JP2006240453A - Sensor failure detector and detection method of sensor failure - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To highly reliably detect sensor failure such as misalignment difficult to identify from the appearance of a vehicle outside monitoring sensor installed on a vehicle, using a configuration suitable for constant detection under an actually running environment with little processing burden. <P>SOLUTION: An impact received by a ranging radar 2a for a vehicle 1a is detected by an accelerator sensor 3. Based on impact detection of a fixed area where sensor failure such as misalignment difficult to identify from the appearance of the sensor 3 occurs, a failure judgement device detects sensor failure of the ranging radar 2a, judging that the ranging radar 2a has some failure. A judgement result processor retains the detection results of the sensor failure as failure diagnosis information and issues a warning output. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a distance measurement radar as an out-of-vehicle monitoring sensor mounted on a vehicle, a sensor abnormality detection device and a sensor abnormality detection method for detecting a sensor abnormality due to an axial deviation of an image sensor.

  Conventionally, vehicles that perform ACC (Adaptive Cruise Control) follow-up control, damage reduction automatic brake control, brake assist control, etc., detect obstacles such as vehicles and pedestrians outside the vehicle (mainly in front of the vehicle). In order to calculate the collision prediction time, a ranging radar such as a laser radar or a millimeter wave radar, or an image sensor such as a CCD monocular camera or a stereo camera is mounted as an outside monitoring sensor.

  These out-of-vehicle monitoring sensors are attached to the vehicle as accurately as possible so that the position and angle of the mounting are the desired position and angle so as not to cause misalignment of the exploration results or captured images during traveling. .

  However, if an object such as a ball hits the outside monitoring sensor while the vehicle is running, a sensor abnormality such as a shaft misalignment may occur in the outside monitoring sensor depending on the impact. It may also occur due to sudden braking of the vehicle.

  If these sensor abnormalities occur, the position and size of the obstacles will not be recognized correctly, so the collision prediction time will be calculated based on this incorrect recognition, and follow-up driving control, damage reduction automatic It is not preferable to perform brake control, brake assist control, or the like.

  Therefore, for the ranging radar mounted on the vehicle, the center position of the preceding vehicle and the center position of the radar search range are determined by statistical processing from the results of multiple searches under traveling conditions in which the vehicle travels straight ahead following the preceding vehicle. It has been proposed to detect the deviation of the axis as the axis deviation or to detect the axis deviation from the amount of left-right movement of the detection position of the roadside reflector under the traveling condition in which the vehicle goes straight (for example, a patent) Reference 1).

Japanese Patent Laid-Open No. 11-142520 (paragraphs [0017]-[0019], [0045]-[0070], [0080]-[0090], FIGS. 1, 2, and 6)

  The detection of sensor abnormality in the conventional proposal (described in Patent Document 1) can only be performed under limited driving conditions in which the vehicle goes straight, and in an actual driving environment where the vehicle is not only going straight, but also turning or turning. It is difficult to apply to the constant detection of sensor abnormalities of the ranging radar and image sensor outside the vehicle, and complicated processing dedicated to sensor abnormality detection such as statistical processing is necessary, increasing the processing burden There is also a problem.

  In addition, since the sensor abnormality is obtained by statistical processing or is detected from the amount of movement of the detection position of the roadside reflector whose position is inaccurate, it is difficult to detect accurately and accurately. There is also a problem of low reliability.

  In particular, it is important and desirable to detect a sensor error of an axis deviation that is difficult to judge from the appearance even if the driver visually observes it.

  The present invention has a configuration that is suitable for continuous detection in an actual driving environment and has a low processing load, and can detect a sensor abnormality such as an axis deviation that is difficult to judge from the appearance of an outside monitoring sensor mounted on a vehicle with high reliability. The purpose is to be able to do it.

  In order to achieve the above-described object, a sensor abnormality detection device according to the present invention is an impact detection device for detecting an impact received by an outside monitoring sensor on a vehicle equipped with at least one of a ranging radar and an image sensor as the outside monitoring sensor. An abnormality determining means for detecting a sensor abnormality by determining that the vehicle exterior monitoring sensor is abnormal by detecting an impact which is greater than or equal to a value generated when an object of the impact sensor collides with and less than a predetermined value; and the abnormality determining means The above-mentioned sensor abnormality detection result is stored as failure diagnosis information and a determination result processing means for outputting an alarm is provided (claim 1).

