GB2513249A

GB2513249A - Method for checking a multi sensor system of a vehicle

Info

Publication number: GB2513249A
Application number: GB1404571.0A
Authority: GB
Inventors: Jasman Malik; Daniel Potter; Justin Kollien
Original assignee: Daimler AG
Current assignee: Mercedes Benz Group AG
Priority date: 2014-03-14
Filing date: 2014-03-14
Publication date: 2014-10-22
Also published as: GB201404571D0

Abstract

The invention relates to a method for checking a multi sensor system 36 having a plurality of temperature sensors 38a-m of a vehicle, configures to detect temperatures of, for example, and engine and or exhaust gas after treatment device. The method comprises: assigning a plurality of the temperature sensors 38a-m to a predeterminable group; determining at least one sensor output of each temperature sensor 38a-m of the group, the respective sensor output characterizing a temperature; calculating at least one mean value of the sensor outputs; and checking the multi sensor system 36 in dependence on the calculated mean value. The method may further comprise determining a difference between the mean value and a respective sensor output and comparing the difference with a predetermined threshold value. Each temperature sensor may also be subjected to a range check.

Description

Method for checking a Multi Sensor System of a Vehicle The invention relates to a method for checking a multi sensor system of a vehicle.

Vehicles having multi sensor systems are well-known from the general prior art. Such a multi sensor system comprises a plurality of temperature sensors which are configured to detect temperatures of, for example, an engine and/or an exhaust gas after treatment device. The temperature sensors, in particular, their signals characterizing respective temperatures are used for, for example, control purposes or otherwise. A drifted temperature sensor being used for control purposes can potentially impact emission and performance behaviour. Moreover, it is known from the general prior art to implement a sensor rationality for checking the multi sensor system, the sensor rationality comprising comparing two sensors with each other. Of the multi sensor rationalities that exist, there are loopholes, which may prevent the detection of a bad sensor, i.e. a faulty sensor in case more than one sensor has drifted. Some multi sensor algorithms employ truth tables and case by case evaluation of sensors which may result in large grey areas.

US 2013/0055802 Al shows a competence diagnosis system for a urea water temperature sensor comprising a competence diagnosis unit that diagnoses a failure of the urea water temperature sensor provided in a urea tank by comparing a temperature of urea water measured by the urea water temperature sensor with an ambient temperature.

The competence diagnosis system comprises a cold start condition determination unit that obtains a difference between an engine cooling water temperature and an ambient temperature, a difference between an engine cooling water temperature and a fuel temperature, and a difference between an ambient temperature and a fuel temperature by using an engine cooling water temperature, an ambient temperature and a fuel temperature measured immediately after engine starting. The competence diagnosis system allows a diagnosis by the competence diagnosis unit when the temperature differences are less than or equal to a predetermined threshold value.

It is an object of the present invention to provide a method for checking a multi sensor system by means of which method a malfunction of the multi sensor system can be detected with certainty even if two or more temperature sensors of the multi sensor system have drifted.

This object is solved by a method having the features of patent claim 1. Advantageous embodiments with expedient and non-trivial developments of the invention are indicated in the other patent claims.

According to the present invention, the method for checking a multi sensor system having a plurality of temperature sensors of a vehicle comprises a step in which a plurality of the temperature sensors of the multi sensor system are assigned to a predeterminable group.

The method according to the present invention comprises a further step in which at least one sensor output of each sensor of the group is determined, the respective sensor output characterizing a temperature measured by the respective sensor. In a further step of the method, at least one mean value of the sensor outputs is calculated. Moreover, the method comprises a step in which the multi sensor system is checked at least in dependence on the calculated mean value. The multi sensor system may additionally be checked in dependence on a range of sensor outputs. The invention solves the problem of identifying one or more faulty sensor output by rationalizing multiple temperature sensors or their sensor outputs respectively, the sensors being installed on, for example, an engine and/or after-treatment apparatus of the vehicle. For this purpose the method preferably comprises a step where the spread of all groups of sensors are probed.

