JP2008520876A - System and method for processing accelerometer signals to assist in flammability control of an internal combustion engine - Google Patents

Abstract

A method for processing accelerometer signals associated with a combustion process in an internal combustion engine in operation. The method measures an accelerometer signal including a combustion acceleration component and a passive acceleration component across a selected crank angle window during one cycle of the engine, and reduces the passive acceleration component. To apply a shape function to the measured accelerometer signal.
[Selection] Figure 1

Description

[Cross-reference with related applications]
This application is related to and claims the priority benefit of US Provisional Application No. 60 / 629,489, filed November 18, 2004, entitled “Accelerometer-Based Combustion Sensor”. This' 489 application is hereby fully incorporated by reference.

  This application is a continuation-in-part of US patent application Ser. No. 10 / 822,333 filed Apr. 12, 2004, entitled “Method and Apparatus for Controlling Internal Combustion Engines Using Accelerometers”. It is. This' 333 application is related to US Provisional Application No. 60 / 483,855, filed June 30, 2003, entitled “Method and Apparatus for Controlling Internal Combustion Engines Using Accelerometers”. And insist on its priority interests. These '333 and' 855 applications are hereby fully incorporated by reference.

[Field of the Invention]
The present invention relates to a method and system for determining the flammability of an internal combustion engine, and more particularly to using an accelerometer in combination with a digital signal processing (DSP) technique for flammability determination. In this case, the flammability indicator is not limited, but (i) the combustion timing, eg, the start of combustion (SOC) indicator, (ii) the combustion rate, (iii) non-ignition or partial combustion, (Iv) including a cylinder pressure increase (knock) that is too fast.

"Background Technology"
Conventional internal combustion engines, such as diesel or spark ignition engines, require control of the SOC for both flammability characteristics, such as efficiency and exhaust control. For example, a diesel engine controls the start of combustion by the timing of fuel injection, while a spark ignition engine controls the start of combustion by the spark timing.

  However, additional combustion control may be desired for normal engines and may be required for other engines. One example of an engine that would benefit greatly from a feedback flammability sensor is one that uses HCCI combustion. Unlike traditional SI or diesel engines, HCCI combustion occurs naturally and generally homogeneously without flame propagation. HCCI combustion is a compression ignition of a relatively well premixed fuel / air mixture. Various combustion strategies based on the HCCI principle have been developed. For example, the combination of HCCI engines and traditional engine injection techniques leads to premixed compression ignition (PCCI) where the fuel / air mixture is premixed but need not be homogeneous. An additional combustion strategy is a strategy that replenishes the energy provided by the PCCI combustion event with a directly injected amount of fuel that is generally provided once combustion is initiated. This type of engine is known as a premixed direct injection (PCDI) engine.

  One problem with HCCI engines is that flammability is susceptible to many parameters, particularly intake manifold temperature, fuel / air ratio, fuel quality, trapped residual gas components and exhaust gas recirculation. Unless parameters that strongly affect flammability are controlled, flammability will encounter large cycle-to-cycle changes. Thus, when compared to conventional diesel and SI engines, misfires and pressure increases at a rate that is too fast are prone to occur. Knowing the start of combustion, that is, the time of combustion, sometimes referred to as SOC or occasionally, helps to provide a control strategy that adjusts the flammability of future engine cycles and enables improved engine performance To do. Thus, an HCCI engine benefits from an accurate and robust determination of SOC and other flammability indicators.

  In addition to the HCCI examples given above, flammability findings, such as the onset and rate of combustion, may be obtained if there is no reason other than monitoring system performance for real-time diagnostic purposes. It is becoming commercially useful for engines. For such engines, adjustments are made to compensate for the measured changes in flammability and maintain high engine fuel savings and low engine external emissions.

