Fault analysis of sensor arrangement in unstable failure

Abstract

According to a method for fault analysis of a sensor arrangement (2, 8, 9) with respect to unstable faults, wherein the sensor arrangement (2, 8, 9) outputs an analog sensor signal (S) depending on a measured variable, a plurality of fault events is identified within a preset time interval, wherein the sensor signal (S) has a first edge and a second edge for each fault event. A diagnostic signal (D) is generated depending on the cumulative height of the first and / or second edge and compared to a first threshold value (SW1). A fault report (F) is generated depending on the comparison result.

Description

[0001] The present application relates to a method for fault analysis of a sensor arrangement with respect to unstable faults, wherein the sensor arrangement outputs an analog sensor signal depending on a measured variable. Furthermore, the present application relates to a diagnostic device for fault analysis of a sensor arrangement with respect to unstable faults.

[0002] A fault of the analog sensor arrangement or a fault of a connection line of the analog sensor arrangement influences the accuracy and reliability of those functions and devices which continue to use the analog sensor output signal. This is the case, for example, in the context of automotive applications which use different analog sensors, such as pressure sensors, temperature sensors, acceleration sensors, etc. Pressure and temperature sensors can be used, for example, for engine control. In particular, an engine control device can control the engine torque depending on the measured pressure or temperature values. In this particular example, a false output signal of an analog sensor has a direct influence on the performance, exhaust emissions or reliability of a motor vehicle. Similar cases also exist in other applications.

[0003] However, not only stable faults, i.e. permanent short circuits or permanent contact breaks, are important here, but also unstable faults, or rather unstable faults, in the case of which the faulty state, i.e. for example a short circuit or a contact break, occurs intermittently. In connection with this, one can speak of intermittent faults, periodic contact release, periodic intermittent short circuits or periodic intermittent contact breaks or colloquially of contact failures. Such unstable faults can also have an influence on performance, fuel consumption or exhaust composition, for example in the context of a motor vehicle. It is therefore desirable in principle to be able to reliably identify unstable faults in analog sensors.

[0004] Depending on the severity or criticality or relevance, or rather importance, of the unstable fault, different measures or consequences can be useful or desirable.

[0005] A method and a device for diagnosing a lambda sensor are described in document WO 2020 / 058001 A1. A diagnostic direct voltage or a diagnostic alternating voltage is input into the lambda sensor, which diagnostic direct voltage or diagnostic alternating voltage is dropped by the Nernst cell of the lambda sensor. Depending on whether a direct voltage or an alternating voltage is detected on the respective connection terminal of the input voltage, and if necessary depending on the amplitude of the detected voltage, it is possible to distinguish between a short circuit of the Nernst cell and a line break.

[0006] However, this method is specifically designed for the diagnosis of lambda sensors and cannot be transferred to any analog sensor. Furthermore, the diagnostic principle of inputting a diagnostic voltage and analyzing the respective response is not suitable for unstable faults.

[0007] Against this background, the technical problem addressed by the present application is to provide an improved design for a fault analysis of an analog sensor arrangement, by which the severity of a detected unstable fault can be taken into account.

[0008] The technical problem is solved by the respective technical solutions of the independent claims. Advantageous developments and preferred embodiments are the technical solutions of the dependent claims.

[0009] The improved design is based on the idea that a cumulative height of signal edges of the analog sensor signal is determined within a preset time interval and a fault report is generated therefrom.

[0010] According to the improved design, a method for a fault analysis of a sensor arrangement with respect to unstable faults is provided. Here, the sensor arrangement outputs an analog sensor signal as a function of a measured variable. A plurality of fault events is identified within a preset time interval by a diagnostic device, wherein the sensor signal has a first edge and a second edge following the first edge for each fault event. A diagnostic signal is generated by the diagnostic device as a function of a cumulative height of the first edges and / or a cumulative height of the second edges. The diagnostic signal is compared by the diagnostic device with a first threshold value and a fault report is generated by the diagnostic device as a function of the result of the comparison.

[0011] Here and in the following, a sensor or a sensor arrangement which outputs an analog sensor signal as a function of a value of a measured variable, wherein the amplitude or absolute value of the sensor signal continuously or substantially continuously changes as a function of the value of the measured variable, is referred to as an analog sensor or an analog sensor arrangement. A sensor arrangement can contain an analog sensor and, if necessary, an analog-digital converter. The analog-digital converter can convert the analog sensor signal into a digitized sensor signal. But according to the improved design, the analog, non-digitized sensor signal is used, inter alia, as described.

[0012] For example, a sensor arrangement can have, in addition to a sensor, a filter unit or other components. The sensor signal can correspond to the direct output signal of the sensor or can correspond to a pre-processed, for example filtered, output signal of the sensor.

[0013] An unstable fault can be understood here and in the following as a fault which is not continuously or permanently present and / or the influence of which on the sensor signal, inter alia, cannot be continuously or permanently measured or the influence of which on the sensor signal, inter alia, is variable over time. An unstable fault can also be regarded and referred to as an intermittent fault, inter alia.

[0014] The sensor signal corresponds, in particular, to an electrical current or a voltage that can be measured at a measurement connection to which the diagnostic device is connected directly or indirectly. The measurement connection is connected to the signal connection of the sensor directly or indirectly, for example, wirelessly. The absolute value or the amplitude of the sensor signal corresponds here to the value of the physical measured variable. Depending on the design of the sensor, this physical measured variable can be a different measured variable, such as pressure, temperature, acceleration, etc. In various embodiments, further electronic components, for example, signal filters or the like, can also be arranged between the sensor output and the measurement connection. In this case, the sensor signal corresponds, for example, to the filtered output signal of the sensor.