  In the sensor abnormality detection device of the present invention, the abnormality determination by the abnormality determination unit is performed when an impact detection that is greater than or equal to a value that occurs when the object of the impact sensor collides and is less than or equal to a predetermined value continues for a predetermined time. (Claim 2).

  Next, a sensor abnormality detection device according to the present invention is provided in a vehicle in which at least one of a ranging radar and an image sensor is mounted as a vehicle outside monitoring sensor, the detection result of the vehicle outside monitoring sensor and the vehicle behavior monitoring means. Collision prediction calculation means for calculating the prediction time of collision between the own vehicle and an object outside the vehicle based on the behavior monitoring result, and when the error time set from the point of time of the collision prediction time elapses in a predetermined collision prediction state, When the collision-free state continues, an abnormality determination unit that determines that the out-of-vehicle monitoring sensor is abnormal and detects a sensor abnormality, and holds the detection result of the sensor abnormality of the abnormality determination unit as failure diagnosis information And a determination result processing means for outputting an alarm (claim 3).

  Furthermore, the sensor abnormality detection device according to the present invention is provided with a use prohibition function for invalidating the detection result of the vehicle monitoring sensor and prohibiting the use by detecting the sensor abnormality of the abnormality determination unit. It is characterized (claim 4).

  Next, in the sensor abnormality detection method of the present invention, a vehicle equipped with at least one of a ranging radar and an image sensor as a vehicle outside monitoring sensor is provided with a shock sensor for detecting a shock received by the vehicle outside monitoring sensor, It is determined that the vehicle monitoring sensor is abnormal by detecting an impact that is greater than or equal to a value that occurs when the sensor object collides and is less than or equal to a predetermined value, detects the sensor abnormality, and stores the detection result of the sensor abnormality as failure diagnosis information. An alarm is output (claim 5).

  The sensor abnormality detection method according to the present invention is based on a detection result of the vehicle monitoring sensor and a behavior monitoring result of the own vehicle by a vehicle equipped with at least one of a ranging radar and an image sensor as the vehicle monitoring sensor. When a predicted collision time between the vehicle and the object outside the vehicle is calculated, and the vehicle is in a collision-free state even when the prediction error time set from the elapsed time of the predicted collision time elapses in a traveling state adapted to a predetermined collision condition The vehicle outside monitoring sensor is determined to be abnormal, a sensor abnormality is detected, and the detection result of the sensor abnormality is held as failure diagnosis information and output as an alarm (Claim 6).

  First, according to the first and fifth aspects of the present invention, the impact received by the outside radar sensor of the ranging radar or the image sensor is directly detected by the obstacle sensor, and the impact which is not less than a predetermined value and not less than a value generated when the object collides. Can be determined that the outside monitoring sensor is abnormal based on the impact detection of the impact sensor, the sensor abnormality can be detected, and an alarm can be output.

  In this application, the “value generated when an object collides” is difficult to judge from the appearance when an object outside the vehicle such as a ball collides with the vehicle outside monitoring sensor, but in reality, an axis deviation occurs in the vehicle outside monitoring sensor. The lower limit magnitude (level) impact value, and the “predetermined value” is an impact value of a magnitude (level) just before the axis deviation of the vehicle exterior monitoring sensor can be clearly seen.

  For this reason, misalignment of the sensor due to an extremely small unnecessary impact that does not cause a misalignment in the outside monitoring sensor is prevented, and the shaft is clearly misaligned due to, for example, a damaged radar mounting portion. Except for the case of receiving an impact that can be recognized, it is possible to reliably detect a sensor error of an axis deviation that is difficult to judge from the appearance.

  In addition, sensor abnormalities can be detected regardless of vehicle driving conditions, etc., making it suitable for continuous detection in actual driving environments, requiring no complicated processing dedicated to sensor abnormality detection such as statistical processing. There is also an advantage of less.

  Therefore, with a configuration that is suitable for continuous detection in an actual driving environment and has a low processing load, reliable and reliable detection of an axis misalignment sensor abnormality that is difficult to judge from the appearance of an out-of-vehicle monitoring sensor mounted on a vehicle is performed. be able to.

