ELECTRONIC CONTROL DEVICE FOR VEHICLE

Abstract

Electronic control device of a vehicle (1) wherein the electronic control device comprises: an action prediction unit that predicts an action of an object around the vehicle according to external information acquired by an external information detection unit (20) that detects external information of the vehicle (1); and a determination unit for a detection unit, which determines whether an anomaly has occurred in the detection unit (20) for external information by comparing the external information captured by the detection unit (20) for external information at the time corresponding to a prediction result of the action prediction unit with the prediction result of the action prediction unit.

Description

Technical field

[0001] The present invention relates to an electronic control device of a vehicle. State of the art

[0002] In the event of a system error in the automated driving system, the system's electronic control unit (ECU), which is a higher-level control device for managing the automated driving, is required to continue the automated driving operation for a specific period of time until the driver takes over. Such an error might include, for example, an anomaly occurring during calculations in an arithmetic processor that performs calculations for controlling the automated driving, or an anomaly occurring in a sensor.

[0003] To continue operation for a specified period, even if a fault occurs, it is necessary to detect the anomaly and switch the control system to the one corresponding to the anomaly. Methods used to detect such a fault or anomaly typically include a procedure for comparing the output by multiplexing arithmetic processing or sensors, and a procedure for verifying the validity of the calculation result or sensor output value using another sensor value or calculation result. However, in the case of multiplexing sensors and arithmetic devices, an increase in the number of sensors and the arithmetic processing load presents a problem, such as increased system configuration complexity, necessitating a separate procedure for verifying validity.

[0004] PTL 1 discloses a device that detects or corrects a deviation in the value of a sensor that detects a state variable of a carrier vehicle, using a different sensor type than the one used in the device. In PTL 1, an anomaly is detected by evaluating the validity of a sensor output value.

[0005] PTL 2 describes a vehicle environment situation estimation device that detects collisions, captures and analyzes the environment before the impact, records this information and, based on this, predicts the environment situation after the collision.

[0006] PTL 3 describes a sensor system that outputs one or more signals and an electronic control unit connected to the sensor system. The external control unit can detect and identify external vehicles, assess their behavior based on evaluation information, and then make vehicle-specific adjustments. List of oppositions patent literature PTL 1: JP 2009-061942 A PTL 2: DE 11 2014 006 071 T5 PTL 3: US 2018 / 0 251 124 A1 Summary of the invention: Technical problem

[0007] The verification method described in PTL 1 makes it possible to detect a fault or anomaly that has occurred in a target sensor using a verification sensor that is different from the target sensor, but no means for detecting the anomaly that occurs in the verification sensor is disclosed.

[0008] In the ECU for automated driving, the position and speed of an environmental object are detected using multiple sensors simultaneously and by combining the properties of the respective sensors. Therefore, supplementary verification processing and anomaly detection processing using various sensors are necessary.

[0009] The present invention was made in view of the above problems, and its objective is to create an electronic control device for a vehicle which, with a relatively simple configuration, can determine whether an anomalous condition has occurred. Solution to the problem

[0010] To solve the above problems, an electronic control device of a vehicle according to the present invention includes an action prediction unit that predicts an action of an object around the vehicle according to external information acquired by an external information detection unit that detects external information of the vehicle, and a determination unit for a detection unit that determines, by comparing the external information acquired by the external information detection unit at the time corresponding to a prediction result of the action prediction unit with the prediction result of the action prediction unit, whether an anomaly has occurred in the external information detection unit. Advantageous effects of the invention