[0015] When the connection between the sensor output and the measurement connection is faulty, for example, interrupted or short-circuited, the signal measured at the measurement connection does not necessarily correspond to the signal present on the sensor output. The signal measured at the measurement connection can also be regarded as an apparent sensor signal, but is referred to here and below as the sensor signal because the diagnostic device cannot distinguish this signal itself from the actual sensor signal.

[0016] The diagnostic device can comprise, for example, an electronic control device, a microcontroller or other computing unit or processor unit and / or other analog and / or digital circuitry, in particular for a motor vehicle.

[0017] If the sensor arrangement is designed or intended for a motor vehicle, the motor vehicle can have, for example, an electronic control device that receives the sensor signal for performing one or more vehicle functions, for example, engine control. The diagnostic device can be part of this electronic control device or can be designed separately with respect to the electronic control device.

[0018] Here and below, an edge can be understood, in particular, as a signal edge of an analog signal, in particular of an analog sensor signal. An edge in an analog signal can be referred to, in particular, when the amplitude of the signal or the value of the signal amplitude changes by a predefined minimum value or more within a predefined time period, in particular in a monotonically rising or monotonically falling manner. Here, the predefined time period or the minimum value or their ratio to each other is to be determined depending on the specific application and, in particular, depending on the design of the sensor and the value range of the sensor signal.

[0019] The height of an edge can correspond to the amplitude or a change in the value of the amplitude over a predefined time period. For example, if a gradient signal of the sensor signal, i.e. a time-dependent signal corresponding to or approximately corresponding to the time derivative of the sensor signal, is considered, the height of an edge can be determined, inter alia, by integrating the gradient signal over a predefined time period. The gradient signal can be determined, for example, by a differentiator, in particular an analog differentiator circuit, by a digital evaluation circuit or another computing unit.

[0020] The cumulative height can be understood, for example, as corresponding to the sum of the determined heights of all first edges of a plurality of fault events or to the sum of all heights of the second edges of these fault events. The cumulative height can also correspond to the sum of all heights of all first and second edges of a plurality of fault events, but in this case either always positive values for these heights or always negative values for these heights should be used.

[0021] The cumulative height can likewise be determined, for example, by summing or integrating, for example, the gradient signal over the respective plurality of time periods.

[0022] The presence of a first and a second edge of the sensor signal can be considered a necessary but not sufficient criterion for the presence of a fault event. In particular, in addition to the mere presence of the respective edge, further conditions can be necessary in order to be able to identify a section of the sensor signal as a fault event. These further conditions can relate to the respective height of the edge, the absolute value of the sensor signal and / or the slope of the edge, etc.

[0023] For certain analog sensors, the sensor signal can take values, for example, in the range from 0 to a few volts, for example, up to 5 V or 10 V. For such sensors, a change in the value of the sensor signal of a few volts within a few tens of milliseconds or a few hundred milliseconds can be referred to as an edge, for example. This numerical example is only for the purpose of explaining the concept of an edge and should by no means be understood restrictively.

[0024] The second edge following the first edge can be understood, inter alia, as meaning that there is no further edge of the sensor signal between the first edge and the second edge. However, this does not necessarily mean that the second edge immediately follows the first edge. Rather, the amplitude or absolute value of the sensor signal can be more or less constant between the first and second edges or change without meeting the above-mentioned conditions for an edge.

[0025] Thus, by means of the improved design, the accumulated height is used as a measure for the severity of the plurality of fault events within the preset time interval. Thereby, it is taken into account that the mere number of fault events identified within the preset time interval alone is not sufficient to state whether an unstable fault has an impact on the subsequent functioning using the sensor signal in a significant or relevant manner. However, the accumulated height and the first threshold value which needs to be defined depending on the application case and the sensor design allows for a precise and individual determination of which fault feature within the preset time interval shall be considered as relevant. This enables, inter alia, a fault analysis to be performed for various different types of sensors or fields of use without adjusting the diagnostic device structure and without adjusting the method principle. If an unstable fault is critical, for example, to safety-relevant functions of a motor vehicle, for example, the engine control, then a fault report can be output, if necessary, at a lower accumulated height than, for example, in the case of pure comfort functions, for example, temperature regulation of the vehicle interior. The flexibility of the fault analysis is thus significantly increased.

[0026] The severity of the plurality of fault events does not necessarily have to be given by the accumulated height alone. Rather, in different embodiments, the diagnostic device can take into account the signal course of the sensor signal and / or further characteristics of the sensor signal. To this end, the diagnostic signal can be generated, for example, depending on the accumulated height and the further characteristics. Alternatively, a further diagnostic signal can also be generated depending on the further characteristics. The above and below explanations regarding the diagnostic signal can be transferred analogously to the further diagnostic signal.

[0027] For example, the value range of the sensor signal can be limited, for example, due to the specific design of the sensor. If the sensor signal reaches the respective upper or lower limit, then the height of the respective edge is likewise limited by this. However, for such events, the severity of the fault can depend on how long the sensor signal stays at the upper or lower limit. Therefore, in different embodiments, the diagnostic signal or the further diagnostic signal can be generated depending on the accumulated staying time of the sensor signal at the upper and / or lower limit. To this end, the sensor signal itself can be aggregated or integrated over a respective plurality of time periods.

[0028] Alternatively, as soon as the sensor signal stays at the upper or lower limit, the gradient signal can be modified such that it assumes a predefined, non-zero value, i.e. in particular the gradient signal is set to this value. In such an embodiment, the staying of the sensor signal at the upper or lower limit is effectively treated as a continued rising of the sensor signal, i.e. the height of the respective edge is artificially increased.

[0029] By taking into account the staying time of the sensor signal at the upper and lower limits, the severity of the fault is more precisely estimated, thereby increasing the reliability of the method.