  Further, according to the configuration of claim 2, the fact that the impact detection of the impact sensor continues for a predetermined time acts as a filter for preventing false detection, so that when the vehicle outside object collides with the impact sensor for some reason. Even if an instantaneous detection of an impact that is greater than the generated value and less than or equal to the predetermined value occurs, this detection does not erroneously detect a sensor abnormality of the vehicle monitoring sensor, and the detection accuracy of the sensor abnormality is further improved.

  Next, according to the configurations of the third and sixth aspects, the detection result of the vehicle monitoring sensor and the behavior monitoring of the own vehicle are detected even though there is no possibility of collision due to the occurrence of a sensor abnormality such as an axial deviation of the vehicle monitoring sensor. If the collision prediction time is erroneously calculated from the vehicle, even if a predetermined error time elapses from the collision prediction time by running as it is (with the predetermined collision prediction state) without performing an avoidance operation etc. Since the non-collision state continues, it is possible to detect a sensor abnormality of the outside monitoring sensor and output an alarm.

  In this case as well, sensor abnormality can be detected regardless of the driving conditions of the vehicle, which is suitable for continuous detection under actual driving environment, and complicated processing dedicated to sensor abnormality detection such as statistical processing is also possible. Unnecessary and less processing burden.

  Furthermore, in order to detect a sensor abnormality based on the actual progress of the collision prediction time calculated from the detection result of the outside monitoring sensor, rather than the stochastic detection by statistical processing, the sensor abnormality such as an axis deviation of the outside monitoring sensor is detected. It can be detected more reliably and accurately than when it is detected by statistical processing or the like.

  Therefore, it is suitable for continuous detection in an actual driving environment, has a low processing burden, and has high detection accuracy and reliability. It can be detected with high accuracy.

  Next, according to the configuration of the fourth aspect, after the sensor abnormality of the outside monitoring sensor is detected as described above, the ACC follow-up running control of the erroneous detection result of the outside monitoring sensor, the damage reduction automatic brake control. In addition, use for brake assist control or the like is prohibited, and safety can be improved.

  Next, in order to describe the present invention in more detail, the embodiment will be described in detail with reference to FIGS.

<First Embodiment>
First, a first embodiment using an impact sensor will be described with reference to FIGS.

  FIG. 1 is a block diagram of a sensor abnormality detection device for a vehicle (own vehicle) 1a, and FIG. 2 is an explanatory diagram of a ranging radar 2a as a vehicle exterior monitoring sensor of FIG.

  In the case of this embodiment in which the front of the vehicle 1a (the front of the host vehicle) is the monitoring range, the ranging radar 2a of FIG. 1 mounted as an out-of-vehicle monitoring sensor on the vehicle 1a is, for example, an appropriate front bumper portion of the vehicle 1a. As shown in FIG. 2, an acceleration sensor 3 as an impact sensor is built in and formed into a built-in acceleration sensor type. The acceleration sensor 3 directly detects at least the impact force in the axial direction (front-rear direction) of the vehicle 1a.

  In FIG. 1, 4a, 5a, 6a are a vehicle speed sensor, a rudder angle sensor, and a yaw rate sensor for detecting the behavior of the vehicle 1a, forming own vehicle behavior detection means, and the vehicle speed sensor 4a is a so-called wheel speed sensor. The vehicle speed signal is output, the steering angle sensor 5a outputs the steering angle signal based on the steering operation (steering operation) of the driver, and the yaw rate sensor 6a outputs the signal of the yaw rate value generated based on the steering operation or the like. .

  Reference numeral 7a denotes an arithmetic processing unit composed of an ECU having a microcomputer configuration, which forms sensor abnormality determination means and determination result processing means described later. An EEPROM 8a as a nonvolatile memory holds the detection result of the sensor abnormality of the sensor abnormality determination means as failure diagnosis information. Reference numeral 9a denotes a display alarm unit that displays the failure diagnosis information (diagnostic information) or the like on the driver of the vehicle 1a and gives an alarm.

  When the vehicle 1a is started (engine start), the ranging radar 2a repeatedly outputs a laser or a millimeter wave while repeatedly scanning the search range R in FIG. Is detected and an obstacle such as a vehicle in front of the host vehicle is searched and detected, and the detection result is taken into the arithmetic processing unit 7a every moment.