[0011] According to the present invention, the occurrence of an anomaly in the external information detection unit can be determined by comparing the result of the object's action around the vehicle, which is predicted by the external information detection unit based on the external information, with the external information from the external information detection unit at the time corresponding to the prediction result. Brief description of the drawings Fig. Figure 1 is a system configuration diagram of an electronic control device of a vehicle. Fig. 2 is a flowchart that depicts the overall process. Fig. Figure 3 is a block diagram representing a sensor fusion processing operation. Fig. Figure 4 is an explanatory view that illustrates the difference between a map of an object and a prediction map of the object around a carrier vehicle, which shows the positions of objects around the carrier vehicle over time. Fig. Figure 5 is a block diagram of a self-diagnostic function that identifies an anomaly of a sensor or microcomputer. Fig. 6 is a flowchart for self-diagnosis processing. Fig. Figure 7 is an explanatory drawing that illustrates a method for performing a total determination based on the determination result of a microcomputer and the determination result of sensor data. Description of embodiments

[0012] The following describes embodiments of the present invention with reference to the accompanying drawings. In the present embodiment, a predetermined calculation is performed based on initial information detected by an information detection unit (detection unit 20 external information), and a first calculation result is compared with second information (where the second information is detected after the first information) detected by the information detection unit, or with a second calculation result obtained by performing a predetermined calculation based on the second information. Thus, the present embodiment can perform a self-diagnosis as to whether an anomaly has occurred in the information detection unit or in the arithmetic processing unit that executes a predetermined operation.

[0013] An example of the present embodiment, as described in detail below, includes the external information detection unit 20, which detects external information from a vehicle 1, and an action prediction unit (S2), which predicts the behavior (an action) of an object located around the vehicle 1. Information M2(T1) obtained by the action prediction unit through predictions of the action of the surrounding object is compared with external information obtained by the external information detection unit 20 to determine the occurrence of an anomaly in the external information detection unit 20.

[0014] According to the present embodiment, it is possible to detect the anomaly occurring in the detection unit 20 of external information by comparing information predicted by previous external information (prediction result of the action prediction unit) with the current external information. Furthermore, according to the present embodiment, it is also possible to detect the anomaly occurring in the overall calculation result by comparing the calculation result of an integrated calculation unit (S1), which performs an integrated calculation based on the external information from the detection unit 20 of external information, with the prediction result of the action prediction unit, the integrated calculation unit.

[0015] Thus, according to the present embodiment, it is possible to diagnose the occurrence of an anomaly in the detection unit 20 of external information or in the integrated computing unit (S1) with a relatively simple configuration, without multiplexing the detection circuit, the arithmetic processing circuit and the computer program. First embodiment

[0016] Based on Fig. 1, Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6 to Fig. Section 7 describes an embodiment. Fig. Figure 1 is a system configuration diagram of an electronic control device of a vehicle according to the present embodiment.

[0017] Vehicle 1 includes, for example, a vehicle body, wheels arranged at the front, rear, right, and left of the vehicle body, an engine mounted on the vehicle body, a driver's cab (not shown), and the like. The ECU 10 for automatic driving is an electronic control device for the automatic driving of vehicle 1. The ECU 10 for automatic driving includes, for example, an arithmetic processing unit (a microcomputer) 11 and a memory 12 used by the microcomputer 11.

[0018] The ECU 10 for automated driving is connected to the external information detection unit 20, which detects external information from the vehicle 1. The external information detection unit 20 includes, for example, a millimeter-wave radar 21 and a camera 22. In addition to these sensors 21 and 22, an ultrasonic sensor, an infrared sensor, or the like may be used. In the following description, the millimeter-wave radar 21 and the camera 22 may be referred to as sensors 21 and 22.

[0019] Furthermore, a vehicle state detection unit 30 is connected to the ECU 10 for automated driving, which detects an internal state of the vehicle 1. The vehicle state detection unit 30 contains, for example, map information 31 and a carrier vehicle position sensor 32. The carrier vehicle position sensor 32 can be a global positioning system (GPS) or a position detection system in which a vehicle speed sensor or an accelerometer is combined with the GPS.

[0020] The ECU 10 for automated driving, using the microcomputer 11 and based on external information supplied by the external information detection unit 20, map information 31, and information from the carrier vehicle position sensor 32, performs arithmetic processing related to automated driving. After calculating a driving trajectory for the carrier vehicle as a result of the calculation, the ECU 10 for automated driving sends a control command value to the lower-level ECU group 40, which contains a brake control device 41, an engine control device 42, and a steering control device 43. As a result, the carrier vehicle 1 drives automatically along a safe route.