  One known approach to assessing engine cycle flammability relies on direct pressure measurement in the combustion chamber, for example, by placing sensors to measure the deflection of the diaphragm in contact with the pressure in the cylinder. It is. The measured pressure signal is correlated to the SOC or other flammability indicator. For example, the feedback control loop is used to adjust engine parameters and affects the SOC of future engine cycles by minimizing the error between the measured SOC and the target SOC. See U.S. Patent No. 6,598,468 and German Patent No. 4,341,796.5 which describe a direct measurement of the signal indicative of pressure in the combustion chamber. Typically, an optical sensor or other direct pressure measuring device is used. Appropriately accurate indicators of flammability, such as sensors, are expensive and / or currently lack the reliability and robustness (due to the harsh environment in the combustion chamber) necessary for many applications. .

  Another technique for evaluating combustion chamber pressure uses a knock sensor (accelerometer) as described in US Pat. No. 6,408,819. Although accelerometers are less expensive and are currently more reliable and robust than direct pressure measurement sensors, this approach is sufficient for many flammability control methods, such as effective SOC control. However, it has the disadvantage of relying on inaccurate pressure signal reproduction methods.

An effective system using an accelerometer is the applicant's co-pending "Method and apparatus for controlling an internal combustion engine using an accelerometer" filed on April 12, 2004. US patent application Ser. No. 10 / 822,333 (“'333”) and is hereby fully incorporated by reference. Rather than reproducing pressure signals from accelerometer data, the '333 application describes a heat release rate reproduction method (HRR) for extracting combustion information from raw accelerometer data.
Research has been done on whether the effectiveness of the above-described method can be improved.

[Summary of Invention]
In light of the foregoing, the presently described technique seeks to provide an improved method and system for determining the flammability, eg, SOC, of an internal combustion engine. In this case, the novel method and system can be used more effectively in various engine applications such as combustion control or combustion diagnostics.
This can be achieved by additional preprocessing of the raw accelerometer data.

One advantage of the presently described technology is that the pre-processing improves the quality of the data introduced into the heat release rate reproduction method, resulting in a more robust, reliable and low cost combustion sensor.
Another advantage of the presently described technology is that it gives a real-time assessment of flammability and SOC.

  In order to enable effective determination of SOC using an accelerometer sensor, Applicants have recreated data relating to combustion, raw heat release rate, and the data from raw accelerometer data as well as SOC. We found that there is an advantage in separating prior to correlating to. This technique is useful. This is because an accelerometer mounted in the proper position of the engine to pick up vibrations related to the combustion process contains data resulting from the motion of the engine components as well as the combustion process.

  Accordingly, the presently described technique provides a method and apparatus for processing accelerometer data and extracting data related to combustion.

According to the presently described technique, a method is provided for processing accelerometer signals associated with combustion processes in an operating internal combustion engine. This method
a. Measuring an accelerometer signal including a combustion acceleration component and a passive acceleration component across a selected crank angle window during a cycle of the engine;
b. Applying a shape function to the measured accelerometer signal to reduce the passive acceleration component;
With.

In another embodiment, the shape function is evaluated by operating the engine in a passive mode.
In another embodiment, the swivel method is applied to the shape function before it is applied to the measured accelerometer signal.
In another embodiment, a reflection method is applied to the difference between the shape function and the measured accelerometer signal.
In another embodiment, a differentiation method is applied to the difference between the shape and the measured accelerometer signal.

  According to another embodiment of the presently described technology, a system for controlling the combustion of a compression ignition engine is provided. In this case, the accelerometer signal is input to a DSP chip based microcontroller, and method steps are performed on the DSP controller to generate combustion information.

