Accident prediction device, accident prediction method, and program

Abstract

To prevent accidents. [Solution] The accident prediction device 10 includes an acquisition unit 101 that acquires driving data measured by sensors installed in the vehicle, an evaluation unit 103 that evaluates driving based on the driving data acquired by the acquisition unit 101, a prediction determination unit 104 that compares evaluation data evaluated by the evaluation unit 103 based on driving data in the immediate vicinity with predetermined comparison data and determines that there is a sign of an impending accident if the evaluation data deviates from the comparison data by a predetermined threshold or more, and a warning unit 105 that warns the driver of the vehicle if the prediction determination unit 104 determines that there is a sign of an impending accident. The comparison data is evaluation data evaluated by the evaluation unit 103 based on driving data from a specific period in the past that is longer than a predetermined period.

Description

Technical Field

[0001] The present invention relates to an accident prediction determination device, an accident prediction determination method, and a program.

Background Art

[0002] A vehicle accident prediction system has been proposed that not only uses a second feature amount representing the state of a vehicle but also a first feature amount representing the attributes of the driver of the vehicle or a third feature amount obtained by combining a plurality of second feature amounts to enhance the input feature amounts themselves and generate a highly accurate learned model for predicting vehicle accidents (for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Although the system described in Patent Document 1 aims to improve the accuracy of accident prediction, a system that can notify the driver of such accident prediction results and prevent the occurrence of accidents is desired.

[0005] In view of the above circumstances, the present invention has been made, and an object thereof is to provide an accident prediction determination device or the like that suppresses the occurrence of accidents by determining a sign of an accident and giving a warning according to the determination result.

Means for Solving the Problems

[0006] To achieve the above object, the accident prediction determination device of the present invention an acquisition unit that acquires driving data measured by sensors provided in a vehicle, an evaluation unit that evaluates driving based on the driving data acquired by the acquisition unit, A precursor determination unit compares evaluation data, which is evaluated by the evaluation unit based on the aforementioned driving data over a fixed period of time, with predetermined comparison data, and determines that there is a precursor to an accident if the evaluation data deviates from the comparison data by a predetermined threshold or more. If the aforementioned warning unit determines that there are signs of an impending accident, the warning unit provides a warning to the driver of the vehicle, and the system is equipped with The comparison data is evaluation data that the evaluation unit has evaluated based on the driving data for a specific period in the past that is longer than the predetermined period. the law of nature, The aforementioned evaluation data includes first evaluation data relating to the frequency of dangerous behavior and second evaluation data relating to the distance traveled. The aforementioned comparative data includes first comparative data relating to the frequency of dangerous behavior and second comparative data relating to the distance traveled. The predictive indicator unit determines that there is a sign of an impending accident if the first evaluation data deviates from the first comparison data by a first threshold or more, and the second evaluation data deviates from the second comparison data by a second threshold or more. .

[0007] The accident prediction method of the present invention is: The accident prediction device, We acquire driving data measured by sensors installed in the vehicle. Based on the acquired driving data, the driving is evaluated. The evaluation data, based on the aforementioned driving data over a fixed period, is compared with predetermined comparison data. If the evaluation data deviates from the comparison data by a predetermined threshold or more, it is determined that there is a sign of an impending accident. If it is determined that there are signs of an impending accident, the driver of the vehicle will be warned. The comparison data is evaluation data evaluated based on the driving data for a specific period in the past that is longer than the predetermined period. the law of nature, The aforementioned evaluation data includes first evaluation data relating to the frequency of dangerous behavior and second evaluation data relating to the distance traveled. The aforementioned comparative data includes first comparative data relating to the frequency of dangerous behavior and second comparative data relating to the distance traveled. The accident prediction device determines that there is an indication of an impending accident if the first evaluation data deviates from the first comparison data by a first threshold or more, and the second evaluation data deviates from the second comparison data by a second threshold or more. .