[0021] The data input by the detection unit 20 of external information into the ECU 10 for automatic driving can be output data directly from sensors 21 and 22, or data pre-processed by a dedicated ECU (not shown) for sensors 21 and 22.

[0022] The schedule from Fig. 2 represents the main processing of the ECU 10 for automated driving. This processing is carried out by the microcomputer 11 in a predefined cycle.

[0023] The microcomputer 11 performs calculations of the sensor fusion processing S1 to integrate various external information provided by the detection unit 20, a processing S2 of the prediction of the action of surrounding objects to predict the action of objects around the carrier vehicle using the map of the object around the carrier vehicle obtained as a result of the sensor fusion processing S1, a carrier vehicle trajectory planning processing S3 to generate a carrier vehicle trajectory based on the action prediction of the surrounding objects around the carrier vehicle, and a self-diagnostic processing S4.

[0024] Fig. Figure 3 provides an overview of the internal processing of the sensor fusion processor S1. The sensor fusion processor S1 is an example of the "integrated computing unit" and performs a time synchronization processor S11 and a sensor integration processor S12.

[0025] In sensor fusion processing S1, sensor data provided by the millimeter-wave radar 21 or the camera 22 to the ECU 10 for automated driving are not time-synchronized. Therefore, the sensor data, which includes a timestamp, is received by the detection unit 20 as external information to perform time synchronization processing S11.

[0026] The data on which the time synchronization processing S11 has been completed are referred to as synchronized millimeter-wave radar data and synchronized camera data. Based on the synchronized millimeter-wave radar data and the synchronized camera data, the sensor integration processing S12 calculates the position coordinates of the surrounding objects around the vehicle 1. The sensor integration processing S12 maps the position of the objects around the carrier vehicle 1 by referencing the map information 31 around the carrier vehicle, using the sensor value from the carrier vehicle position sensor 32 in addition to the position coordinates of each surrounding object.

[0027] Fig. Figure 4, using sensor fusion processing S1, integrates information on the positions, sizes, and speeds of objects around the carrier vehicle M, detected by the millimeter-wave radar 21 and the stereo camera 22, onto a map of an object around the carrier vehicle M, creating an exemplary graphical representation of objects. Besides the carrier vehicle position 101, the map of the object around the carrier vehicle M also depicts the position 102 of another vehicle, a bicycle position 103, and a pedestrian position 104.

[0028] A map of the object around the carrier vehicle M1 contains a map indicating the current state (a map indicating the current situation) and a map predicting a state at a future time. A reference symbol M1(T1) indicates the actual map generated at time T1. A reference symbol M2(T1) indicates a map predicting a future situation at time T2 from time T1. A reference symbol M2(T2) indicates the actual map generated at time T2. The predicted map can be referred to as a forecast map.

[0029] Fig. Figure 4(1) represents the map M1(T1) of the object around the carrier vehicle, indicating the current state at time T1 as generated by the sensor fusion processing S1. The carrier vehicle 101(1) is assumed to be traveling to the right in the drawing. In the direction of travel of the carrier vehicle 101(1), there are objects such as another vehicle 102(1), a bicycle 103(1), and a pedestrian 104(1). As used here, the numbers in parentheses following the object symbols correspond to the numbers in parentheses of Fig. 4(1) to (3). Where the generation time and the prediction time are not differentiated in the following description, the symbols are abbreviated, such as the carrier vehicle 101, the other vehicle 102, the bicycle 103 and the pedestrian 104.

[0030] Fig. 4(2) represents the map M2(T1) of the object around the carrier vehicle, which predicts the situation at time T2, as predicted at time T1 by the action prediction process S2 of surrounding objects. The action prediction process S2 of surrounding objects predicts actions of the surrounding objects around the carrier vehicle 1 in the microcomputer 11 based on the map M1(T1) of the object around the carrier vehicle.

