Sensor arrangement for a vehicle and method for monitoring a sensor
阅读说明：本技术 用于车辆的传感器组件和用于监控传感器的方法 (Sensor arrangement for a vehicle and method for monitoring a sensor ) 是由 S·柯尼希 M·施密德 D·约翰 G·朗 M·克里格 于 2019-03-13 设计创作，主要内容包括：本发明涉及一种用于车辆的传感器组件(1)和一种用于监控传感器(N)的方法。该传感器组件具有：第一传感器(N),其检测至少一个与碰撞相关的物理量并且输出对应的至少一个第一传感器信号(S<Sub>N</Sub>)；至少一个第二传感器(M),其独立于第一传感器(N)地检测至少一个与碰撞相关的物理量并且输出对应的至少一个第二传感器信号(S<Sub>M</Sub>)；以及评估和控制单元(μC),其接收至少一个第一传感器信号(S<Sub>N</Sub>)和至少一个第二传感器信号(S<Sub>M</Sub>)并且进行评估以识别碰撞。在此,评估和控制单元(μC)使用基于至少一个第二传感器(M)的至少一个第二传感器信号(S<Sub>M</Sub>)的至少一个第二比较信号,以监控第一传感器(N),其中如果基于至少一个第一传感器信号(S<Sub>N</Sub>)的至少一个第一比较信号具有高于预定的第一阈值的高信号值,并且第二传感器(M)的用于监控第一传感器(N)的至少一个第二比较信号同时具有低于预定的第二阈值的低信号值,和/或由第一比较信号和第二比较信号产生的偏差函数超过预定的第三阈值,则评估和控制单元(μC)识别出第一传感器(N)有故障。(The invention relates to a sensor arrangement (1) for a vehicle and a method for monitoring a sensor (N). The sensor assembly has: a first sensor (N) which detects at least one physical quantity related to a collision and outputs a corresponding at least one first sensor signal (S) N ) (ii) a At least one second sensor (M) which, independently of the first sensor (N), detects at least one physical quantity related to a crash and outputs a corresponding at least one second sensor signal (S) M ) (ii) a And an evaluation and control unit (μ C) which receives at least one first sensor signal (S) N ) And at least one second sensor signal (S) M ) And evaluated to identify collisions. The evaluation and control unit (μ C) uses at least one sensor (M) based on at least one second sensorA second sensor signal (S) M ) To monitor the first sensor (N), wherein if based on the at least one first sensor signal (S) N ) Has a high signal value above a predetermined first threshold value and at least one second comparison signal of the second sensor (M) for monitoring the first sensor (N) has a low signal value below a predetermined second threshold value and/or a deviation function resulting from the first and second comparison signals exceeds a predetermined third threshold value, the evaluation and control unit (μ C) identifies a malfunction of the first sensor (N).)
1. A sensor assembly (1) for a vehicle, the sensor assembly (1) having: a first sensor (N) which detects at least one physical quantity related to a collision and outputs a corresponding at least one first sensor signal (S)N) (ii) a At least one second sensor (M) which detects at least one physical quantity related to a collision independently of the first sensor (N) and outputs a corresponding at least one second sensor signal (S)M) (ii) a And an evaluation and control unit (μ C) which receives the at least one first sensor signal (S)N) And the at least one second sensor signal (S)M) And an evaluation is made to identify a collision,
it is characterized in that the preparation method is characterized in that,
the evaluation and control unit (μ C) uses the at least one second sensor signal (S) based on the at least one second sensor (M)M) To monitor the first sensor (N), wherein if based on the at least one first sensor signal (S)N) Has a high signal value above a predetermined first threshold value (Thd det) and the at least one second comparison signal of the second sensor (M) for monitoring the first sensor (N) simultaneously has a low signal value below a predetermined second threshold value (Thd drive) and/or a deviation function (f (S) resulting from the first and second comparison signalsN,{SM})) exceeds a predetermined third threshold value (Thd), the evaluation and control unit (μ C) identifies that the first sensor (N) is faulty.
2. Sensor assembly (1) according to claim 1, characterized in that said at least one firstThe comparison signal corresponds to the at least one first sensor signal (S)N) And/or at least one processed first sensor signal (g (S)N) And the at least one second comparison signal corresponds to the at least one second sensor signal (S)M) And/or at least one processed second sensor signal (h)M(SM))。
3. Sensor assembly (1) according to claim 2, characterized in that the evaluation and control unit (μ C) respectively calculates the sensor signals (S)N，SM) Wherein the evaluation and control unit (μ C) performs filtering and/or window integration and/or integration to process the sensor signal (S)N，SM) Or a calculated absolute value function.
