Procedures for identifying and prioritizing monitoring units to be replaced

Abstract

Procedure for identifying and prioritizing monitoring units (22, 28, 30, 32) to be exchanged in the case of existing interactions between the monitoring units (22, 28, 30, 32) comprising the following steps: - Recording and storing driving parameters over a defined number of driving cycles (N) of a vehicle (10), - Acquisition and storage of diagnostic parameters of each monitoring unit (22, 28, 30, 32) from a diagnostic statistic (33) of the monitoring units (22, 28, 30, 32) over the defined number of driving cycles (N) of the vehicle (10) to determine a damage state (SOD) of the respective monitoring unit (22, 28, 30, 32), - Determining and storing an expected increase in the emitted mass as a function of a distance traveled of at least one defined emission component based on the driving parameters, the damage state (SOD) and an aging factor (AF) of the monitoring units (22, 28, 30, 32) of the vehicle (10), - Establishing an initial prioritization list (45) depending on the respective damage status (SOD) of the monitoring units (22, 28, 30, 32), - Storing an interaction matrix (46) which specifies whether a damage state (SOD) of one of the monitoring units (22, 28, 30, 32) affects a damage state (SOD) of another monitoring unit (22, 28, 30, 32), - Storing a correction factor (CF) with which, in the event of such an interaction and exceeding a threshold value of the damage state (SOD) of the influencing monitoring unit (22, 28, 30, 32), the determined expected increase in the emitted mass of the influencing monitoring unit (22, 28, 30, 32) is corrected as a function of the distance traveled, - Creating a second prioritization list (50) depending on the size of the expected increases in the mass emitted for the monitoring units (22, 28, 30, 32) corrected by the correction factor (CF) depending on the distance traveled, depending on which an exchange of the monitoring units (22, 28, 30, 32) takes place.

Description

[0001] The invention relates to a method for identifying and prioritizing monitoring units to be replaced when there are interactions between the monitoring units.

[0002] Modern combustion engines in vehicles have a multitude of monitoring units and sensors that control optimal combustion and subsequent exhaust aftertreatment. Faulty monitoring units and sensors therefore lead to errors in the engine control unit and consequently to increased exhaust emissions and suboptimal combustion.

[0003] For this reason, modern vehicles are equipped with on-board diagnostic systems designed to identify faulty sensors or monitoring units. These diagnostics include, for example, checking the plausibility of measured values, i.e., whether the monitoring unit's reading is within a permissible range for the respective operating state, whether corresponding readings from the monitoring units are consistent, and whether measured increases in these values ​​are within a realistically possible range. These monitoring units serve both system and component diagnostics. These tests are intended to ensure the correct operation of the engine, thereby preventing engine damage and ensuring compliance with legally permitted limits. If corresponding faults are detected, so-called emergency running programs are activated to prevent consequential damage.

[0004] Furthermore, on-board diagnostics are used to simplify maintenance by identifying faulty monitoring units through reading the engine control units in the workshop.

[0005] However, it has become apparent that monitoring units are frequently replaced in workshops even when they are not faulty, necessitating the replacement of further units until the correct fault is identified. The reason for such inaccurate replacements lies primarily in the fact that the individual monitoring units can sometimes interfere with each other, causing one sensor to be flagged as faulty when its reading is actually incorrect due to a fault in another sensor. Examples include the influence of the oxygen sensor or the lambda sensor on the fuel injection system and the detection of misfires.

[0006] Methods for identifying and prioritizing monitoring units to be exchanged in the presence of interactions between the monitoring units are known from DE 10 2011 015 396 A1, DE 43 26 498 A1 and EP 3 158 181 B1.

[0007] The challenge, therefore, is to create a procedure for identifying and prioritizing monitoring units requiring replacement when interactions exist between them. This procedure should significantly increase the probability of correctly identifying and replacing a genuinely defective monitoring unit. Conversely, the replacement of sensors or monitoring units that are still functioning correctly should be avoided as much as possible.

[0008] This problem is solved by a method for identifying and prioritizing monitoring units to be exchanged in the presence of interactions between the monitoring units, with the features of main claim 1.