[0031] In processing S2, the action prediction of surrounding objects predicts the actions of various surrounding objects 102(1) to 104(1) mapped onto the map M1(T1) of the object around the carrier vehicle. A prediction procedure includes, for example, a method of extrapolating and determining the future position of the surrounding objects based on the current position and velocity of each surrounding object.

[0032] The S2 processing step for predicting the actions of surrounding objects provides a prediction map around the carrier vehicle M2(T1) that predicts the future of the objects around the carrier vehicle, as shown in Fig. 4(2) is shown. The prediction map around the carrier vehicle M2(T1) shows a predicted position 101(2) of the carrier vehicle 1, a predicted position 102(2) of the other vehicle, a predicted position 103(2) of the bicycle and a predicted position 104(2) of the pedestrian.

[0033] In Fig. 4(2) The predicted positions of the objects are indicated by dashed lines for explanation by processing S2 of the action prediction of surrounding objects. The position 102(1) of the other vehicle, the bicycle position 103(1), and the pedestrian position 104(1) at the current time T1, resulting from sensor fusion processing S1, are indicated by solid lines. The actual positions are not necessarily included in the result of processing S2 of the action prediction of surrounding objects. Although in Fig. 4(2) Since only one predicted position is represented for each object, a number of predicted positions necessary for the subsequent carrier vehicle trajectory planning processing S3 can be generated for each object.

[0034] For example, assuming that the trajectory of carrier vehicle 1 is planned for 10 seconds every 100 milliseconds in the carrier vehicle trajectory planning processor S3, a maximum of 100 predicted positions will be generated for each object (other vehicles, bicycles, pedestrians, etc.). (100 = 10,000 milliseconds / 100 milliseconds). Using the predicted positions of the objects, the carrier vehicle trajectory planning process S3 is executed in microcomputer 11, and a vehicle trajectory is generated. A control command value for the lower ECU group 40, corresponding to the generated vehicle trajectory, is created and sent to the lower ECU group 40. Thus, the main function processing of ECU 10 for automated driving is completed.

[0035] On the other hand, the self-diagnostic processing S4 in the main function processing sequence of the ECU 10 for automatic driving diagnoses an anomaly in the input values ​​of the sensor fusion processing S1 and the sensors 21 and 22.

[0036] Fig. 5 represents an overall configuration of the self-diagnostic processing S4. It is noted that the in Fig. 5. Self-diagnosis processing (anomaly detection processing) is illustrated by extracting the relevant part from the data in Fig. The main function processing flow of the ECU 10 for automatic driving is shown in section 2.

[0037] The main processing function of the ECU 10 for automated driving is periodic, whereby after the completion of a series of processing steps that began at a specific time T1, similar processing steps begin again at the next time T2. Thus, the sequence of processing steps in the ECU 10 for automated driving is as follows: Fig. 5. This is represented by specifying a portion of the series of processing steps in the horizontal direction and the time at which the series of process steps begins in the vertical direction. Details of the processing will be described later using… Fig. 6 described. Fig. The 5 step numbers are described, which correspond to the process from Fig. 6.

[0038] The sequence of processing steps S1(T1) and S2(T1), which are started at time T1, determines the map M1(T1) of the object around the carrier vehicle at time T1 and determines the environmental prediction map M2(T1) calculated at time T1 at time T2 ( Fig. 4(1), (2)). Furthermore, a series of processing steps S1(T2), which are started at time T2, determine an environment object position M2(T2), which is a map indicating the positions of surrounding objects at time T2 ( Fig. 4(3)).

[0039] In other words, by comparing the predicted map around the carrier vehicle M2(T1) at time T1 with the environmental object map M2(T2) detected at time T2, it is possible to determine whether the sensor fusion processing S1 contains the anomaly (S43). By comparing the predicted map around the carrier vehicle M2(T1) at time T1 with the sensor data acquired by sensors 21 and 22 at time T2 (sensor data after time synchronization processing) (S45), it is possible to determine whether sensors 21 and 22 exhibit an anomaly.

