Noise source identification and mitigation method, device, gas engine using such, and knock detection method

Abstract

The present invention pertains to a noise source identification and mitigation method for a knock detection method of a gas engine, comprising the steps of retrieving (S40F) a false-positive knocking signal, determining (S45) if the false-positive knocking signal (18F) stems from a predetermined noise signal or an unidentified noise signal, and triggering (S47) an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal (18F).

Description

[0001] Description

[0002] Noise Source Identification and Mitigation Method, Device, Gas Engine Using Such, and Knock Detection Method

[0003] Technical Field

[0004] The present invention pertains to a noise source identification and mitigation method for a knock detection method of a gas engine, comprising the steps of retrieving a false-positive knocking signal and determining if the falsepositive knocking signal stems from a predetermined noise signal or an unidentified noise signal. The present invention also pertains to a knock detection method for a gas engine, a noise identification and mitigation device, and a gas engine.

[0005] Technological Background

[0006] Knocking is an abnormal combustion phenomenon, which adversely affects gas engine performance, emissions, and service life. During normal combustion in a gas engine piston, the combustion flame originates from an ignition spark and propagates through compressed combustion gas as a turbulent flame front. In contrast, knocking events occur when combustion gas is ignited before it reaches its designated compression rate and before it is consumed by the flame front. At a knocking event, large pressure waves with amplitudes of several bars are created, causing high-frequency oscillations of the cylinder pressure, which causes vibrations and is audible as sound.

[0007] Generally, the knocking phenomenon is more likely during operation at higher power densities. Pistons operating close to this limit enable the highest possible thermal efficiency, rendering it a desired operating point. However, if this limit is exceeded, knocking occurs, and the combustion knock causes severe damage to the engine components.

[0008] Therefore, detecting and mitigating knocking events are important measures to prevent damage to the engine components, which may be achieved using the detection of noise signals. To improve the efficiency of knock detection methods, noise source identification for noise signals is required for identification and mitigation of noise sources.

[0009] The noise source identification and mitigation method for a knock detection method of a gas engine, the noise identification and mitigation device, and the gas engine of the present disclosure solve one or more problems set forth above.

[0010] Summary of the Invention

[0011] Starting from the prior art, it is an objective to increase accuracy of knock detection methods used in gas engines.

[0012] This objective is solved by means of a noise source identification and mitigation method for a knock detection method of a gas engine with the features of claim 1, a knock detection method for a gas engine with the features of claim 13, a noise source identification and mitigation device with the features of claim 14, and a gas engine with the features of claim 15. Preferred embodiments are set forth in the present specification, the Figures as well as the dependent claims.

[0013] Accordingly, a noise source identification and mitigation method for a knock detection method of a gas engine is provided. The method comprises the steps of retrieving a false-positive knocking signal, determining if the falsepositive knocking signal stems from a predetermined noise signal or an unidentified noise signal, and triggering an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal.

[0014] Furthermore, a knock detection method for a gas engine may be provided. The method comprises the steps of retrieving a gas engine vibration signal, characterizing the retrieved vibration signal to obtain a characterized signal, detecting a knocking signal by comparing the characterized signal to a predetermined threshold signal, qualifying the knocking signal as a false-positive knocking signal, determining if the false-positive knocking signal stems from a predetermined noise signal or an unidentified noise signal, and triggering an action suitable for avoiding detections of noise signals similar to the falsepositive noise signal. Moreover, a noise source identification and mitigation device is provided, suitable for carrying out the noise source identification and mitigation method according to the present disclosure and / or for carrying out the knock detection method according to the present disclosure.

[0015] Further, a gas engine may be provided, comprising at least one cylinder and the noise source identification device according to the present disclosure.

[0016] Brief Description of the Drawings

[0017] The present disclosure will be more readily appreciated by reference to the following detailed description when being considered in connection with the accompanying drawings in which:

[0018] Fig. 1 schematically shows a flow chart of a noise source identification and mitigation method according to a first embodiment;

[0019] Fig. 2 schematically shows a flow chart of a noise source identification and mitigation method according to a further embodiment; and

[0020] Fig. 3 schematically shows a flow chart of a knock detection method according to a first embodiment.

[0021] Detailed Description of Preferred Embodiments

[0022] In the following, the invention will be explained in more detail with reference to the accompanying Figures. In the Figures, like elements are denoted by identical reference numerals and repeated description thereof may be omitted in order to avoid redundancies.

[0023] The present disclosure is generally directed towards noise source identification and mitigation in knock detection methods for a gas engine, the gas engine preferably comprising multiple cylinders. More specifically, the present disclosure is directed towards a noise source identification and mitigation method suitable for avoiding false-positive knocking signals and suitable for a wide range of cylinder configurations and a wide range of combustion gas compositions to enhance accuracy and efficacy of knock detection methods. The underlying principle of the present disclosure is based on a step-wise analysis of gas engine vibration signals obtained from an acceleration sensor, in particular a vibration sensor, and suggests undertaking signal origin considerations for noise signals which had been identified as indicating a falsepositive knocking signal. This approach is in stark contrast to the widely held but herewith overcome opinion that false-positive knocking signals are to be discarded without further post-processing to save computational resources. In any case, the benefits of providing such a false-positive knocking signal postprocessing have been proven outperforming any associated drawbacks associated with increased computational resources.