[0008] The program of the present invention, Computers, An acquisition unit that acquires driving data measured by sensors installed in the vehicle. An evaluation unit evaluates the driving based on the driving data acquired by the acquisition unit. Based on the driving data during the recent fixed period, compare the evaluation data evaluated by the evaluation unit with predetermined comparison data. When the evaluation data deviates from the comparison data by a predetermined threshold or more, a sign determination unit determines that there is a sign of an accident occurring. When it is determined by the sign determination unit that there is a sign of an accident occurring, function as a warning unit that warns the driver of the vehicle. The comparison data is evaluation data evaluated by the evaluation unit based on the driving data during a specific past period that is longer than the predetermined period. the law of nature, The aforementioned evaluation data includes first evaluation data relating to the frequency of dangerous behavior and second evaluation data relating to the distance traveled. The aforementioned comparative data includes first comparative data relating to the frequency of dangerous behavior and second comparative data relating to the distance traveled. The predictive indicator unit determines that there is a sign of an impending accident if the first evaluation data deviates from the first comparison data by a first threshold or more, and the second evaluation data deviates from the second comparison data by a second threshold or more. 。

Advantages of the Invention

[0009] According to the present invention, suppression of the occurrence of an accident can be expected.

Brief Description of the Drawings

[0010] [Figure 1] It is a functional block diagram of an accident sign determination system according to the present embodiment. [Figure 2] It is a diagram showing an example of the hardware configuration of an accident sign determination device according to the present embodiment. [Figure 3] It is a flowchart showing an example of accident sign determination processing. [Figure 4] It is a flowchart showing an example of determination processing. [Figure 5] It is a flowchart showing an example of warning processing.

Embodiments for Carrying Out the Invention

[0011] Hereinafter, an accident sign determination device 10 according to an embodiment of the present invention and an accident sign determination system 1 including the accident sign determination device 10 will be described with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals.

[0012] Figure 1 is a functional block diagram of the accident prediction system 1. The accident prediction system 1 consists of an accident prediction device 10 and an in-vehicle device 21 mounted on a vehicle 20. The accident prediction device 10 is connected to the in-vehicle device 21 via a communication network such as the Internet or a mobile communication network.

[0013] The accident prediction device 10 is a device that determines whether or not there are signs of an impending traffic accident based on driving data of the vehicle 20 acquired from the in-vehicle equipment 21, and issues a warning if there are signs of an impending accident. If the accident prediction device 10 determines that there are signs of an impending traffic accident, it notifies the in-vehicle equipment 21 of this fact, thereby warning the driver of the vehicle 20.

[0014] Vehicle 20 is an automobile used by a user utilizing the accident prediction system 1. The in-vehicle device 21 is mounted on Vehicle 20 and is equipped with various sensors such as GPS (Global Positioning System), acceleration sensors, and cameras to measure driving data such as the vehicle's speed, acceleration, and distance traveled. The in-vehicle device 21 is, for example, an in-vehicle telematics device such as a communication-type drive recorder. The in-vehicle device 21 may also be equipped with a steering angle sensor to detect the steering angle and measure information on driving operations (such as sudden steering). Furthermore, the in-vehicle device 21 may measure driver information such as the driver's gaze and eye opening. The in-vehicle device 21 transmits the measured data to the accident prediction device 10. The driving data measured by the in-vehicle device 21 is stored on a cloud server, and the accident prediction device 10 may acquire the driving data from the cloud server. Additionally, the in-vehicle device 21 is equipped with output devices such as speakers and displays, enabling it to notify the driver of the information transmitted from the accident prediction device 10 via voice or images. In this embodiment, the in-vehicle device 21 notifies the driver and issues a warning when the accident prediction device 10 determines that there are signs of an impending accident.

[0015] The in-vehicle device 21 may be a digital tachograph, a car navigation system, a smartphone owned by the driver of the vehicle 20, or a device that connects to the vehicle 20 to acquire driving data from the vehicle 20, or it may be composed of multiple devices.

[0016] Although Figure 1 shows a single vehicle 20, it may include vehicles 20 belonging to multiple users of the accident prediction system 1, for example, each policyholder of an automobile insurance policy using the in-vehicle equipment 21.

[0017] Functionally, the accident prediction judgment system 1 comprises an acquisition unit 101, a driving data DB (database) 102, an evaluation unit 103, a prediction judgment unit 104, and a warning unit 105, as shown in Figure 1.

[0018] The acquisition unit 101 acquires driving data of the vehicle 20 measured by the in-vehicle device 21, and stores the acquired driving data in the driving data DB 102, associating it with the identification information of the vehicle 20's user (driver). In this embodiment, the in-vehicle device 21 is activated when the vehicle 20 is running, and the driving data measured by the in-vehicle device 21 is sequentially transmitted to the accident prediction judgment device 10. As a result, the acquisition unit 101 receives driving data from the in-vehicle device 21. The acquisition unit 101 may also receive driving data from the in-vehicle device 21 at a predetermined acquisition timing. Alternatively, the acquisition unit 101 may acquire driving data via a cloud server.

