ELECTRONIC CONTROL DEVICE AND OPERATING PROCEDURES

Abstract

Electronic control device (3) comprising: a sensor information acquisition unit (33) that acquires the state variables including a position of a measurement object using an output from multiple sensors or calculates the state variables of the measurement object using the output from the multiple sensors; a storage unit (32) that stores position determination information as a condition of a position-related classification; a positioning unit (311) that classifies the object being measured using the positioning information; and a merging unit (312) which, with reference to the state variables, determines a match among several of the measured objects which are classified as completely identical by the position determination unit (311) and which were measured by different sensors, and which merges the positions of the several measured objects which were determined to be a match, wherein the objects being measured are objects that can be measured by each of the multiple sensors, wherein the storage unit (32) additionally stores: a future classification condition as a condition of a classification that refers to a future position, which is a future position of the object of measurement; and a classified future position as the future position of the object of measurement that has been classified by the future classification condition; the electronic control device (3) additionally includes: an estimation unit (315) of future positions that estimates the future position of the object being measured and classifies the future position using the future classification condition; and the merging unit (312) also incorporates into the objectives of a merging process the measurement objects which are classified as completely identical by the estimation unit (315) of future positions when determining a match of the measurement objects, and merges the positions and future positions of the multiple measurement objects which have been determined to be a match.

Description

TECHNICAL AREA

[0001] The present invention relates to an electronic control device and an operating method. TECHNICAL BACKGROUND

[0002] In recent years, the development of autonomous driving systems has been actively pursued. Autonomous driving often utilizes the combined measurement results of multiple sensors. PTL 1 discloses a mechanism for detecting a vehicle ahead of a driving control device. This mechanism detects a vehicle ahead by performing image processing on images obtained by capturing the conditions in front of the vehicle using camera means. It controls a force machine output to cause the vehicle to drive while following the vehicle ahead. The mechanism for detecting a vehicle ahead employs a laser radar that determines the position of an object in front of the vehicle by emitting a laser beam from the vehicle in a forward direction.While scanning an area in a horizontal plane and receiving and detecting the reflected laser beam, NPTL 1 comprises image processing means that, when performing image processing on the images and detecting the vehicle ahead, define a processing area within a screen corresponding to the position of the object detected by the laser radar, and detect the vehicle ahead by performing image processing only on the images of the processing area. Furthermore, NPTL 1 discloses a method for object tracking for driver assistance systems that fuses data from a radar sensor and a monocular camera. Using filter techniques such as the Kalman filter, the current states (position, velocity) of objects are estimated and tracked. NPTL 2 discloses a method for generating an environment model for a vehicle.where data from multiple sensors are fused into a grid-based representation of the environment. LIST OF COUNTER-POINTS PATENT LITERATURE [PTL 1] JP H07 - 125 567 A [PTL 2] DE 10 2014 103 695 A1 NON-PATENT LITERATURE

[0003] [NPTL 1] Liu, Feng: Object tracking through fusion of radar and monocular camera data for driver assistance systems. Dissertation, Karlsruhe Institute of Technology (KIT). SUMMARY OF THE INVENTION: PROBLEMS TO BE SOLVED BY THE INVENTION

[0004] In the invention described in PTL 1, the amount of calculation becomes enormous when there are multiple measurement objects. MEANS TO SOLVE THE PROBLEMS

[0005] The present invention relates to an electronic control device with the features of claim 1 and an operating method with the features of claim 6. Advantageous further developments are defined in the dependent claims. ADVANTAGEOUS EFFECTS OF THE INVENTION

[0006] According to the present invention, the amount of calculation can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS Fig. Figure 1 is a configuration diagram of vehicle C. Fig. 2(a) is a graphical representation showing the data acquisition area of ​​sensor group 1, and Fig. 2(b) is a perspective view of the defined areas. Fig. Figure 3 is a graphical representation showing an example of the range determination table 3211. Fig. Figure 4 is a graphical representation showing an example of preprocessing table 3212. Fig. Figure 5 is a graphical representation showing an example of post-processing table 3213. Fig. Figure 6 is a graphical representation showing an example of the first range determination table 3221A. Fig. Figure 7 is a flowchart showing the processing to be carried out by processing unit 31. Fig. 8 is a flowchart detailing the S103 according to Fig. 7 shows. Fig. Figure 9 is a graphical representation showing the configuration of vehicle C in the second embodiment. Fig. Figure 10 is a graphic representation showing an example of the pattern identification table 3214. Fig. Figure 11 is a graphical representation showing an example of the range determination table 3211A in the second embodiment. Fig. Figure 12 is a graphical representation showing an example of the sensor area assignment information 3217. Fig. Figure 13 is a flowchart showing the processing of the processing unit 31 in the second embodiment. Fig. Figure 14 is a graphical representation showing the configuration of vehicle C in the third embodiment. Fig. Figure 15 is a graphic representation showing an example of the determination table 3215 of the future areas. Fig. Figure 16 is a graphic representation showing an example of the future Table 3216. Fig. Figure 17 is a flowchart showing the processing of the processing unit 31 in the third embodiment. DESCRIPTION OF THE EXECUTION FORMS - First embodiment -

[0007] The first embodiment of the electronic device according to the present invention is now described with reference to Fig. 1 to Fig. 8 explained. (Configuration diagram)

[0008] Fig. Figure 1 is a configuration diagram of vehicle C. Vehicle C comprises a sensor group 1, an electronic control unit 3, and an actuator group 4. The sensor group 1 and the actuator group 4 are connected to the electronic control unit 3.

[0009] Sensor group 1 is configured with two or more sensors, such as a left camera (LC), a right camera (RC), a left radar (LR), and a right radar (RR). The left camera (LC) is a camera that captures the left front of the vehicle (C), while the right camera (RC) is a camera that captures the right front of the vehicle (C). The left radar (LR) is a radar, such as a millimeter-wave radar, that detects an obstacle on the left front of the vehicle (C), while the right radar (RR) is a radar, such as a millimeter-wave radar, that detects an obstacle on the right front of the vehicle (C).

[0010] The individual sensors that configure sensor group 1 acquire information about a measurement object, which is an object that can be measured by each of the sensors, and calculate the position and velocity of the measurement object from the acquired information. The respective sensors then output the position and velocity as state variables to the electronic control device 3. However, the state variable can contain information other than position and velocity; for example, the state variable can also include the angular velocity. For example, the left camera LC and the right camera RC each analyze the acquired images and calculate the position and velocity of the measurement object from the continuity of the luminance and saturation and their time-series variation. The left radar LR and the right radar RR, for example, calculate...The position and velocity of the object being measured are determined from the continuity of the distance and its time-series variation. The respective sensors can output the position of the object being measured using either their own sensors as the reference or the electronic control device 3 as the reference.