[0030] In addition to generating and, if necessary, outputting the fault report, further measures, for example risk-reducing measures, can also be initiated by the diagnostic device or by other electronic control devices or other further computing units in dependence on the result of the comparison of the diagnostic signal with the first threshold value. For example, upon generation of the fault report, operating parameters or functional parameters of the motor vehicle can be adjusted or limited. This measure can also be initiated on the basis of the fault report. Alternatively or additionally, fault information can be output to the driver or user of the motor vehicle, for example in the form of sound, vision and / or haptics. Alternatively or additionally, the diagnostic device can also store an entry in a fault memory in dependence on the fault report.

[0031] According to at least one embodiment of the method according to the improvement, the counter value is changed by the diagnostic device in dependence on the result of the comparison of the diagnostic signal with the first threshold value. The changed counter value is compared by the diagnostic device with a second threshold value, and the fault report is generated by the diagnostic device in dependence on the result of the comparison of the changed counter value with the second threshold value.

[0032] In particular, the counter value is increased in dependence on the result of the comparison of the accumulated height with the first threshold value, in particular by a predefined increment, or the counter value is not changed. For example, the counter value is increased if the accumulated height is greater than or equal to the first threshold value, and the counter value is not changed otherwise.

[0033] For example, the steps described above with respect to the preset time interval, i.e. the identification of a plurality of fault events, the generation of the diagnostic signal, the comparison of the diagnostic signal with the first threshold value, are repeated for one or more further time intervals.

[0034] Here, the diagnostic signal can be reset, for example, for each repetition. Alternatively, an inertia time (original text: Nachlaufzeit) or a decay time can also be defined, in which the diagnostic signal is continuously or stepwise restored to its original value, in particular zero. The inertia time or the decay time can correspond, for example, to the duration of one or more time intervals. In this way, a certain degree of fault memory can be achieved, so that fault events from previous time intervals can be taken into account in part when evaluating the severity of the fault.

[0035] Thus, the diagnostic signal reaching or exceeding the first threshold value during a time interval is a necessary condition for generating the fault report, but not a sufficient condition. Rather, reaching or exceeding the first threshold value during a single time interval can be interpreted as a corresponding unstable fault being identified in this time interval. However, in order to evaluate or estimate the relevance or criticality of the unstable fault, it can be advantageous or necessary to know how often the corresponding unstable fault occurs or has occurred, depending on the application.

[0036] By dividing the analysis into a first step in which individual time intervals are diagnosed and a second step in which the counter value is compared to the second threshold value, on the one hand it is made possible to take into account various different fault patterns or fault characteristics within one time interval, but on the other hand it is made possible to achieve that the output of the fault report depends on the frequency or the duration with which such a fault occurs, independently of which exact characteristic or which exact fault pattern is present in the individual time intervals.

[0037] The fact that the diagnostic signal reaches or exceeds the first threshold value in a single time interval can be due to the fact that the sensor signal has one very high edge or a few relatively high edges, but also due to the fact that the sensor signal has a plurality of edges which are relatively low in height in the time interval. In both cases, if the first threshold value is exceeded, an unstable fault can be identified in the respective time interval. Subsequently, by comparing the counter value to the second threshold value, it is achieved that individual or a few unstable faults do not lead to the generation of a fault report, since in certain cases there is no significant impairment of the associated function. The availability of the sensor arrangement or function is thereby increased.

[0038] If instead of the two-step check a long time interval is considered and the second step is dispensed with, only the value of the diagnostic signal itself can be used for the evaluation. The number of times that a certain threshold value is exceeded is not taken into account. In contrast, according to the above-described embodiments of the improved design, the fact that the first threshold value is exceeded during a time interval has the same result, independently of how much or how long the first threshold value is exceeded. The relevance of the number of unstable faults therefore has a greater weight than the actual height of the edges.

[0039] According to at least one embodiment, the start point in time and the end point in time of the time interval are preset, or are predetermined.

[0040] This can be achieved, for example, by an explicit presetting of the start and end point in time, or by a presetting of the start point in time and the duration of the time interval or a presetting of the duration of the time interval and the end point in time. The respective repeatedly different time intervals are therefore in particular directly consecutive time intervals, wherein the temporal position of the time intervals is independent of the course of the sensor signal itself. The advantage of this embodiment is that the time intervals can be particularly simply preset.

[0041] According to at least one embodiment, the duration of the time interval is preset, and the start point in time of the time interval corresponds to the first edge of the first fault event of the plurality of fault events.

[0042] In other words, the time interval only starts running when the first edge of a fault event is identified for the first time.

[0043] According to at least one embodiment, all fault events within the time interval are identified by the diagnostic device, wherein the sensor signal has a first edge and a second edge following the first edge for each of the all fault events. The plurality of fault events corresponds here to a subset of the all fault events.

[0044] According to at least one embodiment, for each of the all fault events, one of at least two preset fault types is determined by the diagnostic device from the respective first edge and the respective second edge. The plurality of fault events corresponds to a subset of the all fault events, wherein the same fault type is determined for all fault events of the plurality of fault events or the subset.

[0045] In other words, a single fault type of the two or more preset fault types can be considered and analyzed as described above. This has the advantage that, in the evaluation, different, possibly simultaneously occurring fault types are not mixed, since they can be caused by different causes in certain cases. The reliability and reproducibility of the fault analysis is thereby improved.

[0046] For example, a respective diagnostic signal can be generated by the diagnostic device for each of the at least two preset fault types, for which the above and below explanations apply analogously.

[0047] The at least two fault types can comprise, for example, a fault type corresponding to an open circuit of the signal output of the sensor. Here, an open circuit can be understood as a state in which the electrical connection of the measurement connection to the signal connection is interrupted, i.e. is not connected to a certain extent (English: "floating"). This state can also be referred to as "Open Circuit". The two or more fault types can also comprise one or more fault types each corresponding to a short circuit of the signal output. Depending on the embodiment, different fault types can be specified for short circuits with different reference points or reference potentials, or a common fault type can be specified for short circuits.