[0040] As will be shown later using the following examples: Fig. 6 and Fig. As described in section 7, the self-diagnostic processing unit S4 provides a final overall determination based on the diagnostic result (determination result) from the calculation result of the sensor fusion processing unit S1 and the diagnostic result (determination result) from the sensor data. The ECU 10 for automated driving then executes a control action based on this final overall determination, such as notifying the driver about the anomaly detection or assigning the driver to operate vehicle 1.

[0041] Fig. Figure 6 is a flowchart that presents a detailed example of the self-diagnostic processing S4, which is executed by the microcomputer 11. It is assumed that the current time is T2.

[0042] Microcomputer 11 retrieves the predicted map M2(T1) of surrounding objects from processing S2 of the action prediction process of surrounding objects (S41). Microcomputer 11 then retrieves the map of objects around the carrier vehicle M2(T2) calculated at time T2 (S42). Microcomputer 11 subsequently compares the predicted map M2(T1) retrieved in step S41 with the actual map M2(T2) retrieved in step S42 to determine whether the anomaly occurred in the calculation result of sensor fusion processing S1 (S43).

[0043] An example of the comparison procedure in step S43 is described here. For example, there is a procedure for comparing the deviation range of the position of each surrounding object with a predetermined threshold. Between the predicted position 103(2) of the bicycle at time T2, which was at time T1 in Fig. 4(2) was calculated, and the position 103(3) of the bicycle at time T2 in Fig. 4(3) there is a positional error of less than one grid cell in the drawing.

[0044] Thus, assuming that the threshold for detecting the presence or absence of an anomaly is defined, for example, as a "half grid square," it can be determined that an anomaly occurred at bicycle position 103(3) at time T2. This makes it possible to detect the presence or absence of the anomaly before the sensor fusion processing S1 takes place in the main function processing flow of ECU 10 for automated driving. The determination result of step S43 is sent to the overall determination step S46.

[0045] On the other hand, the microcomputer 11 acquires synchronized millimeter-wave radar data and synchronized camera data from the time synchronization processing S11 in the sensor fusion processing S1 (S44). The microcomputer 11 compares the synchronized data acquired in step S44 (synchronized millimeter-wave radar data, synchronized camera data) with the predicted position M2(T1) of surrounding objects at time T2, calculated at time T1 (S45).

[0046] In the comparison processing step S45, the predicted position M2(T1) of surrounding objects calculated at time T1 is considered correct data for time T2, and the difference between the synchronized millimeter-wave radar data and the actual position of each surrounding object, as indicated by the synchronized camera data, is compared to a predefined threshold. If the difference generated in the sensor data is equal to or greater than the threshold, the microcomputer 11 determines whether the sensor data is anomalous, i.e., whether sensors 21 and 22 are anomalous.

[0047] Similar to the anomaly determination (S43) of sensor fusion processing S1 described above, the presence or absence of an anomaly in the sensor data can be determined using the threshold as described above. The synchronized millimeter-wave radar data and the synchronized camera data used for comparison can be obtained from time synchronization processing S11 to obtain data synchronized with time T2. Alternatively, synchronized data obtained in time synchronization processing S11 at a time closest to time T2 can be used. The determination result of step S45 is sent to the overall determination step S46.

[0048] Based on the determination result of step S43 (the determination result of the sensor fusion processing S1) and the determination result of step S45 (the determination result of the sensor data), the microcomputer 11 performs an overall determination (S46), enabling the microcomputer 11 to detect the anomaly generated in the sensor fusion processing S1, the anomaly generated in the millimeter wave radar 21, and the anomaly generated in the camera 22.

[0049] The overall determination procedure according to the patterns of the determination results in steps S43 and S45 is described by reference to a Table 120 from Fig. 7 described. The overall determination table 120 assigns the determination result S43 of the sensor fusion processing S1 to the determination result S45 of the sensor data and the overall determination result S46. In Fig. Figure 7 shows the determination result S43 of the sensor fusion processing S1 as "calculation result determination".