4. Sensor assembly (1) according to claim 3, characterized in that the evaluation and control unit (μ C) starts integrating when the corresponding signal exceeds a start threshold (Thd start), wherein the evaluation and control unit (μ C) ends integrating and resets the integration value when the corresponding signal is below a reset threshold (Thd reset).
5. Sensor assembly (1) according to one of claims 1 to 4, characterized in that the first threshold value (Thd det) represents a signal value which is significantly higher than the signal value in normal driving situations, wherein the second threshold value (Thd drive) defines the upper limit of the signal value range in normal driving situations.
6. Sensor assembly (1) according to any one of claims 1 to 5, characterized in that the first sensor (N) and the at least one second sensor (M) detect the same physical quantity.
7. Sensor assembly (1) according to any one of claims 1 to 5, characterized in that the first sensor (N) and the at least one second sensor (M) detect different physical quantities.
8. The sensor assembly (1) according to any one of claims 1 to 7, characterized in that the first sensor (N) and the at least one second sensor (M) have the same detection direction or different detection directions.
9. Sensor assembly (1) according to any one of claims 1 to 8, characterized in that the first sensor (N) and the at least one second sensor (M) are arranged in the same mounting position or in different mounting positions.
10. The sensor assembly (1) according to any one of claims 1 to 9, characterized in that the evaluation and control unit (μ C) deactivates the first sensor (N) identified as faulty.
11. Method for monitoring a sensor (N) which detects at least one physical quantity related to a collision and outputs a corresponding at least one first sensor signal (S)N) Wherein the at least one first sensor signal (S)N) Is evaluated for recognizing a collision, wherein at least one second sensor signal (S) of at least one independent further sensor (M) detected for recognizing a collisionM) Is evaluated to identify a fault of the sensor (N), wherein if based on the at least one first sensor signal (S)N) Has a high signal value above a predetermined first threshold value (Thd det) and is based on the at least one second sensor signal (S) of the further sensor (M)M) For monitoring the sensor (N) while having a low signal value below a predetermined second threshold value (Thd drive), and/or a deviation function (f (S) resulting from the first and second comparison signalsN,{SM})) exceeds a predetermined third threshold value (Thd), a malfunction of the sensor (N) is identified.
12. Method according to claim 11, characterized in that the at least one first comparison signal corresponds to the at least one first sensor signal (S)N) And/or at least one processed first sensor signal (g (S)N) And the at least one second comparison signal corresponds to the at least one second sensor signal (S)M) And/or at least one processed second sensor signal (h)M(SM))。
13. Method according to claim 11 or 12, characterized in that, for identifying a fault, the first sensor signal (S)N) And/or the processed first sensor signal (g (S)N) During a predetermined minimum time period (T)min) Is higher than said first threshold value (Thd det), and said second sensor signal (S)M) And/or the processed second sensor signal (h)M(SM) Is lower than said second threshold value (Thd det) over a predetermined time range associated with said first sensor signal (S)N) And/or the processed first sensor signal (g (S)N) Of said minimum time period (T)min) And (4) overlapping.
14. Method according to any one of claims 11 to 13, characterized in that the first threshold value (Thd det) represents a signal value which is significantly higher than the signal value in normal driving situations, wherein the second threshold value (Thd drive) defines the upper limit of the signal value range in normal driving situations.
15. Method according to any one of claims 11 to 14, characterized in that said first sensor (N) and said at least one second sensor (M) detect the same physical quantity and have the same detection direction.
16. Method according to any of claims 11 to 15, characterized in that the first sensor (N) identified as faulty is deactivated.