[0048] According to at least one embodiment, the gradient signal is generated by the diagnostic device, in particular a differentiator, from the sensor signal. For each of the all fault events, a respective first edge is assigned, in particular correspondingly configured, by the diagnostic device to one of at least two preset edge categories, and a respective second edge is assigned to one of the at least two edge categories. The fault type is determined by the diagnostic device from the edge category assigned to the first edge and the edge category assigned to the second edge.

[0049] Thus, in this embodiment, two successive edges of a fault event are analyzed according to the gradient signal, i.e. in particular it is analyzed how steeply the respective edge rises or falls, in order to determine a respective edge category. Subsequently, a fault type is determined on the basis of these two edge categories, e.g. on the basis of the mutual order of the different edge categories. Thereby, the method is in particular suitable for identifying and characterizing unstable faults, such as intermittent short circuits or open circuits or other periodic contact releases or contact malfunctions.

[0050] Since for determining the fault type only sensor signals are processed and evaluated which are inherently generated and used for the underlying application, there is no need to additionally generate dedicated analysis signals and to evaluate the signal response of the sensor system to these analysis signals, for example. Thus, this method can be used widely without fundamental changes for a large number of analog sensors or for any analog sensor. For this purpose, only parameters, limits, etc. have to be adjusted, if necessary.

[0051] According to at least one embodiment, the first edge is assigned an edge category on the basis of the gradient signal during a first time period and the second edge is assigned a respective edge category on the basis of the gradient signal during a second time period. Here, the first time period corresponds to the time period of the first edge and the second time period corresponds to the time period of the second edge, wherein these time periods refer to the respective time periods in which the respective edges occur in the sensor signal.

[0052] According to at least one embodiment, determining the fault type comprises generating an output signal which encodes the fault type according to the edge category assigned to the first edge and according to the edge category assigned to the second edge. Alternatively or additionally, determining the fault type can comprise storing information about the occurrence of the respective fault type. This can be done, for example, according to one or more respective counters or according to other storage methods.

[0053] According to at least one embodiment, the at least two edge categories comprise a first edge category and a second edge category. For example, the first and second edge categories can each correspond to the presence of a fault, i.e. the first and second edge categories can not occur in a fault-free state of the sensor or in a fault-free state of the connection of the signal output to the measurement connection.

[0054] According to at least one embodiment, the at least two edge categories comprise a third edge category and / or a fourth edge category. For example, the third and / or fourth edge category can each correspond to the presence of a fault.

[0055] According to at least one embodiment, the at least two fault types comprise an intermittent short circuit of the signal output of the sensor arrangement to a reference potential connection and / or an intermittent open circuit of the signal output.

[0056] According to at least one embodiment, by the diagnostic device, for each of the total of fault events, only when the gradient signal exceeds a predefined positive first gradient limit value during the respective first edge, the respective first edge is assigned to a first edge category of the at least two edge categories.

[0057] According to at least one embodiment, by the diagnostic device, for each of the total of fault events, only when the gradient signal exceeds the first gradient limit value during the respective second edge, the respective second edge is assigned to the first edge category.

[0058] Here, the expressions "during the first edge" or "during the second edge" can be understood as meaning the course during the respective time period corresponding to the respective edge, during which the edge is present in the sensor signal. The gradient signal exceeding the gradient limit value can be understood in particular as meaning that the gradient signal initially is less than or equal to the gradient limit value and subsequently has (or exhibits) a value greater than the gradient limit value. The gradient signal being below the gradient limit value can be understood in particular as meaning that the gradient signal initially is greater than or equal to the gradient limit value and subsequently has a value less than the gradient limit value. The sensor signal exceeding or being below the signal limit value can likewise be understood analogously.

[0059] In other words, the first edge category is present when the edge in question corresponds to a rising edge (or rising flank) of the sensor signal and this rising at least temporarily has an abruptness exceeding the abruptness defined by the first gradient limit value. Such an edge can also be referred to as a jump-like rising of the sensor signal.

[0060] The value range of the amplitude of the sensor signal can be defined and limited in particular by a first reference potential and a second reference potential. Here, the first reference potential can also be referred to as an upper reference potential and the second reference potential can be referred to as a lower reference potential, wherein the lower reference potential is less than the upper reference potential. For example, the first reference potential can be a positive potential and the second reference potential can be a negative reference potential or a zero potential or a ground potential. But other definitions are also possible. In particular, the upper reference potential can also correspond to a zero potential or a ground potential and the lower reference potential can accordingly be negative.

[0061] The first edge category, i.e. the jump-like rising of the sensor signal, occurs in particular when a short circuit of the signal output of the sensor to the upper reference potential is formed. Furthermore, the first edge category occurs for example when a short circuit of the signal output to the lower reference potential is eliminated.

[0062] Thus, by analyzing the first and second edges with respect to the first edge category a conclusion for classifying or determining a fault type can be drawn. According to at least one embodiment, by the diagnostic device, for each of the total fault events, the first edge is assigned to a second edge category of the at least two edge categories only if the gradient signal is below a negative second gradient limit value during the first edge.

[0063] According to at least one embodiment, by the diagnostic device, for each of the total fault events, the second edge is assigned to the second edge category only if the gradient signal is below the second gradient limit value during the second edge.

[0064] Thus, the second edge category in particular corresponds to a falling edge, or in other words a falling flank, of the sensor signal, wherein the falling at least temporarily has a steepness, in particular a negative steepness, which has a value greater than a value of a slope defined by the second gradient limit value. Such an edge can for example also be referred to as a jump-like falling of the sensor signal.

[0065] Such a jump-like falling of the sensor signal occurs for example when a short circuit of the signal output to the upper reference potential is eliminated or when a short circuit of the signal output to the lower reference potential is formed. Furthermore, a jump-like falling of the sensor signal can also occur when an open circuit of the signal output is ended or eliminated. Thus, according to the second edge category, in particular in combination with the first edge category, a fault type can be determined more specifically.

