DE112016004758T5 - Method and device for determining the internal cylinder pressure of an internal combustion engine - Google Patents

Abstract

The invention relates to methods and devices for processing a signal generated by a sensor (7), which is set up to detect pressure fluctuations generated in a cylinder (61-66) of an internal combustion engine (1). The sensor is mounted outside the cylinder on the engine. The sensor signal is low pass filtered (201) forming at least a portion of an in-cylinder pressure curve, and the in-cylinder pressure curve is scaled using a model of compression in the cylinder, forming a scaled pressure curve of the at least a portion of the in-cylinder pressure curve ,

Description

Field of the invention

The present invention relates to a method and apparatus for determining the in-cylinder pressure of an internal combustion engine.

background

There is a continuing drive to achieve control of an internal combustion engine such that burned fuel therein is burned in the cylinders of the engine, producing a maximum amount of output from the engine and a minimum amount of emissions of environmentally harmful contaminants. In such a quest, it is important to have continuous knowledge of the operating conditions of the internal combustion engine. An important source of information is the pressure in the cylinder (s) of the internal combustion engine.

Further describes US 8,396,649 a method wherein some in-cylinder pressure data can be reconstructed using a vibration signal of a vibration sensor positioned outside the cylinder.

Summary of the invention

It is an object of the present invention to provide methods and apparatus which at least partially solve the above problems and which are improved, at least in some respects, with respect to prior art methods and apparatus.

This object is achieved with the method and the device as defined in the appended claims.

According to one embodiment, there is provided a method of processing a signal generated by a sensor configured to detect pressure variations generated in a cylinder of an internal combustion engine. The method includes low-pass filtering the signal of the sensor, wherein at least a portion of a cylinder internal pressure curve is formed. The in-cylinder pressure curve is then scaled using a model of compression in the cylinder, forming a scaled pressure curve of the at least part of the in-cylinder pressure curve. By first generating an in-cylinder pressure curve and then scaling the curve thus obtained, it is possible to generate an in-cylinder pressure curve which, with a small amount of processing, will produce the in-cylinder pressure curve both in shape and in terms of shape represents absolute pressure values.

The in-cylinder pressure curve thus formed may represent the entire in-cylinder pressure curve of a complete duty cycle of a cylinder or, in some applications, only a portion thereof. The duty cycle may be a duty cycle for a four-stroke engine or for a two-stroke engine.

According to some embodiments, the in-cylinder pressure curve formed is in-phase.

According to some embodiments, a part or parts of the formed in-cylinder pressure curve is replaced by pressure values determined from a model. The replaced part or parts may correspond to parts of the cylinder inner curve that have been determined to include noise above a predetermined threshold.

According to some embodiments, the formed in-cylinder curve is smoothed to produce a curve that can be derived at each point of the in-cylinder curve.

In some embodiments, the sensor detects signals on the long side of the engine. The sensor can detect signals on the long side of the engine at the lower temperature compared to the other long side of the engine. According to other embodiments, the sensor is disposed on the warm side of the engine. The signals may typically be a vibration signal or a displacement signal.

According to some embodiments, the compaction model is an adiabatic compaction model.

According to some embodiments, the sensor detects a vibration or deflection in the engine.

The invention also relates to an apparatus for carrying out the method described above and to a motor vehicle comprising such a device.

The invention also relates to a computer program, a computer program product, an electronic control device and a motor vehicle.

The invention is not limited to any particular type of internal combustion engine, but includes spark-ignited engines as well as pressure-ignited engines, and is not limited to any particular fuel. Not exhaustive Examples include fuel in the form of gasoline, ethanol, diesel and gas.

Likewise, the invention includes internal combustion engines intended for any type of use, such as in industrial applications, crushing machines and various types of motor vehicles, wheeled vehicles, and trucks and buses, and boats and caterpillars or similar vehicles.