  An arithmetic processing unit also detects the detection signal of the impact force that is proportional to the acceleration of the acceleration sensor 3 and the signals of the vehicle speed, the steering angle, and the yaw rate value of the vehicle speed sensor 4a, the steering angle sensor 5a, and the yaw rate sensor 6a. 7a.

  Then, the arithmetic processing unit 7a executes a first sensor abnormality detection processing program set in advance, thereby forming first abnormality determination means and first determination result processing means described below.

(1) First abnormality determination means This means detects the sensor abnormality by determining that the ranging radar 2a is abnormal by detecting an impact which is greater than or equal to a value generated when the object of the acceleration sensor 3 collides and less than a predetermined value. A lower threshold of the magnitude of the minimum impact force value that is difficult to judge from the appearance due to the collision of an object such as a ball but actually causes an axis shift in the ranging radar 2a; An upper limit threshold value corresponding to an impact value of a size (level) that is large and apparently clearly indicates the axial deviation of the vehicle exterior monitoring sensor is preset, and the detected impact force of the acceleration sensor 3, that is, When the impact force received by the ranging radar 2a is greater than or equal to the lower threshold and less than or equal to the upper threshold, the sensor error of the axis deviation of the ranging radar 2a is detected.

  By the way, the detection of the sensor abnormality may be performed immediately when the detected acceleration sensor 3 detects an impact that is not less than the lower limit threshold and not more than the upper limit threshold. However, the acceleration sensor 3 lowers the lower limit for some reason. Even if an impact that is greater than or equal to the threshold value and less than or equal to the upper threshold value is erroneously detected, the detection error of the ranging radar 2a is not detected by this erroneous detection, and the detection accuracy of the sensor abnormality can be improved. Therefore, in this embodiment, the filter condition for preventing false detection is that the acceleration sensor 3 continues to detect the impact of the acceleration sensor 3 not less than the lower threshold and not more than the upper threshold for a predetermined time. Set as

  Then, based on the setting of the filter condition, the first abnormality determination means is actually when the detection of the impact that is greater than or equal to the lower threshold value and less than or equal to the upper threshold value of the acceleration sensor 3 continues for the predetermined time, A sensor abnormality of the axis deviation of the ranging radar 2a is detected.

  The predetermined time is a short time, such as a time when a thrown ball or the like collides and the impact is applied, and is set based on an experiment or the like.

(2) First determination result processing means This means holds the detection result of the axis deviation sensor abnormality of the first abnormality judgment means in the EEPROM 8a as failure diagnosis information, and outputs an alarm from the display / alarm unit 9a. To do.

  Further, in this embodiment, a use prohibition function is added to the first determination result processing means, and an invalid flag is added to the detection result of the ranging radar 2a based on the detection of the sensor abnormality of the first abnormality determination means. The detection result of the ranging radar 2a is invalidated and the erroneous detection result of the ranging radar 2a is prohibited from being used for ACC follow-up control, damage reduction automatic brake control, brake assist control, or the like.

  Based on the above configuration, a certain object actually collides with the ranging radar 2a while the vehicle 1a is traveling, etc., and the ranging radar 2a receives an impact force not less than the lower threshold and not more than the upper threshold. When an axis misalignment sensor abnormality occurs in 2a, the detection of the impact that is greater than or equal to the lower threshold value and less than or equal to the upper threshold value of the acceleration sensor 3 continues for a predetermined time or longer. The first abnormality determination means of the arithmetic processing unit 7a determines that the ranging radar 2a is abnormal and detects a sensor abnormality.

  In this case, the impact received by the ranging radar 2a can be directly detected by the acceleration sensor 3, and further, detection is not required from the result of the direct detection that the distance measurement radar 2a is smaller than the lower limit threshold value and no axis deviation occurs. Erroneous detection based on a large impact is prevented, and an impact that clearly shows an axis deviation of the ranging radar 2a with a larger appearance than the upper limit threshold of the ranging radar 2a is not detected. As described above, it is possible to prevent erroneous detection due to an instantaneous impact force below the upper limit threshold value, and to accurately detect the sensor error of the axis deviation of the distance measuring sensor 2 which is not apparent even when visually recognized by a driver or the like. Can be detected well.