In the method, loopholes that exist in conventional strategies can be avoided. The method according to the present invention is able to detect at least two drifted sensors of the multi sensor system. In other words, the method according to the present invention allows for detecting a malfunction of the multi sensor system even if two or more of the sensors have drifted. Furthermore, the method according to the present invention does not involve the concept of truth tables, thus, minimizing grey areas and reducing implementation complexity. Moreover, the method according to the present invention is hardware independent and, thus, capable of being implemented on an apparatus having, for example, more than four temperature sensors.

The method according to the present invention increases the capability to detect at least one drifted temperature sensor of the multi sensor system and, thus, prevent any warranty issues due to failure that was not detected. Also, due to the proposed calibration guideline, the method will require minimal data collection and, thus, effort to calibrate a monitor configured to check the multi sensor system. Furthermore, since said monitor can be simulated on a computer platform, calibrations can be performed on a computer itself.

Due to the abundance of historic data and continually updated data from, for example, a reliable increase in the number of trucks on the road, analysis can be performed to ensure a large number of cases, ambient conditions and vehicle configurations, in particular, truck and bus configurations, are covered.

The method according to the present invention facilitates the realization of a check performed on a system after the system has been out of use for a set time. In the case of an engine, the method realizes a multi sensor diagnosis at the key on/engine off condition performed after the engine has not been in use for an extended time.

Further advantages, features and details of the invention derive from the following description of a preferred embodiment as well as from the drawings. The features and feature combinations previously mentioned in the description as well as the features and feature combinations mentioned in the following description of the figures and/or shown in the figures alone can be employed not only in the respective indicated combinations but also in any other combination or taken alone without leaving the scope of the invention.

The drawings show in: Fig. 1 a schematic view of an internal combustion engine of a vehicle having a multi sensor system comprising a plurality of temperature sensors; Fig. 2 a flow diagram illustrating a method for checking the multi sensor system; Fig. 3a-b diagrams illustrating a first scenario which can be detected by the method; Fig. 4a-b diagrams illustrating a second scenario which can be detected by the method; Fig. 5a-b diagrams illustrating a third scenario which can be detected by the method; Fig. 6a-b diagrams illustrating a fourth scenario which can be detected by the method; and Fig. 7a-b diagrams illustrating a fifth scenario which can be detected by the method.

In the figures the same elements or elements having the same function are indicated with the same reference sign.

Fig. 1 shows an internal combustion engine 10 for a vehicle which can be, for example, a truck. The internal combustion engine 10 comprises an engine block 12 having six cylinders 14. Moreover, the internal combustion engine 10 has an exhaust system 16 in which a turbine 18 of an exhaust turbo charger 20 is arranged. The exhaust system 16 serves for guiding exhaust gas flowing out of the cylinders 14, the exhaust gas driving the turbine 18. With respect to the flow direction of the exhaust gas, an exhaust gas after treatment device 22 is arranged downstream of the turbine 18 in the exhaust system 16.

The exhaust gas after treatment device 22 comprises an oxidizing catalytic converter 24 which is configured as a diesel oxidizing catalytic converter (DOG) since the internal combustion engine 10 is configured as a diesel engine. Moreover, the exhaust gas after treatment device 22 comprises a diesel particulate filter 25 (DPF) and a SCR catalytic converter 26 (SCR -selective catalytic reduction).

Furthermore, the internal combustion engine 12 comprises an intake system 28 in which a compressor 30 of the exhaust turbo charger 20 is arranged. The compressor 30 can be driven by the turbine 18. The intake system 28 serves for guiding fresh air to the cylinders 14. With respect to the flow direction of the air, an intercooler 32 is arranged in the intake system 28 downstream of the compressor 30. The intercooler 32 is configured to cool the air compressed by means of the compressor 30. Moreover, an intake manifold 34 is arranged in the intake system 28, the intake manifold 34 being configured to distributing the compressed air to the cylinders 14. Furthermore, the internal combustion engine 10 has an exhaust gas recirculation system 37 configured to guide exhaust gas from the exhaust system 16 back to the intake system 28.

As can be seen in Fig. 1, the internal combustion engine 10 further comprises a multi sensor system 36 having a plurality of temperature sensors 38a-m. The temperature sensor 38a is configured as an ambient temperature sensor detecting the ambient temperature, the sensor 38a not being located on the internal combustion engine 10 or the engine block 12.