[0050] The microcomputer 11 determines from the result S43 of the sensor fusion processing S1 whether the anomaly occurred in the sensor fusion processing S1. If the presence of the anomaly in the sensor fusion processing S1 is determined, the microcomputer 11 refers to the result S45 of the sensor data from sensors 21 and 22. Simultaneously, the microcomputer 11 can determine that the anomaly is in the sensor fusion processing S1 if the anomaly is not applicable to either the millimeter-wave radar 21 or the camera 22.

[0051] The determination procedure will be derived from Fig. 7 described. If the presence of the anomaly is determined in step S43 and the anomaly in the millimeter-wave radar data 21 is determined in step S45, the overall determination result is "Anomaly in the millimeter-wave radar". Similarly, the overall determination result is "Anomaly in the camera" if the presence of the anomaly is determined in step S43 and the anomaly in camera 22 is determined in step S45. Similarly, the overall determination result is "Anomaly in the sensor fusion processing" if the presence of the anomaly is determined in step S43 and no anomaly is determined in step S45.

[0052] If no anomaly is detected in step S43, the overall determination result is "no anomaly". Since the sensor fusion processing S1 uses the sensor data from sensors 21 and 22, it can be considered that if no anomaly occurs in the sensor fusion processing S1, then no anomaly is present in the sensor data.

[0053] Reference is made again to Fig. 6. The microcomputer 11 determines whether the overall determination result is “anomalous” (S47) and notifies the driver in vehicle 1 that the anomaly is detected if the anomaly is detected (S47: Yes) (S48). For example, the microcomputer 11 provides a voice output or displays a message such as “An anomaly has been detected in the automated driving system”.

[0054] Since a temporary anomaly in the sensor value may occur due to noise or similar factors, the notification in step S48 can be executed if the "anomaly" is detected in the overall determination result in step S47 at least a predefined number of times. The notification in step S48 serves as a precautionary measure before the transition process from automatic to manual driving begins.

[0055] After notifying the driver in step S48, microcomputer 11 determines whether the anomalous condition persists (S49). For example, microcomputer 11 begins a backup operation (S50) if the overall determination result is still determined to be "anomalous" in at least a further predefined number of instances (S49: Yes).

[0056] Although not shown, the safety operation processing, for example, activates a warning light, modifies the driving control to enter a predefined stop operation, ignores the input from the sensor identified as anomalous, and continues automatic driving using only the remaining normal sensors. After a predetermined period following notification to the driver, it switches back to manual driving. This ensures the occupant's safety when an anomaly occurs in the automatic driving system. The notification processing (S46, S48) not only notifies the driver in vehicle 1 but also, for example, other surrounding vehicles or a server (not shown) that monitors the vehicle's operation.Log information indicating parameter changes during automated driving can be stored in memory 12 and can be saved by mapping the time of anomaly detection and the type of anomaly to the log information. This can also be used to investigate the cause of the anomaly.

[0057] According to the present embodiment configured as described above, the actions of objects around the carrier vehicle are predicted based on sensor data from external information by the detection unit 20. Based on the current sensor data acquired by the detection unit 20, it is determined whether the prediction result is correct or incorrect. Thus, according to the present embodiment, self-diagnosis can be performed with a simple configuration without multiplexing the detection unit 20, the microcomputer 11, and the like, and reliability can be improved without increasing manufacturing costs.

[0058] In other words, according to the present embodiment, the anomaly occurring in the external information detection unit 20 and in the sensor fusion processing S1 can be detected without multiplexing the sensor fusion processing S1 in the ECU 10 for automated driving or in the external information detection unit 20 (millimeter-wave radar 21 and camera 22) used in the ECU 10 for automated driving. The determination method for the presence or absence of an anomaly according to the present embodiment is based on a comparison operation using a map or information obtained in the main function processing process of the ECU 10 for automated driving.For this reason, an increase in the processing load of the ECU 10 for automatic driving can be suppressed, even if the procedure (the self-diagnostic processing) for determining the presence or absence of an anomaly according to the present embodiment is additionally installed in the ECU 10 for automatic driving.

