GB2558604A

GB2558604A - Method to detect faults in boost system of a turbocharged engine

Info

Publication number: GB2558604A
Application number: GB1700347.6A
Authority: GB
Inventors: Carrillo Nicolas
Original assignee: Delphi International Operations Luxembourg SARL
Current assignee: Delphi International Operations Luxembourg SARL
Priority date: 2017-01-09
Filing date: 2017-01-09
Publication date: 2018-07-18
Anticipated expiration: 2037-01-09
Also published as: GB201700347D0; GB2558604B; WO2018127598A1

Abstract

A method of determining a fault in the boost system of a turbocharged or supercharged engine; comprising the steps of a) ascertaining a transient boost period b) determining an actual boost parameter 10 during the transient; c) determining an expected or ideal boost parameter 11 during the transient; and, d) comparing the parameters of steps b) and c) to ascertain if a fault is present. In this way, incorrect boost control can be diagnosed, such as a slow response time. The method may include determining the time taken for the actual boost parameter to reach the value of the ideal boost parameter as steady state, and may also include determining integral values of the actual and expected boost parameters. The expected boost parameter may be determined from a system model, and the actual boost parameter, which may be boost pressure, may be determined from a sensor or system model. The method may include determining a measure of the boost ratio during the transient event and comparing it with a threshold value, and validating any fault determined if it is above said threshold.

Description

(54) Title of the Invention: Method to detect faults in boost system of a turbocharged engine Abstract Title: Fault detection in boost system of a turbocharged engine (57) A method of determining a fault in the boost system of a turbocharged or supercharged engine, comprising the steps of a) ascertaining a transient boost period b) determining an actual boost parameter 10 during the transient; c) determining an expected or ideal boost parameter 11 during the transient; and, d) comparing the parameters of steps b) and c) to ascertain if a fault is present. In this way, incorrect boost control can be diagnosed, such as a slow response time. The method may include determining the time taken for the actual boost parameter to reach the value of the ideal boost parameter as steady state, and may also include determining integral values of the actual and expected boost parameters. The expected boost parameter may be determined from a system model, and the actual boost parameter, which may be boost pressure, may be determined from a sensor or system model. The method may include determining a measure of the boost ratio during the transient event and comparing it with a threshold value, and validating any fault determined if it is above said threshold.

/5

2/5

CM CM

CD

CM

O

N00

CD

CM

CD

O

CD

LO

O

LO

O tT

FIG.

3/5

FIG. 5

4/5

LO

X)

100

Ο

FIG

5/5

FIG

METHOD TO DETECT FAULTS IN BOOST SYSTEM OF A TURBOCHARGED ENGINE

TECHNICAL FIELD

This invention relates to turbocharged and /or supercharged engine system and in particular to a method of detecting faults in the boost system of such engines.

BACKGROUND OF THE INVENTION

Modem engines such as Diesel engines typically may include turbochargers systems. In such systems exhaust gas is used to drive a turbine, connected in turn with a compressor to increase or boost the pressure of air entering the engine. Turbochargers (boost) may be controlled by boost actuators which actuate a wastegate in the turbocharger system to control the proportion of exhaust gas which runs the turbine. In other systems, the turbo may have Variable Gate Geometry (VGT). Failure of VGT or boost control (e.g. wastegate control) can cause problems in efficient control and may result in inefficient operation of the engine and increase noxious gases in the exhaust above regulatory limits. Various other causes adversely affect boost control such as slow responding boost actuator; vacuum pressure leakage on the tubes connected to any of these boost actuators, air leak before and after e.g. charge air intercooler on intake path of the engine; air leak on the exhaust path of the engine before the boost actuator, and such like.

A prior art method of diagnosing such faults in boost control consists of intrusive diagnostic during idle operating conditions of the internal combustion engine by closing or opening the turbocharger/supercharger VGT or wastegate to achieve a noticeable change in the readings of the MAP sensor.

According to CARB OBD II regulation, with regards slow response: for 2013 and subsequent model year vehicles, the OBD II system shall detect a malfunction prior to any failure or deterioration in the boost pressure control system response (e.g., capability to achieve the commanded or expected boost pressure within a manufacturer-specified time) that would cause vehicle's NMHC, CO, NOx, or PM emissions to exceed the applicable emission levels specified. For vehicles in which no failure or deterioration of the boost system response could result in an engine’s emissions exceeding these levels, the OBD II system shall detect a malfunction of the boost system when no detectable response to a commanded or expected change in boost pressure occurs.