[0024] In Figure 1 a flow chart of a noise source identification and mitigation method according to a first embodiment is shown schematically. Therein, the noise source identification and -mitigation method comprises the steps of retrieving S40F a false-positive knocking signal 18F, determining S45 if the false-positive knocking signal 18F stems from a predetermined noise signal or an unidentified noise signal, and triggering S47 an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal 18F.

[0025] In the context of the present disclosure, a noise source identification and -mitigation method may be understood as a noise or vibration signal post-processing method which may be implemented as a stand-alone method or included in a knock detection method.

[0026] In the context of the present disclosure, a knock detection method may be understood as a method suitable for identifying combustion-driven knocking events as they occur in reciprocating piston cylinder engines, in particular gas engines, when combustion gas is ignited before it reaches its designated compression rate and before it is consumed by the flame front.

[0027] In knock detection methods, false-positive knocking signals 18F are bearers of high risk, as a true-positive knocking event would cause an unnecessary halt or power reduction of the gas engine.

[0028] In the context of the present disclosure, a false-positive knocking signal 18F may represent a signal which has previously been qualified as a false positive knocking signal. In other words, the false-positive knocking signal 18F may be a signal which has conclusively been determined as not stemming from a knocking event. Accordingly, any subsequent post-processing based on the falsepositive knocking signal 18F deals with a signal which is not stemming from a knocking event. For example, the false-positive knocking signal 18F may be a signal comprising a time, frequency, and amplitude range comparable to a truepositive knocking event, but which had been qualified as a false-positive knocking signal using further evidence or considerations.

[0029] The person skilled in the art will appreciate that the qualification of the false-positive knocking signal 18F can be achieved in various ways. For example, by qualifying a detected knocking signal as indicating a true knocking event or a false positive knocking event, based on at least one further gas engine parameter. Further, detecting such a knocking signal may be achieved by comparing a characterized signal to a predetermined threshold signal. A characterized signal may be obtained by characterizing a retrieved gas engine vibration signal. Further, a gas engine vibration signal may be retrieved by a sensor device. However, the decision-making process leading to the false-positive knocking signal is not part of this embodiment.

[0030] According to embodiments of the present disclosure, the falsepositive knocking signal 18F may be a characterized signal. In the sense of the present disclosure, a characterized signal may be understood as a signal comprising noise signals recorded for several gas engine cylinders. The characterized signal may consist of, or comprise, several knocking time windows, wherein each knocking time window comprises noise signals for one dedicated cylinder. A knocking time window may be a cylinder-specific time span, during which knocking might occur in that specific gas engine cylinder. Thereby, several noise sources, each carrying noise information for a given gas engine cylinder, can be efficiently patched to one characterized signal comprising only the relevant knocking time window signals for all gas engine cylinders.

[0031] A characterized signal may be achieved using cylinder position information, and / or information pertaining to combustion cycles. For example, a cam shaft signal may be used to identify the current cylinder position and cycle. Based on this information, it is possible to position the knocking time window of this gas engine cylinder for the signal characterization.

[0032] If the false-positive knocking signal 18F is a characterized signal, noise similarity is to be assessed on a time window basis. Accordingly, if the false-positive knocking signal 18F is a characterized signal, triggering S47 an action suitable for avoiding detections of noise signals similar to the falsepositive knocking signal 18F is to be understood as triggering S47 an action suitable for avoiding detections of noise signals within a given time window, hence, noise signals generated at a given cylinder.

[0033] In the context of the present disclosure, retrieving S40F the falsepositive knocking signal 18F may be understood in a broad sense as taking said signal as an input for further processing.

[0034] In the context of the present disclosure, a predetermined noise signal may be any reference noise signal, for example taken from a database. The predetermined noise signal may be a recorded, calculated, and / or simulated noise signal. Further the predetermined noise signal may comprise the noise signal itself as well as references to time and gas engine operation, for example piston angles and combustion cycles, corresponding to the recorded, calculated, and / or simulated time of occurrence. Likewise, an unidentified noise signal may be a time-span and / or a signal characteristic for which no predetermined noise signal exists. The predetermined noise signal may be provided as an electronic file storing signal information, for example an audio file.

[0035] The predetermined noise signals may contain noise information only for a predetermined time window of a gas engine cylinder, for example for a time window during which knocking may occur at a given gas engine cylinder. Thereby, predetermined noise signals may be provided in such a way that they are comparable and compatible with the false-positive knocking signal 18F provided as a characterized signal.

[0036] In the context of the present disclosure, determining S45 if the false-positive knocking signal 18F stems from a predetermined noise signal or an unidentified noise signal may be understood in a broad sense as a decisionmaking process for example based on a signal comparison. In the context of the present disclosure, an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal 18F may be understood as any action that causes, if executed, that the previously false-positive knocking signal 18F would not be considered again during identical or similar circumstances. This may be achieved either on a logical level, by signal processing, or on a physical level, by appropriate physical noise mitigation actions.