[0009] In the inventive method for identifying and prioritizing monitoring units requiring replacement in the presence of interactions between them, driving parameters of a vehicle are first recorded and stored over a defined number of driving cycles. This is typically done in the vehicle's engine control unit. Additionally, diagnostic parameters of each monitoring unit are also recorded and stored from the diagnostic statistics of the monitoring units over the defined number of vehicle driving cycles, and a damage status of the respective monitoring unit is determined from this data. The monitoring units can monitor individual components, entire systems, or simply existing parameters. This is done during on-board monitoring of the vehicle.The following process determines and stores the expected change in the emitted mass of at least one defined emission component as a function of distance traveled. In addition to driving parameters, the damage status and an aging factor of the vehicle's monitoring units are also taken into account. This change is therefore typically an increase, since any damage or aging of the monitoring units is expected to lead to a deterioration of combustion or a reduction in the efficiency of the exhaust aftertreatment units. Furthermore, an initial prioritization list is created from the existing damage statuses of the individual monitoring units. This means that, depending on the frequency and severity of faults during the stored driving cycles, a value is available for each monitoring unit, representing a measure of the damage present in that unit.Furthermore, an interaction matrix is ​​stored and defined, specifying whether a damage state of one monitoring unit influences the damage state of another monitoring unit. This matrix remains unchanged and includes only every possible influence of damage to one monitoring unit on another. Additionally, a correction factor is stored, which, if such an interaction exists and a threshold value of the damage state of the influencing monitoring unit is exceeded, corrects the calculated expected increase in the emitted mass of the affected monitoring unit as a function of distance traveled. This correction factor is thus defined for every possible cross-influence between the monitoring units.A second prioritization list is then created for the monitoring units, based on the magnitude of the adjusted expected increases in the emitted mass. The monitoring units are subsequently replaced according to the order of this second prioritization list. In this step, the previously calculated increase in the emission load of the monitoring unit affected by damage to another monitoring unit is corrected. This affected monitoring unit is then moved accordingly in the prioritization list based on the correction factor. As a result, identified damage states of a monitoring unit no longer directly lead to its replacement if another monitoring unit, which also cross-influences this monitoring unit, exhibits a damage state that exceeds a defined threshold.Since additional influences can be stored in the prioritization lists or correction factors, such as known frequently occurring damage to a corresponding monitoring unit, this provides workshops with a tool that can significantly reduce the frequency of replacing non-faulty monitoring units.

[0010] In addition to the second prioritization list, the costs for replacing monitoring units are preferably taken into account in a repair logic. This can be done via correction factors or as an additional factor in the second prioritization list. This primarily reduces the unnecessary replacement of very expensive spare parts.

[0011] The driving parameters used to calculate emissions are advantageously the vehicle's load state, speed, and distance traveled. These values ​​are usually sufficient to determine reliable emissions levels using the exhaust gas model.

[0012] Preferably, a driving data acquisition matrix is ​​created from the driving parameters and the integrated air mass. This matrix serves as input for calculating the expected change or increase in the emitted mass as a function of the distance traveled for at least one defined emission component. Accordingly, the increase in emitted mass is always derived from the actual driving parameters.

[0013] Furthermore, a misfire matrix is ​​preferably created from the engine speed, load, and number of misfires. This matrix serves as input for calculating the expected increase in emitted mass as a function of the distance traveled by at least one defined emission component. This also serves to calculate the actual emissions present.

[0014] In a further advantageous embodiment, an emission influence map for the emission load is stored for each monitoring unit. This map contains the expected load for the respective emission component at the monitoring unit, assuming a complete failure state and complete aging of the respective monitoring unit, depending on the load state and engine speed. The data are corrected by the determined failure states and aging factors. Subsequently, the increases in emission load for the respective exhaust component at the respective monitoring unit are calculated by matrix multiplication with the driving data acquisition matrix. This emission influence map can be obtained from an exhaust gas model or actual measurements and thus serves as the basis for calculating the following estimated emission effects due to aging or failure states of the monitoring units.