[0019] The driving data DB102 consists of a storage device such as non-volatile memory, an HDD (Hard Disk Drive), and / or an SSD (Solid State Drive), and stores and saves the user-specific driving data acquired by the acquisition unit 101.

[0020] The evaluation unit 103 evaluates the user's driving and generates evaluation data based on the driving data acquired by the acquisition unit 101 or the driving data stored in the driving data DB 102.

[0021] The Precursor Determination Unit 104 determines whether or not there are signs of an impending accident based on the evaluation data evaluated by the Evaluation Unit 103. In other words, the Precursor Determination Unit 104 determines whether or not there are signs of an impending accident involving the vehicle 20 driven by the user corresponding to the evaluation data. Specifically, the Precursor Determination Unit 104 determines that there are signs of an impending accident if the evaluation data based on driving data for a recent fixed period (the past week, a recent fixed time during the current drive, etc.) deviates from the comparison data by a predetermined threshold or more. The comparison data is evaluation data based on driving data for a specific period in the past that is longer than the predetermined period (the past six months, one year, etc.). In other words, the Precursor Determination Unit 104 determines that there are signs of an impending accident if the most recent evaluation data deviates significantly from past evaluation data, as there is a high possibility that there is a problem with the recent driving.

[0022] If the warning unit 105 determines that there are signs of an impending accident by the premonitory sign determination unit 104, it notifies the in-vehicle device 21 of this fact, thereby warning the driver of the vehicle 20 that there are signs of an impending accident.

[0023] Note that the functional configuration shown in Figure 1 is just one example, and the accident prediction device 10 may have configurations other than those shown in Figure 1, or the functional configuration may be integrated or subdivided.

[0024] Figure 2 shows an example of the hardware configuration of the accident prediction and determination device 10. The accident prediction and determination device 10 includes a processor 11 such as a CPU (Central Processing Unit) and a GPU (Graphical Processing Unit), a storage device 12 such as memory, an HDD and / or an SSD, a communication interface 13 for wired or wireless communication, an input device 14 for receiving input operations, and an output device 15 for outputting information. The input device 14 is, for example, a keyboard, a touch panel, a mouse and / or a microphone. The output device 15 is, for example, a display and / or a speaker.

[0025] The processor 11 reads the program stored in the storage device 12 and operates as various functional units as shown in Figure 1. The storage device 12 functions as the driving data DB 102 shown in Figure 1.

[0026] The accident prediction and determination device 10 may be composed of one or more general-purpose information processing devices such as a mainframe, workstation, or personal computer (PC), or it may be a dedicated device. Furthermore, the accident prediction and determination device 10 may be composed of a virtual information processing device operating on a hypervisor, an information processing device using container virtualization, or a cloud server.

[0027] Next, the operation of the accident prediction device 10 will be explained. Figure 3 is a flowchart showing an example of the accident prediction determination process performed by the accident prediction device 10. The processor 11 of the accident prediction device 10 executes the accident prediction determination process by reading the operation program stored in the memory device 12. The accident prediction determination process corresponds to the accident prediction determination method performed by the accident prediction device 10.

[0028] In the accident prediction determination process, first, the processor 11 determines whether or not it has received driving data from the in-vehicle equipment 21 of any vehicle 20 using the accident prediction determination system 1 (step S11). If driving data has not been received (step S11; No), the processor 11 terminates the accident prediction determination process. After terminating the accident prediction determination process, the processor 11 waits for the reception of driving data by repeatedly executing the accident prediction determination process at a predetermined execution cycle.

[0029] If driving data has been received (step S11; Yes), the processor 11 stores the received driving data in the driving data DB 102, associating it with the user identification information of the vehicle 20 (step S12). The user identification information can be determined from the identification information of the in-vehicle equipment 21 included in the communication data of the driving data.

[0030] Next, the processor 11 determines whether it is time for a driving evaluation based on the received driving data (step S13). The timing of the driving evaluation can be set arbitrarily. In this embodiment, the timing of the driving evaluation is set to a predetermined time (for example, 15 minutes) after the start of the current driving. In step S13, the processor 11 determines the start time of the driving from the received driving data and determines whether the predetermined time has elapsed since the start of the driving. In this way, by determining signs of an accident and issuing a warning based on the initial driving data from the start of the driving, accident prevention can be expected.