[0011] It is stated that the "object of measurement" varies depending on the sensor, and that even if the measurable ranges of several sensors overlap, there is an object that can only be measured with a specific sensor. Furthermore, some of the objects of measurement do not impede the vehicle's movement, such as smoke. Even with an object like smoke that does not impede the vehicle's movement, measuring such an object with a sensor capable of measuring it contributes to capturing the external environment.

[0012] Fig. 2(a) is a graphical representation showing the data acquisition area of ​​sensor group 1, and Fig. 2(b) is a perspective view of the defined areas. In Fig. 2(a) However, the reference symbol of the corresponding sensor is assigned to the data acquisition area of ​​each sensor. As in Fig. As shown in Figure 2(a), the data acquisition areas of the left camera LC and the left radar LR overlap on the left side of the graphical representation, the data acquisition areas of two to four sensors near the center of the graphical representation overlap, and the data acquisition areas of the right camera RC and the right radar RR overlap on the right side of the graphical representation.

[0013] In this embodiment, the three areas from Z1 to Z3 are defined as in Fig. Figure 2(b) shows the first area Z1 where the data acquisition areas of the left camera LC and the left radar LR overlap. The second area Z2 is where the data acquisition areas of two to four sensors overlap. The third area Z3 is where the data acquisition areas of the right camera RC and the right radar RR overlap. However, the respective sensors can also acquire data from areas wider than those shown in Figure 2(b). Fig. 2(a) are the data acquisition areas shown. In the preceding case, the information obtained by sensors outside the area shown in Fig. 2(a) shown area can be captured, taken into account, or ignored. For example, if the position of the object being measured, captured by the right camera RC, is the first area Z1, processing can be carried out taking into account the information from this object or by ignoring such information. The explanation now continues with regard to Fig. 1 continued.

[0014] Actuator group 4 consists of several actuators that control the direction or speed of the vehicle and includes, for example, a steering wheel, a brake, and an accelerator pedal. Actuator group 4 is operated based on an operating command from the electronic control device 3. However, actuator group 4 can also be operated based on user input.

[0015] The electronic control device 3 is an electronic control unit comprising a processing unit 31, a storage unit 32, and an I / O unit 33. The processing unit 31 comprises a CPU (central processing unit), a ROM (read-only memory), and a RAM (read / write memory). The CPU performs the following functions: loading programs stored in the ROM into the RAM and executing such programs. The processing unit 31 includes, as its functions, a position determination unit 311, a merging unit 312, and a vehicle control processing unit 313. These functions are described later. The I / O unit 33 is a communication interface that communicates with the sensor group 1 and the actuator group 4. The I / O unit 33 conforms to a communication standard, such as CAN (registered trademark) or IEEE 802.3.

[0016] The fusion unit 312 comprises a first fusion unit 312A, in which the first area Z1 is the processing target, a second fusion unit 312B, in which the second area Z2 is the processing target, and a third fusion unit 312C, in which the third area Z3 is the processing target. The fusion unit 312 determines whether the measurement objects, in which the state variables were acquired by different sensors, are identical and merges the state variables of the multiple measurement objects that have been determined to be identical. In the following explanation, the measurement objects in which the state variables have been merged are referred to as "merged measurement objects." Furthermore, for each fusion unit 312, a sensor in sensor group 1 to be excluded from the processing target can also be specified. The first fusion unit 312A can, for example,only process the measurement objects of the left camera LC and the left radar LR and specify the right camera RC and the right radar RR as the sensors to be excluded from the processing target.

[0017] The storage unit 32 is a non-volatile storage device, such as flash memory. The storage unit 32 is divided into a common storage unit 321, which can be referenced by the entire processing unit 31, and a dedicated storage unit 322, which can only be read from or written to by each merging unit 312. The common storage unit 321 and the dedicated storage unit 322 can be implemented with the same hardware or with different hardware. The common storage unit 321 stores a range determination table 3211, a preprocessing table 3212, and a postprocessing table 3213.

[0018] The dedicated storage unit 322 stores a first area table 3221A, which can only be read from or written to by the first merging unit 312A; a second area table 3221B, which can only be read from or written to by the second merging unit 312B; and a third area table 3221C, which can only be read from or written to by the third merging unit 312C. It is stated that in this embodiment, in the sense that when processing a specific area, it is not necessary to refer to information in another area, e.g.,The declaration is provided stating that "the first area table 3221A may only be read from or written to by the first merging unit 312A." In this embodiment, the first area table 3221A therefore does not require any special access restriction measures, such as an authorization setting or a locking mechanism. (Area determination table 3211)

[0019] Fig. Figure 3 is a graphical representation showing an example of the range determination table 3211. The range determination table 3211 stores a match between the condition and the range number related to the position of the measured object. The position-related condition is represented, for example, based on a condition expression, where in the Fig. In example 3, the condition expression is described as a polar coordinate system based on vehicle C. In other words, a distance R from the reference point provided on vehicle C and a condition of an angle θ of the object being measured with the front of vehicle C set to zero degrees and the right side being true are described. The right side of the Fig. The 3 area numbers shown are those in Fig. 2(b) area numbers Z1 to Z3 shown. In other words, the area determination table 3211 expresses the Fig. 2(b) shown areas of the first area Z1, the second area Z2 and the third area Z3 based on the conditional expressions.

[0020] The position determination unit 311 refers to the range determination table 3211 and determines the position, for example, in the following manner. If, for example, the position of a particular object being measured is expressed in a polar coordinate format, if the distance R from the reference point falls within a range from a predefined threshold R1a to a threshold R1A, and the angle θ from the reference direction lies between a threshold θ1a and a threshold θ1A, the position determination unit 311 determines that the object being measured belongs to the first range Z1. (Preprocessing table 3212)

[0021] Fig. Figure 4 is a graphical representation showing an example of preprocessing table 3212. Preprocessing table 3212 stores the information acquired by I / O unit 33 from sensor group 1 and the information generated by position determination unit 311. Preprocessing table 3212 stores the state variables of the object being measured before the fusion unit 312 performs the processing. Preprocessing table 3212 is configured with multiple records, each containing fields for ID, acquisition sensor, position, velocity, and range. The ID field stores a suitable identifier for identifying the object being measured. The identifier is, for example, a combination of "B," indicating that it is preprocessing, and a sequential number indicating the order in which the object was acquired.