[0066] According to at least one embodiment, by the diagnostic device, for each of the total fault events, the respective first edge is assigned to the first edge category only if the sensor signal exceeds a predefined first signal limit value during the respective first edge, and / or the respective second edge is assigned to the first edge category only if the sensor signal exceeds the first signal limit value during the respective second edge.

[0067] According to at least one embodiment, by the diagnostic device, for each of the total fault events, the first edge is assigned to a second edge category of the at least two edge categories only if the gradient signal is below a negative second gradient limit value during the first edge.

[0068] Here, the value of the first signal limit is in particular greater than the value of the second signal limit. Thus, in such an embodiment, in addition to the gradient signal exceeding the first gradient limit or in addition to the gradient signal falling below the second gradient limit, a further condition is respectively preset in order to assign the edge to the first or second edge category. Thus, in such an embodiment, it is not sufficient that the edge rises or falls steeply enough, but it is also important whether the respective edge also exceeds or falls below the respective value of the sensor signal itself.

[0069] Thereby, it is possible, inter alia, to distinguish an actual fault, such as a short circuit or an open circuit, from other reasons which can lead to a steep rise or fall of the sensor signal. The latter can arise, for example, due to a very fast value of the underlying physical measured variable, without a fault being present. In connection therewith, this is sometimes also referred to as a so-called load shock. But a fault-free change of the physical measured variable with a high rate of change, i.e. with a steep rise or fall of the sensor signal, usually occurs within certain limits which are smaller than the entire value range of the sensor signal. Thus, by respectively adapting the selection of the first and / or second signal limit, it is possible to reliably distinguish between a fault and other reasons for a change of the sensor signal. In other words, the risk of a false-positive fault determination can be reduced.

[0070] According to at least one embodiment, by means of the diagnostic device, for each fault event of all fault events, the respective first edge is assigned to the first edge category just when the gradient signal exceeds the first gradient limit during the respective first edge and the sensor signal exceeds the first signal limit during the first edge.

[0071] According to at least one embodiment, by means of the diagnostic device, for each fault event of all fault events, the respective second edge is assigned to the first edge category just when the gradient signal exceeds the first gradient limit during the respective second edge and the sensor signal exceeds the first signal limit during the second edge.

[0072] According to at least one embodiment, by means of the diagnostic device, for each fault event of all fault events, the respective first edge is assigned to the second edge category just when the gradient signal falls below the second gradient limit during the respective first edge and the sensor signal falls below the second signal limit during the first edge.

[0073] According to at least one embodiment, by means of the diagnostic device, for each fault event of all fault events, the respective second edge is assigned to the second edge category just when the gradient signal falls below the second gradient limit during the respective second edge and the sensor signal falls below the second signal limit during the second edge.

[0074] According to at least one embodiment, the fault type is determined as an intermittent short circuit of the signal output of the sensor to the first reference potential connection, if, in particular exactly if, the respective first edge is assigned to the first edge category and the respective second edge is assigned to the second edge category, wherein the first reference potential connection is in particular at the first reference potential.

[0075] In this case, in other words, a jump-like rise of the sensor signal occurs first and a jump-like fall of the sensor signal occurs thereafter. If a short circuit to the first reference potential connection is formed, the sensor signal assumes a value close to the first reference potential for a short time. If this short circuit is then eliminated again, the sensor signal assumes the original value again or at least approximately the other value which truly corresponds to the physical measured variable, so that the described edge sequence results. In this way, a reliable detection and classification of the presence of an intermittent short circuit to the first reference potential connection is possible.

[0076] According to at least one embodiment, the fault type is determined as an intermittent short circuit of the signal output to the second reference potential connection, if the first edge is assigned to the second edge category and the second edge is assigned to the first edge category, wherein the first reference potential connection and the second reference potential connection are at different electrical reference potentials. In particular, the second reference potential connection is at the second reference potential.

[0077] In this case, in other words, a jump-like fall of the sensor signal occurs first and a jump-like rise of the sensor signal occurs thereafter. If a short circuit to the second reference potential connection is formed, the sensor signal assumes a value close to the second reference potential for a short time. If this short circuit is then eliminated again, the sensor signal assumes the original value again or at least approximately the other value which truly corresponds to the physical measured variable, so that the described edge sequence results.

[0078] According to at least one embodiment, by means of the diagnostic device, for each fault event of the total of fault events, the respective first edge is assigned to a third edge category of the at least two edge categories only, in particular exactly when the gradient signal exceeds a predefined positive third gradient limit value during the respective first edge and does not exceed the first gradient limit value.

[0079] Here, the third gradient limit value is in particular smaller than the first gradient limit value. In other words, in this case a signal rise of the sensor signal occurs which is not as steep as in the case of a jump-like rise according to the above definition, but which is at least as steep as required by the third gradient limit value. According to such an embodiment, it is possible to distinguish between a jump-like rise of the sensor signal and other fault-related features of the sensor signal in the improved design described. Thereby, the fault type can be determined more specifically.

[0080] For example, an edge of the third edge category can occur when an open circuit of the signal output occurs. In this case, the signal output, which is measured as the sensor signal, can be charged to the first reference potential, for example, by a set or parasitic ohmic resistor and / or capacitor, thus producing a slower signal rise than the aforementioned jump-like rise of the sensor signal. It is thereby possible in the respective embodiment to reliably distinguish between a short circuit of the signal output to the first reference potential and an open circuit of the signal output.

[0081] According to at least one embodiment, if the respective first edge is assigned to the third edge category and the respective second edge is assigned to the second edge category, the fault type is determined to be an intermittent open circuit of the signal output.

[0082] According to at least one embodiment, by means of the diagnostic device, for each fault event of the total number of fault events, the respective first edge is assigned to the third edge category only, in particular exactly, when the gradient signal is below a negative fourth gradient limit value and not below the second gradient limit value during the respective first edge.