[0031] Furthermore, the timing of the operational evaluation is not limited to after a predetermined time has elapsed since the start of operation. For example, the operational evaluation may be performed at predetermined intervals (e.g., every hour) after the start of operation. In this case, in step S13, the processor 11 only needs to determine whether a predetermined time has elapsed since the start of operation or the timing of the previous operational evaluation.

[0032] Furthermore, when determining whether it is time for a driving evaluation, the distance traveled may be taken into consideration. For example, if the distance traveled in a predetermined time is less than a predetermined distance, it is assumed that there is insufficient driving data for evaluating the driving, so it may be determined that it is not time for a driving evaluation, and the driving evaluation and accident prediction may not be performed.

[0033] If it is determined that it is not time for driving evaluation (Step S13; No), the processor 11 terminates the accident prediction determination process. If it is determined that it is time for driving evaluation (Step S13; Yes), the processor 11 obtains driving data for the immediate vicinity period, including the driving data received this time, from the driving data DB 102, and generates evaluation data based on said driving data (Step S14). In this embodiment, the current driving is evaluated by generating evaluation data based on driving data from the start of driving to a predetermined time.

[0034] In step S14, the processor 11 generates evaluation data based on information such as acceleration, distance traveled, and travel time included in the driving data. The evaluation data includes numerical data for multiple items, such as the number of dangerous behaviors during a predetermined period (15 minutes from the start of driving), distance traveled, frequency of dangerous behaviors per unit distance (e.g., 1 kilometer), start time of driving, and their average ratio. Dangerous behaviors include sudden braking, sudden steering, and sudden acceleration. Dangerous behaviors can be measured and identified based on acceleration information and steering angle sensor information included in the driving data. The average ratio can be, for example, the ratio of the current evaluation data to the average over the past 30 days.

[0035] The evaluation data only needs to be based on data acquired from the vehicle 20 via the in-vehicle equipment 21 to assess driving safety, and may include evaluations of other driving data (maximum speed, average speed, etc.) and dangerous behaviors (number and frequency of approaching vehicles, driving on one side of the road, drowsiness, distracted driving, etc.). Approaching vehicles, driving on one side of the road, drowsiness, and distracted driving can be identified from the video data of the camera installed in the in-vehicle equipment 21.

[0036] The accident prediction device 10 may also perform driving evaluations based on driving data that includes driving data from the past week or other past driving periods. In this case, for example, in step S13, the processor 11 may determine that the timing for driving evaluation is a predetermined time after the start of driving indicated by the received driving data. Then, in step S14, the processor 11 may generate evaluation data based on driving data going back one week from that timing.

[0037] Following the processing in step S14, the processor 11 retrieves driving data for a specific past period (e.g., the past six months, one year) longer than a predetermined period from the driving data DB 102, and generates comparison data for comparison with the evaluation data based on this driving data (step S15). The comparison data is evaluation data based on the driving data for the specific past period, and can be numerical data for the same items as the evaluation data. In other words, the comparison data is the average value of the evaluation data for the specific past period. If driving data for the specific period is not stored in the driving data DB 102, the comparison data may be generated based on the longest possible period of driving data stored in the driving data DB 102.

[0038] Subsequently, the processor 11 performs a determination process to determine whether or not there are signs of an accident based on the generated evaluation data and comparison data (step S16).

[0039] In the judgment process, if at least one of the numerical values ​​of the items included in the evaluation data deviates by a predetermined threshold or more from the numerical value of the same item in the comparison data, it can be determined that there is a sign of an impending accident. For example, if the items in the evaluation data are the number of dangerous behaviors or the frequency of dangerous behaviors, and the numerical value of the evaluation data is higher than a predetermined threshold or more compared to the comparison data, that is, if the driving evaluation is worse than a predetermined value or more, it can be determined that there is a sign of an impending accident in the current driving.

[0040] The predetermined threshold should be a value determined based on statistical data or machine learning that indicates the presence of an accident precursor. The predetermined threshold may also be, for example, a predetermined percentage of the comparative data (e.g., 10%). Such a predetermined threshold should be defined for each item included in the evaluation data.