  The foregoing and other objects, features and advantages of the presently described technology will become apparent from the following specific description of a preferred embodiment of the presently described technology, as illustrated in the accompanying drawings. In the drawings, like reference characters refer to the same parts throughout the different views. The drawings do not have to be proportional; instead, emphasis is placed on illustrating the principles of the presently described technology

Detailed Description of Preferred Embodiments
The presently described technique is directed to a method and apparatus for determining combustibility in an internal combustion engine using accelerometer data. In general, for the purposes of this disclosure, the disclosed methods can be used to measure flammability, eg, misfire, rate of combustion, peak cylinder pressure, timing of combustion (eg, SOC). One benefit of measuring flammability is that it provides a measure of engine performance, allowing the ability to tune engine behavior in real time, improving efficiency while allowing emissions to be reduced. It should be noted that the apparatus and method according to the presently described technology disclosed below can be used in a variety of different internal combustion engines such as, but not limited to, diesel engines, spark ignition gasoline engines, It can be applied to alternative fuel engines and their variants operating with a modified thermal cycle. The presently described techniques are used in many such engines to promote improved engine efficiency, reduce emissions, or to other applications such as combustion diagnostics by determining and monitoring flammability. used.

  More specifically, the presently described technique relates to an improvement in an accelerometer-based method for determining flammability, eg, SOC, previously described in the '333 application. Embodiments of the presently described technology relate to improved processing of raw accelerometer data to improve accuracy and robustness, as described below.

  Furthermore, while most of this discussion refers to methods directed to monitoring or determining SOC, accelerometer data and the resulting heat release traces resulting from such data are: As will be appreciated by those skilled in the art included herein, flammability forms such as, but not limited to, peak cylinder pressure, location of peak cylinder pressure, rate of combustion, combustion timing, misfire, And can be applied to methods for controlling or assessing premature combustion.

  Referring to FIG. 1, an engine 100 is shown to which an embodiment of the presently described technology can be applied. The engine 100 includes a cylinder block (only one cylinder is shown for simplicity, but the engine typically has two or more cylinders). The cylinder block has a cylinder 14 and a reciprocating piston 16 disposed therein for driving a crankshaft 18. Various intake and exhaust valves (not shown) communicate with the cylinder to deliver air-fuel mixture to the cylinder and exhaust by-products. This is because the engine can also have fuel and air delivery systems and other standard systems associated with internal combustion engines.

  The engine 100 includes an accelerometer 22 that is preferably mounted on the main bearing cap 19 of the engine 100. This accelerometer outputs an electrical signal related to the deflection of the bearing cap due to the pressure on the piston 16. Referring to FIG. 1, an exemplary flammability processing system 12 is also shown in accordance with an embodiment of the presently described technology for processing accelerometer 22 signals as described below. The processing system 12 includes a signal conditioning block 13 and a digital signal processing (DSP) based microcontroller 20 having an input channel for processing data from the accelerometer 22. It should be understood that the SOC refers to the time or position where combustion begins in the cylinder. In general, the SOC is referenced with respect to the angular position of the crankshaft of the engine. As a result, in the disclosed embodiment, engine 100 is provided with an engine position sensor (not shown) and an engine speed sensor (not shown), which provide the angular position and rotation of engine crankshaft 18. Monitor speed. The SOC is referenced with respect to crankshaft position. However, in other embodiments of the presently described technology, the SOC may also be another parameter, such as time, or both time and crankshaft position, or other parameters and various other sensors and options. Reference is made to an internal pressure transducer sensor (not shown). In the preferred embodiment, the DSP controller is capable of sampling each channel with a minimum sampling frequency of 20 kHz and processing a single channel in 7.5 ms. Signals 24 and 26 from the crank and camshaft sensors are provided to the DSP 20 to enable crank synchronous sampling of the signal from the accelerometer 22. Although not a necessary part of the presently described technology, a CAN (Controller Area Network) interface 28 to the CAN bus 29 is shown. A main engine controller (not shown) receives signals from the DSP via the CAN bus and implements a PI controller, as known in the art and as described, for example, in the '333 application. In addition, various parameters of the engine are controlled. It should be noted that the DSP processing system 12 is also integrated into the main engine controller, thereby eliminating the need to send results through the CAN.

In operation, the DSP-based flammability processing system 12 begins analog / digital (AD) sampling of the accelerometer sensor signal based on a user-defined crank degree θ at a sampling frequency f _SR . When sampling is complete, the controller implements a signal processing algorithm. Typically, the calculation time is less than 10 ms / cycle per cylinder. The calculation result is sent to the main controller via the CAN bus. The main controller can then use information for diagnostic purposes, information for closed loop control with respect to the individual cylinder SOC, or many other purposes that appear to be generally related to flammability.