The sensor 38b is configured as an engine oil temperature capable of detecting the temperature of the oil used to lubricate the internal combustion engine 10. The temperature sensor 38c is configured as a coolant temperature sensor configured to detect the temperature of the coolant at an inlet, the coolant being used to cool the internal combustion engine 10. The temperature sensor 38d is configured as a coolant temperature sensor used to detect the temperature of the coolant at an outlet. The temperature sensor 38e is configured as a fuel temperature sensor used to detect the temperature of the fuel of the internal combustion engine 10. The temperature sensor 38f is configured as an ECU temperature sensor used to detect the temperature of an engine control unit (ECU) of the internal combustion engine 10.

The temperature sensor 38g is configured as an SCR temperature sensor used to detect the temperature of the exhaust gas downstream of the SCR catalytic converter 26. The temperature sensor 38h is configured as an SCR temperature sensor used to detect the temperature of the exhaust gas downstream of the diesel particulate filter 25 and upstream of the SCR catalytic converter 26. The temperature sensor 38i is configured as a DPF temperature sensor used to detect the temperature of the exhaust gas downstream of the diesel oxidizing catalytic converter 24 and upstream of the diesel particulate filter 25. The temperature sensor 38j is configured as a DCC temperature sensor used to detect the temperature of the exhaust gas downstream of the turbine 18 and upstream of the oxidizing catalytic converter 24.

The temperature sensor 38k is configured as a compressor inlet temperature sensor used to detect the temperature of the air upstream of the compressor 30. The temperature sensor 381 is configured as an intercooler out temperature sensor used to detect the temperature of the air downstream of the intercooler 32 and upstream of the intake manifold 34. Moreover, the temperature sensor 38m is configured as an intake manifold sensor used to detect the temperature of the air in the intake manifold 34 and upstream of the cylinders 14. The temperature sensors 38a-m may be employed for engine and/or after treatment performance control purposes, diagnostics, and hardware protection or otherwise. The temperature sensors 38a-m are critical to the engine's performance characteristic and emission compliance. For example, the temperature sensors 38a-m can be used for measuring and monitoring temperatures of air systems, the exhaust gas recirculation system 37 and/or a fluid system. Over the lifecycle of the vehicle, the temperature sensors 38a-m installed may drift and alter the performance and emission characteristics of the system. Thus, a challenge arises for on-board diagnosis to detect at least one faulty temperature sensor and, thus, a malfunction of the multi sensor system 36.

Fig. 2 shows a flow diagram illustrating a method for checking the multi sensor system 36. The method is a logic which identifies and groups the system's temperature sensors 38a-m, gathers readings for each temperature sensor 38a-m, and runs a relative comparison of each temperature sensor's temperature reading against its groups mean temperature value. If a temperature sensor reading deviates too far from the group mean, in particular while the readings from other sensors are fairly close to each other, it will be considered faulty. If more than one sensor temperature reading deviates from the group mean, the group will be considered faulty. Additionally or alternatively a group of sensors will be considered faulty if the spread of all sensor outputs, i.e. the temperature readings of the sensors of that group exceeds a predeterminable threshold. In case the logic determines a faulty sensor or sensors, the driver of the vehicle is informed using, for example, a malfunction indicator lamp on a dashboard of the vehicle. The logic can be directly contrasted to systems that compare a sensor's temperature reading only against another sensor either at key-on or when the engine is running.

The method uses a monitor to check the multi sensor system 36. The monitor operates to rationalize the plurality of the temperature sensors 38a-m of the multi sensor system 36.

Based on conditions described below, the method or the monitor determines a system which may consist of multiple groups, wherein each of the groups may comprise a plurality of the temperature sensors 38a-m. Rationality for the temperature sensors 36a-m is performed at the group level and at the level of the individual temperature sensors 38a-m. Rationality of the respective temperature sensor 38a-m may be defined as identifying the output of the respective temperature sensor 38a-m is accurate within bounds of tolerances.

As will be described below, the method comprises a first step in which a plurality of the temperature sensors 38a-m is assigned to a predeterminable group. In a second step, at least one sensor output of each temperature sensor of the group is determined, the respective sensor output characterizing a temperature. In a third step, at least one mean value of the sensor outputs is calculated. In a fourth step, the multi sensor system 36 is checked in dependence on the calculated mean value.