[0022] The Acquisition Sensor field stores the name of the sensor used when I / O Unit 33 acquired the state variables. The Position and Velocity fields store the state variables acquired by I / O Unit 33. The Velocity field stores the name of the range determined by Position Determination Unit 311. In other words, fields other than the Range field are entered by I / O Unit 33 in the Preprocessing Table 3212, while only the Range field is entered by Position Determination Unit 311. (Post-processing table 3213)

[0023] Fig. Figure 5 is a graphical representation showing an example of the post-processing table 3213. The post-processing table 3213 stores the state variables of the merged measurement objects calculated by the first fusion unit 312A, the second fusion unit 312B, and the third fusion unit 312C. The post-processing table 3213 is configured with multiple records, each containing the fields of the new ID, the original ID, the position, and the velocity. The new ID field stores a suitable identifier for identifying the merged measurement objects. The identifier is, for example, a combination of "A," indicating that it is a post-processing operation, and a sequential number indicating the order in which the measurement object was calculated.

[0024] The Original ID field stores an ID of the measurement object used to generate the merged measurement objects; that is, the value of the ID in preprocessing table 3212. The Position and Velocity fields store the position and velocity of the merged measurement objects. It is stated that the New ID and Original ID fields are provided for convenience only, and postprocessing table 3213 does not need to contain these fields. (First area table 3221A)

[0025] Fig. Figure 6 is a graphical representation showing an example of the first area table 3221A. The first area table 3221A primarily stores the state variable of the object being measured, which is present in the first area Z1, detected by sensor group 1. The first area table 3221A is configured with several records, each containing the records of the ID, detection sensor, position, velocity, and adjustment. The ID, detection sensor, position, and velocity fields store the same information as the preprocessing table 3212. The adjustment field stores an internal variable that is used by the first fusion unit 312A when performing the processing described later. The initial value of the adjustment field is "not yet adjusted."The first range table 3221A is created by the first merge unit 312A, which extracts a record from the preprocessing table 3212 where the range is "Z1". It is stated that while the range field in the table is... Fig. In the example shown in section 6, the area field has been deleted, but it can also be left as it is.

[0026] The configuration of the first area table 3221A has been explained above; however, the configuration of the second area table 3221B and the third area table 3221C is also the same as the configuration of the first area table 3221A. The second area table 3221B is generated by the second fusion unit 312B and primarily stores the state variables of the measurement object present in the second area Z2. Furthermore, the third area table 3221C is generated by the third fusion unit 312C and primarily stores the state variables of the measurement object present in the third area Z3. (Operation of processing unit 31)

[0027] Fig. Figure 7 is a flowchart showing the processing to be carried out by processing unit 31. Processing unit 31 performs the tasks described in Figure 7. Fig. Figure 7 shows the processing for each predefined processing cycle, e.g., every 100 ms. In processing unit 31, the I / O unit 33 first acquires the state variables of the object being measured from sensor group 1, generating preprocessing table 3212 (S101). It is stated that at the time of S101, the range field of preprocessing table 3212 is empty. Next, the position determination unit 311 of processing unit 31 determines the range containing each state variable acquired in S101 as the processing target (S102). In other words, the position determination unit 311 determines the range of each data record with reference to the position field of preprocessing table 3212 and the range determination table 3211, storing the name of the determined range in the range field of each data record.

[0028] Subsequently, the first merging unit 312A, as the processing target, performs the merging processing only on the state variables belonging to the first area Z1, among those state variables recorded in S101 (S103). Next, the second merging unit 312B, as the processing target, performs the merging processing only on the state variables belonging to the second area Z2, among those state variables recorded in S101 (S104). Finally, the third merging unit 312C, as the processing target, performs the merging processing only on the state variables belonging to the third area Z3, among those state variables recorded in S101 (S105).

[0029] It is stated that, while the details of S103 to S105 are described later, the first fusion unit 312A, the second fusion unit 312B, and the third fusion unit 312C write the processing results to the post-processing table 3213. Furthermore, while S103 to S105 are executed sequentially, the order of the steps may be reversed, or the steps may be executed in parallel. Finally, the vehicle control processing unit 313 refers to the post-processing table 3213 and determines the control processing of vehicle C, issuing an operating command to actuator group 4. It is stated that the vehicle control processing unit 313 can also delete the information stored in the first area table 3221A, the second area table 3221B and the third area table 3221C before the execution of S106 or after the execution of S106. (Merger processing)

[0030] Fig. 8 is a flowchart detailing the merge processing to be performed by the first merge unit 312A; i.e., the details of S103 according to Fig. Figure 7 shows that the processing unit 31 first refers to the preprocessing table 3212 and generates the first range table 3221A by measuring the measurement objects belonging to the first range Z1 (S210). It is stated that, as described above, the adjustment column of all measurement objects in the first range table 3221A is "not yet adjusted" at the time of generation.

[0031] Next, processing unit 31 designates the measurement object with the smallest ID among those described in the first range table 3221A generated in S210 as the processing target (S211). Next, processing unit 31 extracts from among those described in the first range table 3221A a measurement object where the matching column is "not yet matched," the detection sensor differs from the measurement objects designated as the processing target, and the position difference compared to the processing target is within a predefined value (S212). Next, processing unit 31 extracts from among those extracted in S212 a measurement object where the velocity difference compared to the measurement objects designated as the processing target is within a predefined value (S213).

[0032] Next, processing unit 31 determines whether there is a measurement object extracted in S213 (S214). If processing unit 31 determines that there is a measurement object extracted in S213—that is, that there is another measurement object that satisfies the conditions of S212 and S213, except for the measurement objects being the processing target—processing unit 31 proceeds to S215 upon determining that these measurement objects are a match, and to S218 upon determining that there is no such other measurement object extracted in S213. In S215, processing unit 31 merges the position and velocity of the processing target and the measurement object extracted in S213. The merging is, for example, a simple average.Next, the processing unit 31 writes the calculation result of S215 into the post-processing table 3213, rewriting the matching columns of the measured objects as the processing target of the first range table 23221A and of the measured object extracted in S213 as ‘matched’ (S216).