Technical Field
The present invention relates to a sensor assembly for a vehicle according to the preamble of
Background
Vehicle collision detection in airbag control units is usually based on information from sensors installed in the vehicle, which are implemented, for example, as acceleration sensors and/or pressure sensors. The measurement signals from these sensors are suitably processed (e.g. filtered/integrated) and compared to trigger thresholds in order to make trigger decisions. In order to avoid false triggering by a faulty sensor, which would emit a higher signal value during normal driving operation and possibly exceed the trigger threshold value, a plausibility measure can be used, which requires the second, separate sensor to likewise output a signal value that is significantly higher than the signal value for normal driving situations. The trigger will only be released if the trigger decision of the main sensor is co-pending with the rationality decision of the individual sensor. However, this rationality alone is not sufficient to prevent false triggering. Thus, situations such as driving over a kerb stone with a sufficiently high acceleration signal may occur throughout the service life of the vehicle, wherein the rationality sensor will make a rationality decision. The risk of false triggering is therefore too great if the triggering decision is continuously pending due to a primary sensor failure. For this reason, the rationale solution is to simply provide time for sensor monitoring so that the sensor monitoring can identify the faulty sensor and then can deactivate the faulty sensor.
Methods for monitoring sensors are known from the prior art, which monitor the sensor hardware itself and usually operate in an initialization phase after power-up in the sensor itself. For example, the sensor can be offset in a targeted manner, and it can then be checked whether the sensor signal returns cleanly to "zero" after the offset has been cancelled. Furthermore, it is possible to monitor, for example, whether there is a manchester error in the communication between the sensor and the airbag control unit. Furthermore, a monitoring of the output sensor signal for checking certain plausible properties can also be carried out in the airbag control unit. Including for example offset monitoring and/or jump monitoring or gradient monitoring. The erroneous shift is characterized, for example, by a higher signal amplitude, which occurs over a longer period of time and corresponds to a non-physical higher deceleration. Jump errors or gradient errors are characterized, for example, by the fact that the change in the sensor value from one measured value to the next exceeds the range permitted by the filter curve of the sensor.
Disclosure of Invention
The sensor arrangement for a vehicle having the features of
The signal monitors known from the prior art concentrate on a few special cases (offset, jump) of a clearly faulty sensor signal. In practice, however, it has turned out that the sensor error patterns are so numerous that only a small part of them can be identified with these monitors. Depending on the origin of the sensor fault, the sensor error signal is also very similar to the impact signal and therefore in principle indistinguishable from the impact signal. The essence of the invention is that a faulty sensor signal is not identified by analyzing the one sensor signal itself. Instead, it is intended to use sensors which are always installed in the vehicle only for the purpose of implementing the rational concept described in the prior art. If a good sensor measures a high acceleration signal or pressure signal resulting from an extreme driving operation such as a collision or driving over a curb, other sensors in the vehicle will also measure the event, so that there is no sensor error. Conversely, if the higher sensor signal results from a sensor error, the other sensors only measure weaker signals in normal driving situations, so that a conclusion of the sensor error can be concluded. In an acceleration sensor, the signal value is, for example, of the order of 1g during normal driving operation.
Embodiments of the present invention advantageously reduce the risk of false triggers due to undetected sensor errors coinciding with a rationality determination resulting from aggressive driving maneuvers.
An embodiment of the present invention provides a sensor assembly for a vehicle, the sensor assembly having: a first sensor that detects at least one physical quantity related to a collision and outputs a corresponding at least one first sensor signal; at least one second sensor that detects at least one physical quantity related to the collision independently of the first sensor and outputs a corresponding at least one second sensor signal; and an evaluation and control unit which receives the at least one first sensor signal and the at least one second sensor signal and evaluates them to detect a collision. The evaluation and control unit monitors the first sensor using at least one second comparison signal based on at least one second sensor signal of the at least one second sensor, wherein the evaluation and control unit identifies a malfunction of the first sensor if the at least one first comparison signal based on the first sensor signal has a high signal value above a predetermined first threshold value and the at least one second comparison signal of the second sensor for monitoring the first sensor simultaneously has a low signal value below a predetermined second threshold value and/or a deviation function resulting from the first comparison signal and the second comparison signal exceeds a predetermined third threshold value.
Furthermore, a method for monitoring a sensor is proposed, which detects at least one physical variable associated with a collision and outputs a corresponding at least one first sensor signal, wherein the at least one first sensor signal is evaluated for collision detection. In this case, at least one second sensor signal of at least one separate further sensor, which is detected for detecting a collision, is evaluated in order to detect a malfunction of the sensor, wherein a malfunction of the sensor is detected if at least one first comparison signal based on the first sensor signal has a high signal value above a predetermined first threshold value and at least one second comparison signal based on the at least one second sensor signal of the further sensor for monitoring the first sensor simultaneously has a low signal value below a predetermined second threshold value and/or a deviation function resulting from the first comparison signal and the second comparison signal exceeds a predetermined third threshold value.