  In the following description, the term passive mode simply refers to an engine cycle in which no fuel is injected and no combustion occurs.

  Referring to FIG. 2A, there is shown a flowchart 400 of the general steps of an improved method according to one embodiment of the presently described technique for processing accelerometer 22 signals by a DSP 20 for flammability control. Yes. FIG. 2B graphically shows signals corresponding to each process. The method includes obtaining 402 sensor data, pre-processing 403 the data, and applying a septal smoothing and transfer function to the pre-processed sensor data to calculate the HRR trace signal. And 404 for extracting combustibility information from the HRR trace signal. The preprocessing step 403 is the focus of this disclosure.

  FIG. 3 shows a plot of a typical raw accelerometer signal 300 measured at step 402 by an accelerometer 22 mounted on the main bearing cap of engine 100. The heat release rate plot 310 is assumed to have occurred during the engine cycle. As shown, the accelerometer signal 300 is set over a crank angle window between -50 dATDC and +60 dATDC. This crank angle window focuses on the combustion event timing for one cycle of the engine. The raw accelerometer signal 300 is considered to have two main components. First, there is a ground signal 302 related to the acceleration caused by the passive pressure change in the cylinder, which is called the passive acceleration component. Second, there is an acceleration related to the additional pressure produced on the piston by the combustion signal 304, which is called the combustion acceleration component, from which information such as the heat release rate 310 can be extracted. Other components are also found in the data, but the passive acceleration component and the combustion acceleration component are believed to dominate the signal within the selected sample range. In general, combustion is completed at some point after TDC. From this point on, combustion does not contribute to a pressure increase higher than the pressure associated with piston motion during the power stroke. The raw accelerometer 300 and the passive acceleration component 302 converge. In this case, the decay rate of the raw accelerometer signal 300 is slightly higher than the decay rate of the passive acceleration component 302. Subtracting the passive acceleration component from the raw accelerometer signal results in a signal called the difference signal 304. This difference signal 304 is related to the acceleration due to the combustion acceleration component. Note that the substantially flat differential acceleration component 304 response after combustion 306 is nearly complete. The information contained in the accelerometer signal 300 prior to the combustion end point 306 (shown in this embodiment at +20 dATDC) is a measure of the influence of the total combustion acceleration component. Information regarding the rate of combustion after the end of combustion 306 is limited. Prior to applying the septal smoothing and reproduction technique described and referenced in the '333 application, a processing method has been developed that removes much of the non-burning rate information from the raw accelerometer signal 300. There are benefits to using the pretreatment solution described above. First, the combustion acceleration component to noise ratio is significantly improved. The second byproduct is that the application of preprocessing techniques eliminates the need to process large sets of data samples (from well before TDC to well after TDC), reducing the effects of non-zero edges. It is. Such a reduced set reduces the computation time.

  The presently described embodiment of the presently described technique illustrates a pre-processing method for removing substrate trends in data associated with piston motion. The pre-processing method of the technology currently being described defines a shape function that emulates the passive acceleration component 302 shown in FIG. This shape function generally matches the accelerometer signal 300 when the engine is in passive mode. Resulting raw accelerometer signal 300 in an accelerometer data set with a matched shape function indicating a passive acceleration component and a dominant combustion acceleration component for a given cylinder, ie, the underlying piston motion signal, many of the passive acceleration components The difference 304 is removed.

  The passive acceleration component is acquired during a calibration process for a given engine to determine an appropriate shape function. This data is determined from one cylinder at different engine operating conditions. Different engine operating conditions include different engine rotational speeds, loads, or cylinder pressures. The phase of the passive acceleration component can be shifted by a number of factors depending on several factors such as cylinder pressure, engine speed and engine load. For most engine operating conditions, one shape function is preferably used per cylinder.