  Next, the first determination result processing means of the arithmetic processing unit 7a stores and holds the detection result of the axis deviation sensor abnormality of the first abnormality determination means in the EEPROM 8a as nonvolatile diagnosis information. The failure diagnosis information of the detection result of the sensor abnormality can be output as a warning from the display / alarm unit 9a, and the detected sensor abnormality of the axis deviation of the distance measuring sensor 2a can be notified to the driver or the like.

  Further, after detecting the sensor error of the axis deviation of the ranging radar 2a, the ACC of the ranging result (detection result) including the axis deviation error of the ranging radar 2a by the use prohibition function of the first determination result processing means. Can be prohibited from being used for follow-up running control, damage-reducing automatic brake control, brake assist control, etc., and safety of the vehicle 1a can be improved.

  Based on the alarm output of the display / alarm unit 9a, a driver or the like carries the vehicle 1a into a repair shop or the like to repair the axis deviation of the distance measuring sensor 2a, and at that time, the fault diagnosis information of sensor abnormality in the EEPROM 8a is deleted. Thus, the state before the occurrence of the abnormality (initial state) is restored.

  In this case, since the sensor abnormality of the ranging radar 2a can be detected regardless of the traveling state of the vehicle 1a, etc., it is suitable for continuous detection of the sensor abnormality in the actual traveling environment. Complex processing dedicated to sensor abnormality detection is unnecessary, and the processing load on the arithmetic processing unit 7a can be reduced.

  Therefore, with a configuration that is suitable for continuous detection in an actual driving environment and has a small processing load, even if the driver of the ranging radar 2a as an out-of-vehicle monitoring sensor mounted on the vehicle 1a is visually checked, it is difficult to make a judgment from the appearance. A reliable and highly reliable detection of an axis misalignment sensor abnormality can be performed.

<Second Embodiment>
Next, a second embodiment will be described with reference to FIGS.

  FIG. 3 is a block diagram of the sensor abnormality detection device for the vehicle (own vehicle) 1b, FIG. 4 is a flowchart for explaining the operation of FIG. 3, and FIG.

  The vehicle 1b is equipped with a ranging radar 2b such as a general laser radar or a millimeter wave radar that does not include the acceleration sensor 3 as a vehicle exterior monitoring sensor in front of the vehicle, and as a vehicle behavior detection means, FIG. A vehicle speed sensor 4b, a steering angle sensor 5b, and a yaw rate sensor 6b similar to the sensors 4a to 6a are provided.

  Moreover, it comprises the arithmetic processing unit 7b which consists of ECU of the same microcomputer structure as the arithmetic processing unit 7a of FIG. 1, and this arithmetic processing unit 7b executes the preset 2nd sensor abnormality detection processing program, Next, a collision prediction calculation unit, a second abnormality determination unit, and a second determination result processing unit are formed.

(1) Collision prediction calculation means This means is also used for calculation of collision prediction time, such as follow-up running control, damage reduction automatic brake control, brake assist control, etc., and the detection result of the ranging radar 2b and own vehicle behavior monitoring means Based on the behavior monitoring result of the vehicle 1b of each of the sensors 4b to 6b, a predicted collision time Ti between the vehicle 1b and an external object such as a preceding vehicle or an obstacle is calculated.

  Specifically, the collision prediction is performed every moment based on the distance Li from the vehicle 1b to the object outside the vehicle obtained from the transmission / reception time difference of the ranging radar 2b and the like and the relative speed Vi of the object outside the vehicle obtained from the time change of the distance Li. Time Ti is calculated from the calculation of Ti = Li / Vi.

(2) Second abnormality determination means This means is that the vehicle 1b is in a collision-free state when a set error time Te of about 0.5 seconds has elapsed from the time when the collision prediction time Ti has elapsed in a predetermined collision prediction state. Is maintained, the ranging radar 2b is determined to be abnormal and the sensor abnormality is detected.

  In other words, the collision prediction time Ti is calculated from the detection result of the ranging radar 2b and the behavior monitoring result of the own vehicle behavior monitoring means, and the collision between the own vehicle 1b and an outside vehicle such as a preceding vehicle or an obstacle is predicted. In such a case, even if the collision prediction time Ti + the error time Te has elapsed from the time of calculation of the collision prediction time Ti in a predetermined collision prediction state without performing a collision avoidance operation or the like, If this collision does not occur, it is determined that the ranging radar 2b is abnormal, and the sensor abnormality is detected.