[0033] Next, processing unit 31 determines whether there is a measurement object described in range table 3221A that is not a processing target and for which the matching column is "not yet matched" (S217). If processing unit 31 receives a positive determination in S217, it designates a measurement object that is not yet a processing target and for which the matching column is "not yet matched" as a new processing target (S218), returning to S212. Processing unit 31 then completes the process described in S217. Fig. 8 Processing shown when receiving a negative result in S217.

[0034] While the fusion processing to be carried out by the first fusion unit 312A in Fig. As explained in section 8, the fusion processing to be performed by the second fusion unit 312B and the third fusion unit 312C is the same, and its explanation is omitted. However, there are the following differences: specifically, the second fusion unit 312B is to process only the measurement objects belonging to the second area Z2, and the third fusion unit 312C is to process only the measurement objects belonging to the third area Z3.

[0035] According to the preceding first embodiment, the following operation and effect result.

[0036] (1) An electronic control device comprises an I / O unit 33 that acquires the state variables, including the position of a measurement object, using an output from several sensors contained in a sensor group 1; a storage unit 32 that stores a range determination table 3211 as a condition of a position-related classification; a position determination unit 311 that classifies the measurement object using the classification condition; and a merging unit 312 that, with reference to the state variables, determines a match between several of the measurement objects that have been classified as completely identical by the position determination unit 311 and that were measured by different sensors, and merges the positions of the several measurement objects that have been determined to be a match. The measurement objects are objects that can be measured by each of the several sensors.

[0037] The merging unit 312 determines the conformity of the measured objects by targeting those objects classified as completely identical by the positioning unit 311. Because the targets for which conformity with the measured objects is to be determined by the merging unit 312 are limited to the classification by the positioning unit 311, the processing load of the merging unit 312 can consequently be reduced. Furthermore, it is possible to increase the upper limit of the measured objects that can be processed by the electronic control device 3.

[0038] Here, a case in which the positioning unit 311 does not perform the classification is referred to as a "comparative example," and the effect of the positioning unit 311, with which the electronic control device 3 is equipped, is explained. For example, a case is considered in which the left camera LC, the right camera RC, the left radar LR, and the right radar RR each measure 10 objects; that is, a case in which a total of 40 objects are measured. In this comparative example, a match is determined for a specific object measured by the left camera LC with a total of 30 objects measured by the other three sensors. Because a simple calculation determines a match for each of the 40 objects with 30 other objects, 1200 combinations must be considered in the comparative example.Meanwhile, when performing the classification based on the position determination unit 311, the total number of combinations decreases because the combinations of the measured objects in the respective areas from the first area Z1 to the third area Z3 are taken into account.

[0039] Because the processing of processing unit 31 is executed for each predefined processing cycle, there is also an upper limit to the number of combinations that can be considered. Even if the total number of objects measured with the respective sensors is the same, the number of combinations to be considered in electronic control device 3 is smaller compared to the comparison example. In other words, if the number of combinations that can be considered is the same in the comparison example and in electronic control device 3, the upper limit of the objects that can be processed by electronic control device 3 is greater. (Modified Example 1)

[0040] The type and number of sensors that configure sensor group 1 in the first embodiment, the data acquisition range of each sensor, and the defined ranges are merely illustrations and may differ from the first embodiment. Sensor group 1 need only contain at least two sensors. The data acquisition ranges of the respective sensors need only overlap at at least one point. The ranges may be defined based on the data acquisition ranges of the respective sensors, in particular based on the number of overlaps of the data acquisition ranges with the other sensors. Moreover, the range determination table 3211 may be expressed in a format other than a table, and the position condition may also be specified as a mathematical expression. Furthermore, the range determination table 3211 may express the position using the coordinates of a rectangular coordinate system. (Modified Example 2)

[0041] The individual sensors contained in sensor group 1 can also directly output their sensor data to the electronic control device 3. In the preceding case, the electronic control device 3 additionally includes a state variable calculation unit that calculates the state variables of the measured objects using the respective sensors. The state variable calculation unit can be implemented with the processing unit 31 or with a hardware circuit. (Modified Example 3)

[0042] The electronic control device 3 can also be connected to an input device, such as a mouse or a keyboard, whereby the operator's input, which was entered into the input device via the I / O unit 33, can be used when rewriting the information stored in the memory unit 32. (Modified Example 4)

[0043] If the merging process is to be performed outside the electronic control device 3 and its result is to be input into the electronic control device 3, the electronic control device 3 can omit such merging processing. For example, a disparity image is generated using the images output by the left camera LC and the right camera RC, and the generated disparity image is then input into the electronic control device 3. In the preceding case, the disparity image can be generated either by the left camera LC, the right camera RC, or by another device.

[0044] According to the modified Example 4, by omitting the merging process to be performed by the electronic control device 3, the processing volume can be reduced, thereby also reducing the processing delay. Furthermore, because the processing power required by the electronic control device 3 decreases, the manufacturing costs of the electronic control device 3 can be reduced. - Second embodiment -

[0045] The second embodiment of the electronic control device according to the present invention is now described with regard to Fig. 9 to Fig. 12. In the following explanation, the same constituent elements as in the first embodiment are given the same reference numeral, with the main focus being on explaining the differences between the second and first embodiments. All points not specifically explained are the same as in the first embodiment. This embodiment differs from the first embodiment mainly in that several patterns are described in the range determination table, and the pattern to be used is changed depending on the sensor's performance degradation. Furthermore, in this embodiment, the sensors are assigned to the respective ranges, which may overlap. (Configuration)

[0046] Fig. Figure 9 is a graphical representation showing the configuration of vehicle C in the second embodiment. In this embodiment, vehicle C comprises all the configurations of the first embodiment, additionally including an illuminance sensor 21, which measures the ambient brightness, and a wiper sensor 22, which monitors the operation of the vehicle C's wiper. In the following explanation, however, the illuminance sensor 21 and the wiper sensor 22 may optionally be referred to collectively as an ambient monitoring device 2. The electronic control device 3 acquires the measured value of the ambient monitoring device 2 using the I / O unit 33. Furthermore, in this embodiment, the common storage unit 321 additionally stores a pattern determination table 3214 and the sensor range assignment information 3217, storing a range determination table 3211A as a replacement for the range determination table 3211.The processing unit 31 additionally includes a range change unit 314 as its function.