Other advantageous features and advantages of the invention will be explained in the description below.

list of figures

• 1 eine schematische Ansicht ist, die einen Teil eines Verbrennungsmotors zeigt,
• 2 ein Ablaufdiagramm ist, das einige Schritte darstellt, die durchgeführt werden, wenn ein Sensorsignal verarbeitet wird,
• 3 ein Diagramm einer elektronischen Steuervorrichtung ist,
• 4 eine mögliche Anordnung eines Sensorelements zeigt,
• 5 ein Sensorsignal, einen Zylinderdruck und einen Wärmeausstoß für einen Arbeitszyklus darstellt,
• 6 ähnlich zu 5 ist mit gemittelten und gefilterten Signaldaten,
• 7 ein Betragsspektrum eines gemessenen Drucks und eines detektierten Signals darstellt,
• 8 und 9 eine typische adiabatische Verdichtungskurve darstellen,
• 10 ein gefiltertes und skaliertes Sensorsignal darstellt,
• 11 ein Signal darstellt, das unter Verwendung eines Vorverdichtungsmodells modelliert wurde,
• 12 eine geglättete Kurve darstellt,
• 13 ein Hinzufügen einer Phasenkompensation zu dem Signal darstellt,
• 14 die Verwendung eines Post-Power-Modells in der Signalverarbeitung darstellt, und
• 15 das endgültige Ergebnis unter Verwendung eines fortgeschrittenen Modells darstellt.

Below are descriptions of exemplary embodiments of the invention with reference to the accompanying drawings, in which:

• 1 is a schematic view showing a part of an internal combustion engine,
• 2 FIG. 3 is a flowchart illustrating some steps performed when a sensor signal is processed; FIG.
• 3 is a diagram of an electronic control device,
• 4 shows a possible arrangement of a sensor element,
• 5 represents a sensor signal, a cylinder pressure and a heat output for a work cycle,
• 6 similar to 5 is with averaged and filtered signal data,
• 7 represents an amount spectrum of a measured pressure and a detected signal,
• 8th and 9 represent a typical adiabatic compression curve,
• 10 represents a filtered and scaled sensor signal,
• 11 represents a signal modeled using a precompression model
• 12 represents a smoothed curve,
• 13 represents an addition of a phase compensation to the signal,
• 14 represents the use of a post-power model in signal processing, and
• 15 represents the final result using an advanced model.

Detailed description of embodiments

1 schematically shows an internal combustion engine 1 who is here in an implied automobile 2 is arranged, for example in a truck. The engine is with a device 3 equipped with a dashed line, and which is adapted to detect operating conditions of the engine, and such a device has a schematically drawn device 4 , which is adapted to pressure changes in the cylinder chambers 5 the cylinder 61 - 66 of the internal combustion engine, of which six are present in this case, but of which any number may be present.

The device 4 has a sensor element in this example 7 per cylinder 61 - 66 this being outside the associated cylinder chamber 5 is arranged. The sensor elements 5 are adapted to detect vibrations or deflections, and in particular are adapted to detect deflections and vibrations in the engine resulting from pressure fluctuations in the cylinders of the same. The terms vibration / displacement as used herein refer to any movement in the engine that originates from the sensor element 7 can be detected. The sensors may be piezoresistive or piezoelectric sensors or other types of vibration or deflection detecting sensor elements such as a strain gauge, for example a strain gauge.

The device 3 also includes a unit 9 coming from the electronic control device of the vehicle 2 may exist, which is adapted to information about the detected movements of the / the sensor element / sensor elements 7 and to compare such information or information calculated based on such sensor information with stored values and, in turn, measurements of the condition of the engine 1 and / or to provide operations in the engine. Therefore, based on the detection of the sensor elements 7 Information about the operating conditions of the engine or deviations thereof are obtained, which form the basis for the control of different components in the engine expediently.

As was recognized, pressure fluctuations in the cylinder chamber 5 be detected by a sensor outside the cylinder chamber to produce high quality signals which require simple filtering or processing to form a cylinder internal pressure curve or a portion thereof.

2 FIG. 10 is a flowchart illustrating one embodiment of a method of processing a signal generated by a sensor configured to detect vibrations or deflections generated in a cylinder of an internal combustion engine. The sensor is attached to the engine outside the cylinder. First, in one step 201 the signal from the sensor is low-pass filtered to form at least a portion of an in-cylinder pressure curve. Then in one step 203 scales the in-cylinder pressure curve using a model of compression in the cylinder, forming a scaled pressure curve of the at least a portion of the in-cylinder pressure curve. According to one embodiment, that is in step 203 Model used an adiabatic compression model. In addition, additional processing and modeling steps may be performed, as by step 205 is indicated. The order in which the steps 201 - 205 may be different in different processing implementations to process the vibration signal to form an in-cylinder pressure curve or a portion thereof. Processing and modeling steps that may be performed are described in greater detail below.