[0083] In such an embodiment, the above explanations regarding the third edge category apply analogously, wherein a polarity reversal of the reference potential connection is assumed.

[0084] According to at least one embodiment, if the respective first edge is assigned to the third edge category and the respective second edge is assigned to the first edge category, the fault type is determined to be an open circuit of the signal output.

[0085] According to the improved design, a diagnostic device for fault analysis of a sensor arrangement with respect to unstable faults is also provided. The sensor arrangement is provided for outputting an analog sensor signal as a function of a measured variable. The diagnostic device has an evaluation unit, in particular an analysis unit, which is provided for identifying a plurality of fault events within a predetermined time interval, wherein the sensor signal has, for each fault event, a first edge and a second edge following the first edge. The diagnostic device has a cumulater unit or an integrator unit, which is provided for generating a diagnostic signal as a function of a cumulative height of the first edge and / or a cumulative height of the second edge. The diagnostic device has a diagnostic unit, which is provided for comparing the diagnostic signal to a first threshold value and for generating a fault report as a function of the comparison result.

[0086] Here, the sensor arrangement is not necessarily part of the diagnostic device. The diagnostic device can have, for example, connection terminals for connecting the sensor arrangement or the sensor, in order to connect, in particular, the signal output of the sensor to the diagnostic device.

[0087] According to at least one embodiment, the diagnostic device has a differentiator which is designed to generate a gradient signal from the sensor signal.

[0088] According to at least one embodiment, the accumulator unit or integrator unit is designed to generate the diagnostic signal from the gradient signal.

[0089] Further embodiments of the diagnostic device according to the improved design are generated directly from the various design options of the method according to the improved design and vice versa. In particular, the diagnostic device according to the improved design can be designed or programmed to implement the method according to the improved design or the diagnostic device implements such a method.

[0090] According to the improved design, a sensor device is also provided. Here, the sensor device has an analog sensor arrangement or an analog sensor and a diagnostic device according to the improved design, wherein the analog sensor arrangement and / or the analog sensor is coupled or connected with the diagnostic device, wherein the connection can be designed in particular as a direct connection or as an indirect connection via one or more further components of the sensor arrangement.

[0091] According to at least one embodiment of the sensor device, the sensor is designed as a pressure sensor, for example as an intake pipe pressure sensor for a motor vehicle, or as a temperature sensor, for example as an intake air temperature sensor for a motor vehicle.

[0092] According to at least one embodiment, the sensor device or the diagnostic device comprises a low-pass filter, for example an RC element. The low-pass filter is designed to generate a filtered output signal as a sensor signal from the output signal of the sensor. The low-pass filter can be arranged for example between the analog sensor and the diagnostic device.

[0093] According to the improved design, a motor vehicle is also provided, which has a diagnostic device and / or a sensor device according to the improved design.

[0094] According to at least one embodiment of the motor vehicle according to the improved design, the motor vehicle has an engine control device and the engine control device comprises the diagnostic device or the evaluation unit of the diagnostic device.

[0095] The invention also comprises combinations of features of the described embodiments.

[0096] In the drawings:

[0097] Figure 1 schematic diagram showing exemplary embodiments of a diagnostic device and of a sensor device according to the improved design;

[0098] Figure 2 diagnostic device and sensor arrangement of Figure 1 a first fault type;

[0099] Figure 3 diagnostic device and sensor arrangement of Figure 2 a first fault type;

[0100] Figure 4 diagnostic device and sensor arrangement of Figure 1 a second fault type;

[0101] Figure 5 diagnostic device and sensor arrangement of Figure 4 a second fault type;

[0102] Figure 6 diagnostic device and sensor arrangement of Figure 1 a third fault type;

[0103] Figure 7 diagnostic device and sensor arrangement of Figure 6 a third fault type; and

[0104] Figure 8 diagnostic device and sensor arrangement of

[0105] The embodiments set forth below represent the presently preferred embodiments of the application. In these embodiments, the components described are presented with reference to their function or their combination in the present application. As such, the component configurations can differ depending on the desired implementation. For example, the same component can be used to achieve substantially the same results in different embodiments. Furthermore, not all of the components described are required, and some components can be utilized independently, while some others can be combined, in either other components or in a different manner. Additionally, the description of the embodiments is provided as an example. Numerous other embodiments will be apparent to those skilled in the art, and are intended to be included within the scope of the application.

[0106] In the drawings, like reference numerals refer to like elements throughout.

[0107] An exemplary embodiment of a diagnostic device 1 according to the improved design is schematically shown in Figure 1 . Further, a sensor arrangement comprising an analog sensor 2 is shown. The diagnostic device 1 and the sensor arrangement can be, for example, part of a sensor arrangement according to the improved design. The sensor 2 has a signal output 2' at which the sensor can output an analog output signal S', in particular a sensor voltage. The output signal S' can be provided at a sensor connection 4 of the sensor arrangement or the diagnostic device 1, which is connected with the signal output 2' by one or more electrical lines 3.

[0108] InFigure 1 The sensor arrangement also shows a low-pass filter 8, which is shown schematically as an RC element with a resistor 8a and a capacitor 8b. The capacitor 8b is connected here with one connection to the second reference potential connection 10' and with the other connection to one connection of the resistor 8a. The other connection of the resistor 8a is coupled, in particular via a pull-up resistor 9, to the first reference potential connection 10. At the output of the low-pass filter 8, a filtered output signal can accordingly be output, which can be regarded as a sensor signal S. The output signal S' can also be regarded as a sensor signal in alternative embodiments.

[0109] The low-pass filter 8 and / or the pull-up resistor 9 can be part of a control device for a motor vehicle, for example. It should be noted that other embodiments of the low-pass filter 8 can also be provided. In the fault-free operation of the sensor 2, the sensor provides the output signal S' via the signal connection 4 to the low-pass filter 8, which accordingly generates the sensor signal S.