[0037] In the context of the present disclosure, the step or triggering S47 an action suitable for avoiding detections of noise signals similar to the falsepositive knocking signal 18F may be understood as outputting a signal indicative of an action. For example, the outputted signal may be a reference for an action stored in a lookup table. Depending on the system architecture, the triggering S47 may comprise a Boolean information or a byte.

[0038] The determination step S45 may comprise mapping a falsepositive knocking signal 18F, or a signal profile thereof, to a predetermined noise signal profile, or a signal profile thereof. Further, the false-positive knocking signal profile and the predetermined noise signal profile may be mapped based on signal frequencies and / or amplitudes.

[0039] Further, the determination step S45 may comprise the falsepositive knocking signal 18F to a predetermined noise signal on a time window basis.

[0040] In the context of the present disclosure, a time window may be understood in its literal sense as a time interval comprising a time reference, for example a time stamp. In other words, a time window may be understood as a time-cutout of a signal. For example, the false-positive knocking signal time window may be understood as the time window containing the false-positive knocking signal.

[0041] In the context of the present disclosure, the comparison may be understood as comparing two signals, the false-positive knocking signal 18F and the predetermined noise signal.

[0042] The false-positive knocking signal time window may be different than a time window, or length of the predetermined noise signal. More specifically, the comparison may aim to find a match between the false-positive knocking signal 18F contained in the false-positive knocking signal time window and a signal contained in the predetermined noise signal. If an ad-hoc comparison of the false-positive knocking signal and the predetermined noise signal does not lead to a match, a slightly adjusted false-positive knocking signal time window may be used in the comparison against the predetermined noise signal.

[0043] Preferably, this selection and comparison loop may comprise a finite number of attempts and may be based on predetermined mapping parameters.

[0044] For example, selecting the time window may comprise the step of moving the false-positive knocking signal time window by changing the start and / or the end of the false-positive knocking signal time window. Thereby, the false-positive knocking signal 18F may be swept in time, shortened and / or lengthened.

[0045] Further, selecting the time window may comprise the step of moving the false-positive knocking signal time window as a function of a piston movement, preferably wherein the piston movement is represented by a time- resolved piston angle.

[0046] In the context of the present disclosure, a time-resolved piston angle may be understood as a piston rotation documented as piston angle at a given time. The underlying concept is to narrow the time window selection down to time-spans corresponding to specific piston movements.

[0047] For example, if in the process of mapping the false-positive knocking signal to a predetermined nose signal the predetermined noise signal is known to only occur during a specific gas engine cycle and / or a specific piston angle range, the selection of the false-positive knocking signal time window may be adjusted such that it matches the specific gas engine cycle and / or specific piston angle range.

[0048] For example, the time window may have a constant length covering a piston angle of 5° and is selected by moving the time window incrementally by a piston angle increment of 1°. The predetermined noise signal may for example be known to occur at a specific gas engine cycle and / or a specific piston angle range. However, the data may be susceptible to change, for example to a drift in piston angle due to wear, changed boundary conditions, and / or due to a slightly different cylinder configuration. Further, it may be known that the predetermined noise signal may have a time-length equivalent to one or more piston angles. To match the false-positive knocking signal to the predetermined noise signal, an iterative process may be conducted, wherein the time window, hence, the false-positive knocking signal contained therein, is incrementally changed by a piston angle increment of 1°. After each movement of the time window by the piston angle increment of 1°, a subsequent comparison may be conducted. Further, the number of increments may be limited by a predetermined number.

[0049] Further, the predetermined nose signal may comprise a valve closure event, and / or a piston slap event. In the context of the present disclosure, a valve closure event may be a vibration or noise emission triggered by the physical event of a valve closure. This noise emission may be stored as an electronic file, for example as an audio file, and may comprise information about time, gas engine cycle, and / or the piston angle. Likewise, a piston slap event may be a vibration or noise emission triggered by the physical event of a piston slap. This noise emission may be stored as an electronic file, for example as an audio file, and may comprise information about time, gas engine cycle, and / or the piston angle.

[0050] Comparing the false-positive knocking signal 18F to the predetermined noise signal may comprise comparing signal maxima, signal differences, and / or signal integrals. Maxima and differences may correspond to signal frequencies, amplitudes. Signal integrals may comprise an integration of a signal function to obtain the area under the function in between a predetermined time-step according to the literal meaning of the term integration.

[0051] Further, comparing the false-positive knocking signal to the predetermined noise signal may comprise utilizing a signal filter. A signal filter according to the present disclosure may be understood as an audio signal postprocessing filter, for example noise filter, a low pass filter, a high pass filter, a compressor, expander, limiter, polarity inverter and / or a combination thereof. In Figure 2, a noise source identification and -mitigation method according to a further embodiment is shown schematically by a flow chart. The embodiment of Figure 2 is based on the embodiment shown in Figure 1. Accordingly, the same principles, definitions, and explanations provided in the context of Figure 1 also apply to the embodiment shown in Figure 2 where applicable.