The at least one first comparison signal may correspond, for example, to the at least one first sensor signal and/or the at least one processed first sensor signal. The at least one second comparison signal may correspond, for example, to the at least one second sensor signal and/or the at least one processed second sensor signal.
In this context, an evaluation and control unit is understood to be an electrical device, for example a control unit, in particular an airbag control unit, which processes or evaluates sensor signals detected by the sensors. The evaluation and control unit can have at least one interface which is constructed on the basis of hardware and/or software. In a hardware-based configuration, the interface can be, for example, a part of a so-called ASIC system that contains the various functions of the evaluation and control unit. It is also possible that the interface is an integrated circuit of its own or is at least partly composed of discrete components. In a software-based configuration, the interface may be, for example, a software module that is present on the microcontroller together with other software modules. Also advantageous is a computer program product with a program code, which can be stored on a machine-readable carrier, such as a semiconductor memory, a hard disk memory or an optical memory, and which is used to carry out the evaluation when the program is run by the evaluation and control unit.
In this context, a sensor is understood to be an assembly comprising at least one sensor element which directly or indirectly detects a physical quantity or a change in a physical quantity and preferably converts it into a sensor electrical signal. The sensors may comprise, for example, pressure-sensitive sensor elements which determine the impact area on the vehicle and/or acceleration sensor elements which detect acceleration-related information of the vehicle and/or sensor elements which determine objects and/or obstacles and/or other impact-related vehicle surroundings data and provide these for evaluation. Such sensor elements may be based, for example, on video and/or radar and/or lidar and/or PMD and/or ultrasound technology. Furthermore, signals and information of the existing ABS sensor system and variables derived in a control unit provided for this purpose can also be evaluated.
The sensor arrangement for a vehicle in
It is particularly advantageous that the evaluation and control unit can calculate an absolute value function of the sensor signal, respectively, wherein the evaluation and control unit can perform a filtering and/or a window integration and/or an integration to process the sensor signal or the calculated absolute value function. There are many possibilities for signal processing. Typically, some kind of filtering will be used on the first sensor signal, since one does not want to identify an error based on a single high sensor value. For this purpose, the first sensor signal or the absolute value of the sensor signal can be filtered, windowed or integrated. The absolute value has the advantage that a strongly oscillating error signal also leads to a higher processed first sensor signal. When the first sensor signal exceeds a starting threshold, integration may then begin. The integration value may be reset or reset when the first sensor signal again falls below a reset threshold, which may be equal to the starting threshold.
In an advantageous embodiment of the sensor arrangement, the first threshold value can represent a signal value that is significantly higher than a signal value in normal driving situations. Further, the second threshold value may define an upper limit of the signal value range in the normal running condition. Thus, a triggering-critical sensor error of the first sensor to be monitored can be detected by the fact that the sensor to be monitored measures a high signal value which is significantly higher than in the normal driving situation, while the second sensor used for monitoring simultaneously measures a low signal value in the normal driving situation range.
In a further advantageous embodiment of the sensor arrangement, the first sensor and the at least one second sensor can detect the same physical variable. Alternatively, the first sensor and the at least one second sensor may detect different physical quantities. Furthermore, the first sensor and the at least one second sensor may have the same detection direction or different detection directions. Furthermore, the first sensor and the at least one second sensor may be arranged at the same mounting location or at different mounting locations. Thus, depending on the sensor configuration available in the vehicle, different schemes may be advantageously used to identify faulty sensors in different completion phases.
In a further advantageous embodiment of the sensor arrangement, the evaluation and control unit can deactivate the first sensor identified as faulty. The risk of false triggering due to a not yet discovered sensor error coinciding with a plausibility decision resulting from a violent driving maneuver is thereby significantly reduced.
In an advantageous embodiment of the method for monitoring the sensor, for detecting a fault, the first sensor signal and/or the processed first sensor signal may be above a first threshold value for a predetermined minimum time period, and the second sensor signal and/or the processed second sensor signal may be below a second threshold value for a predetermined time range, which may overlap the minimum time period for the first sensor signal and/or the processed first sensor signal. This means that the first sensor signal can be identified as possibly faulty if the first sensor signal and/or the processed first sensor signal is above the first threshold value within a predetermined minimum time period. This prevents a single and short-term erroneous measurement from leading to a deactivation of the sensor.
Drawings
Embodiments of the invention are illustrated in the drawings and are explained in more detail in the following description.