[0041] Furthermore, in the judgment process, if a predetermined number of items included in the evaluation data deviate from the comparison data by a predetermined threshold or more, it may be determined that there is a precursor to an accident. Alternatively, the determination of a precursor to an accident may be made based on the sum or average value of the degree of deviation of each item. In other words, any method may be used to determine whether there is a precursor to an accident by comparing the evaluation data with the comparison data, as long as it can suitably determine whether there is a precursor to an accident.

[0042] Figure 4 is a flowchart showing an example of such a judgment process algorithm. When the applicant verified the correlation between data from actual accidents and evaluation data, it was found that among several items of the evaluation data, the frequency of dangerous behavior and the distance traveled had a high correlation with accidents. In particular, the frequency of sudden braking, the distance traveled, and the frequency of sudden steering were found to have a high correlation with accidents in that order. In the judgment process shown in Figure 4, the signs of an accident are determined based on the frequency of sudden braking, the distance traveled, and the frequency of sudden steering, which have a high correlation with accidents.

[0043] In the determination process shown in Figure 4, the processor 11 first determines whether the difference obtained by subtracting the sudden braking frequency of the comparison data from the sudden braking frequency of the evaluation data is greater than or equal to threshold a (step S101). That is, it determines whether the sudden braking frequency of the evaluation data is greater than or equal to threshold a than the sudden braking frequency of the comparison data. Threshold a can be any positive value determined from the sudden braking frequency of the comparison data.

[0044] If the difference in the frequency of sudden braking is greater than or equal to threshold a (step S101; Yes), the processor 11 determines whether the difference obtained by subtracting the mileage of the evaluation data from the mileage of the comparison data is greater than or equal to threshold b (step S102). That is, it determines whether the mileage of the comparison data is greater than or equal to threshold b than the mileage of the evaluation data. This determination is based on the assumption that if the mileage traveled in the current predetermined time is less than the past average, there is a high possibility that smooth driving is not possible due to factors such as road shape (narrow road, mountain road, etc.) or the driver's condition (fatigue, lack of sleep, etc.), and that there is a sign of an impending accident. Threshold b can be, for example, any positive value determined from the mileage of the comparison data.

[0045] If the difference in mileage is greater than or equal to threshold b (step S102; Yes), the processor 11 determines whether the difference obtained by subtracting the sudden steering frequency of the comparison data from the sudden steering frequency of the evaluation data is greater than or equal to threshold c (step S103). That is, it determines whether the sudden steering frequency of the evaluation data is greater than or equal to threshold c than the sudden steering frequency of the comparison data. The threshold c can be any positive value determined from the sudden steering frequency of the comparison data.

[0046] If the difference in the frequency of sudden steering is greater than or equal to threshold c (step S103; Yes), the processor 11 determines that there is an indication of an accident in the driving being judged (step S104).

[0047] If the difference in the frequency of sudden braking is less than threshold a (step S101; No), if the difference in the distance traveled is less than threshold b (step S102; No), or if the difference in the frequency of sudden steering is less than threshold c (step S103; No), the processor 11 determines that there are no signs of an accident in the driving being judged (step S105).

[0048] As described above, the accident prediction device 10 of this embodiment determines that there is an impending accident if the frequency of sudden braking, driving distance, and sudden steering, which have a high correlation with accidents, deviate from the comparison data by a predetermined threshold or more. This allows for the appropriate determination of accident precursors. In addition, the device may determine the presence of an impending accident using only two items, the frequency of sudden braking and driving distance, which have a particularly high correlation with accidents. Furthermore, the items used to determine an impending accident are not limited to these, as the results of the correlation calculation will differ depending on the number of accident data points, the verification method, etc.

[0049] After executing step S104 or S105, the processor 11 terminates the determination process. Upon termination of the determination process, the system returns to Figure 3, and the processor 11 determines whether or not an indication of an accident was determined in the determination process (step S17).

[0050] If it is determined that there are no signs of an accident (Step S17; No), the processor 11 terminates the accident sign determination process. If it is determined that there are signs of an accident (Step S17; Yes), the processor 11 executes a warning process to warn the driver that there are signs of an accident (Step S18).

