JP2017102795A - Traveling capability evaluation device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a traveling capability evaluation device capable of accurately evaluating traveling capability.SOLUTION: A traveling capability evaluation device 1 for evaluating a traveling capability of a vehicle comprises: an own vehicle scene recognition part 41 for recognizing a driving scene representing a traveling state including at least one of a vehicle speed and a traveling direction as for an own vehicle as a vehicle loaded with the traveling capability evaluation device 1; a preceding vehicle scene recognition part 42 for recognizing a driving scene of a preceding vehicle; a coincidence degree calculation part 43 for calculating a coincidence degree between the driving scene of the own vehicle recognized by the own vehicle scene recognition part 41 and a driving scene of the preceding vehicle recognized by the preceding vehicle scene recognition part 42; and a traveling capability evaluation part 44 for successively evaluating a traveling capability of the own vehicle on the basis of a coincidence degree M calculated by the coincidence degree calculation part 43.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a traveling ability evaluation device that evaluates the ability of a vehicle to travel.

  There is known a technique for setting control related to the running behavior of a vehicle based on driving operation characteristics of a driver. For example, in Patent Document 1, the alarm output timing is set based on the driving operation characteristic unique to the driver.

  There is also known a technique for setting control related to the running behavior of a vehicle based on vehicle characteristics. For example, in Patent Document 2, the notification timing is set based on the collision avoidance performance of the vehicle. Further, a driving scene recognition device that recognizes various driving scenes of a vehicle is known (for example, Patent Document 3).

JP 2007-249539 A JP 2012-168811 A JP 2013-250663 A

  In Patent Literature 1, the alarm output timing is set based on the driving characteristics of the driver. However, even if the driver has the same driving operation characteristics, the behavior of the vehicle is not the same if the vehicle characteristics are different.

  In Patent Document 2, the notification timing is set based on the collision avoidance performance of the vehicle, but even if the collision avoidance performance of the vehicle is the same, if the driving ability of the driver is different, the behavior of the vehicle is not the same. Don't be.

  In order to estimate the behavior of the vehicle, it is necessary to evaluate the driving characteristics of the driver together with the characteristics of the vehicle related to the behavior. In this specification, the driving characteristics of the driver and the characteristics of the vehicle related to the behavior are combined to determine the driving ability.

  The technique of Patent Document 1 does not evaluate the characteristics of the vehicle, and the technique of Patent Document 2 does not evaluate the characteristics of the driver. Therefore, none of the techniques of Patent Documents 1 and 2 have been able to accurately evaluate the running ability.

  The present invention has been made based on this situation, and an object of the present invention is to provide a travel capability evaluation apparatus that can accurately evaluate the travel capability.

  The above object is achieved by a combination of the features described in the independent claims, and the subclaims define further advantageous embodiments of the invention. Reference numerals in parentheses described in the claims indicate a correspondence relationship with specific means described in the embodiments described later as one aspect, and do not limit the technical scope of the present invention. .

  The present invention for achieving the above object is a travel capability evaluation device for evaluating the travel capability of a vehicle, and includes at least one of a vehicle speed and a traveling direction of the own vehicle, which is a vehicle equipped with the travel capability evaluation device. A host vehicle scene recognition unit (41) for recognizing a driving scene representing a traveling state, and a comparison vehicle scene recognition unit for recognizing a driving scene of a comparison vehicle that is one of a preceding vehicle and a nearest succeeding vehicle for the host vehicle ( 42), a coincidence degree calculation unit (43) for calculating a coincidence between the driving scene of the own vehicle recognized by the own vehicle scene recognition unit and the driving scene of the comparison vehicle recognized by the comparison vehicle scene recognition unit, A running ability evaluation unit (44) that sequentially evaluates the running ability of the target vehicle, which is one of the host vehicle and the comparison vehicle, based on the degree of coincidence calculated by the calculation unit.

  In the present invention, the degree of coincidence between the driving scene of the host vehicle and the driving scene of the comparative vehicle is calculated. The driving scene represents a traveling state including at least one of the vehicle speed and the traveling direction. The method of acceleration and deceleration varies depending on the driver, and the steering wheel operation varies depending on the driver. Therefore, the driving characteristics of the driver are reflected in the driving scene.