The prior art is capable of diagnosing a slow responding boost pressure control system only during idle operating conditions of the combustion engine. It is an object of the invention to provide an improved method of diagnosing incorrect boost control such as slow responding boost pressure control systems. The diagnostic of a slow responding boost pressure control system is mandatory for obtaining certification.

It is an object of the invention to provide an improved method of diagnosing faulty boost systems and faulty boost response.

SUMMARY OF THE INVENTION

In one aspect is provided a method of determining a fault in the boost system of a turbocharged engine, comprising the steps of: a) ascertaining a transient boost period; b) determining an actual boost parameter during the transient; c) determining an expected or ideal boost parameter during the transient; and, d) comparing the parameters of steps) b) and c) to ascertain if a fault is present.

Step d) may comprise determining the time taken for the actual boost parameter to reach the value of the ideal boost parameter at steady state.

Step d) may compare the value of the actual boost parameter with the ideal boost parameter value at a time within the transient period.

Step d) may comprise determining integral values of the actual and expected boost parameters during a time period within the transient, and comparing these.

The expected boost parameter may be determined from a system model.

The actual boost parameter may be determined from a pressure sensor or from a system model.

Said boost parameter may be is boost pressure.

Step a) may include determining if the requested torque demand or rate of change of torque demand exceeds a predetermined value.

Step a) may include determining if one or more of the following exceeds predetermined value(s): minimum IMEP, minimum enthalpy and minimum rate of change of the boost pressure or ratio during transient conditions.

Step a) may include determining if one or more operating conditions of the engine are present.

Said transient period may be defined as any time period where there is a positive boost demand.

The method may include determining a measure of the boost ratio during said transient and comparing this with a threshold value, and if above said threshold, validating any fault from step d).

Said measure of boost ratio may be the average boost ratio during the transient.

The method may include the initial step of determining if the accumulated value of velocity multiplied by positive acceleration (VxA) or Positive Kinetic Energy (PKE) increases at a rate above a threshold value and implementing the step of claim 12 is this condition if fulfilled.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention is now described by way of example with reference to the accompanying drawings in which:

Figure 1 shows a schematic diagram of a turbocharger system;

Figure 2 shows the typical system response;

Figure 3 shows plots of a measured actual boost pressure with time along with the expected boost pressure;

Figure 4 shows a portion of figure 3 in more detail;

Figure 5a shows a block diagram showing simple and refined embodiments of the methodology of the inventions; and

Figure 5b shows a yet further refinement of the methodology.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Figure 1 (DP-321059) shows a schematic diagram of a turbocharger system for a combustion engine 5 showing on air inlet 1 to a compressor 2. An air filter 3 is typically located between the air inlet and the compressor. Compressed air is fed into the engine from the compressor outlet and may be via an intercooler 6. A turbine 4, connected to the compressor, is located downstream of the engine e.g. downstream of the exhaust manifold/port. The turbine includes a wastegate arrangement (not shown) which is typically controlled dependent on pressure in the air inlet by means of a wastegate bypass regulator. Typically an intercooler 6 is located between the compressor outlet and engine. The compressor pressure ratio is the compressor outlet pressure divided by the compressor inlet pressure. As mentioned the exhaust arrangement may include a wastegate with the turbine 4. The wastegate is typically controlled dependent on pressure in the air inlet by means of wastegate bypass regulator (boost controller) not shown.

In one aspect the methodology of determining a faulty (e.g. slow responding) boost pressure system is generally a two-step process. In the first step, the method determines whether there is a valid transient condition; e.g. if there is an increase (e.g. sharp) in boost demand. Under such conditions e.g. when there is a driver request for acceleration (torque demand) a quick response of the boost pressure control system is expected. In the second step the response of the system is analysed; this may be performed in several ways.

Generally engine systems include engine models which can provide outputs such as a predicted boost pressure; that is the pressure on the high side of the compressor or the pressure entering the engine, or any point between them. Typically, such models have an input of torque demand which provides a boost pressure demand from which the model can convert or provide a modelled expected boost pressure. In one aspect the output from an actual (boost) pressure sensor located between the compressor outlet and engine is compared with that from the modelled expected boost pressure. In this way, the second part of the method looks at the boost pressure control system response.