[0051] Figure 5 is a flowchart showing an example of warning processing. In warning processing, the processor 11 determines whether the current evaluation data is the lowest-rated in the most recent period (e.g., one month) (step S201). In step S201, for example, the processor 11 calculates the evaluation value of each evaluation data by referring to the history of evaluation data for the most recent period, and determines whether the current evaluation data is the lowest-rated in the most recent period. The evaluation value can be, for example, a numerical representation of the deviation of the evaluation data from the comparison data, and can be the sum of the deviations of each item included in the evaluation data from the comparison data, the average value, etc.

[0052] If the evaluation data for this time is the lowest possible (step S201; Yes), the processor 11 configures itself to output a warning that there are signs of an impending accident when the vehicle 20 is stopped next time (step S202). In step S202, for example, the processor 11 sends a command to the in-vehicle equipment 21 instructing it to issue a warning that there are signs of an impending accident when the vehicle is stopped next time. The warning should notify the driver that there are signs of an impending accident while driving, or prompt the driver to stop or interrupt (take a break) driving.

[0053] After executing the process in step S202, or if the evaluation data is not the lowest rating this time (step S201; No), the processor 11 terminates the warning process. Once the warning process is terminated, the processor 11 terminates the accident prediction judgment process.

[0054] In this way, the accident prediction device 10 compares evaluation data of driving in the immediate vicinity with evaluation data of a specific past period to determine whether or not there are signs of an accident in the current driving situation, and if there are signs of an accident, it issues a warning to the driver. This helps to suppress and prevent the occurrence of accidents.

[0055] Furthermore, in this embodiment, even if the accident prediction device 10 determines that there is an impending accident based on the evaluation data, it only issues a warning to the driver if the evaluation data is the lowest rating in a recent period. If the current evaluation data is not the lowest rating in a recent period, the driver will be warned during the driving period in which the lowest rating data was generated in a recent period. Therefore, the accident prediction device 10 prevents the driver from being frequently warned of impending accidents and becoming annoyed by not issuing a warning if the current evaluation data is not the lowest rating in a recent period.

[0056] Furthermore, in this embodiment, the accident prediction device 10 is configured to issue a warning to the vehicle 20 at the next time it stops if it determines that there are signs of an accident. Therefore, the accident prediction device 10 can effectively warn of signs of an accident while driving, without causing distress to the driver, and is expected to suppress and prevent accidents.

[0057] (modified version) Furthermore, this invention is not limited to the above embodiments, and various modifications and applications are possible. For example, some parts of the above embodiments can be omitted, replaced, or any configurations can be added.

[0058] In the accident prediction system 1 of the above embodiment, the accident prediction device 10 communicates with the in-vehicle equipment 21 to determine signs of an accident, and the in-vehicle equipment 21 issues a warning. However, the accident prediction device 10 may be mounted on the vehicle 20. Alternatively, the functions of the accident prediction device 10 may be provided by the in-vehicle equipment 21. That is, the acquisition unit 101, driving data DB 102, evaluation unit 103, prediction unit 104, and warning unit 105 shown in Figure 1 may be provided by the in-vehicle equipment 21.

[0059] In the judgment process shown in Figure 4, the evaluation data was compared with the comparison data in the order of sudden braking frequency, driving distance, and sudden steering frequency, but the order of comparison may be changed. That is, the order of steps S101 to S103 may be arbitrarily changed.

[0060] In the judgment process shown in Figure 4, an accident precursor is determined if the frequency of sudden braking, driving distance, and sudden steering included in the evaluation data deviates by more than a threshold from the comparison data. However, it is also possible to determine an accident precursor if two or fewer of these items deviate by more than a threshold from the comparison data.

[0061] For example, in the determination process shown in Figure 4, the process in step S103 may be omitted. That is, the evaluation data includes first evaluation data relating to the frequency of sudden braking and second evaluation data relating to the distance traveled, and the comparison data includes first comparison data relating to the frequency of sudden braking and second comparison data relating to the distance traveled. The predictive determination unit 104 may determine that there is a warning sign of an accident occurring if the first evaluation data deviates from the first comparison data by a first threshold or more, and compares the second evaluation data with the second comparison data, and if the second evaluation data deviates from the second comparison data by a second threshold or more. In this case, the frequency of sudden braking may be the number of sudden braking incidents over a predetermined period.