The method uses an algorithm for checking the multi sensor system 36. Said algorithm employs a block heater detection logic which indicates the presence of a block heater or any other source of heat not integral to the vehicle. For example, a block heater may be an external source of heat that is used in extreme cold weather to heat up the engine block 12 and to keep the engine block 12 warm to permit ease in starting the internal combustion engine 10. Usually, such a block heater is an electric heater and placed on the engine block 12 in the coolant circuit and may be powered by a power source such as a wall outlet. Alternatively, such an external heat source may be configured as a fuel tired heater. The block heater may or may not be a part of the standard equipment ot the internal combustion engine 10 and the vehicle and could be, for example, an aftermarket installation requested by the owner of the vehicle. The output ot the block heater detection logic can return either a "true" it a block heater is detected or a "talse" if a block heater is not detected.

Using a block heater creates an offset in the expected temperatures detected by the temperature sensors 38a-m and, thus, needs to be known in the on-board diagnostic system, i.e. in the method tor correct diagnosis. For example, in case the ambient temperature is -10, atter several (for example seven to eight) hours ot soak in this temperature, the temperature of the internal combustion engine 10 can be expected to be very close to -1 0'C. In case a block heater is plug ged into the coolant circuit, the block heater will heat up the coolant and as a result, the coolant temperatures will no longer be reading very close to -1 0'C, but instead the temper ature of the coolant will be ot the order of, tor example, 25 or 30'C atter four hours of the block heater being plugged into the coolant circuit. This trend (increase in the temperature) can be seen by other temperature sensors 38a-m as well. Now, it a rationality of temperature sensors is performed comparing the coolant temperature at the inlet with, tor example, the temperature ot the air detected by the temperature sensor 38a, there will be a visible difference in their respective values. In a conventional approach, by comparing one sensor against the other after a long soak, one of the two sensors will have to be deemed taulty. However, detecting the presence of a block heater tacilitates avoidance ot getting into such a situation so that it is possible to separate the sensors attected by a block heater trom the sensors which are not affected by a block heater.

For example, other non-integral sources ot heat can be in the torm ot pan heaters which heat up the engine's oil pan or any other source that is not communicated to the ECU.

The assignment of the temperature sensors 38a-m to a particular group is preterably a dynamic assignment. Such a dynamic assignment ot the temperature sensors 38a-m to a particular group may be based upon factors such as, but not limited to, physical location, ambient conditions, material to which the respective temperature sensor SSa-m is attached or in contact with, correlation to temperature profiles of other temperature sensors in the group, variation permitted in the group, tault thresholds, etc. In case the presence of an external heat source is not detected, the temperature sensors 38a-m may be divided into up to three groups or up to two groups if a block heater is detected. The

S

assignment will be determined by, for example, a calibration engineer responsible for the function. One sensor can be assigned to one group only under each block heater detection condition.

The capability of Dynamic assignment" means that one sensor can be assigned to different groups under different initial conditions. For example, the condition determining the dynamic assignment for the sensors is based on the detection of a block heater or not. However, this can be extended to another condition if the need arises (using a software change and deterministic estimation of the sources present). For example, if a block heater is not present, the temperature sensor 38k and the temperature sensor 38c and/or the temperature sensor 38d can belong to the same group because these sensors will read close to the same temperature, for example, close to -1O'C. In case a block heater is detected, the coolant and compressor inlet temperature sensors will belong to two different groups based on the effect the block heater has on various sensors. This assignment to different groups will be determined by a calibrator by looking at the sensors affected by the block heater and those not affected. Thus, the assignment remains dynamic based on the presence of the block heater but is known to the calibrator as to the groups each sensor can go to. Similarly, if required, there may be a condition to group sensors differently in case the ambient temperature is too high or too low. For example, in case the presence of a block heater is detected, the temperature sensors affected by the presence of the block heater will be grouped together while the ones not affected by the block heater can be grouped separately.