  The change in the vehicle speed changes depending on the acceleration / deceleration performance of the vehicle, and the change in the traveling direction of the vehicle changes depending on the steering performance of the vehicle. Therefore, the driving scene is affected by both the driving characteristics of the driver and the vehicle characteristics related to the behavior.

  When this driving scene is compared between the own vehicle and a comparison vehicle that is one of the preceding vehicle and the nearest succeeding vehicle for the own vehicle, the driving ability of the driver behind the own vehicle and the comparison vehicle is high. The driving scenes tend to coincide with each other than when the driving ability of the driver on the rear side is low. In addition, the higher the ability related to the behavior of the rear vehicle of the host vehicle and the comparative vehicle, the higher the possibility that the rear vehicle can follow the driving scene at an early stage. Therefore, the degree of coincidence between driving scenes reflects the high driving ability of the driver and the high ability of the vehicle related to the behavior.

  Since the running ability evaluation unit evaluates the running ability of the target vehicle based on the degree of coincidence of the driving scenes, the running ability can be accurately evaluated.

2 is a diagram showing a state in which a host vehicle 2 equipped with a travel capability evaluation device 1 is traveling as a subsequent vehicle of a preceding vehicle 3. FIG. It is a block diagram which shows the structure of the driving capability evaluation apparatus 1 of FIG. It is a figure which shows the detailed structure of the own vehicle scene recognition part 41 of FIG. It is a figure which illustrates the relationship of the own vehicle behavior value, the own vehicle differential value, a driving symbol, and a driving word. It is a figure which shows the detailed structure of the preceding vehicle scene recognition part 42 of FIG. It is a figure which shows the detailed structure of the coincidence calculation part 43 of FIG. It is a figure which illustrates the driving symbol coincidence table used in the content comparison part 434 of FIG. It is a figure explaining the change of a driving scene and a response | compatibility of a driving scene. It is a figure which displays the coefficient f ((DELTA) t) calculated in the calculation part 435 of FIG. 6 on a graph. It is a figure which shows the detailed structure of the driving capability evaluation part 44 of FIG. It is a figure which illustrates the driving capability increase / decrease table used by the driving capability update part 442 of FIG. It is a figure which shows the relationship used in order to determine whether the driving assistance part 46 of FIG. 2 performs driving assistance. It is a figure which shows the driving scene and conditions set in 2nd Embodiment. It is a coincidence degree table | surface used in order to determine the coincidence degree s in the content comparison part 434 in 2nd Embodiment. It is a table | surface used in order to determine a driving | running level in the driving assistance part 46 in 2nd Embodiment. It is a coincidence degree table | surface used in order to determine the coincidence degree s in the content comparison part 434 in 3rd Embodiment.

<First Embodiment>
Hereinafter, embodiments of the present invention will be described with reference to the drawings. As shown in FIG. 1, the traveling ability evaluation device 1 according to the first embodiment is mounted on the host vehicle 2 and includes a front sensor 10, a host vehicle behavior sensor 20, a communication unit 30, and a control unit 40. In addition, the own vehicle 2 means the vehicle in which this driving capability evaluation apparatus 1 is mounted. In the state shown in FIG. 1, the host vehicle 2 is traveling immediately after the preceding vehicle 3 that is a comparative vehicle in the present embodiment.

  In the present embodiment, the preceding vehicle 3 is equipped with a traveling ability evaluation device 100. The driving ability evaluation device 100 includes a front sensor 110, a vehicle behavior sensor 120, a communication unit 130, and a control unit 140. The components included in the driving capability evaluation device 100 and the driving capability evaluation device 100 have different reference numerals for convenience of explanation, but the components having the same names in the components included in the driving capability evaluation device 1 and the driving capability evaluation device 1 Is the same.

[Configuration of Traveling Capacity Evaluation Device 1]
With reference to the block diagram shown in FIG. 2, the configuration of the traveling ability evaluation apparatus 1 will be described. The front sensor 10 is a sensor for detecting a value related to the relative behavior of the preceding vehicle 3 traveling immediately in front of the host vehicle 2 with respect to the host vehicle 2. Since the value related to the relative behavior of the preceding vehicle 3 with respect to the host vehicle 2 corresponds to the comparative vehicle behavior value in the claims, the front sensor 10 corresponds to the comparative vehicle behavior sensor in the claims.