[0047] The range change unit 314 estimates the performance degradation of one or more sensors based on information about the state variables of the object being measured, acquired by sensor group 1, and the external environment, acquired by environmental monitoring device 2, and determines whether the estimated performance degradation corresponds to any of several predefined patterns. This pattern is a pattern that defines the range. The performance degradation includes a degradation in the amount and quality of information received and a degradation in sensitivity. In other words, the range change unit 314 acquires the performance estimation information for estimating the performance degradation of sensor group 1 from sensor group 1 and environmental monitoring device 2. A higher-level conceptualization of the operation of the range change unit 314 is as follows.In other words, the range change unit 314 identifies a sensor whose performance has deteriorated, using the performance estimation information acquired by the I / O unit 33, and changes the coordinate condition of the range determination table 3211A so that the range assigned to that sensor is narrowed.

[0048] The range determination table 3211A in the second embodiment describes the coordinate condition that corresponds to the respective patterns determined by the range change unit 314. It is noted that in the following explanation, the coordinate condition may also be referred to as the "position condition." In this embodiment, the position determination unit 311 therefore refers to the range determination table 3211A after the range change unit 314 has first determined the pattern. In other words, the position determination unit 311 determines the ranges of the respective measurement objects based on the coordinate condition described in the range determination table 3211A, which corresponds to the pattern determined by the range change unit 314.Furthermore, in this embodiment, the six areas of a second A-area Z2A up to a second F-area Z2F are defined as a replacement for the second area Z2 in the first embodiment, with a total of eight areas defined.

[0049] It is stated that, while in Fig. Figure 9 illustrates only a second merging unit 312B, which can process the six areas of the second A-area Z2A up to the second F-area Z2F, or the merging units for processing the preceding six areas can be independent of each other. In the preceding case, the processing unit 31 comprises a total of eight merging units.

[0050] Pattern determination table 3214 describes the information for determining which of the several patterns for defining the area corresponds to the state of sensor group 1. Because several patterns are described in area determination table 3211A in this embodiment, the position determination unit 311 refers to pattern determination table 3214, as described above, to determine which of these patterns should be used.

[0051] The sensor area assignment information 3217 shows the assignment of the sensors with respect to the respective areas, with at least two sensors assigned to each area. The fusion processing unit 312 refers to the sensor area assignment information 3217 and processes only the measurement objects of the sensors that are assigned to each area. The details are described later. (Sample determination table 3214)

[0052] Fig. Figure 10 is a graphical representation showing an example of pattern identification table 3214. Pattern identification table 3214 describes a combination of a condition and an area pattern. The condition field describes the condition regarding the sensor's performance, while the area pattern field describes the name of the corresponding area pattern. The first data record after Fig. Figure 10 shows, for example, that the area pattern is "Pattern B" when the illuminance E detected by the illuminance sensor 21 is less than a predefined threshold Ea. The next data set shows that the area pattern is "Pattern C" when the output of the wiper sensor 22 indicates that the wiper switch has been turned ON.

[0053] The next data set shows that the area pattern is "pattern D" when the velocity of the measured object output by the left camera LC is discontinuous. The continuity of the velocity of the measured object output by the left camera LC can be evaluated, for example, as follows. With respect to a particular measured object output by the left camera LC, it is assumed that the previously measured time t0 is t0, the previously measured position is R(t0), the previously measured velocity is V(t0), the currently measured time t1 is t1, the currently measured position is R(t1), and the currently measured velocity is V(t1). If the following formula (1) is continuously satisfied for a given number N of times with respect to all measured objects of the left camera LC, then the velocity is determined to be discontinuous. |R(t1)−{R(t0)+V(t0)(t1−t0)}|>ΔR

[0054] The next data set shows that when it is determined that the left camera LC has failed, the area pattern is "Pattern D," which is the same as the immediately preceding data set. The last data set shows that the area pattern is "Pattern A" when the condition does not correspond to either of the cases; that is, when there is no particular problem with sensor group 1. It is stated that the area pattern in the first embodiment corresponds to "Pattern A." While in the Fig. In the example shown in 10, where several patterns are specified, there is also no restriction on the number of patterns, as long as the number of patterns is two or more patterns. (Area determination table 3211A)

[0055] Fig. Figure 11 is a graphical representation showing an example of the range determination table 3211A in its second embodiment. The range determination table 3211A records several patterns with respect to the position condition. In the Fig. In the 11 examples shown, the two positional conditions of pattern A and pattern B are specified in the area determination table 3211A, while the specific values ​​of the other patterns are omitted. The area determination table 3211A contains the coordinate conditions of all patterns described in the pattern determination table 3214. (Sensor area assignment information 3217)

[0056] Fig. Figure 12 is a graphical representation showing an example of the sensor area assignment information 3217. This is in Fig. Example 12 shows that the left camera LC and the left radar LR are assigned to the first area Z1, two sensors selected from four sensors, specifically the left camera LC, the right camera RC, the left radar LR and the right radar RR, are assigned to the second A area Z2A to the second F area Z2F, and the right camera RC and the right radar RR are assigned to the third area Z3.

[0057] It is stated that with regard to pattern D, which is in Fig. As not illustrated in Figure 11, the size of the first area Z1, the second A area Z2A, the second B area Z2B, and the second C area Z2C, which are assigned to the left camera LC, is equal to or smaller than the size of the areas in Pattern A, and the size of the areas increases, with the exception of the preceding areas not assigned to the left camera LC. In Pattern D, for example, the first area Z1 is narrower compared to Pattern A, the planar dimensions of the second A area Z2A do not change up to the second F area Z2F, and the planar dimensions of the third area Z3 are wider. (Operation of processing unit 31)

[0058] Fig. Figure 13 is a flowchart showing the processing of processing unit 31 in the second embodiment. The differences compared to the one in Fig. The differences in the processing shown in section 7 of the first embodiment are that the S101A and the S101B have been added between the S101 and the S102, and that the S102 has been changed to the S102A. The differences compared to the first embodiment are now explained.

[0059] In S101A, the I / O unit 33 receives the output from the environmental monitoring device 2. In the subsequent S101B, the range change unit 314 refers to the information acquired in S101 and S101A and the pattern determination table 3214 and determines a range pattern. In S102A, the position determination unit 311 uses the position condition identified based on the range pattern determined in S101B and the range determination table 3211A to determine the range to which the processing target acquired in S101 belongs. (Merger processing)

[0060] The operation of the fusion unit 312 in the second embodiment differs from that of the first embodiment with respect to the determination of the processing target by also taking into account the sensor area assignment information 3217. For example, the first fusion unit 312A defines the measurement object as the processing target if the position of the processing target is the first area Z1 and if the sensor that detected the processing target is assigned to the first area in the sensor area assignment information 3217. When the operational flow of the first fusion unit 312A in the second embodiment is described, the S210 can be followed by Fig. 8 will be amended as follows.