As can be seen from Fig. 2, there are monitoring conditions. The respective group of sensors is checked for a common set of monitoring conditions before the rationality of the temperature sensors has begun. For example, for each group formed above due to the dynamic assignment, the following conditions are checked to be true: a) The system has been shut down for more than a calibratable or predeterminable amount of time, for example six to seven hours.

b) Once the chock starts, the system or auxiliary units have not been switched on for more than a calibratable or predeterminable amount of time, for example no more than one hour.

c) Battery voltage is above a minimum threshold, for example 10.8 volts.

d) None of the temperature sensors 38a-m in the group have been diagnosed to have a fault.

The following monitor logic is run after the above mentioned monitoring conditions are true. In the following, the monitoring logic will be described. The following process is repeated for all groups in the system.

At first, a preprocessing is conducted. A first step of said preprocessing comprises averaging of sensor outputs. This means an output of each sensor in the system is averaged for a calibratable time period. For example, the temperature sensors are read in to the ECU at 40 milliseconds or 25 Hertz and averaged for 10 seconds. At the end of the calibrated time period, the output of a sensor is represented by an average value. The aim of averaging sensor output is to reduce noise in the signal and obtain a steady, representative output for every sensor. In other words, each temperature sensor of the group provides a plurality of sensor values, the sensor values characterizing a temperature detected by the respective temperature 38a-m. In the first step of the preprocessing process, the plurality of sensor values of each temperature sensor 38a-m is averaged thereby calculating an average value. Methods other than averaging such as weighted means, low pass filtering, etc. may be employed to achieve the same objective.

In a second step the sensors on the system are assigned group numbers based on predefined conditions such as high ambient temperatures or detection of an external heat source such as the block heater. Steps three to five mentioned below are performed for all the groups of sensors formed in step two.

As a third step, for each group, the representative values obtained using the above mentioned preprocessing step are averaged to obtain a mean value of all sensors in that group. Thus obtained mean value has been termed as the group mean'.

As a fourth step, the individual sensor checks are conducted. In said individual sensor checks two checks are performed on each temperature sensor of the group. A first step of the individual sensor checks comprises a sensor drift check'. In said sensor drift check, the monitor compares the difference between the sensor output and the mean value of all temperature sensors in that group, i.e. the group mean, against a threshold. If the difference is greater than the threshold, the check is said to have failed. In other words, in the first step of the individual sensor checks, one of the sensors of the group is chosen and the chosen sensor is subjected to the sensor drift check. This check is preferably repeated for all sensors in that group.

In a second step of the individual sensor checks, a range check' is conducted. The monitor compares the range of all sensors other than the sensor being chosen and checked in the sensor drift check against a threshold. If the range of all sensors, except the sensor being subjected to individual drift check, in the group is less than a threshold, the check is said to have passed. This means that each time a sensor is being chosen and checked for drift, the other sensors in the group will be subjected to the range check.

For example, the range of the sensors other than the chosen sensor is the absolute difference between the lowest sensor output and the highest output of the sensors other than the chosen sensor of the group. Each sensor of the group may undergo the same number of checks (drift or range) as the number of sensors in the group.

As can be seen from Fig. 2, the method further comprises a group check'. The group check sums up the pass and fail results of range checks performed on sensors in that group and and compares the result with a threshold. If more than a calibrated number of range checks for the sensors of the group fail, the group check fails.

Moreover, the method comprises setting fault conditions. A sensor specific fault is set if, for a particular sensor, the individual sensor drift check fails and the individual sensor range check passes. A group specific fault is set if the group check fails. This means there is both the potential for the logic to relay an individual sensor fault, and/or a group fault.

Fig. 3a and 3b show diagrams illustrating a first scenario that can be detected by the logic. The solid vertical line shown in Fig. 3a is an expected mean or mean value of sensors present in the system. For the purpose of this check, this expected mean is close to the ambient conditions in which the system is placed. The process of determining the mean or mean value comprises the step of preprocessing in which the respective output of all sensors of the group is averaged over a calibrated time, for example, to reduce any noise in the respective signal. Moreover, in order to determine the mean of all sensors in the group, a mean calculation is conducted in which, based on the sensors in the group, all (preprocessed) average values are summed up and divided by the number of sensors in that group thereby calculating the mean. The sensor drift check is performed between the sensor reading and the actual mean of all sensors in that group including the sensor being analysed.