  Specifically, the information detected by the front sensor 10 is the relative position of the preceding vehicle 3 with respect to the host vehicle 2. Since the relative position of the preceding vehicle 3 with respect to the own vehicle 2 is detected, the front sensor 10 also detects the inter-vehicle distance between the own vehicle 2 and the preceding vehicle 3. As the front sensor 10, for example, a radar using a millimeter wave or a laser as a measurement medium can be used. A camera may be used as the front sensor 10.

  The host vehicle behavior sensor 20 is a sensor that detects a host vehicle behavior value that is a value related to the behavior of the host vehicle 2. The behavior of the host vehicle 2 can be defined by the vehicle speed and the relative traveling direction with respect to the traveling locus up to now. Further, the relative traveling direction changes depending on the steering operation. Accordingly, in the present embodiment, the host vehicle behavior sensor 20 includes a vehicle speed sensor and a steering angle sensor, and the vehicle speed and the steering angle are sequentially detected as the host vehicle behavior value. Further, the host vehicle behavior sensor 20 includes an accelerator opening sensor and a brake hydraulic pressure sensor, and the accelerator opening and the brake master cylinder pressure are sequentially detected as the host vehicle behavior value. The own vehicle behavior sensor 20 corresponds to the own vehicle behavior detection unit in the claims.

  The communication unit 30 has a function of performing vehicle-to-vehicle communication by short-range wireless communication. When the communication device is mounted on another vehicle such as the preceding vehicle 3, the communication unit 30 performs communication with the communication device. The frequency used by the communication unit 30 is, for example, a 5.8 GHz band or a 700 MHz band. The communication unit 30 receives a preceding vehicle behavior value described later. The preceding vehicle behavior value corresponds to the comparative vehicle behavior value in the claims. Therefore, the communication unit 30 corresponds to a receiving unit in the claims. In addition, the communication unit 30 sequentially transmits the own vehicle behavior value to the surroundings. Further, in addition to the own vehicle behavior value, a later-described own vehicle differential value may be sequentially transmitted to the surroundings.

  The control unit 40 is a computer that includes a CPU, a ROM, a RAM, and the like. The CPU executes a program stored in a non-transitory tangible storage medium such as a ROM while using the temporary storage function of the RAM. By executing this program, the control unit 40 causes the host vehicle scene recognition unit 41, the preceding vehicle scene recognition unit 42, the coincidence degree calculation unit 43, the driving ability evaluation unit 44, the collision predicted value calculation unit 45, and the driving support unit 46. Function as. When these functions are executed, a method corresponding to the program is executed. Note that part or all of the functions executed by the control unit 40 may be configured by hardware using one or a plurality of ICs.

  The own vehicle scene recognition unit 41 recognizes the driving scene of the own vehicle 2. The driving scene represents a traveling state of the vehicle including at least one of the vehicle speed and the traveling direction. In addition, the advancing direction here is a relative advancing direction mentioned above.

  The own vehicle scene recognition unit 41 sequentially acquires signals detected by the own vehicle behavior sensor 20 in order to recognize this driving scene. The driving scene in the present embodiment is represented by a driving word composed of one or a plurality of driving symbols.

  As shown in FIG. 3, the host vehicle scene recognition unit 41 includes a host vehicle differential value calculation unit 411, a symbolization unit 412, and a symbol integration unit 413 as a detailed configuration. The own vehicle differential value calculation unit 411 sequentially calculates a differential value (hereinafter referred to as an own vehicle differential value) of a preset type of own vehicle behavior value among the own vehicle behavior values detected by the own vehicle behavior sensor 20. For example, the own vehicle differential value calculation unit 411 calculates the differential value of the brake master cylinder pressure and the differential value of the steering angle.