  For this reason, the second abnormality determination means is a function used for collision prediction such as follow-up running control, damage reduction automatic brake control, and brake assist control, for example, from the time when the collision prediction time Ti is calculated to the collision prediction time Ti + error time. A function for measuring the time of Te and the detection position P1, P2, P3, P4 of the detection object P1, P2, P3, and P4 of the outside object at each detection time as shown in FIG. After the collision prediction time Ti has elapsed in the figure, the relative movement prediction trajectory of the vehicle-like object as shown by the solid line α in the figure is obtained from the data (past data) held at the time of calculation of the object by estimation calculation such as the least square method. The deceleration based on the change in the vehicle speed, the steering angle, and the yaw rate are within a predetermined range based on the function for obtaining the predicted collision position Pi of the vehicle and the detection results of the sensors 4b to 6b of the vehicle behavior detection means. Whether, i.e., and a function of detecting the presence or absence of the avoidance operation. Note that the z axis and the x axis in FIG. 5 are coordinate axes in the front-rear direction (travel direction) and the lateral direction with the vehicle center β in the lateral width direction of the vehicle 1b as the origin.

  Then, based on the results of timing and detection of these functions, the condition “predicted collision time Ti + error time Te has elapsed, and the predicted collision position Pi calculated from past data is within a predetermined range from the vehicle center β in FIG. Is detected, and the collision avoidance operation has not been detected from the behavior of the vehicle, for example, there is no collision detection by the pressure sensor or the like of the vehicle 1b, and the vehicle 1 continues running and the like. Therefore, when it is recognized that there is no collision in the vehicle 1b, it is determined that an abnormality such as an axis deviation has occurred in the distance measuring sensor 2b, and the sensor abnormality is detected.

(3) Second determination result processing means This means has substantially the same configuration as the first determination result processing means of the first embodiment, and the sensor abnormality detection result of the second abnormality determination means is determined as a failure. It is stored in the EEPROM 8b as diagnostic information, an alarm is output from the display / alarm unit 9b, and the detection result of the ranging radar 2b is obtained by the added use prohibition function based on the detection of the sensor abnormality of the second abnormality determination means. An invalid flag is added to invalidate the detection result of the ranging radar 2b and prohibit the use of the erroneous detection result of the ranging radar 2b for ACC follow-up control, damage reduction automatic brake control, brake assist control, and the like.

  Based on the above configuration, when the vehicle 1b starts (engine starts), the arithmetic processing unit 7b executes, for example, the processes of steps S1 to S10 in FIG. Note that steps Q1 to Q4 in FIG. 4 are processes in the case of combining detection of an impact sensor described later.

  Then, the detection results of the distance measuring radar 2b in front of the vehicle and the signals of the vehicle speed, the steering angle, and the yaw rate value of the sensors 4b, 5b, 6b from time to time are taken into the arithmetic processing unit 7b. The collision prediction calculation means calculates the collision prediction time Ti, and the second abnormality determination means calculates the collision prediction position Pi, the steering angle, the yaw rate, the deceleration from the own vehicle speed change, etc. in steps S2 to S5.

  Further, the process proceeds from step S5 to step S6, and in steps S6 to S8, whether or not the predicted collision time Ti + error time Te has elapsed, whether the predicted collision position Pi is within a predetermined range from the vehicle center β, deceleration It is determined whether the steering angle and the yaw rate are within a predetermined range. When all of these determinations are affirmative (YES) and the above conditions are satisfied, the distance measuring sensor 2b is abnormal such as an axis deviation in step S9. Is detected, and sensor abnormality is detected.

Then, the process proceeds from step S9 to step S10, and the second determination result processing means
The sensor abnormality detection result is held in the EEPROM 8b as failure diagnosis information, and an alarm is output from the display / alarm unit 9b.

  At this time, an invalid flag is added to the detection result of the ranging radar 2b and the detection result of the ranging radar 2b is invalidated by the added use prohibition function based on the detection of the sensor abnormality of the second abnormality determining means. Use of the erroneous detection result of the ranging radar 2b for ACC follow-up control, damage reduction automatic brake control, brake assist control, etc. is prohibited.