[0015] Preferably, the monitoring units comprise at least one monitoring unit for measuring the oxygen concentration, one monitoring unit for the fuel supply system, one monitoring unit for detecting misfires or ignitions, and one monitoring unit for determining the air-fuel ratio per cylinder. These can be formed, in particular, by an oxygen sensor or a lambda sensor, a misfire detector for detecting misfires or ignitions, which can be used to determine irregularities between the cylinders, and a lambda sensor with corresponding detectors for determining irregularities in engine operation, wherein, in the latter case, irregularities in the air-fuel ratio in the cylinders can be determined from the data.These sensors and monitoring units are particularly important for calculating emissions and for correctly controlling the combustion engine.

[0016] It is also advantageous if the repair logic is connected via a network to a repair statistics system that collects error frequencies for the monitoring units. These can then also be considered in the prioritization list, allowing known errors to be taken into account, which leads to further optimization of the order in which the monitoring units are to be replaced.

[0017] Furthermore, it is preferable to store multiple driving cycles in a single storage unit, with the stored data including at least the driving data acquisition matrix and the damage status data from the on-board diagnostics. This results in a larger amount of data being available, which again improves the priority list.

[0018] In a particularly preferred embodiment, a defined number of driving cycles are stored in which at least one of the emission measurement values ​​is exceeded. If the defined number of driving cycles is exceeded, the oldest driving cycle is overwritten by the newest one. This ensures that only the driving cycles in which errors occurred are considered. As a result, the total number of driving cycles to be stored can be reduced, thus reducing the storage space required.

[0019] In a further advantageous embodiment, the aging factor is determined as a function of the oxygen storage capacity of the catalyst and taken into account accordingly.

[0020] Preferably, the driving data acquisition matrix, the emission influence maps, the interaction matrix, and a correction factor table with the correction factors are stored on a computer's repair logic or in a web-based environment. This allows workshops to access this data on their computers and collect it, potentially enabling the correction factors to be improved using artificial intelligence.

[0021] This provides a method for identifying and prioritizing monitoring units to be replaced in the presence of interactions between the monitoring units, taking into account existing interactions between the individual damage states of the monitoring units, thereby enabling more precise identification of actually defective monitoring units.

[0022] An embodiment of a method according to the invention for identifying and prioritizing monitoring units to be replaced in the presence of interactions between the monitoring units is described below with reference to the figures.

[0023] The Fig. Figure 1 schematically shows the process of the method according to the invention.

[0024] The Fig. Figure 2 schematically shows an example of how to create a prioritization list when an error occurs in an oxygen sensor.

[0025] The Fig. Figure 3 schematically shows an example of creating a prioritization list when two errors occur on monitoring units and there is a multiple dependency.

[0026] In the method according to the invention, driving parameters are first recorded and stored in the vehicle 10 during a driving cycle N via an engine control unit 12. In the present embodiment, these include the load state 14 of the internal combustion engine, the speed 16, the distances traveled 18, and consequently the driving performance. Furthermore, misfire events 20 are detected via a monitoring unit 22 for misfire detection. A driving data acquisition matrix 24 is created from the driving parameters, which contains an engine speed 21 and the load state 14 against the associated integrated air mass. For monitoring the misfires, such a matrix 26 can also be applied to the detected misfires.

[0027] In addition, the combustion engine and the exhaust system are monitored by a multitude of monitoring units during on-board monitoring. In the present embodiment, these include, besides the aforementioned monitoring unit 22 for misfire detection, an oxygen sensor or lambda probe 28, a fuel system monitoring unit 30, which may include, for example, a fuel pressure sensor and sensors for measuring fuel consumption or valve opening times, and a monitoring unit 32 for monitoring for inconsistencies in the air-fuel ratio per cylinder. If one or more of the monitoring units 22, 28, 30, or 32 output an implausible value, this is stored as a fault condition (SOD) as a diagnostic parameter in a diagnostic statistic 33.Plausibility involves checking whether a value from the monitoring unit lies within a defined range in the respective operating state, whether corresponding values ​​are consistent, and whether measured increases in values ​​are within the realistically possible range. The more frequently implausible values ​​are output by a monitoring unit 22, 28, 30, 32, and the greater the deviation from the realistically expected value, the higher the determined damage state (SOD) of the respective monitoring unit 22, 28, 30, 32.

[0028] The information stored in the engine control unit 10 in this way is saved for each driving cycle N. For example, a storage unit 34 always contains ten driving cycles in which an exceedance of the emission limit values ​​was detected, thus indicating a fault condition SOD of one of the monitoring units 22, 28, 30, 32. After more than ten driving cycles N have been stored, the oldest driving cycle is deleted and replaced by the most recent.