A computer program code for implementing a method according to the invention is expediently contained in a computer program which can be loaded into an internal memory of a computer, such as an internal memory of an electronic control device of an internal combustion engine. Such a computer program is conveniently provided via a computer program product comprising a non-volatile data storage medium readable by an electronic control device, the electronic data storage medium carrying the computer program stored thereon. The data storage medium is, for example, an optical data storage medium in the form of a CD-ROM, a DVD, etc., a magnetic data storage medium in the form of a hard disk, a floppy disk, a cartridge, etc., or a flash memory or a ROM, PROM memory , EPROM or EEPROM type.

3 schematically illustrates an electronic control device 9 the execution means 17 for executing computer software, such as a central processing unit (CPU). The execution means 17 communicate with a memory 18 , for example a RAM memory, via a data bus 19 , The control device 9 also includes a data storage medium 20 For example, in the form of a flash memory or a memory of the ROM, PROM, EPROM or EEPROM type. The execution means 17 communicate with the data storage means 20 over the data bus 19 , A computer program comprising computer program code for implementing a method according to the invention is on the data storage medium 20 saved.

4 shows a possible sensor arrangement. The sensor element 7 is located here on the engine. In particular, the sensor element can be arranged on a section of the cylinder head. The sensor elements / sensors 7 may be of a suitable type, for example piezoresistive or piezoelectric sensor elements or optical sensors. The sensor elements may be disposed on the engine in an area adjacent to the outlet of the exhaust passage from a cylinder. For example, it may be disposed on a surface of the engine adjacent to the outlet, on the engine, of the exhaust passage of a cylinder. The surface on which the sensor 7 can be arranged substantially vertically. The sensor may be configured to detect vibrations and deflections that are perpendicular to the movements of the piston. The sensor may also be adapted to detect vibrations or deflections which are perpendicular both relative to the direction of movement of the piston and relative to the longitudinal direction of the motor. In one embodiment, the sensor is located on the long side of the engine. The sensor may be configured to detect vibrations or deflections in a direction perpendicular to the surface on which it is disposed.

In another embodiment (not shown), the sensor element 7 in a corresponding manner, as when it is disposed on the engine at the outlet of the exhaust passage of a cylinder, but instead at a corresponding point on the engine at the inlet of the intake passage to a cylinder.

That of the sensor element 7 detected signal can be handled in various ways, as described below. The signal from the sensor detects vibrations that are low-pass filtered to produce an in-cylinder pressure curve or at least a portion thereof. The in-cylinder pressure curve may typically be a continuous curve. The pressure curve thus formed may be used to calculate different values in engine control. To increase the accuracy of the in-cylinder pressure curve, the in-cylinder pressure curve may be further processed in one or more modeling steps and refinement steps.

Below, some of such modeling steps and refinement steps are described by detailed embodiments. Of the The invention is in no way limited to the embodiments described, but numerous possible modifications thereof are conceivable. In particular, steps may be omitted or steps of different embodiments combined or performed in a different order than those described.

Signal content identification

An exemplary, recurring appearance is in 5 to recognize where the sensor signal (knock signal) is plotted along with the cylinder pressure. The data in this example is based on 1200 rpm operation at 100% load. Of course, other settings are possible. The sensor data may be averaged according to some embodiments. For example, the data may be averaged over ten cycles and Savitzky-Golay-smoothed (optimized at polynomial degree 2 and frame size 111 ), the result of the operation in 6 is shown.

The sensor signal can be compared for different operating points to verify that the same appearance is evident in all modes.

model development

The act of obtaining an estimated pressure model may include at least one of filtering out high frequency noise, adjusting phase shifts, scaling, and replacing noisy parts of the signal with known physical models and assumptions. The model may be based on a Crank Angle Degree CAD (Crank Angle Degree) signal, for example, 0.1 degrees. Other resolutions can be used, such as a 6 degree resolution. Values lying between the recorded sample data can be modeled. The modeled values can be generated, for example, by a virtual sensor.

To address robustness, different models can be used. The models can be combined. Here are two models described. A first model with light signal processing to minimize model dependencies and focus on achieving a small average deviation of the maximum pressure amplitude from the measurement data. A second, more advanced model will involve more extensive data processing, including phase alignment, power stroke modeling, high engine load dependencies, and a stronger emphasis on achieving complete pressure signal correlation.

compaction model

A compression model can be used for scaling purposes. According to some embodiments, the compression model is based on the ideal adiabatic equations for compressing a gas to compensate for engine / load changes as well as possibly non-linear heat transfer loss. For example, a heat transfer model based on the Woschni / Hohenberger model can be used. The heat transfer model can, for example, with a coefficient in the range 1 - 10 , such as in the field 1 - 5 , can be multiplied to compensate for all modeled losses, and the wall temperature coefficient can be adjusted depending on engine speed and engine load to get a proper exponential rise during the compression stroke. These variables can be tested during tolerance analysis to investigate the impact these decisions have on the model.