The dotted vertical lines in the diagram shown in Fig. 3a are or bound the expected spread inclusive of specification tolerances of the sensors. The expected mean of the sensors is likely to be close to the temperature of the system surroundings, and any difference can be attributed to reasons discussed above (physical location of the vehicle, material on which the sensors are attached, etc.). In other words, the vertical dotted lines indicate the range boundaries. If, during the individual drift check, the range of the output of other sensors is less than the spread covered by the dotted lines, the range check is passed. In other words, the vertical lines are indicative of the normal expected range of the sensors and will be similar to the range threshold for a group of sensors.

Fig. 3b shows a left diagram 40 having an ordinate 42 and an abscissa 44. The ordinate 42 shows the respective number of the temperature sensor of the group. The abscissa 44 shows the respective average value of the sensors of the group. In Fig. 3b, the mean value of the temperature sensors of the group is designated with M. The other asterisks show the average value of the respective sensors. A diagram 46 shows the differences between the respective average values and the mean value M. A diagram 48 illustrates the range of the sensors other than the sensor being chosen and subjected to the drift check. A diagram 50 illustrates the range check status. Moreover, red bars in a diagram 52 indicate a sensor fault, i.e. an individual sensor fault. Moreover, red bars in a diagram 54 indicate a group fault. As can be seen in Fig. 3a and 3b, in the first scenario illustrated by Fig. 3a and 3b, an individual fault of sensor number 4 is detected.

Fig. 4a and 4b show a second scenario in which an individual fault of sensor number 4 is detected. Fig. 5a and 5b show a third scenario in which an individual fault of sensor number 4 is detected.

Fig. Ba and Sb illustrate a fourth scenario in which the group check fails so that a group error is detected. Furthermore, Fig. 7a and 7b show a fifth scenario in which a group fault is detected. As can be seen in Fig. 4a, the far right sensor (sensor number 4) or the average value of sensor number 4 is too far outside of the mean value or mean temperature of the group. Therefore, sensor number 4 is recognized as faulty. The same applies to Fig. 3a and sensor number 4, whose average value is too far outside of the mean value of the group. This also applies to sensor number 4 in Fig. 5a.

List of reference numbers internal combustion engine 12 engine block 14 cylinder 1 6 exhaust system 18 turbine exhaust turbo charger 22 exhaust gas after treatment device 24 oxidizing catalytic convertor diesel particulate filter 26 SCR catalytic converter 28 inlet system compressor 32 intercooler 34 intake manifold 36 multi sensor system 37 exhaust gas recirculation system 38a-m temperature sensor diagram 42 ordinate 44 abscissa 46 diagram 48 diagram diagram 52 diagram 54 diagram M mean value

Claims

Claims A method for checking a multi sensor system (36) comprising a plurality of temperature sensors (38a-m) of a vehicle, the method comprising: -assigning a plurality of the temperature sensors (38a-m) to a predeterminable group; -determining at least one sensor output of each temperature sensor (38a-m) of the group, the respective sensor output characterizing a temperature; -calculating at least one mean value of the sensor outputs; and -checking the multi sensor system (36) at least in dependence on the calculated mean value.
2. The method of claim 1, characterized in that the respective sensor output comprises at least one average value calculated by averaging a plurality of sensor values provided by the respective temperature sensor (38a-m) of the group, the sensor values characterizing a temperature detected by the respective temperature sensor (38a-m).
3. The method according to any one of claims 1 or 2, characterized in that the method comprises: -determining a difference between the mean value and the respective sensor output; -comparing the difference with a predeterminable threshold value; and -checking the multi sensor system (36) in dependence on the comparison of the difference with the predeterminable threshold value.
4. The method according to any one of the preceding claims, characterized in that each temperature sensor (38a-m) of the group is subjected to a range check comprising: -choosing one of the temperature sensors (38a-m) of the group; -comparing the range between the lowest sensor output and the highest output of the other temperature sensors (38a-m) of the group with a predeterminable threshold value; and -checking the multi sensor system (36) in dependence on the comparison of the respective range with the predeterminable threshold value.