So according to a basic example, the method diagnoses a slow responding boost pressure control system by recognizing two possible statuses (transient and steady state) of the turbo system, and determining boost system characteristics during the transient states.

The method in one example gathers information to diagnose the boost pressure control system by comparing the actual boost pressure (characteristics) with an expected boost pressure (characteristics). In refined embodiments, those transient conditions for the best results are determined. If such methodology indicates a potential fault, in further refined embodiments, the fault is further validated or not validated according to other criteria; which includes depending on the aggressiveness of the driving profile and subsequent analysis of the boost ratio.

Technical Background

From the control theory the response of a control system to a change over its operating condition is well known. Figure 2 shows the typical system response. The top plot 8 shows a step change in a (demand) variable which is input into the system and the bottom plot shows the response with time; in other words it shows the response behavior of a control system to a change, in this case, a step change. There are two stages/ statuses of this control system: transient and steady state. A system is said to be in a transient state when a process variable has been changed and the system has not yet reached steady-state (from tl -13). During steady-state the dynamic changes on the system are mostly damped so that no perceptible change on the dynamics of the system will be visible and this same behavior will continue into the future.

In looking at a boost control system for turbocharged engines, once the steady state in a boost pressure control system has been determined, the dynamics of boost response (measured boost in the system) should have been mostly damped, i.e., not showing strong oscillating behaviour. However, if the dynamics of the boost response have not yet been damped even after the steady-state has been determined, the boost pressure control system is said to be slow responding and may indicate a fault.

Example 1

Figure 3 shows plots of a measured actual boost pressure 10 with time along with the expected boost pressure 11, the latter being determined from an engine model based on the input parameter of e.g. torque/boost demand. At point A the driver of the vehicles puts his foot down on the accelerator pedal which is registered as an increase in torque demand. So plot 11 indicates the ideal response of the boost pressure in a normally operating engine /boost control systems. The plot 10 shows the example of slow responding boost pressure control system caused by e.g. a leak in the exhaust system just in front of the turbine. The valid transient regions 13 are shown by in timeslots 12. According to aspects, the actual and ideal transient response are compared to determine if the boost control system is operating normally. At point B the driver lifts his foot from the accelerator so the boost/torque demand falls. It should be understood that the comparison may be alternatively determined subsequent to a decrease in torque/boos demand.

In one aspect the time taken for the actual boost pressure to achieve a generally steady state level is compared to the time taken for the modelled expected boost pressure to achieve this level. The modelled boost pressure reaches a steady state value of 250 at time of t=64 seconds, so 2 seconds after the increase in demand. It actually takes nearly 4 seconds for the actual pressure to reach this value at just before t=66. Thus, in one example, the time is takes for the actual boost pressure to reach a percentage values of the modelled pressure (preferably a steady state level) compared to the time can be compared to give an evaluation of a slow responding sytems. Other methods of comparing the two signals/parameters will be described in more detail.

Figure 4 shows a portion of figure 3 in more detail. Again shows the value of the actual boost pressure (Pa) 10 (e.g. measured by a pressure sensor at the intercooler outlet) and modelled ideal boost pressure (Pi) 11 in the transient region. At any point, preferably in the transient region, the values of the ideal boost pressure (Pi) and actual boost pressures (Pa) may be compared such as at points Z or Y, the latter point being where the ideal boost pressure should have reached steady state. So the pressure Pi at Z (PiZ) and Pa at Y (piY) may be compared tot he pressure Pa at Z (PaZ)and Pi at Y (PiY) respectively . Alternatively the gradients 13 and 14 may be compared.

The inventors have determined that a particular and accurate method of determing a slows responding or faulty system is by looking at the areas under the curves in a respective (e,g. transient) periods; in other words integrating the values of the ideal and actual pressure over a time period/slot, and comparing them. In the example the area 15 bounded with respect to the expected/ideal boost pressure and time (shown by single hatching in the figure) may be compared to the area 16 with respect to the area of the actual pressure with respect to time (shown by cross-hatching in the figure). The time slot for the integration may be any time slot during the transient response. It should be noted that in the examples the values of boost pressure (ideal and actual) are compared but any boost parameters such as boost ratios (ideal versus actual) may be compared. Although the above describes looking at the actual transient response and ideal trasnient response by looking at the expected and actual boost pressures, any boost parameters may be used such as the boost ratios (ideal and actual) so in other words the boost ratio computed during transient conditions of the boost pressure control system is used as evaluation parameters. The boost ratio is obtained dividing the area under the curve of the boost demand by the area under the curve of the boost response.