According to one embodiment, each step of the process is iteratively calculated based on the first law of thermodynamics, first calculating the number of moles using the intake valve pressure and the intake valve temperature as starting values for the model and the cylinder volume calculation at CAD-180 degrees, ie, bottom dead center (BDC) before top dead center for combustion. A first estimate of the index, in particular an adiabatic index, is then calculated using an index function. In case the model is adiabatic, the index will be an adiabatic index. The main dependence of this function may be the current temperature and lambda. The heat loss of the heat transfer is calculated, which then provides the information needed to calculate the derivative of the pressure, the heat transfer being used primarily to compensate for the heat loss / heat gain due to the temperature in the cylinder wall. The temperature, the (adiabatic) index and the cylinder pressure can be up to CAD 160 iteratively calculated to obtain the complete cycle.

Below, two models (simple and advanced) for a typical test implementation are described in detail.

Print model - easy

Filter

The signal-related signal content of the signal related to the pressure Vibration sensor has a relatively low frequency content compared to the entire frequency spectrum; The high-frequency noise is reduced by applying a digital low-pass filter to the signal. The magnitude spectrum of the relevant frequency range is in 7 to detect which the detected signal (knock signal) together with a measured cylinder pressure (pressure reference) represents.

Different realtime filters like Butterworth, Elliptic, Chebyl and Cheby2 can be used. Typically, there will be a trade-off between filtering too much information near the beginning of the combustion and obtaining a comparatively significant peak (depending on the operating point). According to some embodiments of the invention, minimizing roll-off effects may be performed by real-time filters with a Fourier transform filter. In some embodiments, a tenth order Butterworth filter with a normalized cutoff frequency of about 0.04 may be used. Phase shifting may be avoided in some embodiments by applying a zero phase filter (forward and backward filtering) so that the signal is then filtered, for example, with two successive 10th order butterworth filters.

scaling

The obtained sensor signal can be received at any amplitude. According to some embodiments, data of the order of magnitude 1 - 10 V, but this could also be done in other ranges of amplitude, depending on the amplifier and settings used, as well as engine speed and load. A pressure model, particularly an adiabatic pressure model, can be used to adjust the scaling to the correct level, regardless of the amplifier, the settings, and the current operating point.

Proper scaling may be achieved by minimizing the average difference between the filtered sensor signal and the pressure model, in accordance with some embodiments. For example, the minimization can be done between two CADs with negative values, for example -28 CAD and -3 CAD, that is, H. during the compression phase. Here, all operating points have approximately the same appearance (but not necessarily the same amplitude, so individual scaling factors are used for each operating point).

Vorverdichtungsmodell

The early phase of the compression stroke is overloaded and distorted by various noise; To reduce the amount of filtering needed, and since the beginning of the compression stroke can be idealized as a compression, the model can be replaced by some negative values, such as starting in the range -150 to -110 CAD and ending in the range of -30 to -5 CAD. For example, -130 CAD to -15 CAD can be used with the above scaling model. This is done on the assumption that the state of the signal in this region coincides with the model used.

Intake emissions model

The opening of the inlet valve will level the cylinder pressure at approximately the main line pressure so that the inlet stroke of the signal may be replaced by an average value of the main line pressure for the particular operating point to minimize signal filtering and smoothing; in order to reduce the number of required parameters, the exhaust stroke must be set to the same level. For example, based on measurements from an experimental series, it is true that the average inlet main line pressure and discharge main line pressure are approximately identical, but with available sensors this assumption can be replaced by the actual discharge main line pressure, if necessary. The replaced inlet area is between some negative CAD values, for example, -360 CAD to -130 CAD may be used, and the replaced ejection area is between some positive CAD values, for example, 190 CAD to 360 CAD may be used.

smoothing

The sensor will typically consistently detect slightly higher peak values than the measured pressure curve. To obtain a smoother tip and to minimize discontinuities between the replaced precompression model and the inlet / ejection model, the model can be smoothed. This may be done in accordance with some embodiments by using a filter, for example a Savitzky-Golay first order smoothing filter (for all operating points). Maintaining the content surrounding Start of Combustion (SOC) is critical so that a second signal with a smaller frame size can be smoothed. A smooth transition can be obtained in some embodiments by iteratively comparing the two smoothed signals and choosing the lower value of them from the range around 0 CAD, for example between -5 CAD and +5 CAD, or from another area around 0 CAD.