  As shown in FIG. 4, the symbolizing unit 412 divides the traveling state of the host vehicle 2 into segmented travels based on the host vehicle behavior values and the host vehicle differential values continuously detected or calculated over time. Divide into states. Then, the segmented traveling state is represented by a corresponding driving symbol. As a result, a symbol string of driving symbols is generated. This symbol string corresponds to the host vehicle symbol string in the claims, and the symbolizing unit 412 corresponds to the host vehicle symbolizing unit in the claims.

  In the classification into the traveling state, the various traveling states grasped from the respective values are clustered in a multidimensional space composed of the own vehicle behavior value and the own vehicle differential value. Then, by statistically processing which cluster each value belongs to, each value is classified into each segment running state (that is, each cluster) as a segment unit.

  For example, the encoding unit 412 uses a hierarchical Dirichlet process hidden Markov model (hereinafter referred to as HDP-HMM), which is one of models expressed by a hidden state and a stochastic transition between the states, as in Patent Document 3. To generate a symbol string. The HDP-HMM determines the number of hidden states according to each input value by assuming an infinite dimensional hidden state in the HMM, and it is not necessary to design the number of hidden states in advance.

  In particular, it is preferable to use sticky HDP-HMM as the HDP-HMM. The sticky HDP-HMM is obtained by adding a bias to the self-transition probability of the HDP-HMM. By increasing the self-transition probability, the sticky HDP-HMM can suppress the excessive transition of the hidden state, and can efficiently perform modeling assuming the continuity of the operation.

  The symbolization unit 412 is not limited to the HDP-HMM described above, and can generate a symbol string using another model. For example, a general HMM, an N-shaped Markov model, a hierarchical Markov model, a switching AR model, a switching Marman filter, or the like can be used. In this case, for example, the number of hidden states of each model is designed in advance, and the transition probabilities between the hidden states are stored in the symbol transition database. Then, the most likely hidden state can be obtained by calculating the posterior probability of the hidden state based on the model and the transition probability of the hidden state. When HDP-HMM is used, the transition probability between hidden states is automatically calculated and updated, so there is no need to determine the hidden state transition probability in advance.

  Furthermore, the following method may be employed as a simpler method for generating a symbol string without using the above-described model. That is, for each value that is continuously detected or calculated over time, a threshold for classifying the magnitude of each value is set. Further, different driving symbols are associated with all combinations of the size categories. Then, a symbol string is generated by determining which category each detected or calculated value belongs to, and repeating a process of assigning a driving symbol corresponding to the combination of the categories.

  The symbol integration unit 413 corresponds to the own vehicle segmentation unit in the claims. The symbol integration unit 413 integrates the symbol string output from the symbolization unit 412 into a driving word that represents a predetermined traveling state. As a result, the symbol string is segmented into driving words as shown in FIG. Then, the symbol integration unit 413 recognizes the driving word as a driving scene. In this embodiment, Nested Pitman-Yor Language Model (NPYLM), which is an example of an unsupervised chunking technique for discrete character strings using statistical information, is used as the symbol integration unit 413. A plurality of types of driving words are generated by using the statistical information, and a plurality of types of driving states are represented by the plurality of types of driving words. In addition, since the driving word is determined using the statistical information, the driving word, that is, the driving scene, is determined from a plurality of types of driving words (that is, a plurality of types of running states).

  The NPYLM is expanded by incorporating an N-gram model of words into the Hierarchical Pitman-Yor Language Model (HPYLM), and realizes morphological analysis without dictionary data. HPYLM improves the robustness against unknown words and low-frequency words by smoothing the N-gram language model by the Pitman-Yor process, and enables appropriate word segmentation.

  NPYLM and HPYLM are described in detail in “Analysis of Behavior Using Hierarchical Pitman-Yor Language Model” (The 25th Annual Conference of the Japanese Society for Artificial Intelligence, 3B1-OS22c-8. (2011)). Description is omitted.

  When the symbol integration unit 413 segments the symbol string into driving words using the NPYLM described above, the generation probability of a sentence composed of the driving words is maximized, and the size of the dictionary (that is, the number of driving words) is increased. To be as small as possible. This makes it possible to segment the symbol string while reducing the amount of calculation. The transition probability between the living words and the generation probability of the driving word are learned during training and stored in the word transition database 14.