[0052] According to the embodiment shown in Figure 2, the determination step S45 comprises a source analysis step, in which it is determined if the false-positive knocking signal 18F stems from a predetermined noise signal or an unidentified noise signal. Herein, the predetermined noises may comprise a valve closure event, and a piston slap event, which is indicated by the left and right paths and boxes. In all other cases, it is determined that the falsepositive knocking signal 18F stems from an unidentified noise signal, which is indicated by the default middle path and box.

[0053] If it is determined that the false-positive knocking signal stems from a valve closure event or a piston slap event, an optional window mitigation adaption may occur. Subsequently, the noise source identification method proceeds to the step of triggering S47 an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal 18F.

[0054] The trigger step S47 may comprise setting a valve lash if the falsepositive knocking signal 18F is determined as stemming from a valve closure event. Likewise, the trigger step S47 may comprise setting a piston-ring-liner, PRL, if the false-positive knocking signal 18F is determined as stemming from a piston slap event. Further, the trigger step S47 may comprise desensitizing a cylinder if the false-positive knocking signal 18F is determined as stemming from an unidentified source.

[0055] Further, the noise source identification and mitigation method may further comprise the step of broadcasting S60 an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal, via a CAN bus, a modbus, and / or an ethemet for further usage.

[0056] In Figure 3, a knock detection method is schematically shown by a flow chart according to an embodiment of the present disclosure. The knock detection method shown in Figure 3 comprises the noise identification and - mitigation method of the present disclosure. For method steps pertaining to the noise identification and -mitigation method, the same principles, definitions, and explanations provided in the context of Figures 1 and 2 also apply to the knock detection method where applicable.

[0057] Accordingly, the knock detection method may be suitable for a gas engine and comprises the steps of retrieving S10 a gas engine vibration signal 12, characterizing S20 the retrieved vibration signal 12 to obtain a threshold signal 16, detecting S30 a knocking signal 18 by comparing the characterized vibration signal 14 to a predetermined threshold signal 16, qualifying S41 the knocking signal as a false-positive knocking signal 18F, determining S45 if the falsepositive knocking signal 18F stems from a predetermined noise signal or an unidentified noise signal, and triggering S47 an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal 18.

[0058] In the context of the present disclosure, a gas engine vibration signal 12 may represent a signal suitable for carrying information about vibrations as they are typically observed during a knocking event, for example in the shape of frequencies or frequency bands. The information carried by the gas engine vibration signal 12 may also comprise vibration ranges beyond vibrations observed during a knocking event.

[0059] The retrieval step S10 may comprise retrieving the vibration signal 12 from a sensor device 20. The retrieval step S10 may comprise retrieving the vibration signal 12 from several sensor devices 20. The retrieval step S10 may comprise retrieving several vibration signals 12 from individual sensor devices 20.

[0060] In the sense of the present disclosure, during the characterizing S20 step, the one or more retrieved gas engine vibration signals 12 are assigned to gas engine cylinders, reduced in length, and patched to a new signal, the characterized signal 14.

[0061] Preferably, the characterization step may comprise identifying knocking time windows during which knocking may occur, knocking time window signals contained in said knocking time windows, and patching the knocking time window signals to one continuous characterized signal 14.

[0062] The predetermined threshold signal 16 may comprise, for each gas engine cylinder, a threshold amplitude for a given frequency, frequency range, or frequency band. The predetermined threshold signal 16 is to be defined such that a characterized signal 14 having an amplitude surpassing said threshold signal 16 amplitude is likely indicative of a knocking event.

[0063] The predetermined threshold signal 16 may be based on noise measurements and / or simulation data. Further, the predetermined threshold signal may comprise knocking time window information and / or a time stamp for synchronizing the predetermined threshold signal 16 with the characterized signal 14. Alternatively, or additionally, the predetermined threshold signal 16 may consist of several predetermined threshold signals 16, some of which may comprise threshold signals pertaining to one or just some cylinders of the total number of cylinders. In any case, the predetermined threshold signal 16 must comprise threshold information for all operated gas engine cylinders.

[0064] The threshold signal 16 preferably comprises a frequency sequence that allows direct comparison with the characterized signal 14. To this end, the sequence of the predetermined threshold signal 16 may be identical to the knocking time window sequence of the characterized signal 14.

[0065] The predetermined threshold signal 16 may be stored as a library or database. Preferably, the predetermined threshold signal 16 may be stored with references to the gas engine cylinders considered therein and a time stamp. Further, the predetermined threshold signal 16 may be stored in a look-up table.

[0066] In the context of the present disclosure, the gas engine vibration signal 12, the characterized signal 14, and the predetermined threshold signal 16 may have the same format and may be provided such that they are comparable to each other, for example by frequency.

[0067] The retrieved vibration signal 12 may have the form of a vibration signal profile. The vibration signal profile may extend over the time duration of the time window. The nominal operation vibration signal may also have the form of a vibration signal profile, of individual peaks, or it may be zero. Likewise, the predetermined threshold signal 16 may have the form of a signal profile or a constant signal.