  In one embodiment, it is desirable to adaptively adjust the shape function to correctly match the passive acceleration component to most engine operating conditions. The method of adaptively adjusting the timing of the shape function or predicting what shape function should be used for a given raw accelerometer signal 300 is applied as described below. obtain.

  In another embodiment, the shape function is determined dynamically or on a real-time basis. For example, the engine is periodically misfired by the operator and a passive measurement component is taken. This is typically done by delaying combustion in a controlled manner and determining during the compression stroke where the peak of the measured motor function is. This can be done in a way that is easy for the operator to understand.

  In another different embodiment, determining the shape function on a real-time basis is used in combination with an adaptive adjustment of the shape function timing.

  Referring to FIG. 4A, a first such technique for preprocessing the accelerometer signal is demonstrated and is described herein as a turning technique. It is assumed that both the measured raw acceleration signal 300 data and the passive acceleration component 302 are determined as described above. Signal preprocessing occurs at step 403 in FIG. First, the shape function is preferably matched to the raw signal before it is subtracted from the raw accelerometer signal 300. The accelerometer 22 generates a data set that spans a range of piston motion from about -40 dATDC to TDC, where almost all of the data originates from piston motion alone, i.e., generally occurs prior to combustion in the combustion chamber. Thus, the crank angle when the shape function is matched to the raw signal is made within the compression stroke, more specifically between about -40 dATDC and TDC. This matching is performed as follows.

  i. The raw accelerometer signal representing the passive acceleration component and the shape function are stabilized to zero at a fixed crank degree (eg, -30 dATDC).

  ii. The raw accelerometer signal data and shape function slope during the compression process are calculated (eg, between -25 and -15 dATDC). The zero cross point of the raw accelerometer and the zero cross point of the shape function are then calculated. These zero cross points are calculated by extrapolating from one point within the range of crank angles for which the respective slopes were calculated. For example, if the slope is calculated between −25 and −15 dATDC, the line 312 starting from −20 dATDC is extrapolated to cross the zero line.

  iii. The shape function is phase shifted so that the raw accelerometer signal data matches the zero cross point of the shape function. Regarding the error strength, the phase shift of the shape function is limited. Typical limits of ± 2 or ± 3 crank degrees are used.

  vi. The raw accelerometer signal data and shape function are again stabilized to zero at a fixed crank degree (eg, -30 dATDC).

  v. Gain is applied so that the shape function matches the magnitude of the raw accelerometer signal data at a fixed angle before TDC (eg, -10 dATDC).

  The difference between the raw accelerometer signal data and the matched shape function is then calculated. An example of the results from shape function matching is demonstrated in FIG. 4A. The result shows matching during the compression stroke. The resulting leading edge of the difference signal 304 contains information related to the combustion rate, while the trailing edge of the difference signal 304 contains mostly information related to piston movement. The processing technique disclosed here extracts information related to the combustion rate from this difference signal. The swivel method is used to cross the shape function with the raw accelerometer signal at 40 dATDC, indicated by reference numeral 314 in FIG. 4A. Interpolation is used to maintain a smooth shape function. In this case, a turning point 316 on the shape function 302 close to TDC is identified. Interpolation is used to fill between the turning point 316 (TDC) and the cross point 314 (40 dATDC) of the shape function vector. The resulting swirl difference curve 318 (difference between raw data and swirled shape function curve) is shown in FIG. 4A, which further represents the target heat release rate curve 310. However, in some examples, information related to non-combustion, such as passive acceleration components, is also removed by using a second method, referred to herein as a reflection method. Here, the leading edge of the accelerometer combustion component is reflected around a predefined point on the difference signal data. The result is a symmetrical accelerometer combustion component profile against which the heat release rate can be reproduced. Therefore, FIG. 4B graphically illustrates this reflection method. Again, it is assumed that both the raw accelerometer signal 300, the passive signal 302, and the shape function are obtained in the same manner as described above. A peak 322 on the difference curve 304 (difference between the measured signal 300 and the passive signal 302) is determined. Note that the peak search is done within a limited pre-selected crank angle range. A limited window range is used to improve computational efficiency and to prevent noise from affecting the results by identifying “false” peaks. This window is preferably wide enough to cover the expected range of combustion times. The range from -10 dATDC to +20 to +30 dATDC is generally a suitable range, but as those skilled in the art will appreciate, the range can vary from engine to engine or from application to application. The time position (crank angle) associated with the fraction of the value associated with peak 322 is calculated. The selected fractional value is (a) small enough to remove the uncertainty of the peak position associated with a relatively flat signal with noise, and (b) a significant amount of combustion information is not lost. Should be large enough. A suitable fraction can be in the range of 50% to 99%. In this case, a value of 90% is preferably chosen. Points on the difference curve 302 that exceed the 90% point are reflected around this point, resulting in an acceleration curve associated with contrast combustion. In addition, if an averaging filter is applied around the top peak of the reflected curve, the peak can be rounded. This method works well for HCCI type engines.