Fig. 1 shows a schematic block diagram of an embodiment of a sensor assembly for a vehicle according to the present invention.
Detailed Description
As can be seen from fig. 1, the illustrated embodiment of the
The at least one first comparison signal may correspond to the at least one first sensor signal SNAnd/or at least one processed first sensor signal (g (S)N)). The at least one second comparison signal may correspond to the at least one second sensor signal SMAnd/or at least one processed second sensor signal hM(SM)。
A sensor error critical for triggering of the first sensor N can thus be detected by the fact that the first sensor N to be monitored measures a high signal value S which is significantly higher than in normal driving situationsNAnd the sensor { M } (M ≠ N) used for monitoring has at the same time a low signal value S in the normal driving situation rangeM. Thus, that is to say, if the deviation function f (S) of the individual sensor signals is evaluated, as can be seen from the inequality (1)N,{SM) }) exceeds the third threshold value Thd, the sensor signal S may be usedMTo identify a sensor error of the first sensor N.
f(SN,{SM})>Thd (1)
The evaluation and control unit μ C deactivates the first sensor N identified as faulty.
This solution can be used for different finishing phases, depending on the sensor configurations available in the vehicle.
In the case of local monitoring, in addition to the first sensor N to be monitored, other sensors { M } are located in the same security zone in the vehicleAnd (5) installing the position. This is the case, for example, in airbag control units, in which there are usually a plurality of acceleration sensors which measure in the longitudinal and transverse directions and optionally also a rotational speed sensor which, for example, measures the rotational speed about a longitudinal or vertical axis. First sensor signal S measured by first sensor N to be monitored during a collision or in the event of a severe malfunction (e.g. driving over a curb stone)NAt large, the second sensor signal S is also always measured by the other sensors { M }M. This generally applies also to the case of measuring other physical quantities. Thus, for example, the first sensor N can measure a longitudinal acceleration and the at least one second sensor M can measure a lateral acceleration or a rotational speed about the longitudinal axis. However, it is of course advantageous that the physical quantities measured are as uniform as possible. In the ideal case, the sensor system is implemented in the airbag control unit in a redundant manner to achieve the plausibility described in the prior art, so that in addition to the main or first sensor N, there is a second acceleration sensor or second sensor M, which has the same measuring direction and very similar sensor properties, such as measuring range, resolution, filter properties.
In the first exemplary embodiment, the at least one second sensor M cannot be directly compared with the first sensor N, for example due to different detection directions or due to different sensor characteristics, so that the two sensor signals S cannot be evaluatedN、SMThe actual signal deviation of (2). It can also be seen from fig. 1 that at least one sensor M is represented by a set of second sensors { M }, which includes k second sensors M1. However, it is also possible to use only one second sensor M.
The embodiment of the
In the first embodiment, when the first signal g (S) is processedN) Exceeds a predetermined first threshold value Thd det and the processed second sensor signal hM(SM) If it is lower than the second threshold value Thd drive, a sensor error critical to the triggering of the first sensor N is identified. In this variant, the processed sensor signal { h) of each sensor MM(SM) Has to be below a corresponding sensor-specific second threshold. However, in a simplified embodiment, it is also possible to combine a set of processed signals h, for example by selecting the maximum valueM(SM) Combine into a single processed signal H ({ S })M}). The query is then reduced to two threshold comparisons. In the first embodiment, the first deviation function f (S)N,{SM}) is a boolean function which the evaluation and control unit μ C uses to evaluate at least one first sensor signal SNAnd at least one second sensor signal SMOr at least one processed first sensor signal g (S)N) And at least one processed second sensor signal hM(SM). Then, equation (1) becomes equation (2).
f(SN,{SM})＝[g(SN)>Thd_Det]&[H({SM})<Thd_drive](2)
Here, the general condition for the error recognition is f (S)N,{SM})>0。
There are many possibilities for the signal processing functions g and h. The processed first sensor signal g (S) will typically be comparedN) Some filtering is used because one does not want to identify errors based on a single high sensor value. For this purpose, the sensor signal S is sensedNOr the absolute value of the sensor signal | SN | is filtered, window integrated or integrated. The absolute value has the advantage that even strongly oscillating error signals lead to higher processed second ordersA sensor signal g (S)N). When the sensor signal SNBeyond the start threshold Thd start, the integration may be started. When the sensor signal SNAnd falls below a reset threshold value Thd _ reset, which may be equal to the start threshold value Thd _ start, the integration value may be reset or reset.