[0029] If, for example, the driver of vehicle 10 is alerted to a fault in the system by the engine light coming on, he takes his vehicle to a workshop.

[0030] In the workshop, the data from the stored driving cycles N to N+9 are read out on a computer, and the emission load of the monitoring units 22, 28, 30, 32 for the stored driving cycles N to N+9 is calculated using an exhaust gas model. First, the emissions are calculated based on the driving data acquisition matrix 24. For this purpose, emission influence maps 38, 40, 42, 44 are stored for each monitoring unit 22, 28, 30, 32, from which the expected emissions in such a driving cycle N to N+9 can be read. These emission influence maps 38, 40, 42, 44 are based on the assumption that the respective monitoring unit 22, 28, 30, 32 is no longer functional, exhibits a 100% damage condition, or is completely worn out.These emissions of a defined emission component are now corrected depending on the damage state (SOD) of the respective monitoring unit 22, 28, 30, 32 and an aging factor (AF). This means that if the monitoring unit 22, 28, 30, 32 has a damage state of less than 100%, the corresponding emission load value at this monitoring unit is corrected downwards. A subsequent matrix multiplication with the driving data acquisition matrix 24 calculates an estimated emission impact (EMI) of a defined emission component in mg / km. From this, an emission table is created, which shows, for example, by how many mg / km of nitrogen oxide the emission load of monitoring units 22, 28, 30, 32 increases as a result of the preceding calculations.From this, an initial prioritization list 45 is created, with monitoring units 22, 28, 30, and 32, which show the greatest increase according to the emissions table, occupying the first position on priority list 45. The following monitoring units 22, 28, 30, and 32 follow in descending order of emissions increase.

[0031] Furthermore, the computer contains a repair logic 36 with an interaction matrix 46, which specifies the extent to which a damage state SOD of one monitoring unit 22, 28, 30, 32 influences the determined damage state SOD of another monitoring unit 22, 28, 30, 32. This interaction matrix 46 represents the damage threshold above which one monitoring unit influences another. The y-axis represents the monitoring units 22, 28, 30, 32 in a damage state SOD, and the x-axis represents those monitoring units 22, 28, 30, 32 that can be influenced. If the damage state SOD exceeds the threshold specified in the interaction matrix 46, the corresponding monitoring unit 22, 28, 30, 32 indicated on the x-axis is considered to be influenced.

[0032] A corresponding example is in the Fig. Figure 2 shows the first prioritization list 45, which contains the determined damage states SOD and the calculated increases in nitrogen oxide emissions for the four monitoring units 22, 28, 30, 32 given here as examples. A threshold value for a damage state SOD of a monitoring unit 22, 28, 30, 32, above which another monitoring unit 22, 28, 30, 32 is affected, is given in an interaction matrix 46. The only monitoring unit that exceeds one of the specified threshold values ​​is the lambda sensor 28 at 80%, which exceeds the threshold value 75 for a possible influence on the monitoring unit 22 for the detection of misfires. Thus, in each row of the interaction matrix 46, it is read whether one of the threshold values ​​specified there is exceeded by the determined damage state of the respective monitoring unit 22, 28, 30, 32.This only applies to the misfire monitoring unit 22, which has a threshold value of 75%, a value exceeded by the 80% fault condition of the lambda sensor 28. Accordingly, the value of the misfire monitoring unit 22 must be corrected.

[0033] For this purpose, a correction factor table 48 with correction factors KF is used, which can depend on various factors and are measured or defined during the development of the vehicle 10. This can later be modified in the field using artificial intelligence based on new field data, since the associated software does not run in the car, but on the repair logic 36 of the computer or via the web. This also allows consideration of whether the field data indicates that a monitoring unit 22, 28, 30, 32 is more prone to failure or subject to faster aging.