So in general it is also possible to observe that during the transient state the boost response (PSE_Intercooler_out_press) (actual boost pressure) is not able to follow the boost demand (ACM_Vgt_boost_est_ctrl) (computed ideal boost pressure/demand thereof) due to the leak before the turbine.

Example 2

In further examples the initial step of the method determines that an adequate transient state of the boost pressure control system is present only during positive boost demand.

So as mentioned, in aspects of the methods, the diagnostic method is determined under transient state. In preferred embodiments, a method step entails determining suitable transient states, and further ideal transient states likely to give the most accurate results are determined by further considerations - thus in preferred aspects, the transient state is validated by additionally considering extra parameters for this purpose, such as minimum IMEP, minimum enthalpy and minimum rate of change of the boost during transient conditions, as well as the aggressiveness of the drive.

In simple embodiments transient periods can be determined from the accelerator/pedal input or torque demand (e.g. torque demand variable output form the ECU). If there is a sharp increase e.g. rate of change of this variable, then transient conditions are determined to be present. So such parameters or derivatives of parameters can be compared with thresholds, said threshold may be variable. These thresholds and other parameters in this stage can be formulated from test or calibration data and it would be obvious for the skilled person to provide such a determinations in various ways.

In examples the values of one or more of the parameters of IMEP, enthalpy and (requested) rate of change of the boost during transient conditions, are determined and compared with minimum values. If above the minimum then transient conditions may be determined as being present.

Other factors may be considered in determining adequate transient condition. These include index of aggressiveness of the current drive and/or accumulated time of transient conditions of positive boost demand. Additionally the operating condition of boost control (e.g. boost ratio)may be determined and analysed as a factor in determining if transient conditions are present to further validate any initial indication of a fault. It should be noted that any one or more or combinations of such factors may be performed or validated before the further step of comparison is implemented.

So preferably therefore also the diagnostics ensure an improved robustness when the transient is validated by under aggressive driving conditions. Under gentle driving conditions, as NEDC or FTP75 for example, the results of the methodology are not as accurate as they can be. During NEDC there is only one phase with a representative acceleration event in the extra urban portion of the cycle. In “aggressive” cycles, as RTS95, UDC92 or US06 there are much more opportunities to obtain representative acceleration phases. Thus the driving profile or torque demand profile can be considered to determine the best transient time to perform the diagnostics. These can depend on the “aggressiveness” of such profiles and can take parameters of Positive Kinetic Energy (PKE) or the cumulated VxA (velocity multiplied by acceleration) as the aggressiveness evaluation factor to select times for diagnostics. With this particular methodology, where the determined aggressiveness is above a level, the average boost ratio during the transient is calculated (e.g. depending on the number of available valid events) and dependent on this the preliminary indication of fault is validated or not.

Depending on the aggressiveness of the driving profile, the diagnostic may adapt itself to gather the necessary information to diagnose the boost pressure control system. If the aggressiveness of the driving profile is high, less information about the behavior of the boost pressure control system is necessary' to perform the diagnostic. If the aggressiveness of the driving profile is low, more information about the behavior of the boost pressure control system is necessary to perform the diagnostic. This adaptive way of doing diagnostics allows to launch the diagnostic during all driving conditions and aggressiveness profiles, as city, extra urban or motorway off-cycle drive profiles, or (among others) on-cycle driving profiles.

Example 3

Figure 5a shows a block diagram showing simple and refined embodiments of the methodology of the inventions e.g. diagnosing a slow responding boost pressure control system. It is to be noted that one or more input parameters or stages can be omitted, in that they are preferred features. In block 20 the input is boost demand 21 (equivalent to expected boost pressure 11 of figure3), and actual boost pressure 22 (so equivalent to measured actual boost pressure 10 of figure 3). The boost response 22 parameters may be actual boost pressure (measured by a sensor or modelled). The boost demand may be the boost pressure demand or an ideal boost pressure if the response was ideal; this parameter may be obtained from a system model. In block 20 these parameters are compared to determine if there is a problem with the boost control e.g. slow responding, according to the various methods referred to above. The output of block 20 that is 40, is a possible indication of a slow or faulty boost response. This may be validated further in more refined embodiments to improve accuracy as will be explained later.