Printed model - advanced

phase alignment

The low-pass filtered signal may be phase shifted prior to scaling to raise the clock rate of the signal to the correct level. According to some embodiments, the phase-shift model used may be a delay equalizer, i. H. a delay equalizer of second order. In an exemplary embodiment, the position of the pole is p = 1e-12 for all operating points and the position of zero is varied between z = 4.25e-04 and z = 3e-04 or some other suitable values with increasing speed and cylinder used. The value can be optimized by comparing a measured pressure curve and the phase shift model.

The phase delay equalizer will typically only minimize a certain amount of phase differences between the signal and the reference pressure signal. In some examples, minimizing further phase differences may be performed using higher orders of models. However, a simple second-order model typically works with acceptable results for all operating points without overly adjusting its parameters.

scaling

The filtered sensor signal can be scaled with the same adiabatic compression model used for the above simple model.

Post-Power model

The phase alignment typically does not correct all irregularities in the power stroke, so a region can be replaced while opening the valve with a similar model that was used to scale the signal. The range can be around 80 CAD, such as between CAD 20 and CAD 135 , The required initial temperature can be obtained by the first law of thermodynamics with a total amount of moles obtained from the estimated value in the scaling model, the initial pressure, and the initial volume, which is at about CAD 20 or any CAD value in this area, such as CAD 10 - 30 , Instead of balancing heat transfer losses with a model, a simple measurement based on displacement can be used; the displacement can be based on a sensor that measures the amount of fuel injected and can be optimized by comparing the injected amount with the actual relative load (from measurements). The displacement for 75% relative load and above is 0.10, 0.07 for 50%, 0.03 for 25%, and 0 for 0% relative load and motorized cycles, in accordance with some embodiments. A linear interpolation can be between the last value, for example CAD 135 , and the replaced main line pressure value, for example CAD 200 ,

Smooth

The signal can be smoothed to accommodate scaling issues. This will remove irregularities in the transition between models and assumptions. For example, a first Savitzky-Golay smoothing filter with a frame size of 111 may be used.

Combined model

A combined model can be developed to get a better fit in the working stroke and around the pressure peak. For example, the developed simple printing model can be used as the basis, with the advanced printing model replacing the simple printing model between some positive CAD values. The range can start at about 0-5 CAD and end at about 180-200 CAD. For example, from CAD 2.5 to CAD 195 , Here, a smooth transition is obtained by choosing the higher value between the two models in the above-mentioned range.

Smooth

The signal can be smoothed one last time to remove any irregularities in the transition between the models. In one embodiment, based on measurements, a Savitzky-Golay smoothing filter with a frame size of 35 for non-motorized cycles and a frame size of 51 for motorized cycles may be used. The motorized cycles can be displayed by the available sensor, which measures injected fuel.

test results

Adiabatic pressure model

The end result of a typical adiabatic pressure curve is in 8th and 9 to recognize, where 9 a detail of 8th shows. In 8th and 9 becomes the model curve (adiabatic model) compared with the measured pressure (reference pressure).

Advantageously, the adiabatic scaling model can be calculated individually for each operating point prior to scaling to obtain correct pressure levels from the current sensor readings.

Filter

The filtered signal is in 10 to recognize. The high frequency noise from the sensor signal and the remaining variations are related to lower frequency measurements. 10 shows the filtered and scaled sensor signal (filtered and scaled). The signal in 10 is also scaled using the scaling model (here an adiabatic scaling model).

Vorverdichtungsmodell

It can be clearly seen that the result of the replaced precompression content smoothes the curve during the beginning of compaction, as can be seen in Fig. 11 for the pressure model (precompression model).

Intake emissions model

The curve in 11 is the replaced inlet and outlet contents, which closely follows the measured pressure; the curve (intake-exhaust model) shows the noise decomposition.

smoothing

The smoothed curve and the end result of the simple pressure model are in 12 which shows the model (pressure - simple model) compared to the cylinder pressure (pressure reference). From the beginning of the compression cycle to near the pressure peak, the model precisely follows the measured pressure.