[0061] In other words, in the S210, the first fusion unit 312A refers to the preprocessing table 3212 and the sensor area mapping information 3217, generating a first area table 3221A by extracting a measurement object belonging to the first area Z1 where the detection sensor is the left camera LC or the left laser LR. It is stated that in the first embodiment, if the position of the measurement object detected by the right camera RC is the first area Z1, the processing can be performed by considering or ignoring the information about this measurement object. However, in this embodiment, the measurement object in the first area Z1 detected by the right camera RC is ignored.

[0062] According to the preceding second embodiment, the following operation and effect result.

[0063] (2) The corresponding areas classified on the basis of the area determination table 3211A are assigned to at least two sensors among several sensors. The merging unit 312 determines, with reference to the state variables, a match between the several measurement objects that are classified by the position determination unit 311 as being in the same area and that were measured by different sensors assigned to the same area, and merges the positions of the several measurement objects that have been determined to be a match.The electronic control device 3 comprises an I / O unit 33 that acquires the power estimation information for estimating the deterioration of the performance of each of the multiple sensors, and a range change unit 314 that changes the range to which reference is made by the range determination table 3211A, so that the range associated with the sensor in which the performance has deteriorated is narrowed using the power estimation information.

[0064] For example, because pattern D is selected as the area pattern when it is determined that the performance of the left camera LC has deteriorated, the first area Z1, which is assigned to the left camera LC, is narrowed, while the third area Z3, which is not assigned to the left camera LC, is widened. Consequently, the size of the area using the left camera LC, where the performance has deteriorated, is reduced, making fusion processing using the detection result from a sensor with relatively high sensor performance more likely compared to the sensor performance before the area change. Accordingly, the electronic control device 3 can select fusion processing using a detection result from a sensor with relatively high sensor performance without increasing the overall processing load.Consequently, a high level of security can be ensured with a smaller amount of processing, even if there is a change in sensor performance.

[0065] Furthermore, pattern C is selected when the wiper switch is ON, narrowing the area assigned to at least either the left camera LC or the right camera RC. Specifically, the following three configurations can be considered. The first configuration sets only the second F-area Z2F, assigned to the left radar LR and the right radar RR, as a non-empty set, while the other areas are set as empty sets. In other words, in the first configuration, the planar dimension of the areas, except for the second F-area Z2F, is zero. In the second configuration, only the second C-area Z2C, assigned to the left camera LC and the right camera RC, is set as empty sets, while the other areas are set to be the same as pattern A.The third configuration is a setting that lies between the first and second configurations. The reason for this is explained below.

[0066] If the wiper switch is ON, it is most likely raining. While a camera is slightly affected by rain, a millimeter-wave radar is not easily affected by rain. Consequently, the three configurations described above can be used to utilize the left radar (LR) and the right radar (RR), which are not easily affected by rain.

[0067] (3) The storage unit 32 stores a pattern determination table 3214, which prescribes a match between the degradation of performance of the multiple sensors contained in sensor group 1 and the classification condition. The range change unit 314 determines a pattern contained in the pattern determination table 3214 using the performance estimation information. Consequently, the range can be easily changed. (Modified example 1 of the second embodiment)

[0068] The environmental monitoring device 2 can be replaced by the sensor group 1. In other words, if the conditions described in pattern determination table 3214 can be determined using the output of sensor group 1, the electronic control device 3 need not include the environmental monitoring device 2. If the CPU that configures the processing unit 31 has the function of a clock or calendar, the electronic control device 3 can still use the output of the processing unit 31 for determining the pattern determination table 3214. (Modified example 2 of the second embodiment)

[0069] In the preceding second embodiment, several patterns were pre-registered in the range determination table 3211A. The electronic control device 3 can nevertheless also determine the respective ranges based on operation without including the range determination table 3211A. The processing unit 31 can, for example, additionally include a Kalman filter that estimates an error contained in the observed set of the respective sensors included in the sensor group 1, wherein the range change unit 314 can determine the respective ranges based on the magnitude of the error estimated by the Kalman filter. In the preceding case, based on the same concept as the second embodiment, the range associated with the sensor in which the estimated error is large is narrowed, while the range associated with the sensor in which the estimated error is small is widened.In other words, in this modified example, the size of the area can be flexibly changed according to the size of the error. (Modified example 3 of the second embodiment)

[0070] In the second embodiment, the sensor and the area were assigned to the sensor area assignment information 3217. However, the sensor can also be assigned to the corresponding fusion units 312. In the previous case, the fusion units 312 are configured, for example, from eight fusion units (the first fusion unit to the eighth fusion unit), with each fusion unit being assigned to a different area. - Third embodiment -

[0071] The third embodiment of the electronic control device according to the present invention is now described with regard to Fig. 14 to Fig. 17. In the following explanation, the same constituent elements as in the first embodiment are given the same reference numeral, with the main focus being on explaining the differences between the third embodiment and the first embodiment. All points not specifically explained are the same as in the first embodiment. This embodiment differs from the first embodiment mainly in that the future positions of the fused measurement objects are estimated, and the estimated future positions are further subjected to the fusion process. (Configuration)

[0072] Fig. Figure 14 is a graphical representation showing the configuration of vehicle C in the third embodiment. In this embodiment, the electronic control device 3, in addition to the configuration of the first embodiment, further comprises an estimation unit 315 of the future positions in the processing unit 31. Moreover, the storage unit 32 additionally stores a determination table 3215 of the future areas and a future table 3216.

[0073] The Future Positions Estimator Unit 315 estimates the future state variables based on the merged state variables described in the Post-Processing Table 3213 and writes the result to the future table 3216. In the following explanation, the position contained in the state variables estimated by the Future Positions Estimator Unit 315 is referred to as the "future position." For example, the Future Positions Estimator Unit 315 calculates the difference between the positions using the time interval for each processing cycle and the velocity of the merged state variables, and estimates the future position based on the position of the merged state variables and the calculated difference. The Future Positions Estimator Unit 315 can, for example, define the future velocity as the velocity of the merged state variables.Furthermore, the estimation unit 315 of future positions determines whether the estimated position corresponds to any area described in the determination table 3215 of future ranges, storing the determination result in the determination table 3215 of future ranges.