In preferred embodiments, one or more other considerations are taken into account to determine valid transient conditions. Thus inputs of (requested) IMEP 24, and/or (requested) enthalpy 25 and/or (requested) rate of change of the boost during transient conditions 26 may be input to block 27 which compares these values or rates of change thereof with a minimum or thresholds. The output of this may validate the transient conditions in such a way that one or more of the conditions (e.g. above said minimums) have to be present for a valid/adequate transient to be present before any results from block 20 are performed or validated.

A further consideration may be one or more operating conditions of the boost control or engine 28 which is input to block 29 where these conditions are processed e.g. compared with threshold data in order to validate whether a transient condition is valid This may entail e.g. determining transient conditions with positive boost demand or any other situation where certain conditions have to be in place for the calculation in block 20 to take place before and output is made. The output of block 27 and 29 is to block 30 where one or both of the set of conditions in blocks 29 and 27 have to be valid before the output of block 30 indicates a positive boost transient which is valid. Of course blocks 29 or 27 may be dispensed with in simpler systems and methods and any one or more of the parameters 24, 25, 26 and 28 may be utilised.

If one or more of these conditions are valid then the calculation stage of block 20 may be performed.

Example 4

Figure 5b shows a yet further refinement of the methodology which may be performed in addition to any of the methodology described above and which includes extra steps or stages. These stages may include further ensuring the best transient times are used in the diagnostics. To box 31 is input an index of aggressiveness of the current drive 32 and/or the accumulated time of transient conditions with positive boost demand 33. These terms will be explained hereinafter. Input to box 31 is also the output from Figure 5a, so a potential fault indicated. Thus the figure 5b can be interpreted as the indication of a preliminary flagged-up fault as also dependent on further conditions being fulfilled, in particular dependent on the input of one or more parameters from 32 and 33. Box 31 represents a method which determines the additional requirements to be valid in order for the boost system failure to be triggered. In one aspect the box 31 compares one or more parameters of the index of aggressiveness/and or the transient time where there is a positive boost demand with thresholds and if these inputs satisfies certain requirements

i.e, the parameters are above a threshold). The aggressiveness is a measure of how hard the vehicle is driven and can be determined form parameters such as Positive Kinetic Energy (PKE) or (changes thereof) or the cumulated VxA (velocity multiplied by acceleration) over the transient period as the aggressiveness evaluation parameter to select times for diagnostics. So , the box 31 compares one or more parameters of the index of aggressiveness/and or the transient time where there is a positive boost demand and if these inputs satisfies certain requirements (i.e, the parameters are above a threshold) box 31 also determines the average boost ratio, during the transient period in question e.g. where the boost demand is above a certain level or positive or where at least one of the parameters of aggressiveness is above a threshold). The output of box 31 is the determination of average boost ratio during t this time.

The boost ratio is then compared at 34 with threshold 35. If the average boost ratio 36 is greater than the threshold then the fault at 40 is validated and a fault indicated. If the average boost ratio is below a threshold value the potential fault at 40 is not validated and a pass registered. So thus in this example effectively a transient (period timeslot) which is used e.g. in figure 5a methodology is designated as accurate and highly reliable for the diagnostic to be marked as valid if the boost ratio is above a certain level. This may further be determined as being sufficient to perform the further validation. In summary box 31 determines the needed valid events (dependent on driving profile /aggressiveness) before calculating the average boost ratio 36.

Depending on e.g. the aggressiveness of the driving profile, the diagnostic method will (e.g. by learning in an adaptive way) determine the number of valid transient states necessary' for the diagnostic. So for example there may only be needed one valid transient event i.e. one calculation of the methodology Use of the VxApos method (Velocity x Positive Acceleration) to determine, in real time, the aggressiveness of the driving profile

Figure 6 shows plots of various parameters during a vehicle journey against time. In the top chart 6 a), plot 51 shows a plot of vehicle speed. The plot in figure 5b shows the number of valid events calculated by box 31 /and plot of figure 6c shows the cumulated VxA (velocity multiplied by acceleration). Area indicated by the arrow A shows the point where the speed rapidly increases as a result of high acceleration. This can be considered as a time of aggressive driving. The value of the cumulated VxA grows quickly and in an example reaches the minimum required value for the transient to be valid. Boost ratio (boost pressure/inlet pressure) is then computed at one or more intervals when the VxA is at or above the minimum, (i.e. during the transient). In one aspect the (average during the transient) boost ratio is compared to a minimum and if at or above minimum, the diagnostic from box 20 is validated i.e. the fault is deemed valid. Area of “gentle” driving conditions shown by arrow B. Cumulated VxA is still not high enough to allow the execution of the diagnostic to take a decision on boost ratio.