  Therefore, in the case of this embodiment, by calculating the predicted collision time Ti and the like, the axis of the ranging radar 2b as the vehicle monitoring sensor is used without using an impact sensor such as the acceleration sensor 3 of the first embodiment. Sensor abnormality such as deviation can be detected.

  Also in this embodiment, the sensor abnormality of the ranging radar 2b can be detected regardless of the traveling condition of the vehicle 1b, etc., which is suitable for continuous detection in an actual traveling environment, and statistical processing or the like. No complicated processing dedicated to sensor abnormality detection is required and the processing load is small, and the sensor is not based on stochastic detection by statistical processing, but based on the actual progress of the predicted collision time Ti calculated from the detection result of the ranging radar 2b. In order to detect an abnormality, it is possible to detect a sensor abnormality including an axis misalignment that is difficult to judge from the appearance even if the driver of the ranging radar 2b visually observes the sensor more accurately than when it is detected by statistical processing or the like. it can.

  Therefore, without providing a collision sensor such as the acceleration sensor 3, it is suitable for continuous detection in an actual driving environment, has a low processing load, and has a high detection accuracy and reliability. It is possible to detect a sensor abnormality including an axis deviation of the ranging radar 2b.

  The present invention is not limited to the above-described embodiment, and various modifications other than those described above can be made without departing from the spirit of the present invention.

  For example, in the second embodiment, the detection of the impact sensor is combined, and the ranging radar 2b is formed by the ranging radar 2b including the acceleration sensor 3 similar to the ranging radar 2a of the first embodiment, and the arithmetic processing When the first abnormality determining means of the arithmetic processing unit 7a is added to the apparatus 7b, and an object outside the vehicle actually collides with the ranging radar 2b to cause a shaft misalignment, the sensor abnormality is detected from the impact detection result of the acceleration sensor 3. You may make it detect.

  In this case, the processing unit 7b performs, for example, steps S1 to S10 and steps Q1 to Q4 in FIG. 4, and the detected impact force of the acceleration sensor 3 is detected by step Q1 provided between steps S5 and S6. When it is determined that it is greater than or equal to the lower threshold value and less than or equal to the upper threshold value, it is only necessary to detect an axis deviation sensor abnormality of the ranging radar 2b through steps Q2 to Q4 in the same manner as in the first embodiment. By doing so, in addition to the effect of the second embodiment, an axis misalignment sensor abnormality that is difficult to judge even if the driver of the ranging radar 2b or the like when the object collides is visually observed. Thus, there is an effect that it can be detected more reliably by actual impact detection.

  Next, in both the embodiments, the vehicle outside monitoring sensor is the ranging radar 2a, 2b, but the vehicle outside monitoring sensor can be similarly applied even when the vehicle outside monitoring sensor is an image sensor such as a CCD monocular camera or a stereo camera. The present invention can be applied in the same manner even when the vehicle monitoring sensor is both a ranging radar and an image sensor.

  In addition, when monitoring the front, the image sensor is provided, for example, by adjusting the angle on the center mirror support portion in the vehicle, and the position, distance, shape, and the like of the object outside the vehicle are detected by well-known image recognition processing of the captured image. The In addition, when the vehicle monitoring sensor is both a ranging radar and an image sensor, the position, distance, shape, etc. of the object outside the vehicle are detected by a well-known sensor fusion based on the detection result of the ranging radar and the imaging result of the image sensor, for example. Is done.

  Further, the vehicle outside monitoring sensor may be one that monitors and detects the rear, side, or entire circumference of the vehicles 1a and 1b, and the present invention can be similarly applied to these cases.

  Next, for example, in the first embodiment, the acceleration sensor 3 only needs to detect the impact received by the ranging radar 2a, and may not be built in the ranging radar 2a, and may be in the vicinity of the ranging radar 2a. It may be provided.

  In addition, the impact sensor of the present invention is not limited to the acceleration sensor 3 and may be various sensors that detect an impact received by an external monitoring sensor such as a pressure sensor. In addition, when the vehicle monitoring sensor is both a ranging radar and an image sensor, an impact sensor is provided separately to detect each impact received, and the larger detected impact is preferentially adopted. For example, a common impact sensor may be provided on the ranging radar side, and the detected impact may be detected as an impact received by the ranging radar and the image sensor.