In another embodiment of the presently described technology, information related to non-combustion is also removed by using a differential method. As with the other two methods described above, this method is applied to the difference curve 304. When the difference curve 304 shown in FIG. 3 is differentiated, a curve similar to the integrated heat release rate 310 is obtained. The result of this method is shown graphically in FIG. 4C. When the difference curve 304 is differentiated, a differential curve 324 is obtained. In this case, the differential method removes the influence of the end effect described above. It should be noted that the differential method has a number of steps to ensure error strength. First, a filter is applied to the raw accelerometer signal data to ensure a relatively smooth result when the difference signal is differentiated. In order to reduce the influence of the filter on the phase information, a finite impulse response type filter is preferred. The number of poles and values associated with this filter pole creates a trade-off between cutoff frequency, transition region length, bandwidth ripple, and computation time. As an example, the 19-pole filter has little influence on the phase information and functions well. At the same time, the differential signal remains relatively smooth. In this case, the value of the filter tap is based on Gaussian attenuation (ie, the attenuation is proportional to exp {− (x / σ) ² }, where x is the distance from the center of the filter to the tap, and σ is selected to ensure that the filter edge is associated with an x value that is 2-3 times the value of σ).

  Second, it has been found that the mismatch between the shape function and the raw accelerometer signal data introduces errors in the early stage heat release rate assessment. As these problems appear, a single shape function is used for almost all speed and load conditions. In this case, the slope of the shape function does not reliably match the slope near the top dead center of the raw accelerometer data signal. For the swivel and reflection methods, this mismatch is simply a reduction in the signal to noise ratio. However, for differential methods this problem is exacerbated. This is because differentiation amplifies the effect of mismatch. One method has been developed to avoid this problem. This method uses the peak position of the differentiated signal as a reference point. Information at a fixed time point or a crank degree interval before the peak position is set to zero. This method works well when used in combination with the septal smoothing method shown in the '333 application. Other methods can be used to solve this problem. For example, the entire differentiated signal can be offset downward based on a threshold to eliminate the effects of early mismatch. A method using the peak position as a reference point is a preferred method.

  The above approach can introduce error strength problems. For example, under nominal combustion conditions, the size of the first peak associated with the mismatch is part of the size of the main peak corresponding to combustion. Generally, the first peak is within 10% of this main peak. However, as the combustion time is slowed or advanced relative to nominal combustion, the magnitude of the main peak begins to decrease. In contrast, the size of the mismatch peak remains relatively constant or the main peak additionally interferes with the first peak. In either case, the peaks associated with combustion are more difficult to identify. Generally, if the absolute magnitude of each peak is less than a threshold, the engine is considered non-ignited. However, if the absolute magnitude is greater than this threshold and the absolute magnitude of the first peak is less than a predetermined percentage of the second peak magnitude, the second peak is selected as the reference position as described above. It is. As an example, a first peak whose relative magnitude is typically 75% or less of the magnitude of the second peak is such that the second peak is selected as the reference position. However, if the relative magnitude is greater than this threshold level (75% in the given example), the first peak is usually selected.