  Note that the above-described NPYLM is merely an example for segmentation of a symbol string, and segmentation may be performed by other methods. For example, driving words are set in advance based on symbol strings generated when the vehicle is driven in various driving conditions, a driving word dictionary is created, and a database of transition probabilities and generation probabilities of each driving word is created. To do. These driving word dictionaries and databases are stored in a predetermined storage unit. Then, referring to these driving word dictionaries, transition probability database, and generation probability database, the input symbol string may be segmented for each most probable driving word.

  The preceding vehicle scene recognition unit 42 recognizes the driving scene of the preceding vehicle 3. The meaning of the driving scene is the same as the driving scene in the vehicle scene recognition unit 41. In the present embodiment, the preceding vehicle 3 corresponds to the comparative vehicle in the claims, and therefore the preceding vehicle scene recognition unit 42 corresponds to the comparative vehicle scene recognition unit in the claims. As shown in FIG. 5, the preceding vehicle scene recognizing unit 42 includes a preceding vehicle differential value calculating unit 421, a symbolizing unit 422, and a symbol combining unit 423 as a detailed configuration.

  The preceding vehicle differential value calculation unit 421 is a preceding vehicle behavior value that is a value related to the behavior of the preceding vehicle 3, and uses the same type of value as that acquired by the own vehicle differential value calculation unit 411. Acquired using one or both of the front sensors 10. And the differential value (henceforth, preceding vehicle differential value) of the acquired preceding vehicle behavior value is calculated sequentially. Since the preceding vehicle 3 corresponds to a comparative vehicle, the peripheral vehicle differential value corresponds to the comparative vehicle differential value in the claims, and the preceding vehicle differential value calculation unit 421 corresponds to the comparative vehicle differential value calculation unit in the claims.

  The symbolization unit 422 corresponding to the comparison vehicle symbol string generation unit in the claims sequentially acquires the same type of values as the symbolization unit 412 included in the host vehicle scene recognition unit 41 as the preceding vehicle behavior value. Based on the acquired preceding vehicle behavior value, the symbolizing unit 422 divides the traveling state of the preceding vehicle 3 into segmented traveling states, and represents the segmented segmented traveling states by corresponding driving symbols. The segmented traveling state is the same as that described in the symbolizing unit 412 of the own vehicle scene recognizing unit 41, and the classification method for the segmented traveling state is also the same method as the symbolizing unit 412 of the own vehicle scene recognizing unit 41. . Since the symbolizing unit 422 sequentially outputs driving symbols, the symbolizing unit 422 outputs a symbol string of driving symbols. This symbol string corresponds to the comparative vehicle symbol string in the claims.

  The symbol combination unit 423 corresponding to the comparative vehicle segmentation unit in the claims integrates the symbol string output from the symbolization unit 422 into a driving word meaning a predetermined traveling state. The integration method is the same as the symbol combination unit 423 provided in the host vehicle scene recognition unit 41.

  The coincidence calculation unit 43 calculates the coincidence M between the driving scene of the host vehicle 2 recognized by the host vehicle scene recognition unit 41 and the driving scene of the preceding vehicle 3 recognized by the preceding vehicle scene recognition unit 42. The degree-of-match calculation unit 43 of the present embodiment has a detailed configuration shown in FIG.

  Specifically, the coincidence degree calculation unit 43 of the present embodiment includes a driving scene storage unit 431, a switching detection unit 432, a time difference calculation unit 433, a content comparison unit 434, and a calculation unit 435.

  The driving scene storage unit 431 is a writable storage medium, and the driving scene of the host vehicle 2 and the driving scene of the preceding vehicle 3 are sequentially stored. When the switching detection unit 432 detects that the driving scene of the host vehicle 2 has been switched, the driving scene of the host vehicle 2 before switching is stored in the driving scene storage unit 431 by the switching detection unit 432. Since the driving scene before switching is determined when the driving scene is switched, the driving scene before switching is stored when the driving scene is switched. When the switching detection unit 432 detects that the driving scene of the preceding vehicle 3 has been switched, the driving scene of the preceding vehicle 3 before the switching is stored in the driving scene storage unit 431 by the switching detection unit 432.