[0062] Furthermore, in the judgment process shown in Figure 4, the frequency of sudden braking and the frequency of sudden steering were compared with comparison data to determine signs of an accident. However, the frequency of sudden braking and the frequency of sudden steering may be combined as a dangerous behavior frequency, and signs of an accident may be determined by comparing it with the dangerous behavior frequency of the comparison data. That is, the evaluation data includes first evaluation data related to the dangerous behavior frequency (sudden braking, sudden steering) and second evaluation data related to the distance traveled, the comparison data includes first comparison data related to the dangerous behavior frequency and second comparison data related to the distance traveled, and the sign determination unit 104 may determine that there are signs of an accident occurring if the first evaluation data deviates from the first comparison data by a first threshold or more, and the second evaluation data deviates from the second comparison data by a second threshold or more. In this case, the dangerous behavior may also include sudden acceleration. In this case, the dangerous behavior frequency may be the number of dangerous behaviors over a predetermined period.

[0063] In the warning process shown in Figure 5, even if an accident precursor is determined based on the current evaluation data, a warning is issued to the driver only if the evaluation data is the lowest rating in a recent period. However, if an accident precursor is determined based on the current evaluation data, a warning may be issued always, regardless of other evaluation data in that period. Alternatively, it may be possible to configure whether or not to always issue a warning in such cases.

[0064] Furthermore, as a condition for reducing the frequency of accident warnings, a predetermined grace period may be set as the condition for issuing a warning if a warning has been issued since the previous warning. That is, if it is determined that there is an accident based on the current evaluation data, the warning process may be set to issue a warning if a grace period has elapsed since the previous warning.

[0065] Furthermore, in the warning process shown in Figure 5, if it is determined that there are signs of an impending accident, a warning will be issued when the vehicle 20 next stops. However, the warning may also be issued while the vehicle 20 is in motion. If a warning is issued while the vehicle is in motion, the nature of the warning may be less pronounced than when the vehicle is stopped. For example, a warning using only a lamp may be issued while the vehicle is in motion, while a warning using both a lamp and an audible alert may be issued when the vehicle is stopped. It may also be possible to set whether a warning is issued while the vehicle is in motion or when the vehicle is stopped.

[0066] The accident prediction and determination device 10 can be implemented using a regular computer, without requiring a dedicated device. For example, the accident prediction and determination device 10 that performs the above-mentioned processing may be configured by installing a program for performing any of the above-mentioned actions from a recording medium to the computer. Alternatively, multiple computers may work together to form a single accident prediction and determination device 10.

[0067] Furthermore, the method for supplying programs to computers is arbitrary. For example, they may be supplied via communication lines, communication networks, communication systems, etc.

[0068] Furthermore, if the OS (Operating System) provides some of the above-mentioned functions, then the parts not provided by the OS should be provided by the program.

[0069] The embodiments described above are provided to facilitate understanding of the present invention and are not intended to limit its interpretation. The flowcharts, sequences, elements, and their arrangement, materials, conditions, shapes, and sizes described in the embodiments are not limited to those exemplified and can be modified as appropriate. Furthermore, configurations shown in different embodiments can be partially substituted or combined. [Explanation of symbols]

[0070] 1... Accident prediction system, 10... Accident prediction device, 11... Processor, 12... Memory device, 13... Communication interface, 14... Input device, 15... Output device, 20... Vehicle, 21... In-vehicle equipment, 101... Acquisition unit, 102... Driving data DB, 103... Evaluation unit, 104... Prediction unit, 105... Warning unit

Claims

1. An acquisition unit that acquires driving data measured by sensors installed in the vehicle, An evaluation unit that evaluates the driving based on the driving data acquired by the acquisition unit, A precursor determination unit compares evaluation data, which is evaluated by the evaluation unit based on the aforementioned driving data over a fixed period of time, with predetermined comparison data, and determines that there is a precursor to an accident if the evaluation data deviates from the comparison data by a predetermined threshold or more. If the aforementioned warning unit determines that there are signs of an impending accident, the warning unit provides a warning to the driver of the vehicle, and the system is equipped with The comparison data is evaluation data that the evaluation unit has evaluated based on the driving data for a specific period in the past that is longer than the predetermined period. The evaluation data includes a first evaluation data relating to the frequency of dangerous behavior and a second evaluation data relating to the distance traveled. The aforementioned comparative data includes first comparative data relating to the frequency of dangerous behavior and second comparative data relating to the distance traveled. The predictive indicator unit determines that there is a sign of an impending accident if the first evaluation data deviates from the first comparison data by a first threshold or more, and the second evaluation data deviates from the second comparison data by a second threshold or more. Accident prediction device.