[0034] In the present example, the relevant correction factor KF for the affected monitoring unit for misfire detection 22 is 0.4. For monitoring unit 22 for misfire detection, an increase of 10 mg / km was calculated for the emission component nitrogen oxide, which places this value in rank 1 in the first prioritization list 45. Multiplying this by the correction factor KF of 0.4 reduces this increase to 4 mg / km, thereby moving monitoring unit 22 for misfire detection to rank 3 in a newly created second prioritization list 50. Meanwhile, monitoring unit 32 for determining the air-fuel ratio per cylinder, with an increase of 8 mg / km due to the damage conditions, is moved to rank 1.Thus, the monitoring unit 32 for determining the air-fuel ratio per cylinder would be the first monitoring unit to be replaced in the workshop in order to rectify the fault.

[0035] Another example of how to create and calculate the second priority list 50 is in the Fig.Figure 3 illustrates this. The calculated increase in nitrogen oxides based on the measurement data is 10 mg / km for monitoring unit 22 (for detecting misfires), 8 mg / km for monitoring unit 32 (for the air-fuel balance), 5 mg / km for monitoring unit 28 (for measuring oxygen concentration, i.e., the oxygen sensor or lambda probe), and 1 mg / km for monitoring unit 30 (for the fuel system). Here, the damage status of both monitoring unit 28 (for measuring oxygen concentration) and monitoring unit 32 (for air-fuel balance) partially exceeds the specified threshold values. Both values ​​are 80%. This means that, as in the first embodiment, the calculated increase in nitrogen oxides for monitoring unit 22 (for detecting misfires) must be corrected by the SOD (Safety of Discharge) damage status of monitoring unit 28 (for measuring oxygen concentration).The SOD (State of Death) damage status of monitoring unit 32 for measuring the air-fuel ratio per cylinder necessitates corrections not only to the measured increase in nitrogen oxides (NOx) from monitoring unit 22 for misfire detection, but also to the measured increases in NOx from monitoring unit 28 for oxygen concentration and from monitoring unit 30 of the fuel system. Thus, only the increase value from monitoring unit 32 for measuring the air-fuel ratio remains unchanged, while the other three values ​​are corrected using correction factors KF.This value is thus multiplied by the correction factor 0.5 for monitoring unit 28 for measuring oxygen concentration (2.5 mg / km), by the correction factor 0.2 for monitoring unit 30 of the fuel system (0.2 mg / km), and by the correction factor 0.4 for monitoring unit 22 for determining misfires (4 mg / km). As a result, monitoring unit 32 for measuring the air-fuel ratio is moved from second to first place in the second priority list (50), monitoring unit 22 for determining misfires is moved from first to second place, and monitoring units 28 and 30 of the fuel system and for determining oxygen concentration remain unchanged in third and fourth place, respectively.

[0036] This method allows for a shift in the priorities for replacing monitoring units by considering the mutual influence of the various monitoring units in the workshop, in addition to the existing, identified damage states. This is done based on the available driving data and, where applicable, existing experience with occurring errors, if additional statistics are used for the correction factor or elsewhere.

[0037] It should also be clear that, in addition to the monitoring units described, all other monitoring units and sensors can be taken into account. Furthermore, besides the impact of an increase in the nitrogen oxide content, an increase for any other emission component can also be calculated and considered.

[0038] The ammonia content, in particular, should be mentioned here, as it is also suitable for this process. Additionally, the costs incurred by a replacement can also be taken into account when making the final prioritization.