In principle, the function hMMay be used in a similar manner. However, in general, only weak filtering or no filtering at all will be performed here, since a sufficiently high signal value at one of the second sensors { M } already indicates that this signal value is a real driving situation. This is especially true if the second sensor { M } does not measure the same physical quantity as the first sensor N to be monitored.
The existing algorithmic features and thresholds of the evaluation and control unit μ C may also be used for both the signal processing functions g and h and the thresholds Thd _ det and Thd _ drive.
In the second embodiment there is a redundant sensor system, which means that the sensor N, M measures the same physical quantity, e.g. longitudinal acceleration. In this case, the actual driving situation or the collision should result in a comparable physical measurement of the sensor N, M within the tolerance range, i.e. the deviation between the sensors N, M can be evaluated directly. In general, there is only one second sensor M here, which is redundant for the first sensor N. Similar situations exist for pressure hose sensors that can be used to detect pedestrian impacts. Here, a pressure hose is integrated in the front of the vehicle, which pressure hose is closed on both sides by pressure sensors. The impact may deform the hose and cause the pressure at the two pressure sensors to rise. Depending on the impact location, the pressure signals at the two sensors differ from one another (for example due to the time-of-flight effect), but always have a similar order of magnitude.
Since in the second embodiment the sensor signal SN、SMAre directly compared with each other and it is therefore advantageous to minimize any nuances in the measurement range or filter characteristics or the influence of sensor noise by suitable pre-processing, such as signal definition or weak pre-filtering.
The deviation metric k (S) can then be directly calculatedN-SM). Likewise, an error or malfunction of the first sensor N is only recognized if there may be a significant deviation over a longer period of time. For this purpose, it is proposed to measure the first sensor signal SNAnd a second sensor signal SMSignal difference S betweenN-SMOr at the first sensor signal SNAnd a second sensor signal SMAbsolute value of signal difference between | SN-SML is filtered, windowed or integrated. In order to avoid the accumulation of small signal differences over a long time to the total threshold value, it is proposed to only apply the value | S of the signal differenceN-SMThe integration is started when l exceeds the start threshold value Thd start. When again below the reset threshold Thd _ reset (which may be equal to the starting threshold), the integration value may be reset or reset. If the calculation of the deviation measure k is to be limited to "normal driving situations" and large deviations that may occur in a collision are to be excluded, a condition according to inequality (3) can also be used as a starting condition for the filtering or summing of the difference signals:
|SN|>Thd1&|SM|<Thd2 (3)
wherein the thresholds Thd1 and Thd2 are higher than the first sensor signal SNAnd a second sensor signal SMAnd the first threshold value Thd1 is significantly higher than the second threshold value Thd2 to ensure a sufficient deviation. The accumulation is then only carried out if the monitoring sensor or the second sensor M measures a value in the normal driving range and the sensor to be monitored or the first sensor N measures a value outside the normal driving range due to a sensor error. The accumulated value may be reset or reset as long as one of the two conditions is no longer true. It is also particularly advantageous to evaluate the filtered or integrated signal difference not absolutely, but relatively, for example with reference to a similarly filtered or integrated first sensor signal SN。
The deviation metric k (S) calculated in this wayN-SM) The threshold comparison can now be used directly for error detection. In this case, equation (1) becomes equation (4).
f(SN,{SM})＝k(SN–SM)>Thd (4)
The solutions of the two embodiments can also be combined with one another particularly advantageously. Then, a sensor error is identified by a logical and combination of the three conditions. The first sensor N has a higher (processed) first sensor signal SN(g(SN)>Thd det) and the second sensor M has a lower (processed) second sensor signal SM(h(SM)<Thd drive) and there is a large signal deviation (k (S) between sensors N and MN-SM)>Thd)。
In this case, equations (1) to (5) and the deviation function are described by a boolean function having a zero threshold.
f(SN,{SM})＝[g(SN)>Thd_Det]&[h(SM)<Thd_drive]&[k(SN-SM)>Thd]>0 (5)
In the case of non-local monitoring, no other second sensor M is provided at the installation location of the first sensor N to be monitored. But these second sensors M are installed at other positions of the vehicle. This applies, for example, to peripheral acceleration sensors or pressure sensors.
