtual sensors for NOx and particulate matter could be used.

However, according to a further embodiment of the invention they form respectively virtual sensors corresponding to the sensors hOC and hOD on Fig.3 and second virtual sensors hh2C and hh2D for NOx and particulate -27 -matter are produced in a similar manner to that described above but using different variables namely fuel rail pressure and fuel injection timing.

As before, the first step to produce sensors 112C and 112D is to select a predetermined number of speed and load combinations such as those shown in Figs.6A and 6B.

Then, as before, for each speed and load combination, one variable is set to a predetermined level and the other variable is swept through its normal operating range while capturing data forming the subject of the sensor.

For example and with reference to Fig.8A, fuel injector timing is held at level K while fuel rail pressure L' is swept from L to L5 and the data (NOx in this case) is stored

as a look up table.

The process is then repeated for predetermined levels K2 to K5 representing steps in the normal fuel injection timing range and in each case the data is stored as a look up table. Although these look up tables are sufficient to meet the needs of the invention it is preferred if a second sweep as detailed below is performed in order to confirm the values obtained from the first sweep. The other variable namely fuel rail pressure is set to a predetermined level L and the fuel timing is swept from K to K5 corresponding to its normal fuel injection timing range while capturing data representing the NOx level which in this case forms the subject of the sensor, and the process is repeated for predetermined levels L2 to L5 of the normal range of fuel line pressure and the data is saved as a series of look up

tables.

-28 -Then, with reference to Fig.8B a sensor for particulate matter is produced using the same technique. For each of the points XY to X5Y5 shown on Fig.6B fuel injector timing is held at level K while fuel rail pressure L' is swept from L to L5 and the data (particulate matter in this case) is stored as a look up table.

The process is then repeated for predetermined levels K2 to K5 representing steps in the normal fuel injection timing range and in each case the data is stored as a look up table. Although these look up tables are sufficient to meet the needs of the invention it is preferred if a second sweep as detailed below is performed in order to confirm the values obtained from the first sweep. The other variable namely fuel rail pressure is set to a predetermined level L and the fuel timing is swept from K to K5 corresponding to its normal fuel injection timing range while capturing data representing the particulate matter level which in this case forms the subject of the sensor, and the process is repeated for predetermined levels L2 to L5 of the normal range of fuel line pressure and the data is saved as a series of look up

tables.

After this process is complete a second series of look up tables providing NOx output for each of the points XY to X5Y5 on Fig.6A has been produced and a second series of look up tables providing particulate matter output for each of the points XY to X5Y5 on Fig.6B has been produced.

The plurality of look up tables can then be saved in the memory of the ECU 24 to form four virtual sensors hOC, hh2C and hOD, hh2D that are able for various combinations -29 -of engine speed and load to provide a value for NOx or particulate matter.

As shown in Fig.3 the output values from the virtual sensors hOC and 112C are combined to provide the composite sensor 115C and the outputs from the virtual sensors hOD and 112D are combined to produce the composite sensor 115D.

However, although these may be combined in a simple manner such as; 115C = O.5(11OC Value + 112C value); and 115D = O.5(11OD Value + 112D value); it will be appreciated that other combinations are possible. For example, if the major transient change to operation of the engine 10 is a change in fuel injector timing or fuel rail pressure then bias may be given to the values of NOx and particulate matter provided by the sensors 112C and 112D whereas if the major change is EGR flow or boost pressure then bias may be given to the sensors hOC and hOD and in this way the transient levels of NOx and particulate matter are modelled more accurately on the actual changes occurring during use of the engine 10.

The process of generating data by varying airpath (EGR valve versus turbocharger boost position) or combustion variables (fuel rail pressure versus fuel injection timing) may be improved by applying design of experiment, statistical or CAE based methodologies.

-30 -Although the invention has been described with reference to an engine having a throttle valve it will be appreciated that it can also be applied to a diesel engine having no throttle valve.

It will be appreciated by those skilled in the art that although the invention has been described by way of example with reference to one or more embodiments it is not limited to the disclosed embodiments and that one or more modifications to the disclosed embodiments or alternative embodiments could be constructed without departing from the scope of the invention as set out in the appended claims.