[0110] The diagnostic device 1 has, for example, a differentiator 5, which is connected to the low-pass filter 8 in order to receive the sensor signal S. The differentiator 5 is provided for differentiating the sensor signal S and generating a gradient signal G by the differentiation. The diagnostic device 1 also has an evaluation unit 6, which is connected to the low-pass filter 8 in order to receive the sensor signal S and which is connected to the output of the differentiator 5 in order to receive the gradient signal G.

[0111] Furthermore, the diagnostic device 1 has an integrator 7, which is connected to the output of the differentiator 5 in order to receive the gradient signal G and which can be actuated by the evaluation unit 6, for example. The integrator 7 can generate a diagnostic signal D on the basis of the gradient signal. The diagnostic device 1 also has a diagnostic unit 11, which is connected to the integrator 7 in order to receive the diagnostic signal D and which can generate a fault report F from the diagnostic signal D.

[0112] The first reference potential connection 10 is at a first potential, which is positive, for example. The reference potential connection 10 can correspond to the output potential of a voltage source, for example. The second reference potential connection 10' is at a second potential, for example a ground potential.

[0113] Reference will be made in the following to Figures 2 to 8 The working principle of the diagnostic device 1 or of the sensor arrangement will be explained in more detail depending on various different cases.

[0114] To this end, in Figure 8A flow chart of an exemplary embodiment of a method for fault analysis of a sensor arrangement with respect to instability faults according to the improved design is schematically shown in FIG. In step S1 of the method, diagnostic device 1, in particular evaluation unit 6, identifies a plurality of fault events within a predetermined time interval based on sensor signal S. Each fault event has a first edge and a second edge following the first edge. Integrator 7 receives gradient signal G and is controlled by evaluation unit 6 such that, based on gradient signal G, it generates diagnostic signal D, which corresponds to the cumulative height of the second edges or the cumulative height of the first edges of the fault events.

[0115] In step S2, diagnostic unit 11 compares diagnostic signal D with a predefined first threshold value SW1. If diagnostic signal D is greater than or equal to first threshold value SW1, diagnostic unit 11 increases a counter by a corresponding increment. In step S3, diagnostic unit 11 compares the increased counter value with a predefined second threshold value SW2. Based on the result of the comparison with second threshold value SW2, diagnostic unit 11 generates a fault message F. In particular, diagnostic unit 11 generates fault message F when the counter value or the increased counter value is greater than or equal to second threshold value SW2.

[0116] Thus, in step S2 it is checked whether a relevant fault occurs during a single time interval, and in step S3 it is determined whether the severity of the fault justifies generating a fault message F. The severity of the fault is understood to be the severity of the effect of the respective faulty component on the system.

[0117] exist Figure 2 Shown in Figure 1 Sensor arrangement and diagnostic device 1 in which there is an intermittent opening of the signal output 2', as indicated by two parallel lines between the signal output 2' and the sensor connection 4. Figure 3 A corresponding signal diagram is shown in , which schematically illustrates the sensor signal S, the gradient signal G and the diagnostic signal D.

[0118] exist Figure 3 , three consecutive fault events are shown, each of which is characterized by a first and a second consecutive edge of the sensor signal S. The rise of the sensor signal S during the first edge of the fault event is not as steep as, for example, in the case of a sudden rise of the sensor signal S. This is because, as in Figure 2As can be seen, the capacitor 8b of the low-pass filter 8 is delayed in its occurrence by the coupling with the first reference potential connection 10 via the resistor 8a and the pull-up resistor 9. The first edge of the sensor signal S is followed by a step-like drop as second edge, respectively. The different edge categories can also be identified from the course of the gradient signal G. The value of the gradient signal G thus remains relatively small, while a spike occurs during the second edge.

[0119] The diagnostic signal D corresponds to the output of the integrator 7, which integrates the gradient signal G under the control of the evaluation unit 6, for example during the second edge. For other fault types, in particular a short circuit of the signal output 2' to one of the reference potential connections 10, 10', a corresponding diagnostic signal can be generated.

[0120] In Figure 4 the diagnostic device 1 or the sensor arrangement is shown, in which an intermittent short circuit between the signal output 2' of the sensor 2 and the first reference potential connection 10 is shown by a dash connection, or rather a dashed connection. In this case, the current flows from the first reference potential connection 10 via the sensor connection 4 into the low-pass filter 8.

[0121] In Figure 5 the relevant sensor signal S and the relevant gradient signal G are shown schematically as a function of time. The first edge of the sensor signal S is reflected in the gradient signal G as a positive signal pulse, and the second edge as a negative signal pulse. The described short circuit to the reference potential connection 10 is formed at the beginning of the first edge. The sensor signal S thus shows a step-like rise. At the beginning of the second edge, the short circuit is again eliminated, so the sensor signal S shows a step-like drop of approximately the same magnitude. The duration of the step-like rise or step-like drop can be in the order of magnitude of a few milliseconds, for example less than 5 ms. The maximum or minimum value of the gradient signal G can have a value in the range of several hundred V / s or several thousand V / s.

[0122] In Figure 6 the diagnostic device 1 and the sensor arrangement are shown, which corresponds to an intermittent short circuit of the signal output 2' to the second reference potential connection 10'. In Figure 7 the relevant signal diagram is shown. Here, the signal course is to some extent contrary to Figure 4 and Figure 5 the case. Thus, first a short circuit is formed, which is accompanied by a step-like drop of the sensor signal S, then the short circuit is again eliminated, which causes a corresponding step-like rise of the sensor signal S. This characteristic and the intermittent short circuit of the signal output 2' to the second reference potential connection 10' can be viewed similarly to the above.

[0123] As described especially with respect to the figures, according to the improved design it is possible to estimate and accordingly take into account the severity or relevance of intermittent faults.