  On the other hand, when the calculation unit 435 calculates the degree of coincidence M, the driving scene of the host vehicle 2 and the driving scene of the preceding vehicle 3 are deleted from the driving scene storage unit 431 by the calculating unit 435. As a result, only the driving scene for which the matching degree M has not been calculated remains in the driving scene storage unit 431, and therefore the matching degree M is calculated for driving scenes for which the matching degree M has not been calculated.

  The switch detection unit 432 sequentially acquires the driving scene of the host vehicle 2 recognized by the host vehicle scene recognition unit 41. When a driving scene is newly acquired from the own vehicle scene recognition unit 41, it is assumed that the driving scene of the own vehicle 2 is switched.

  Further, when the switching detection unit 432 sequentially acquires the driving scene of the preceding vehicle 3 recognized by the preceding vehicle scene recognition unit 42 and newly acquires the driving scene from the preceding vehicle scene recognition unit 42, the switching scene of the preceding vehicle 3 is acquired. Is switched.

  When the switching detection unit 432 detects that the driving scene has been switched, the switching detection unit 432 stores the driving scene newly acquired from the host vehicle scene recognition unit 41 or the preceding vehicle scene recognition unit 42 in the driving scene storage unit 431. The driving scene in the present embodiment is specifically a driving word, but in addition to the driving word, the driving symbols constituting the driving word are also stored in the driving scene storage unit 431. Furthermore, when the switching detection unit 432 detects that the driving scene has been switched, the switching detection unit 432 outputs this to the time difference calculation unit 433.

  When the switching detection unit 432 detects switching of the driving scene and the other driving scene is stored in the driving scene storage unit 431, the time difference calculating unit 433 is configured to detect the time difference Δt between the times when the two driving scenes are switched. Is calculated.

  In the present embodiment, the time difference Δt is set to 0 when the driving scene of the host vehicle 2 that is the rear vehicle is switched before the driving scene of the preceding vehicle 3 that is the front vehicle. By doing in this way, it can suppress that driving ability will be evaluated low although the driver of the own vehicle 2 which is a back side vehicle predicts the behavior of the preceding vehicle 3, and carried out advance driving. An example in which the driving scene of the host vehicle 2 is switched before the driving scene of the preceding vehicle 3 will be described later with reference to FIG.

  The content comparison unit 434 compares the contents of the two driving scenes for which the time difference calculation unit 433 has calculated the time difference Δt, and determines the degree of coincidence s between the two driving scenes. The degree of coincidence s of contents is determined using a coincidence degree table of contents set in advance. More specifically, in the present embodiment, a driving word is recognized as a driving scene, and the driving word is configured by one or a plurality of driving symbols. Therefore, the content comparison unit 434 of the present embodiment creates in advance a matching degree table (hereinafter referred to as a driving symbol matching degree table) between the driving symbols constituting the driving word. FIG. 7 is an example of the driving symbol coincidence table. Using this driving symbol coincidence table, the coincidence s of the contents of the two driving scenes is calculated.

  For the driving symbols constituting the two driving words to be compared, the matching score s is calculated using the driving symbol matching table in order from the top. The number of driving symbols constituting the driving word differs depending on the content thereof. Therefore, a column without symbols is provided in the driving symbol coincidence table. “A” described in the column without a symbol is a preset constant.

  For each driving symbol, the degree of coincidence is calculated using the driving symbol coincidence table, and the value obtained by multiplying all the calculated values is defined as the degree of coincidence s of the contents of the two driving scenes. Alternatively, after the degree of coincidence is calculated for each driving symbol, the average value of all the calculated degrees of coincidence may be used as the degree of coincidence s of the contents of the two driving scenes.

  The calculation unit 435 calculates the degree of coincidence M between the driving scene in which the switching is detected and the corresponding driving scene each time the switching detection unit 432 detects the switching of the driving scene. The corresponding driving scene here is the driving scene of the preceding vehicle 3 if the driving scene detected by the switching detection unit 432 is the driving scene of the host vehicle 2. As these driving scenes, those stored in the driving scene storage unit 431 are used.