[0074] If the time of the current processing cycle is t1, the time of the next processing cycle is t2, the position of the merged state variables is R(t1) and the velocity of the merged state variables is V(t1), the estimation unit 315 of future positions can, for example, also calculate the future position R(t2), as shown in formula (2) below. R(t2)=R(t1)+V(t1)(t2−t1). (Determination table 3215 of the future areas)

[0075] Fig. Figure 15 is a graphical representation showing an example of the future area determination table 3215. The future area determination table 3215 stores the future positions of the merged measurement objects; that is, it stores the correspondence between the state and the area number relative to the position estimated by the future position estimation unit 315. The configuration of the future area determination table 3215 is the same as that of the area determination table 3211. The area numbers contained in the future area determination table 3215 are the same as those in the area determination table 3211, although the position conditions may be the same as or different from those of the area determination table 3211. (Future Table 3216)

[0076] Fig. Figure 16 is a graphical representation showing an example of future table 3216. Future table 3216 stores the future state variables of the merged measurement objects and the classification of the future positions of the merged measurement objects. Future table 3216 is written using the estimation unit 315 of the future positions. (Operation of processing unit 31)

[0077] Fig. Figure 17 is a flowchart showing the processing of the processing unit 31 in the third embodiment. The differences compared to the one in Fig. The differences in the processing shown in section 7 of the first embodiment are that S102B has been added after S102, and S107 and S108 have been added after S106. The differences compared to the first embodiment will now be explained.

[0078] In S102B, the estimation unit 315 of future positions rewrites the future table 3216; that is, it reads the entire future table 3216 and records the corresponding records in the appropriate range table according to the value of the range number field. Because in the Fig. In the example shown in Figure 16, if the two topmost records both have a future range of "Z2", the information from these records is recorded in the second range table 3221B. However, the value of the capture sensor field to be written to the range table in S102B is fixed so that it is a value different from the respective sensors, such as "Merger". Once the future position estimation unit 315 has read the entire future table 3216, it deletes all the information that was stored in the future table 3216.

[0079] In S107, the Future Positions Estimating Unit 315 calculates the state variables in the next processing cycle, including the future positions of all merged processing targets described in Post-Processing Table 3213, and describes the result in Future Table 3216. In the subsequent S108, the Future Positions Estimating Unit 315 refers to Future Table 3216 and Future Range Determination Table 3215, classifying the future positions and entering the result into the Range Number field of Future Table 3216. More specifically, the Future Positions Estimating Unit 315 determines whether the future positions correspond to any range described in Future Range Determination Table 3215, and enters the determination result into the Range Number field of Future Table 3216. This concludes the explanation of the Fig. 17 and up.

[0080] It is stated that the merging process in the respective areas shown in S103 to S105 is the same as in the first embodiment, but varies considerably due to the influence of S102B as follows. For example, the first merging unit 312A extracts the state variables, including the future position, in S212 according to Fig. 8, wherein in S214 it includes the future positions that meet the conditions of S212 and S213 as the merger target. In S214, the processing unit 31 can use a simple average as in the first embodiment or a weighted average in which the weight between the estimated positions and the other positions is changed.

[0081] According to the preceding third embodiment, the following operation and effect result.

[0082] (4) The storage unit 32 stores a determination table 3215 of future ranges as a condition of a future position-related classification, which is a future position of the object being measured, and a future table 3216 containing the future position at which the future range was classified by the determination table 3215 of future ranges. The processing unit 31 comprises an estimation unit 315 of future positions, which estimates the future position of the object being measured and classifies the future position using the determination table 3215 of future ranges.The merging unit 312 also contains in the targets to be determined the measurement objects which are classified as completely identical by the estimation unit 315 of future positions when determining a match of the measurement objects, and merges the positions and the future positions of the several measurement objects which have been determined to be a match.

[0083] To improve the accuracy of the merged measurement objects, it would generally be effective to include estimates based on past measurements in the merging processing targets. However, the number of combinations increases significantly with the increasing number of processing targets, creating the problem that the processing load increases while the number of measurement objects that can be processed within the processing cycle decreases. The electronic control device 3 therefore determines the range of calculated future positions and restricts the combinations so that they remain within the same range. Consequently, the accuracy of the state variables of the merged measurement objects can be improved while suppressing the increase in processing load.

[0084] (5) The future position estimation unit 315 estimates the future positions of the measurement objects merged by the fusion unit 312. Consequently, the state variables of the measurement objects can be calculated with high accuracy in the subsequent processing cycle using the future positions of the state variables calculated using multiple sensor outputs. (Modified example 1 of the third embodiment)

[0085] As an alternative to writing the estimated future positions to future table 3216, the future positions estimation unit 315 can also write the estimated future positions to one of the first range table 3221A, the second range table 3221B, and the third range table 3221C. In other words, the future positions estimation unit 315 writes the future positions, according to the range into which the calculated future positions are classified, to one of the first range table 3221A, the second range table 3221B, and the third range table 3221C. In the preceding case, S102B can be written to... Fig. 17 will be deleted. (Modified example 2 of the third embodiment)

[0086] In the preceding third embodiment, the estimation target for the state variables was only the merged measurement objects. However, the estimation unit 315 of the future positions can also specify the estimation target for the state variables so that it is only the measurement objects that have not been merged, or specify the estimation target for the state variables to include both the merged and the non-merged measurement objects.

[0087] Furthermore, the estimation unit 315 of future positions can perform the following processing as a substitute for writing the estimated state variables to the future table 3216. In other words, the estimation unit 315 of future positions can also classify the future positions of the measurement objects that have not been merged, with reference to the determination table 3215 of future areas, and write the result to one of the first area table 3221A, the second area table 3221B, and the third area table 3221C that corresponds to the classified area. It could be said that the processing in the preceding case is the processing of deleting the measurement objects that were determined to be moving to another area from the area table of the movement source and moving them to the area table of the movement target.

[0088] According to this modified example, the following operation and effect result.

[0089] (6) The position determination unit 311 and the estimation unit 315 of future positions store the results of the classification in an assigned area of ​​the dedicated storage unit 322 for each result of the classification determined with respect to the area table 3211A and the future table 3216. Consequently, there is no need to rewrite the state variables from the future table 3216 into the dedicated storage unit 322, thereby simplifying processing.