Claims

-31 -Claims 1. A method for controlling a diesel engine in which NOx emissions are controlled by varying exhaust gas recirculation flow through an exhaust gas recirculation system and particulate matter emissions are controlled by varying the boost from a turbocharger wherein the method comprises prioritising control of the exhaust gas flow and the turbocharger boost based upon which of the two emissions is currently the most significant.
2. A method as claimed in claim 1 wherein, if NOx emissions are currently the most significant, then priority is given to increasing exhaust gas circulation flow.
3. A method as claimed in claim 1 or in claim 2 wherein, if particulate matter emissions are currently the most significant, then priority is given to increasing turbocharger boost.
4. A method as claimed in any of claims 1 to 3 wherein the diesel engine has an exhaust manifold and an inlet manifold, the turbocharger has a turbine arranged so as to be driven by the exhaust gases from the engine, a compressor for selectively increasing the pressure in the inlet manifold of the engine and an electronically controlled inlet flow control device to regulate the exhaust gas entering the turbine and the exhaust gas recirculation system selectively recirculates gas from a position in the exhaust manifold located upstream from the turbocharger to the inlet manifold of the engine and includes an electronically controlled exhaust gas recirculation control valve to control the flow of exhaust gas passing from the -32 -exhaust manifold to the inlet manifold wherein the method further comprises prioritising control of the exhaust gas recirculation control valve or the inlet flow control device based upon which is the larger of the differences between (i) a target NOx output value and a current NOx output value, and (ii) a target particulate matter output value and a current particulate matter output value.
5. A method as claimed in claim 4 wherein the method further comprises, if the larger difference is the NOx difference, adjusting the position of the exhaust gas recirculation valve so as to drive the NOx output from the engine towards the target NOx output value.
6. A method as claimed in claim 5 wherein the method further comprises adjusting the position of the exhaust gas recirculation valve so as to drive the NOx output from the engine towards the target output value based upon the magnitude of the NOx difference.
7. A method as claimed in any of claims 4 to 6 wherein the method further comprises, if the larger difference is the particulate matter difference, adjusting the position of the inlet flow control device so as to drive the particulate matter output from the engine towards the target particulate matter output value.
8. A method as claimed in claim 7 wherein the method further comprises adjusting the position of the inlet flow control device so as to drive the particulate matter output from the engine towards the target output value based upon the magnitude of the particulate matter difference.

-33 -
9. A method as claimed in any of claims 4 to 8 wherein the method further comprises summing the NOx and particulate matter differences and, if the sum of the NOx and the particulate matter differences is greater than a predetermined value, reducing the fuel supplied to the engine.
10. A method as claimed in any of claims 4 to 9 wherein the method further comprises summing the NOx and particulate matter differences and, if the sum of the NOx and the particulate matter differences is greater than a predetermined value, providing a warning to a user of the engine.
11. A method as claimed in claim 9 or in claim 10 wherein the predetermined value varies based upon a function of NOx output magnitude and particulate matter output magnitude.
12. A control system for controlling a diesel engine in which NOx emissions are controlled by varying exhaust gas recirculation flow through an exhaust gas recirculation system and particulate matter emissions are controlled by varying the boost from a turbocharger wherein the system includes an electronic controller arranged to prioritise control of the exhaust gas flow and the turbocharger boost based upon which of the two emissions is currently the most significant.
13. A system as claimed in claim 12 wherein, if NOx emissions are currently the most significant, then the electronic controller is operable to give priority to increasing exhaust gas circulation flow.