Advanced printing model

phase alignment

The result of adding a phase alignment to the signal is in 13 which shows the model (pressure model in-phase and scaled) in comparison with the cylinder pressure (pressure reference). The signal now has a better correlation with the measured cylinder pressure during the power stroke and the exhaust stroke.

scaling

The model is scaled in the same way using the same adiabatic scaling model as the simple model. The results are in 13 to recognize.

Post-Power model

The replaced curve is less noisy and much smoother. The pressure peak may be increased for most operating points, in accordance with some embodiments. Smoothing the curve with a larger frame will cause the peak amplitude gain to be lowered slightly, leading to a better correlation with the measured pressure trace; the result is in 14 which shows a smoothed curve for the model curve (pressure model postpower and smoothed) compared to the cylinder pressure (pressure reference).

Advanced model

A final smoothing can be done to remove any discontinuities in the advanced model. Thus, motorized cycles can be smoothed. According to some embodiments, a first-order Savitzky-Golay smoothing filter with a frame size of 51 may be used, and non-motorized cycles use a frame size of 25. The end result is in 15 which shows a resulting model curve (pressure-advanced model) compared to the cylinder pressure (pressure reference).

The processing of a vibration signal from a sensor outside a cylinder of an internal combustion engine to produce an in-cylinder pressure curve is performed using a model. The overall model may involve several steps, but is nonetheless simple. Filtering can be done with simple butter-worth filters and can be performed using only a single cutoff frequency for all the operating points, fuels and cylinders tested. The inlet-exhaust model shows minimal differences between the measured pressure at the inlet main line pressure for the measurements compared herein. According to some embodiments, individual ejection sensor readings may be added. The smoothing steps are primarily intended to minimize discontinuities between the different models that are applied and performed with comparatively few frames. The smoothing performed with the larger frame is mainly done to minimize the peaks of certain pressure spikes. This can be done to the counteract that the sensor signal detects higher peak values than those measured with the pressure sensor. The phase alignment provides an increase in pressure relationship for all operating points at various degrees.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• US 8396649 [0003]

Claims (14)

A method of processing a signal generated by a sensor (7) adapted to detect pressure variations generated in a cylinder (61-66) of an internal combustion engine (1), the sensor being mounted outside the cylinder on the internal combustion engine, characterized by Lowpass filtering (201) the signal of the sensor forming at least part of an in-cylinder pressure curve; and scaling (203) the in-cylinder pressure curve using a model of compression in the cylinder, wherein a scaled pressure curve of at least a portion of the in-cylinder pressure curve is formed. Method according to Claim 1 wherein the entire in-cylinder pressure curve of a complete duty cycle of a cylinder is formed. Method according to one of Claims 1 or 2 , wherein the formed in-cylinder pressure curve is in phase. Method according to one of Claims 1 to 3 wherein a part or parts of the formed in-cylinder pressure curve are replaced by pressure values determined from a model. Method according to Claim 4 wherein the part or parts that are replaced correspond to parts of the cylinder inner curve that have been determined to include noise above a predetermined threshold. Method according to one of Claims 4 to 5 wherein the formed in-cylinder curve is smoothed to produce a curve that can be derived at each point of the in-cylinder curve. Method according to one of Claims 1 to 6 wherein the sensor detects signals on the long side of the engine. Method according to one of Claims 1 to 7 , where the compaction model is an adiabatic compaction model. Method according to one of Claims 1 to 8th wherein the sensor detects a vibration or deflection in the engine. Apparatus (3) for processing a signal generated by a sensor adapted to detect pressure variations generated in a cylinder of an internal combustion engine, the sensor being mounted outside the cylinder on the internal combustion engine, characterized by : - a low-pass filter configured to lowpass-filter the signal from the sensor and configured to output at least a portion of an in-cylinder pressure curve, and a scaling module configured to increase the in-cylinder pressure curve using a model of compression in the cylinder scale, and which is adapted to form a scaled pressure curve of the at least part of the in-cylinder pressure curve. Internal combustion engine (1), characterized in that it comprises a device according to Claim 10 includes. A computer program that may be downloaded into an internal memory of a computer, the computer program including computer program code configured to cause the computer to follow the steps of one of Claims 1 to 9 to control when the computer program is running in the computer. A computer program product comprising a non-transitory data storage medium readable by a computer and the computer program code according to Claim 12 bears on it. Motor vehicle (2), characterized in that it has an internal combustion engine (1) according to Claim 11 includes.

Images