In this case, in the event of a collision or misuse event, the different sensor signals S are spatially separatedN、SMThere may be large differences in amplitude and time characteristics between them, which makes direct comparison impossible. It is applicable, however, that each vehicle position and sensor mounting location experiences the same speed change after the end of an event, such as a crash, for example, a misuse event. In practice, certain deviations occur in the sensor, particularly in the area of intrusion, which can be attributed to clipping effects in the sensor or to a rotation of the sensor away from its original measuring direction. However, all sensors measure similar speed drops within a certain tolerance. Thus, if the signals of the acceleration sensors having the same measuring direction are integrated over the entire event, it is possible to obtain availability on a usual time scaleThe value of the comparison. In the event of a collision, sensors inside or near the intrusion zone detect this decrease in velocity faster than sensors outside the intrusion zone. In the event of a misuse event (e.g. local hammering) nearby sensors will measure significant velocity changes for a short time, however these changes will disappear soon.
Thus, if a higher speed drop value occurs only at the first sensor N over a certain period of time, while the other sensors M do not detect any significant speed drop over a similar time range, this indicates a sensor error of the first sensor N. In practice, in addition to the speed drop at the installation location of the first sensor N, other sufficiently filtered "macroscopic" first signals g (S) may also be usedN) The time scale of the filtering or integration is preferably selected to be long enough here to record the entire event. Similarly, the speed drop at the mounting position of the second sensor { M } may also be by the second sensor signal { h } processed in other waysM(SM) Description of the drawings. It is even advantageous here that the filtering is only carried out weakly or not at all, since a high signal value at the second sensor M already indicates the presence of a real event. This applies in particular to the case in which the second sensor M does not have the same detection direction as the first sensor N or measures a different physical quantity.
A general identification scheme for a sensor error of the first sensor N is then: processed first sensor signal g (S)N) During a predetermined minimum time period [ t; t + Tmin]Is higher than the first threshold value Thd det and the processed second sensor signal hM(SM) In a time interval encompassing a minimum time period [ t; t + Tmin]Is determined for a predetermined time range T-T1; t + T2]Is lower than a second threshold value Thd _ drive. The optional minimum time period Tmin may mask the first sensor signal g (S) in a local misuse event (e.g., hammer blow)N) A short-time phenomenon in which the first threshold value Thd _ det is exceeded. Alternatively, this can also be ensured by selecting a sufficiently high first threshold value Thd det. The two optional time periods TI and T2 take into account different time characteristics at different measurement positions. This means that if the first passProcessed sensor signal g (S) of sensor NN) If, for a sufficiently long time, the first threshold value Thd _ det is exceeded and if no abnormal signals, i.e. only signals in the normal driving range, are measured by the other sensors M in the vehicle for a certain period of time before and after the first sensor N exceeds this threshold value, a sensor error or fault is present in the first sensor N and a sensor error or fault of the first sensor N is detected.
According to an embodiment of the method according to the invention for monitoring a sensor N, at least one second sensor signal S of at least one individual further sensor M, which is detected for detecting a collision, is evaluatedMTo identify a malfunction of the sensor N, wherein the sensor N detects at least one physical quantity related to a collision and outputs a corresponding at least one first sensor signal SNWherein the at least one first sensor signal SNIs evaluated for identifying a collision. Here, if based on the first sensor signal SNHas a high signal value above a predetermined first threshold value Thd det and is based on at least one second sensor signal S of the other sensor MMHas a low signal value below a predetermined second threshold value Thd drive, a malfunction of the sensor N is identified. Additionally or alternatively, if the deviation function f (S) is generated by the first comparison signal and the second comparison signalN,{SM}) exceeds a predetermined third threshold value Thd, a malfunction of the sensor N is identified.
As mentioned above, the at least one first comparison signal may correspond to the at least one first sensor signal SNAnd/or at least one processed first sensor signal g (S)N) And the at least one second comparison signal may correspond to the at least one second sensor signal SMAnd/or at least one processed second sensor signal hM(SM)。
Here, the first threshold value Thd _ det represents a signal value that is significantly higher than the signal value in the normal driving situation. The second threshold value Thd _ drive defines the upper limit of the signal value range in the normal running condition.
If the second sensor does not verify the signal of the first sensor, embodiments of the present invention identify the faulty sensor as faulty and deactivate the sensor identified as faulty.
- 上一篇：一种医用注射器针头装配设备
- 下一篇：电动车辆的前端面板组件