-34 -
14. A method as claimed in claim 12 or in claim 13 wherein, if particulate matter emissions are currently the most significant, then the electronic controller is arranged to give priority to increasing turbocharger boost.
15. A system as claimed in any of claims 12 to 14 in which the diesel engine has an exhaust manifold and an inlet manifold, the turbocharger has a turbine arranged so as to be driven by the exhaust gases from the engine and a compressor for selectively increasing the pressure in the inlet manifold of the engine and the exhaust gas recirculation system selectively recirculates gas from a position in the exhaust manifold located upstream from the turbocharger to the inlet manifold of the engine, wherein the system further comprises an inlet flow control device controlled by the electronic controller to regulate the exhaust gas entering the turbine of the turbocharger, an exhaust gas recirculation control valve controlled by the electronic controller to control the flow of exhaust gas passing from the exhaust manifold to the inlet manifold though the exhaust gas recirculation system, the electronic controller being operable to prioritise control of the exhaust gas recirculation control valve or the inlet flow control device based upon which is the larger of the differences between (i) a target NOx output value and a current NOx output value, and (ii) a target particulate matter output value and a current particulate matter output value.
16. A system as claimed in claim 15 wherein the electronic controller is further operable if the larger difference is the NOx difference to adjust the position of -35 -the exhaust gas recirculation valve so as to drive the NOx output from the engine towards the target NOx output value.
17. A system as claimed in claim 16 wherein the electronic controller is further operable to adjust the position of the exhaust gas recirculation valve so as to drive the NOx output from the engine towards the target output value based upon the magnitude of the NOx difference.
18. A system as claimed in any of claims 15 to 17 wherein the electronic controller is further operable if the larger difference is the particulate matter difference to adjust the position of the inlet flow control device so as to drive the particulate matter output from the engine towards the target particulate matter output value.
19. A system as claimed in claim 18 wherein the electronic controller is further operable to adjust the position of the inlet flow control device so as to drive the particulate matter output from the engine towards the target output value based upon the magnitude of the particulate matter difference.
20. A system as claimed in any of claims 15 to 19 wherein the electronic controller is further operable to sum the NOx and particulate matter differences and, if the sum of the NOx and the particulate matter differences is greater than a predetermined value, reduce the fuel supplied to the engine.
21. A system as claimed in any of claims 15 to 19 wherein the electronic controller is further operable to sum the NOx and particulate matter differences and, if the sum -36 -of the NOx and the particulate matter differences is greater than a predetermined value, provide a warning to a user of the engine.
22. A system as claimed in claim 20 or in claim 21 wherein the predetermined value varies based upon a function of NOx magnitude and particulate matter magnitude.
23. A system as claimed in any of claims 15 to 22 wherein determining the current NOx output level comprises measuring the current NOx output level using a NOx sensor located in the exhaust manifold.
24. A system as claimed in any of claims 15 to 22 wherein determining the current NOx output level comprises using a virtual NOx sensor generated using a number of look up tables stored in the electronic controller to provide a value for current NOx.
25. A system as claimed in any of claims 15 to 24 wherein, determining the current particulate matter output level comprises measuring the current output level of particulate matter using a particulate matter sensor located in the exhaust manifold.
26. A system as claimed in any of claims 15 to 24 wherein determining the current NOx output level comprises using a virtual particulate matter sensor generated using a number of look up tables stored in the electronic controller to provide a value for current NOx.
27. A motor vehicle having a control system as claimed in any of claims 12 to 26.

-37 -
28. A method of producing a virtual sensor for an engine having a pair of interrelated variables wherein the method comprises selecting a predetermined number of speed and load combinations, for each speed and load combination, setting one variable to a predetermined operating level and controlling other variable so as to sweep through its normal operating range while capturing data forming the subject of the sensor, repeating the process for predetermined operating levels throughout the normal operating range of the one variable and storing the data as a series of look uptables.
29. A method as claimed in claim 28 wherein the method further comprises, setting the other variable to a predetermined operating level and controlling said one variable so as to sweep through its normal operating range while capturing data forming the subject of the sensor, repeating the process for predetermined operating levels throughout the normal operating range of the other variable and storing the data as a series of look up tables.
30. A method as claimed in claim 28 or in claim 29 wherein the method further comprises producing a first virtual sensor for the subject of the sensor using a method as claimed in claim 22 or in claim 23 producing a second virtual sensor for the same subject using a method as claimed in claim 22 or in claim 23 with different variables to those used to produce the first virtual sensor and combining the values from the first and second virtual sensors to produce a compensated sensor value for the subject matter of the sensor.

-38 -
31. A method as claimed in any of claims 28 to 30 wherein one pair of variables is turbocharger boost position and exhaust gas recirculation flow valve position.
32. A method as claimed in any of claims 28 to 30 wherein one pair of variables is fuel rail pressure and fuel injection timing.
33. A method as claimed in any of claims 28 to 32 wherein the virtual sensor is a virtual NOx sensor, the subject is NOx and the data stored is values of NOx produced.
34. A method as claimed in any of claims 28 to 32 wherein the virtual sensor is a virtual particulate matter sensor, the subject is particulate matter and the data stored is values of particulate matter produced.
35. A method for controlling a diesel engine substantially as described herein with reference to Figs. 3 to 9 of the accompanying drawing.
36. A control system for controlling a diesel engine substantially as described herein with reference to Figs. 1 and 3 to 9 of the accompanying drawing.
37. A motor vehicle substantially as described herein with reference to Figs.1 and 3 to 9 of the accompanying drawing.
38. A method of producing a virtual sensor for an engine having two controllable interrelated variables -39 -substantially as described herein with reference to Figs. 3 to 8B of the accompanying drawing.