  However, when one of the driving scenes of the host vehicle 2 and the driving scene of the preceding vehicle 3 is stored in the driving scene storage unit 431, the following is performed. For the two or more driving scenes stored in the driving scene storage unit 431, only the driving scene stored at the oldest time is calculated as the degree of coincidence M with the other driving scene. Of the two or more driving scenes stored in the driving scene storage unit 431, the driving scene coincidence M is calculated assuming that there is no corresponding driving scene other than the driving scene stored at the oldest time. .

  The situation where there is no corresponding driving scene is a situation where the other driving scene is switched a plurality of times before one driving scene is switched to the next driving scene. For example, as shown in FIG. 8, when the driving scene of the host vehicle 2 is switched from the scene C to the scene D and further to the scene E while the driving scene of the preceding vehicle 3 is the scene A, the driving corresponding to the scene A is performed. The scene becomes scene C. For the scene D, the matching degree M is calculated assuming that there is no corresponding driving scene. After calculating the coincidence degree M, the calculation unit 435 deletes the scene A, the scene C, and the scene D from the driving scene storage unit 431. Thereby, next time, the scene E and the scene B become driving scenes corresponding to each other. When the time when the scene E starts and the time when the scene B starts are compared, the scene E which is the driving scene of the own vehicle 2 which is the succeeding vehicle starts first. Therefore, when calculating the degree of coincidence M between the scene E and the scene B, the time difference Δt is set to zero.

When there is a corresponding driving scene, the degree of coincidence M is calculated by setting the degree of coincidence s determined by the content comparison unit 434 to a coefficient f (Δt) determined based on the time difference Δt calculated by the time difference calculation unit 433. The value multiplied. Therefore, the degree of coincidence M can be calculated from Equation 1.
(Formula 1) M = f (Δt) × s
As shown in FIG. 9, the coefficient f (Δt) is a value smaller than 1 regardless of the value of the time difference Δt, and the coefficient f decreases as the time difference Δt increases. Since the coefficient f (Δt) with respect to the time difference Δt has such a tendency, the degree of coincidence M decreases as the time difference Δt increases.

  On the other hand, when there is no corresponding driving scene as in the scene D of FIG. 8, the matching degree M is set to a predetermined value θ that can be determined without comparing the driving scenes. This predetermined value θ may be, for example, a fixed value set in advance, or may be the latest value of the calculated matching degrees M. When there is no corresponding driving scene, the coincidence degree M is set to the predetermined value θ, so that the influence of the matching degree M determined when there is no corresponding driving scene on the driving ability can be reduced.

  The coincidence degree M calculated by the coincidence degree calculation unit 43 is input to the driving ability evaluation unit 44 shown in FIG. In addition to the degree of coincidence M, the travel capability evaluation unit 44 also acquires the inter-vehicle distance detected by the front sensor 10. Then, from the coincidence degree M and the inter-vehicle distance, the traveling ability of the host vehicle 2 that is the target vehicle in the present embodiment is sequentially evaluated. In the following description, the traveling ability is expressed by a numerical value, and it is assumed that the higher the numerical value is, the higher the traveling ability is. In addition, there is an upper limit and a lower limit in the value representing the driving ability, and a value exceeding the range is not taken.

  As shown in FIG. 10, the travel capability evaluation unit 44 includes a travel capability storage unit 441 and a travel capability update unit 442 as detailed configurations. The travel capability storage unit 441 stores the latest travel capability that is sequentially updated by the travel capability update unit 442. The travel capability update unit 442 acquires the coincidence degree M and the inter-vehicle distance, and determines an increase / decrease value of the travel capability from the coincidence degree M, the inter-vehicle distance, and the travel capability increase / decrease table shown in FIG.

  The running ability increase / decrease table shown in FIG. 11 is set in advance. As shown in FIG. 11, the increase / decrease value of the travel capability determined from this travel capability increase / decrease table is as long as the inter-vehicle distance is shorter as the inter-vehicle distance is shorter than the threshold distance THd. The absolute value of the increase / decrease value of is large. The threshold distance THd is a distance at which the following vehicle does not behave so much as following the preceding vehicle, and is determined in advance based on experiments.