Plots d) of figures 6 shows the activation events of the diagnostic during the driving cycle and plot e) shows passes (indicated by spikes) where the further validation of any fault is not made (because the boost ratio is below the threshold); this can be considered the output of 37.. So plots d) show indication of passes and plot f) shows indications of fail. So spikes in plots d) indicate passes where the boost ratio are within the prescribed limits. In figure 6 there are no failed diagnostics even with several valid transients.

Figure 7 shows plots similar to figure 6 with a different example; it shows the situation where the diagnostic properly identified a failed part indicated by arrow C in figure 7f at the time when the diagnostic was activated during the driving cycle in plot 7d. Here the average boost ratio is above the threshold. The spike in plot f) indicates that the boost ratio is above prescribed limits and the preliminary indication of a fault is validated.

Methods according to the invention provide reliable identification of transient and steadystate operating conditions of the boost pressure control system in order to properly differentiate between overboost/underboost and slow responding boost failure modes. No need of idle for launching the diagnostic. Robustness when diagnosing a slow responding boost pressure control system even during low aggressiveness driving conditions.

Ihe methods can be implemented with non-intmsive execution. There is no need for of DFCO or engine idle conditions to launch the diagnostic. Proper detection of the two possible states of a boost pressure control system (transient and steady state) in order to differentiate between under-boost and boost slow response failures. The methodology determines e.g. boost ratio between desired and actual (measured) boost computed during validated transient state conditions as diagnostic parameter. Depending on the aggressiveness of the driving profile, the diagnostic will adapt itself to determine the number of valid transient states necessary for the diagnostic. The use of the VxA method (Velocity x Positive Acceleration) can be used, in real time, to ascertain the aggressiveness of the driving profile. Aspects of the invention are applicable to single or dual stage boost pressure control systems.

Claims

CLAIMS:

1. A method of determining a fault in the boost system of a turbocharged engine, comprising the steps of:

a) ascertaining a transient boost period;

b) determining an actual boost parameter during the transient;

c) determining an expected or ideal boost parameter during the transient; and,

d) comparing the parameters of steps) b) and c) to ascertain if a fault is present.

2. A method as claimed in claim 1 where step d) comprises determining the time taken for the actual boost parameter to reach the value of the ideal boost parameter at steady state.

3. A method as claimed wherein step d) compares the value of the actual boost parameter with the ideal boost parameter value at a time within the transient period.

4. A method as claimed step d) comprises determining integral values of the actual and expected boost parameters during a time period within the transient, and comparing these.

5. A method as claimed in any preceding claim where the expected boost parameter is determined from a system model.

6. A method as claimed in any preceding claim where the actual boost parameter is determined from a pressure sensor or from a system model.

7. A method as claimed in any preceding claim wherein said boost parameter is boost pressure.

8. A method as claimed in any preceding claim step a) includes determining if the requested torque demand or rate of change of torque demand exceeds a predetermined value.

9. A method as claimed in any preceding claim step a) includes determining if one or more of the following exceeds predetermined value(s): minimum IMEP, minimum enthalpy and minimum rate of change of the boost pressure or ratio during transient conditions.

10. A method as claimed in any preceding claim where step a) includes determining 5 if one or more operating conditions of the engine are present.

11. A method as claimed in any preceding claim where said transient period is defined as any time period where there is a positive boost demand.

10

12. A method as claimed in claims 1 to 11 including determining a measure of the boost ratio during said transient and comparing this with a threshold value, and if above said threshold, validating any fault from step d).

13. A method as claimed in claim 12 wherein said measure of boost ratio is the

15 average boost ratio during the transient.

14. A method as claimed in claims 12 or 13 including the initial step of determining if the accumulated value of velocity multiplied by positive acceleration (VxA) or Positive Kinetic Energy (PKE) increases at a rate above a threshold value and

20 implementing the step of claim 12 is this condition if fulfdled.

Intellectual

Property

Office

Application No: GB1700347.6