  Next, it is a matter of course that the arithmetic processing devices 7a and 7b may perform processing in a procedure different from that shown in FIG. 4, and for example, only the lower limit value is set in the arithmetic processing device 7a to measure the distance. An abnormality in the sensor of the axis deviation of the radar 2a may be detected because an impact sensor such as the acceleration sensor 3 receives an impact force equal to or higher than a lower limit threshold value. Further, the EEPROMs 8a and 8b as nonvolatile storage means may be provided in the ranging radars 2a and 2b, and the nonvolatile storage means may be formed by a storage element other than the EEPROM.

  By the way, in order to reduce the number of equipment parts of the vehicles 1a and 1b, the present invention can be applied to the case where each sensor 2a, 2b, 3a, 3b, 4a, 4b, etc. is also used as another control sensor.

It is a block diagram of a 1st embodiment of this invention. It is explanatory drawing of the ranging radar of FIG. It is a block diagram of 2nd Embodiment of this invention. It is a flowchart for operation | movement description of FIG. It is explanatory drawing of collision prediction.

Explanation of symbols

1a, 1b Vehicle 2a, 2b Ranging radar 3 Acceleration sensor 4a, 4b Vehicle speed sensor 5a, 5b Steering angle sensor 6a, 6b Yaw rate sensor 7a, 7b Arithmetic processing unit 8a, 8b EEPROM
9a, 9b Display alarm unit

Claims (6)

In vehicles equipped with at least one of ranging radar and image sensor as a vehicle monitoring sensor,
An impact sensor for detecting an impact received by the vehicle exterior monitoring sensor;
An abnormality determining means for determining that the vehicle exterior monitoring sensor is abnormal by detecting an impact that is greater than or equal to a value that occurs when the object of the impact sensor collides and less than or equal to a predetermined value;
A sensor abnormality detection device, comprising: a determination result processing unit that holds the detection result of the sensor abnormality of the abnormality determination unit as failure diagnosis information and outputs an alarm. In the sensor abnormality detection device according to claim 1,
A sensor abnormality detection device, wherein abnormality determination by an abnormality determination means is performed when an impact detection that is greater than or equal to a value generated when the object of the impact sensor collides and less than or equal to a predetermined value continues for a predetermined time. In vehicles equipped with at least one of ranging radar and image sensor as a vehicle monitoring sensor,
A collision prediction calculation means for calculating a collision prediction time between the own vehicle and an object outside the vehicle based on the detection result of the outside monitoring sensor and the behavior monitoring result of the own vehicle behavior monitoring means;
When the error time set from the time when the collision prediction time has elapsed in a predetermined collision prediction state has elapsed and the vehicle's no-collision state continues, it is determined that the vehicle monitoring sensor is abnormal and a sensor abnormality is detected. An abnormality determination means to detect;
A sensor abnormality detection device, comprising: a determination result processing unit that holds the detection result of the sensor abnormality of the abnormality determination unit as failure diagnosis information and outputs an alarm. In the sensor abnormality detection device according to any one of claims 1 to 3,
A sensor abnormality detection device, wherein a use prohibition function is added to a determination result processing means to invalidate a detection result of a vehicle-external monitoring sensor by detecting a sensor abnormality of an abnormality determination means and to prohibit use. A vehicle equipped with at least one of a ranging radar and an image sensor as an outside monitoring sensor is provided with an impact sensor for detecting an impact received by the outside monitoring sensor,
A sensor abnormality is detected by determining that the vehicle exterior monitoring sensor is abnormal by detecting an impact that is greater than or equal to a value that occurs when an object of the impact sensor collides, and less than a predetermined value;
A sensor abnormality detection method, wherein the detection result of the sensor abnormality is held as failure diagnosis information and an alarm is output. By a vehicle equipped with at least one of ranging radar and image sensor as a vehicle monitoring sensor,
Based on the detection result of the outside monitoring sensor and the behavior monitoring result of the own vehicle, the collision prediction time between the own vehicle and the outside object is calculated,
A sensor that determines that the outside monitoring sensor is abnormal when the vehicle is in a collision-free state even when the prediction error time set from the elapsed time of the collision prediction time elapses in a traveling state that conforms to a predetermined collision condition. Detect anomalies,
A sensor abnormality detection method, wherein the detection result of the sensor abnormality is held as failure diagnosis information and an alarm is output.

Images