[0124] According to different embodiments it is possible to determine the presence of intermittent faults of an analog sensor particularly reliably and to distinguish and evaluate different faults reliably. Thereby it is possible, among other things, to effectively monitor so-called dither signals, which are detected as repeated rising and falling of the sensor signal. Thereby it is possible, among other things, to meet regulatory requirements for monitoring such signals in motor vehicles.

[0125] In different embodiments two edges of the sensor signal are analyzed, thus avoiding false diagnoses, for example due to load impacts or the like.

[0126] The improved design can also be applied to other analog electrical or electronic components having an analog output signal, which are not sensors. These analog electronic components can be designed, for example, as control units. Then in different embodiments the sensor signal should be replaced by an analog output signal of an analog electronic component.

[0127] List of reference signs

[0128] 1 diagnosis device

[0129] 2 sensor

[0130] 2' signal output

[0131] 3 line

[0132] 4 sensor connection

[0133] 5 differentiator

[0134] 6 evaluation unit

[0135] 7 integrator

[0136] 8 low-pass filter

[0137] 8a resistor

[0138] 8b capacitor

[0139] 9 pull-up resistor

[0140] 10, 10' reference potential connection

[0141] 11 diagnosis unit

[0142] S sensor signal

[0143] S' output signal

[0144] G gradient signal

[0145] D diagnostic signal

[0146] SW1, SW2 threshold value

[0147] F failure report

[0148] S1 to S3 method steps

Claims

1. A method for fault analysis of a sensor arrangement (2, 8, 9) with respect to unstable faults, wherein: The sensor arrangement (2, 8, 9) outputs an analog sensor signal (S) as a function of the measured variable, It is characterized in that - detecting a plurality of fault events within a predetermined time interval by means of the diagnostic device (1), wherein the sensor signal (S) has, for each fault event, a first edge and a second edge following the first edge; - generating a diagnostic signal (D) by the diagnostic device (1) based on the cumulative height of the first edge and / or the cumulative height of the second edge; - comparing the diagnostic signal (D) with a first threshold value (SW1) by means of a diagnostic device (1), wherein the cumulative height is used as a measure of the severity of a plurality of fault events within a preset time interval; and - generating a fault message (F) by the diagnostic device (1) based on the comparison result.

2. The method according to claim 1, characterized in that - changing the counter value by the diagnostic device (1) according to the comparison result of the diagnostic signal (D) and the first threshold value (SW1); - comparing the changed counter value with a second threshold value (SW2) by the diagnostic device (1); and The fault message (F) is generated as a result of a comparison of the changed counter value with a second threshold value (SW2).

3. The method according to any one of the preceding claims, characterized in that - presetting the start time point and the end time point of the time interval; or - A duration of the time interval is preset, and a starting time point of the time interval corresponds to a first edge of a first fault event among the multiple fault events.

4. The method according to any one of the preceding claims, characterized in that - detecting all fault events within the time interval by means of the diagnostic device (1), wherein the sensor signal (S) has, for each of all fault events, a first edge and a second edge following the first edge; - for each fault event in all fault events, determining one fault type from at least two predefined fault types by the diagnostic device (1) based on the respective first edge and the respective second edge; and The plurality of fault events corresponds to a subset of all fault events, wherein a same fault type is determined for all fault events of the plurality of fault events.

5. The method according to claim 4, characterized in that - generating a gradient signal (G) based on the sensor signal (S); - For each fault event among all fault events, the diagnostic device (1) - assigning one edge category of at least two predefined edge categories to the respective first edge; - assigning one of the at least two edge categories to the respective second edge; and The fault type is determined as a function of the edge class assigned to the first edge and the edge class assigned to the second edge.

6. The method according to claim 4 or 5, characterized in that The at least two fault types include an intermittent short circuit between a signal output (2') of the sensor arrangement (2, 8, 9) and a reference potential connection (10, 10') and / or an intermittent open circuit of the signal output (2').

7. The method according to any one of claims 4 to 6, characterized in that For each fault event in all fault events, the diagnostic device (1) - assigning the respective first edge to the first edge class of the at least two edge classes only if the gradient signal (G) exceeds a positive first gradient limit value during the respective first edge; or - the respective second edge is assigned to the first edge class only if the gradient signal (G) exceeds the first gradient limit value during the respective second edge.

8. The method according to claim 7, characterized in that For each fault event in all fault events, the diagnostic device (1) - assigning the respective first edge to the second edge class of the at least two edge classes only if the gradient signal (G) falls below a negative second gradient limit value (GG2) during the respective first edge; or - assigning the respective second edge to the second edge class only if the gradient signal (G) falls below the second gradient limit (GG2) during the respective second edge.

9. The method according to any one of the preceding claims, characterized in that The diagnostic signal (D) is generated based on a cumulative dwell time of the sensor signal (S) at a preset upper limit of the sensor signal (S) and / or at a preset lower limit of the sensor signal (S).

10. A diagnostic device for fault analysis of a sensor arrangement (2, 8, 9) with respect to unstable faults, wherein: The sensor arrangement (2, 8, 9) is configured to output an analog sensor signal (S) as a function of a measured variable. It is characterized in that The diagnostic device (1) has an evaluation unit (6) which is configured to detect a plurality of fault events within a predetermined time interval, wherein the sensor signal (S) has a first edge and a second edge following the first edge for each fault event; The diagnostic device (1) comprises an accumulator unit or an integrator unit (7) which is configured to generate a diagnostic signal (D) as a function of the accumulated height of the first edge and / or the accumulated height of the second edge; and The diagnostic device (1) comprises a diagnostic unit (11) which is configured to compare the diagnostic signal (D) with a first threshold value (SW1), wherein the cumulative height is used as a measure of the severity of a plurality of fault events within a predetermined time interval, and to generate a fault message (F) as a function of the comparison result.

Images