[0059] It is noted that the present invention is not limited to the embodiment described above. A person skilled in the art can make various additions and modifications within the scope of protection of the present invention. In the embodiment described above, the present invention is not limited to the configuration example shown in the accompanying drawings. It is possible to suitably modify the structure and processing method of the embodiment to the extent necessary to solve the problem of the present invention.

[0060] Furthermore, each component of the present invention can be chosen and arranged arbitrarily, and the invention with any such chosen and arranged configuration is also contained within the present invention. In addition, the configurations described in the appended claims can be combined with any combination besides those specified in the claims.

[0061] In the embodiment described above, the method for determining the presence or absence of an anomaly is described as comparing two pieces of map information at time T1 and time T2. Alternatively, the presence or absence of an anomaly can be determined as an anomaly detection method, for example, by comparing position information and predicted position information at three or more time points, such as times T1, T2, and T3.

[0062] The anomaly detection method (self-diagnostic method) described in the embodiment above verifies the validity of the ECU's processing for automatic driving at the next time point T2, assuming that the ECU's processing for automatic driving is normal at a given time point T1. Thus, it is preferred, for example, to confirm that the ECU's processing for automatic driving is normal at a key-on time point, when the vehicle key is switched on, or at the start of the ECU's processing.

[0063] The confirmation procedure includes, for example, a method of executing the same processing both at the start of the automatic driving process and the next time, and confirming whether the result of the processing is correct. Furthermore, a position detection is performed on an object, such as a location point whose position is known, like a building, etc., and the amount of movement of the vehicle 1 is estimated using the map information 31 and the carrier vehicle position sensor 32 while the ECU 10 is operating for automatic driving. The validity of the external information detection unit 20 and the sensor fusion processing S1 can also be confirmed by comparing the value with the position of an object, such as a location point. Reference symbol list 1 vehicle 10 ECUs for automated driving 11 microcomputers 20 Detection unit for external information 21 millimeter wave radar 22 Camera 30 Vehicle Condition Detection Unit 31 map information 32 Carrier vehicle position sensor 40 ECU group lower level

Claims

Electronic control device of a vehicle (1), wherein the electronic control device comprises: an action prediction unit that predicts an action of an object around the vehicle according to external information acquired by an external information detection unit (20) that detects external information of the vehicle (1); and a determination unit for a detection unit that determines, by comparing the external information acquired by the external information detection unit (20) at the time corresponding to a prediction result of the action prediction unit with the prediction result of the action prediction unit, whether an anomaly has occurred in the external information detection unit (20). Electronic control device of a vehicle (1) according to claim 1, further comprising: an integrated computing unit that holistically calculates the external information; and a determination unit for a computing unit that determines, by comparing a calculation result of the integrated computing unit at the time corresponding to the prediction result of the action prediction unit with the prediction result of the action prediction unit, whether an anomaly has occurred in the integrated computing unit. Electronic control device of a vehicle (1) according to claim 2, which further comprises a total determination unit which performs a total determination according to a determination result of the determination unit for the detection unit and a determination result of the determination unit for the calculation unit. Electronic control device of a vehicle (1) according to claim 3, which further comprises a notification unit which issues an anomaly notification to a driver of the vehicle (1) indicating the detection of an anomalous condition when the overall determination unit determines the presence of an anomaly. Electronic control device of a vehicle (1) according to claim 4, wherein the notification unit issues the anomaly notification when the overall determination unit determines the presence of an anomaly at least in a preset predetermined number. Electronic control device of a vehicle (1) according to claim 5, wherein the processing to shift the operation of the vehicle (1) to the driver is started when the overall determination unit determines the presence of an anomaly in at least one further predetermined number, which is preset, after the notification unit has issued the anomaly notification. Electronic control device of a vehicle (1) according to claim 1, wherein the detection unit (20) for external information includes a radar and / or a camera (22).

Images