Claims

[1] Procedure for identifying and prioritizing monitoring units (22, 28, 30, 32) to be exchanged in the presence of interactions between the monitoring units (22, 28, 30, 32) comprising the following steps: - Recording and storing driving parameters over a defined number of driving cycles (N) of a vehicle (10), - Acquisition and storage of diagnostic parameters of each monitoring unit (22, 28, 30, 32) from a diagnostic statistic (33) of the monitoring units (22, 28, 30, 32) over the defined number of driving cycles (N) of the vehicle (10) to determine a damage state (SOD) of the respective monitoring unit (22, 28, 30, 32), - Determining and storing an expected increase in the emitted mass as a function of a distance traveled of at least one defined emission component based on the driving parameters, the damage state (SOD) and an aging factor (AF) of the monitoring units (22, 28, 30, 32) of the vehicle (10), - Establishing an initial prioritization list (45) depending on the respective damage status (SOD) of the monitoring units (22, 28, 30, 32), - Storing an interaction matrix (46) which specifies whether a damage state (SOD) of one of the monitoring units (22, 28, 30, 32) affects a damage state (SOD) of another monitoring unit (22, 28, 30, 32), - Storing a correction factor (CF) with which, in the event of such an interaction and exceeding a threshold value of the damage state (SOD) of the influencing monitoring unit (22, 28, 30, 32), the determined expected increase in the emitted mass of the influencing monitoring unit (22, 28, 30, 32) is corrected as a function of the distance traveled, - Creating a second prioritization list (50) depending on the size of the expected increases in the mass emitted for the monitoring units (22, 28, 30, 32) corrected by the correction factor (CF) depending on the distance traveled, depending on which an exchange of the monitoring units (22, 28, 30, 32) takes place. [2] Method for identifying and prioritizing monitoring units to be replaced according to claim 1, characterized by, that in addition to the second prioritization list (50) in a repair logic (36) the costs for replacing the monitoring units (22, 28, 30, 32) are taken into account. [3] Method for identifying and prioritizing monitoring units to be replaced according to claim 1 or 2, characterized by , that the driving parameters from which the emission load is calculated are the load state (14) of the vehicle (10), the speed (16) and the distance traveled (18). [4] Method for identifying and prioritizing monitoring units to be replaced according to any of the preceding claims, characterized by , that a driving data acquisition matrix (24) is created from the driving parameters and the integrated air mass, which is used as an input value to calculate the expected increase in the emitted mass as a function of a distance traveled of at least one defined emission component. [5] Method for identifying and prioritizing monitoring units to be replaced according to any of the preceding claims, characterized by , that a misfire matrix (26) is created from the engine speed (21), the load condition (14) and the number of misfires, which is used as an input value to calculate the expected increase in the emitted mass as a function of a distance traveled by at least one defined emission component. [6] Method for identifying and prioritizing monitoring units to be replaced according to any of the preceding claims, characterized by, that for each monitoring unit (22, 28, 30, 32) an emission influence map (38, 40, 42, 44) is stored for the emission load, in which the load to be expected at the respective monitoring unit (22, 28, 30, 32) with the respective emission component is stored assuming a complete damage state (SOD) and a complete aging of the respective monitoring unit (22, 28, 30, 32) as a function of the load state (14) and the engine speed (21), whose data are corrected by the determined damage states (SOD) and aging factors (AF), and subsequently the increases in the emission load with the respective exhaust component at the respective monitoring unit (22, 28, 30, 32) are calculated by matrix multiplication with the driving data acquisition matrix (24). [7] Method for identifying and prioritizing monitoring units to be replaced according to any of the preceding claims, characterized by, that the monitoring units (22, 28, 30, 32) include at least one monitoring unit (28) for measuring the oxygen concentration, one monitoring unit (30) of the fuel system, one monitoring unit (22) for detecting misfires or ignitions, and one monitoring unit (32) for determining the air-fuel ratio per cylinder. [8] Method for identifying and prioritizing monitoring units to be replaced according to any of the preceding claims, characterized by , that the repair logic (36) is connected via a network to a repair statistics system in which error frequencies of the monitoring units (22, 28, 30, 32) are collected. [9] Method for identifying and prioritizing monitoring units to be replaced according to any of the preceding claims, characterized by, that a plurality of driving cycles are stored in a storage unit (34), wherein the stored data includes the driving data acquisition matrix (24) and the damage status data (SOD) from the on-board diagnostics. [10] Method for identifying and prioritizing monitoring units to be replaced according to any of the preceding claims, characterized by , that a defined number of driving cycles are stored in which at least one of the emission measurement values ​​is exceeded, whereby if the defined number of driving cycles is exceeded, the oldest driving cycle is overwritten by the newest driving cycle. [11] Method for identifying and prioritizing monitoring units to be replaced according to any of the preceding claims, characterized by that the aging factor (AF) is determined as a function of the oxygen storage capacity of a catalyst. [12] Method for identifying and prioritizing monitoring units to be replaced according to any of the preceding claims, characterized by , that the driving data acquisition matrix (24), the emission influence maps (38, 40, 42, 44), the interaction matrix (46) and a correction factor table (48) with the correction factors (KF) are stored on a repair logic (36) of a computer or web-based.

Images