  Other approaches can be used to compensate for acceleration information associated with non-combustion. For example, model-based methods can be used to predict the decay rate for a given accelerometer signal.

  The next step in this process is to build the HRR from the signal determined by one of the pre-processing methods described above, ie, the turning method, the reflection method and the differentiation method. A schematic diagram of the main bearing cap and cylinder is shown in FIG. There are two bearing caps available for each cylinder. One potential problem is whether the firing sequence has a strong impact on cylinder bearing cap pairing. It is generally known that most bearing cap / cylinder combinations provide usable flammability information. The selection of the shape function is preferably performed for each bearing cap / cylinder combination. In addition, care should be taken to ensure that the signal quality is acceptable, ie, a relatively high signal to noise ratio. For example, noise spikes may be observed in combustion events when observed for certain cylinder / bearing cap combinations. In the second engine, the signal from the special bearing cap has in-phase low frequency fluctuations that are not related to passive or combustion processes. This low frequency variation can be in phase, but with some phase jitter. In each of the above cases, the result is usually an unacceptably low signal to noise ratio. Thus, care should be taken when selecting a bearing cap / cylinder combination. For redundant purposes, it is advantageous to mount two sensors on the bearing cap that provide acceptable combustion information for two adjacent cylinders.

  In summary, a four-step method has been described for extracting combustion information from raw accelerometer data as follows. First, the measured accelerometer signal is decomposed into combustion and passive acceleration (shape function) components. Second, the heat release rate is reproduced from the combustion components using the transfer function. The last two steps include extracting flammability information from the reproduced heat release. Pre-processing methods are used to remove much of the information related to non-combustion, resulting in significant improvements in SOC prediction. The ability of an accelerometer-based SOC control system has been demonstrated with an HCCI engine. The accelerometer based control system has the same capabilities as the in-cylinder pressure sensor based SOC control system. The developed algorithm yields a usable and robust assessment of combustion timing.

  While particular elements and embodiments of the presently described technology have been illustrated and described, it will be understood that the presently described technology is not limited thereto. This is because variations can be made by those skilled in the art, particularly in light of the above teachings, without departing from the scope of the present disclosure.

It is a schematic diagram of the combustion control system which concerns on one Embodiment of the technique now in description. 6 is a flowchart of steps for processing accelerometer signals according to one embodiment of the presently described technology. The signal of each process of FIG. 2A is shown. It is a graph showing accelerometer data. 6 is a graph representing processed accelerometer data according to one embodiment of the presently described technology. 6 is a graph representing processed accelerometer data according to another embodiment of the presently described technology. 6 is a graph representing processed accelerometer data according to another embodiment of the presently described technology. It is a mimetic diagram showing bearing cap selection concerning one embodiment of this art under present explanation.

Explanation of symbols

100 engine,
10: Main bearing cap 12 Exemplary flammability treatment system 13 Signal conditioning block 16 Piston 18 Crankshaft 20 Microcontroller 22 Accelerometer 28 Interface 29 CAN bus

Claims (3)

A method for processing accelerometer signals associated with a combustion process in an internal combustion engine in operation, comprising:
a) measuring an accelerometer signal including a combustion acceleration component and a passive acceleration component across a selected crank angle window during a cycle of the engine;
b) applying a shape function to the measured accelerometer signal to reduce the passive acceleration component. a) swiveling the shape function prior to the applying step;
b) determining, in the applying step, a difference signal between the measured accelerometer signal and the shape function, and extracting the combustion component reflecting a part of the difference signal relating to a predetermined crank angle; ,
c) one preprocessing selected from the step of determining in the applying step a difference signal between the measured accelerometer signal and the shape function and differentiating the difference signal to extract the combustion component; The method of claim 1, comprising a step.   3. The method of claim 2, including the step of providing a phase shift to the shape function prior to the preprocessing step.

Images