[0068] The step of detecting S30 a knocking signal 18 by comparing the characterized signal 14 to the predetermined threshold signal 16 may be understood as a quantitative signal filtering step. The threshold signal 16 may be defined such a comparison of the characterized signal 14 with the threshold 16 indicates the possibility or likelihood of a knocking event. A detection of a knocking signal 18 may occur if the comparison gives the result that the predetermined threshold signal 16 was surpassed. In this case, the characterized signal 14 is graduated to a knocking signal 18. For example, if the comparison of the characterized signal 14 with the threshold signal 16 gives that the characterized signal 14 comprises a signal magnitude greater than the threshold 16, the characterized signal 14 is detected as a knocking signal 18. Likewise, if the comparison of the characterized signal 14 with the threshold signal 16 gives that the characterized signal 14 has a signal magnitude smaller than or equal to the threshold 16, the characterized signal 14 may be detected as a false knocking signal. In this case, according to the embodiment shown in Figure 1, the method is looped back to the retrieval step S10, also called signal acquisition.

[0069] The knocking signal detection step S30 may be understood as a comparative step, comparing the characterized signal 14 to the predetermined threshold signal 16.

[0070] In the qualification step S40, the detected knocking signal 18 is qualified as a false positive knocking event, for example based on at least one further gas engine operation parameter.

[0071] In this case, the determination step S45 and the trigger step S47 are initiated. Hereto, it is referred to the embodiments of Figures 1 and 2.

[0072] It will be obvious for a person skilled in the art that these embodiments and items only depict examples of a plurality of possibilities. Hence, the embodiments shown here should not be understood to form a limitation of these features and configurations. Any possible combination and configuration of the described features can be chosen according to the scope of the invention. The noise source identification and mitigation method of the present disclosure is suitable for being used in a knock detection method, in particular a knock detection method of a gas engine. The noise source identification and -mitigation method generally aims to avoid detections of noise signals which are false-positive knocking signals, in order to declutter knock detection methods, thereby increasing the accuracy of knock detection methods.

[0073] To address this, a noise source identification and mitigation method for a knock detection method of a gas engine is provided, comprising the steps of retrieving a noise signal qualified as a false-positive knocking signal, determining if the false-positive knocking signal stems from a predetermined noise signal or an unidentified noise signal, and triggering an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal.

[0074] In the sense of the present disclosure, a false-positive knocking signal may be understood in its literal sense as a noise signal which has previously been identified as a false-positive knocking signal.

[0075] The qualification of a noise signal as a false-positive knocking signal is not part of the noise source identification and mitigation method in the sense of claim 1.

[0076] By retrieving and processing a false-positive knocking signal, future occurrences of similar signals may be reduced or avoided altogether. Thereby, resources can be saved, and a slimmer, faster knock detection may be achieved.

[0077] The step of determining if the false-positive knocking signal stems from a predetermined noise signal or an unidentified noise signal may be understood as an attempt to explain the retrieved false-positive knocking signal by a known, predetermined, noise signal. This can for example be done by successfully mapping the false-positive knocking signal to one or more predetermined noise signals. Likewise, the step of triggering an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal may be understood as the mitigation of the noise source responsible for the noise signal qualified as a false-positive knocking signal. Thereby, known false-positive noise signals may be identified as such. In addition, the detection of such known false-positive noise signals may be mitigated either on a logical signal processing level, or on a physical level. Thereby, the noises to be considered in the knock detection method may be reduced. Hence, the knock detection method input may be decluttered. Thereby, the knock detection method may operate at higher accuracy and at reduced computational requirements. Ultimately, the gas engine may thereby be operated close to its knocking value with higher confidence, providing more work at less down time.

[0078] The determination step may comprise one or more attempts to match the false-positive knocking signal to one, or one of the predetermined noise signal(s). If such a match is found, a successful determination occurred that the false-positive knocking signal stems from a predetermined noise signal. Hence, determining that the false-positive knocking signal stems from an unidentified noise signal may be understood as the default result, corresponding to an unsuccessful mapping attempt of mapping the false-positive knocking signal to one, or one of the predetermined noise signal(s). To achieve a successful match between the false-positive knocking signal and a predetermined noise signal, several consecutive mapping attempts may be conducted.

[0079] According to a preferred embodiment, the retrieval step may comprise acquiring the noise signal from a sensor device. The sensor device may for example be a vibration sensor. Thereby, a simple, non-combustion chamber internal noise signal may be retrieved. Thereby, the noise source identification and mitigation method may be suitable for an indirect knock detection method, meaning a method based on measuring gas engine housing accelerations, including accelerations caused by a knocking event, are measured instead of measuring parameters within the combustion chamber. Such an indirect knock event detection method may be more cost-effective while at the same time overcoming the draw-back of inaccuracy as observed in conventional indirect knock detection methods.

[0080] According to a preferred embodiment, the determination step may comprise mapping a false-positive knocking signal, or signal profile, to a predetermined noise signal, or signal profile. Thereby, the false-positive knocking signal, or signal profile, may be identified conveniently based on a signal, or signal profile, match. It will be appreciated by the person skilled in the art that the mapping itself is to be based on predetermined match quality parameters which are to be selected on a case-by-case basis using simple, straightforward experiments or quality considerations.