[0090] In each of the embodiments and modified examples described above, the programs of the electronic control device 3 are stored in ROM, but the programs may also be stored in the memory unit 32. Furthermore, the programs can also be read from another device via the I / O interface and a medium that can be used by the electronic control device 3. This medium could be, for example, a storage medium that can be connected to and disconnected from an I / O interface, or a communication medium such as a wired, wireless, or optical network, or carrier waves or digital signals transmitted over such a network. In addition, some or all of the functions implemented by the programs may be implemented by a hardware circuit or an FPGA.

[0091] Each of the embodiments and modified examples described above can be combined. While various embodiments and modified examples have been explained above, the present invention is not limited to its subject matter. Other modes considered to fall within the technical concept of the present invention are also within its scope of protection.

[0092] The disclosure of the following priority application is fully incorporated here by reference.

[0093] Japanese patent application no. 2018-84266 (filed on April 25, 2018) LIST OF REFERENCE MARKS 1 sensor group 2 Environmental monitoring device 3 electronic control device 4 Actuator group 31 processing units 32 storage units 33 I / O units 311 Position determination unit 312 Merger unit 313 Vehicle control processing unit 314 Area Change Unit 315 Estimation unit of future positions 321 common storage unit 322 dedicated storage units 3211 Area determination table 3212 Preprocessing table 3213 Post-processing table 3214 Sample determination table 3215 Determination table of future areas 3216 future table 3221A first area table 3221B second area table 3221C third area table C vehicle

Claims

[1] Electronic control device (3) comprising: a sensor information acquisition unit (33) that acquires the state variables including a position of a measurement object using an output from multiple sensors or calculates the state variables of the measurement object using the output from the multiple sensors; a storage unit (32) that stores position determination information as a condition of a position-related classification; a positioning unit (311) that classifies the object being measured using the positioning information; and a merging unit (312) which, with reference to the state variables, determines a match among several of the measured objects which are classified as completely identical by the position determination unit (311) and which were measured by different sensors, and which merges the positions of the several measured objects which were determined to be a match, wherein the objects being measured are objects that can be measured by each of the multiple sensors, wherein the storage unit (32) additionally stores: a future classification condition as a condition of a classification that refers to a future position, which is a future position of the object of measurement; and a classified future position as the future position of the object of measurement that has been classified by the future classification condition; the electronic control device (3) additionally includes: an estimation unit (315) of future positions that estimates the future position of the object being measured and classifies the future position using the future classification condition; and the merging unit (312) also incorporates into the objectives of a merging process the measurement objects which are classified as completely identical by the estimation unit (315) of future positions when determining a match of the measurement objects, and merges the positions and future positions of the multiple measurement objects which have been determined to be a match. [2] Electronic control device (3) according to claim 1, wherein: Each area, which is classified based on the position determination information, has at least two sensors assigned to it among the multiple sensors; The merging unit (312) determines, with reference to the state variables, a match between the multiple measurement objects that are classified by the position determination unit (311) as being in the same area and that were measured by different sensors assigned to the same area, and merges the positions of the multiple measurement objects that have been determined to be a match; the merger unit (312) comprises: a performance estimation information acquisition unit that acquires performance estimation information to estimate the degradation of performance of each of the multiple sensors; and a range change unit (314) that modifies the position determination information so that the area associated with the sensor in which performance has deteriorated is narrowed using the power estimation information. [3] Electronic control device (3) according to claim 2, wherein: the storage unit (32) additionally stores pattern determination information that prescribes several of the classification conditions and a correspondence between the deterioration of the performance of the multiple sensors and the classification conditions; and The range change unit (314) uses the performance estimation information to determine which of the classification conditions contained in the pattern determination information should be used. [4] Electronic control device (3) according to claim 1, wherein: The estimation unit (315) of future positions estimates at least one of either the future positions of the measurement objects merged by the fusion unit (312) or the future positions of the measurement objects for which the state variable was acquired by the sensor information acquisition unit (33). [5] Electronic control device (3) according to claim 1, wherein: The positioning unit (311) and the estimation unit (315) of the future positions store the results of the classification in an assigned area of ​​the storage unit (32) for each result of the classification. [6] Operating method to be carried out by an electronic control device (3) comprising a sensor information acquisition unit (33) which acquires the state variables containing a position of a measurement object using an output from multiple sensors or calculates the state variables of the measurement object using the output from the multiple sensors, and a storage unit (32) which stores position determination information as a condition of a position-related classification, wherein the operating method comprises: Classifying the object being measured using the position determination information; and Determining, with reference to the state variables, a match among several of the measured objects, which are classified as completely identical based on the classification and which were measured by different sensors, and merging the positions of the several measured objects that were determined to be a match, where the objects being measured are objects that can be measured by each of the multiple sensors, wherein: the storage unit (32) additionally stores: a future classification condition as a condition of a classification that refers to a future position, which is a future position of the object of measurement; and a classified future position as the future position of the object of measurement that has been classified by the future classification condition; the operating procedure additionally includes: Estimating the future position of the object being measured and classifying the future position using the future classification condition; wherein In the classification process, the future positions that are classified as completely identical are included in the objectives of a merging process, which are determined when determining a match between the measurement objects, and the positions and future positions of the multiple measurement objects that have been determined to be a match are merged. [7] Operating method according to claim 6, wherein: the storage unit (32) additionally stores sensor area assignment information as an assignment to at least two sensors among the multiple sensors with respect to respective areas classified based on the condition of classification; During the merging process, a match is determined among the multiple measurement objects that are classified as being in the same area based on the classification and that were measured by different sensors that are assigned to the same area based on the sensor area assignment information, and the positions of the multiple measurement objects that have been determined to be a match are merged; the operating procedure additionally includes: Gathering performance estimation information to estimate the degradation of performance of each of the multiple sensors; and Modify the positioning information so that the area associated with the sensor where performance has deteriorated is narrowed using the power estimation information. [8] Operating method according to claim 7, wherein: the storage unit (32) additionally stores pattern determination information that prescribes several of the classification conditions and a correspondence between the deterioration of the performance of the multiple sensors and the classification conditions; and the operating procedure additionally includes: Using the performance estimation information, determine which of the classification conditions contained in the pattern determination information should be designated as the new position determination information. [9] Operating method according to claim 6, wherein: When estimating the future position, at least one of either the future positions of the objects that were merged or the future positions of the objects for which the state variable was recorded must be estimated. [10] Operating method according to claim 6, wherein: The set of states and the future position that were subjected to classification are stored for each result of the classification in an assigned area of ​​the storage unit (32).

Images