2. An acquisition unit that acquires driving data measured by sensors installed in the vehicle, An evaluation unit that evaluates the driving based on the driving data acquired by the acquisition unit, A precursor determination unit compares evaluation data, which is evaluated by the evaluation unit based on the aforementioned driving data over a fixed period of time, with predetermined comparison data, and determines that there is a precursor to an accident if the evaluation data deviates from the comparison data by a predetermined threshold or more. If the aforementioned warning unit determines that there are signs of an impending accident, the warning unit provides a warning to the driver of the vehicle, and the system is equipped with The comparison data is evaluation data that the evaluation unit has evaluated based on the driving data for a specific period in the past that is longer than the predetermined period. The evaluation data includes first evaluation data relating to the frequency of sudden braking and second evaluation data relating to the distance traveled. The aforementioned comparative data includes first comparative data relating to the frequency of sudden braking and second comparative data relating to the distance traveled. The predictive indicator unit, if the first evaluation data deviates from the first comparison data by a first threshold or more, compares the second evaluation data with the second comparison data, and if the second evaluation data deviates from the second comparison data by a second threshold or more, determines that there is a sign of an impending accident. Accident prediction device.

3. An acquisition unit that acquires driving data measured by sensors installed in the vehicle, An evaluation unit that evaluates the driving based on the driving data acquired by the acquisition unit, A precursor determination unit compares evaluation data, which is evaluated by the evaluation unit based on the aforementioned driving data over a fixed period of time, with predetermined comparison data, and determines that there is a precursor to an accident if the evaluation data deviates from the comparison data by a predetermined threshold or more. If the aforementioned warning unit determines that there are signs of an impending accident, the warning unit provides a warning to the driver of the vehicle, and the system is equipped with The comparison data is evaluation data that the evaluation unit has evaluated based on the driving data for a specific period in the past that is longer than the predetermined period. The warning unit issues the warning if the discrepancy between the evaluation data and the comparison data is greatest during a certain period including the predetermined period. Accident prediction device.

4. The start date of the aforementioned predetermined period is the start of this operation. An accident prediction device according to any one of claims 1 to 3.

5. The warning unit issues the warning when the vehicle is stopped. An accident prediction device according to any one of claims 1 to 3.

6. The accident prediction device, We acquire driving data measured by sensors installed in the vehicle. Based on the acquired driving data, the driving is evaluated. The evaluation data, based on the aforementioned driving data over a fixed period, is compared with predetermined comparison data. If the evaluation data deviates from the comparison data by a predetermined threshold or more, it is determined that there is a sign of an impending accident. If it is determined that there are signs of an impending accident, the driver of the vehicle will be warned. The comparison data is evaluation data evaluated based on the driving data for a specific period in the past that is longer than the predetermined period. The evaluation data includes a first evaluation data relating to the frequency of dangerous behavior and a second evaluation data relating to the distance traveled. The aforementioned comparative data includes first comparative data relating to the frequency of dangerous behavior and second comparative data relating to the distance traveled. The accident prediction device determines that there is an indication of an impending accident if the first evaluation data deviates from the first comparison data by a first threshold or more, and the second evaluation data deviates from the second comparison data by a second threshold or more. Accident prediction method.

7. Computers, An acquisition unit that acquires driving data measured by sensors installed in the vehicle. An evaluation unit evaluates the driving based on the driving data acquired by the acquisition unit. A precursor determination unit compares evaluation data, which is evaluated by the evaluation unit based on the aforementioned driving data over a fixed period of time, with predetermined comparison data, and determines that there is a precursor to an accident if the evaluation data deviates from the comparison data by a predetermined threshold or more. If the aforementioned precursor detection unit determines that there are signs of an impending accident, it functions as a warning unit that warns the driver of the vehicle. The comparison data is evaluation data that the evaluation unit has evaluated based on the driving data for a specific period in the past that is longer than the predetermined period. The evaluation data includes a first evaluation data relating to the frequency of dangerous behavior and a second evaluation data relating to the distance traveled. The aforementioned comparative data includes first comparative data relating to the frequency of dangerous behavior and second comparative data relating to the distance traveled. The predictive indicator unit determines that there is a sign of an impending accident if the first evaluation data deviates from the first comparison data by a first threshold or more, and the second evaluation data deviates from the second comparison data by a second threshold or more. program.

Images