[0081] According to a preferred embodiment, the false-positive knocking signal and the predetermined noise signal are mapped based on signal frequencies and / or signal amplitudes. Thereby, a simple and robust signal comparison may be achieved. Preferably, signal intensities of predetermined noise signals and falsepositive knocking signals are calibrated to warrant a reliable comparison.

[0082] According to a preferred embodiment, the determination step may comprise the step of comparing the false-positive knocking signal to a predetermined noise signal on a time window basis. Thereby, a large, falsepositive knocking signal may be reduced to a short signal snippet, or used as such, for example if the false-positive knocking signal is a characterized signal. According to the time window approach, only the signal contained in the time window will be considered in the determination if the false-positive knocking signal stems from a predetermined noise signal or an unidentified noise signal. Thereby, the computational requirements of the noise identification and mitigation method can be reduced effectively, allowing a more cost-effective execution of the knock detection method and ultimately, a more cost-effective operation of the gas engine at no losses of accuracy.

[0083] According to a preferred embodiment, the determination step may further comprise the step moving a false-positive knocking signal time window by changing the start and / or the end of the false-positive knocking signal time window. Thereby, it is possible to try mapping the false-positive knocking signal to several predetermined noise signals of different noise lengths at a minimum of computational requirements.

[0084] According to a preferred embodiment, the determination step may further comprise the step of moving the false-positive knocking signal time window as a function of a piston movement, preferably wherein the piston movement is represented by a time-resolved piston angle. Thereby, the falsepositive knocking signal may be reduced to certain time windows during which a potential match would theoretically be possible. Hence, all parts of the falsepositive knocking signal outside of the time window of interest may be discarded. Thereby, it is possible to try mapping the false-positive knocking signal to a predetermined noise signal at a minimum of computational requirements.

[0085] According to a preferred embodiment, the false-positive knocking signal time window may have a constant length covering a piston angle change of 5° to 6°, specifically 5°, wherein the time window is selected by moving the time window incrementally by a time step representing a piston angle change of 1° to 2°, specifically 1°. Thereby, noise drifts can be accounted for. Noise drifts can occur, if a cylinder-piston configuration is slightly different to a cylinder-piston configuration used as the basis for the predetermined noise signal, or if operation conditions differ. More specifically, a noise signal may be created during operation at a slightly different piston angle compared to the angle of its predetermined reference noise signal. Thereby, the chances of determining that the noise signal stems from a predetermined noise signal may be increased further, allowing a more targeted nose mitigation.

[0086] According to a preferred embodiment, the predetermined noise signal may comprise a valve closure event, and / or a piston slap event. A valve closure event can be conveniently correlated to a given gas engine cycle and a piston angle range. Further, a valve closure event can be conveniently recorded, calculated and / or simulated, and stored in a database together with time, gas engine cycle, and piston angle information. A false-positive knocking signal may therefore be conveniently shortened to the given time window around the valve closure event and a mapping attempt may be conducted during the determination step. Likewise, a piston slap event may be conveniently correlated to a given piston angle range. Further, a piston slap event can also be recorded, calculated, and / or simulated, and stored in a database together with time, and piston angle information. A false-positive knocking signal may therefore be conveniently shortened to the given time window around a potential piston slap event and a mapping attempt may be conducted during the determination step. According to a preferred embodiment, comparing the falsepositive knocking signal to a predetermined noise signal may comprise comparing signal maxima, signal differences, and or signal integrals. Thereby, an efficient, repeatable, and easy to implement comparison routine may be implemented into the noise source identification method.

[0087] According to a preferred embodiment, comparing the falsepositive knocking signal to the predetermined noise signal comprises utilizing a signal filter. Thereby, the false-positive knocking signals and / or the predetermined noise signals may be transformed in such a way that the comparison can be achieved in a simpler fashion. Alternatively, or additionally, by utilizing the signal filter, a more complex comparison routine may be implemented in the form of a filter to achieve repeatable results.

[0088] According to a preferred embodiment, the trigger step may comprise setting a valve lash, if the false-positive knocking signal is determined as stemming from a valve closure event, checking a piston-ring-liner, PRL, if the false-positive knocking signal is determined as stemming from a piston slap event, and / or desensitizing a cylinder, if the false-positive knocking signal is determined as stemming from an unidentified source. By triggering setting a valve lash, it is possible to effectively mitigate the occurrence of such a noise signal as a false-positive knocking signal. Once the valve lash is adjusted appropriately, the valve closure event will be softer and will subsequently not be detected as a potential noise signal indicative of a knocking event. Likewise, by triggering checking a piston-ring-liner, PRL, it is possible to effectively mitigate the occurrence of such a noise signal as a false-positive knocking signal. Once the piston-ring-liner, PRL, is serviced, the risk of subsequent piston slaps can be reduced. Further, by triggering desensitizing a cylinder, it is possible to effectively mitigate the occurrence of the unidentified noise signal as a falsepositive knocking signal. Once the sensitivity of the cylinder is reduced, the sensed signal intensity will be lower and such signals will be less likely to be detected as a potential noise signal indicative of a knocking event.

[0089] According to a preferred embodiment, the method may further comprise the step of broadcasting an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal, via a CAN bus, a modbus, and / or an ethernet for further usage. Thereby, the obtained information may be stored, post-processed and / or used for subsequent operation gas engine operation cycles.

[0090] A knock detection method for a gas engine may be provided, comprising the steps of retrieving a gas engine vibration signal, characterizing the retrieved vibration signal to obtain a threshold signal, detecting a knocking signal by comparing the characterized signal to a predetermined threshold signal, qualifying the knocking signal as a noise signal indicating a false-positive knocking event, determining if the noise signal stems from a predetermined noise signal or an unidentified noise signal, and triggering an action suitable for avoiding detections of noise signals similar to the noise signal.

[0091] The knock detection method for a gas engine according to the present disclosure may be understood as an indirect knock detection method, meaning that instead of measuring parameters within the combustion chamber, gas engine housing accelerations, including accelerations caused by a knocking event, are measured.

[0092] By this, existing gas engines may be retrofitted with an efficient knocking detection method according to the present disclosure. Further, utilizing such an indirect knock event detection method is cost-effective while at the same time overcoming the draw-back of inaccuracy as observed in conventional indirect knock detection methods.

[0093] The knock detection method of the present disclosure utilizes a stepwise approach of signal retrieving, signal characterization, knock signal detection, and knock signal qualification, in combination with a smart falsepositive knock detection signal detection and mitigation. In other words, this approach allows to qualify whether a detected knocking signal is indicating a true knocking event or indicating a false positive knocking event and to improve the knock detection method based on false positive knocking events by identification and mitigation thereof. The detected knocking signal must pass a further investigative qualification step before a true or a false knocking event is determined. Thereby, the detection of a true increased and the risk of detecting a false positive knocking event is reduced. Thereby, the gas engine may be operated close to its knocking value with higher confidence, providing more work at less down time. In addition, by being able to identify and mitigate falsepositive knocking signals, the knock detection method may be more accurate and may require less computational resources.

[0094] A noise source identification and mitigation device may be provided, suitable for carrying out the noise source identification and mitigation method according to the present disclosure, and / or for carrying out the knock detection method according to the present disclosure.

[0095] The noise identification and mitigation device may comprise a sensor device, a signal processing device, and a control device. Pertaining to the noise identification and mitigation device, the same principles, explanations, and definitions provided in the context of the noise identification and mitigation method apply where applicable. By the provision of a noise identification and mitigation device according to the present disclosure, an existing gas engine, or gas engine infrastructure may be upgraded without introducing substantial changes to the gas engine.

[0096] A gas engine may be provided, comprising at least one cylinder and the noise source identification and mitigation device according to the present disclosure.

[0097] Pertaining to gas engine, the same principles, explanations, and definitions provided in the context of the noise identification and mitigation method and the noise identification and mitigation device apply where applicable.

[0098] The sensor device may comprise a vibration sensor, which may be mounted to a gas engine surface, in particular the gas engine piston. Preferably, the sensor device may comprise a plurality of vibration sensors.

[0099] The control device may be configured to receive the signals and to execute the method steps. Further, the control device may comprise a data processing device suitable for post-processing signals.

[0100] According to an embodiment of the gas engine, the gas engine may further comprise at least one cooling circuit and means to monitor fluid temperatures within the cooling circuit. Further, the gas engine may comprise a water jacket having an inlet and an outlet as well as means to detect water jacket inlet and outlet temperatures.

[0101] According to a further embodiment, the gas engine may comprise an intercooler and an air intake comprising an intake manifold and means to detect an intercooler inlet temperature and an intake air temperature.

[0102] According to a further embodiment, the gas engine may comprise a CAN bus, a modbus, and / or an ethernet configured to broadcast information from the control device for further processing.

[0103] Industrial Applicability

[0104] With reference to the Figures, a noise source identification and mitigation method for knock detection method for a gas engine, a knock detection method, a noise source identification and mitigation device, and a gas engine are provided.

[0105] In practice, a noise source identification and mitigation device and a gas engine according to the present disclosure may be manufactured, bought, or sold to retrofit a gas engine, or a gas engine already in the field in an aftermarket context, or alternatively may be manufactured, bought, sold, or otherwise obtained in an OEM (original equipment manufacturer) context. Likewise, the knock detection method according to the present disclosure can be implemented in an such a gas engine.

[0106] As alluded to previously herein, the aforementioned developments may provide a simple, cost-effective and more reliably operating noise source identification and mitigation method for a knock detection method of a gas engine.

[0107] Referring to Figure 1, there is an embodiment shown disclosing a knock detection method for a gas engine, comprising the steps of retrieving a false-positive knocking signal, determining if the false-positive knocking signal stems from a predetermined noise signal or an unidentified noise signal, and triggering an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal. One skilled in the art will expect that various developments of the present disclosure will cause an increased knock detection accuracy and a reduced risk of false positive knock detection incidents when operating a gas engine. Thereby, necessitating less maintenance and allowing prolonged service life of the gas engine.

[0108] The same advantages apply to the remaining Figures.

[0109] The present description is for illustrative purposes only and should not be construed to narrow the breadth of the present disclosure in any way. Thus, those skilled in the art will appreciate that various modifications might be made to the presently disclosed developments without departing from the full and fair scope and spirit of the present disclosure. Other aspects, features and advantages will be apparent upon an examination of the attached drawings and appended claims. As used herein, the articles “a” and “an” are intended to include one or more items and may be used interchangeably with “one or more.” Where only one item is intended, the term “one” or similar language is used. Also, as used herein, the terms “has,” “have,” “having,” “include”, “includes”, “including”, or the like are intended to be open-ended terms. Further, the phrase “based on” is intended to mean “based, at least in part, on” unless explicitly stated otherwise.

[0110] All references to the disclosure or examples thereof are intended to reference the particular example being discussed at that point and are not intended to imply any limitation as to the scope of the disclosure more generally. All language of distinction and disparagement with respect to certain features is intended to indicate a lack of preference for those features, but not to exclude such from the scope of the disclosure entirely unless otherwise indicated.

[0111] Recitation of ranges of values herein are merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, unless otherwise indicated herein, and each separate value is incorporated into the specification as if it were individually recited herein.

[0112] Certain steps of any method may be omitted, performed in an order that is different than what has been specifically mentioned or in some cases performed simultaneously or in sub-steps. Furthermore, variations or modifications to certain aspects or features of various developments may be made to create further developments and features and aspects of various developments may be added to or substituted for other features or aspects of other developments in order to provide still further developments.

[0113] Accordingly, this disclosure includes all modifications and equivalents of the subject matter recited in the claims appended hereto as permitted by applicable law. Moreover, any combination of the above-described elements in all possible variations thereof is encompassed by the disclosure unless otherwise indicated herein or otherwise clearly contradicted by context.

Claims

Claims1. A noise source identification and mitigation method for a knock detection method of a gas engine, comprising the steps of retrieving (S40F) a false-positive knocking signal (18F); determining (S45) if the false-positive knocking signal (18F) stems from a predetermined noise signal or an unidentified noise signal; and triggering (S47) an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal (18F).

2. The noise source identification and mitigation method, wherein the determination step (S45) comprises mapping a false-positive knocking signal (18F) to a predetermined noise signal.

3. The noise source identification and mitigation method according to claim 2, wherein the false-positive knocking signal (18F) and the predetermined noise signal are mapped based on signal frequencies and / or amplitudes.

4. The noise source identification method according to any of the previous claims, wherein the determination step (S45) comprises comparing the false-positive knocking signal (18F) to a predetermined noise signal on a time window basis.

5. The noise source identification and mitigation method according to claim 4, wherein the determination step (S45) comprises the step of moving a false-positive knocking signal time window by changing the start and / or the end of the false-positive knocking signal time window.

6. The noise source identification and mitigation method according to any of the previous claims 4-5, wherein the determination step (S45) further comprises the step of moving the false-positive knocking signal time window as a function of a piston movement, preferably wherein the piston movement is represented by a time-resolved piston angle.

7. The noise source identification and mitigation method according to claim 6, wherein the false-positive knocking signal time window has a constant length covering a piston angle of 5° to 6°, specifically 5°, and is selected by moving the time window incrementally by a piston angle increment of 1° to 2°, specifically 1°.

8. The noise source identification and mitigation method according to any of the previous claims, wherein the predetermined noise signal comprises a valve closure event, and / or a piston slap event.

9. The noise source identification and mitigation method according to any of the previous claims 4-8, wherein comparing the false-positive knocking signal (18F) to the predetermined noise signal comprises comparing signal maxima, signal differences, and / or signal integrals.

10. The noise source identification and mitigation method according to any of the previous claims, wherein comparing the false-positive knocking signal (18F) to the predetermined noise signal comprises utilizing a signal filter.

11. The noise source identification and mitigation method, wherein the trigger step (S47) comprises setting a valve lash, if the false-positive knocking signal (18F) is determined as stemming from a valve closure event,checking a piston-ring -liner, PRL, if the false-positive knocking signal (18F) is determined as stemming from a piston slap event, and / or desensitizing a cylinder, if the false-positive knocking signal(18F) is determined as stemming from an unidentified source.

12. The noise source identification and mitigation method, further comprising the step of broadcasting (S60) an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal, via a CAN bus, a modbus, and / or an ethemet for further usage.

13. A knock detection method for a gas engine, comprising the steps of: retrieving (S10) a gas engine vibration signal (12); characterizing (S20) the retrieved gas engine vibration signal (12) to obtain a characterized signal (14); detecting (S30) a knocking signal (18) by comparing the characterized vibration signal (12) to a predetermined threshold signal (16); qualifying (S41) the knocking signal (18) as a false-positive knocking signal (18F); determining (S45) if the false-positive knocking signal (18F) stems from a predetermined noise signal or an unidentified noise signal; and triggering (S47) an action suitable for avoiding detections of noise signals similar to the false-positive knocking signal (18F).

14. A noise source identification and mitigation device suitable for carrying out the noise source identification and mitigation method according to any of the previous claims 1-12 and / or for carrying out the knock detection method according to claim 13.

15. A gas engine comprising at least one cylinder and the noise source identification and mitigation device according to claim 14.

Images