  When the inter-vehicle distance is short, the maximum value of the increase / decrease value of the driving ability is 0, and the lower the coincidence M, the smaller the increase / decrease value of the driving ability, that is, a negative value with a larger absolute value. It has become.

  The reason why the increase / decrease value of the driving ability is made smaller as the matching degree M is lower is as follows. The degree of coincidence M is a value that represents the degree to which the driving scene of the preceding vehicle 3 and the driving scene of the host vehicle 2 that is the following vehicle coincide. Therefore, the lower the coincidence degree M, the more the own vehicle 2 that is the succeeding vehicle cannot travel in accordance with the preceding vehicle 3. Thus, the lower the coincidence M, the smaller the increase / decrease value of the driving ability. In addition, the reason why the maximum value of the increase / decrease value of the driving ability when the inter-vehicle distance is short is 0, even if the degree of coincidence M is high, if the inter-vehicle distance is too short, it is an element that affects the driving ability. This is because the driving characteristics of the driver cannot be highly evaluated.

  Looking at the same degree of coincidence M, if the increase / decrease value of the smallest driving ability is negative, the increase / decrease value of the driving ability becomes closer to 0 as the inter-vehicle distance increases, and if the inter-vehicle distance is the longest, Increase / decrease value of ability is 0. The reason for this is that the longer the inter-vehicle distance, the lower the need to travel following the preceding vehicle 3, so even if the driving ability is high, the driving scene does not match if the inter-vehicle distance is long. The possibility increases. Therefore, when the inter-vehicle distance is long, the traveling ability is not updated so much.

  When the inter-vehicle distance is medium, when the degree of coincidence M is the highest, the increase / decrease value of the driving ability is made positive. This is because, if the inter-vehicle distance is medium, unlike the case where the inter-vehicle distance is too short, it is not necessary to evaluate the driving characteristics of the driver low because of the inter-vehicle distance, and the matching degree M of the driving scene is high. Therefore, the increase / decrease value of the driving ability is made positive.

  When the increase / decrease value of the travel capability is determined, the travel capability is acquired from the travel capability storage unit 441, and the determined increase / decrease value is added to the acquired travel capability. Then, the running ability stored in the running ability storage unit 441 is updated to the added running ability.

  Returning to FIG. The collision predicted value calculation unit 45 sequentially acquires the inter-vehicle distance between the host vehicle 2 and the preceding vehicle 3 from the front sensor 10. From the change in the inter-vehicle distance, the relative speed of the preceding vehicle 3 with respect to the host vehicle 2 is calculated. Then, a predicted collision value representing the possibility of a collision is calculated from the relative speed of the preceding vehicle 3 and the inter-vehicle distance. In the present embodiment, TTC (Time-To-Collision) is calculated as the predicted collision value.

  The driving support unit 46 determines whether or not to execute the driving support control based on the predicted collision value calculated by the predicted collision value calculation unit 45 and the latest driving capability evaluated by the driving capability evaluation unit 44. In order to determine whether or not to perform the driving support control, in addition to the predicted collision value and the latest driving capability, whether or not to perform the driving support is determined in advance from the driving capability and the predicted collision value shown in FIG. Use a stored relationship. The relationship shown in FIG. 12 is a relationship in which, when the traveling capability is within a certain range, the predicted collision value determined to execute driving assistance increases as the traveling capability increases. In other words, the relationship shown in FIG. 12 indicates that when the driving ability is within a certain range, the lower the driving ability, the lower the predicted collision value determined to execute driving assistance, that is, the assistance execution condition is satisfied. It is a relationship that makes it easier to do.

  The reason for this relationship is that the lower the running ability, the more difficult it is to achieve a state to be achieved by driving assistance, such as collision avoidance, unless driving assistance is executed earlier. Further, the reason why the relationship is as shown in FIG. 12 is that when the driving ability is high, the behavior of lowering the predicted collision value can be quickly executed, so that unnecessary control is not executed. It can be said that.

[Effect of first embodiment diameter embodiment]
As described above, in the first embodiment described above, the degree of coincidence M between the driving scene of the host vehicle 2 and the driving scene of the preceding vehicle 3 is calculated. The driving scene of the host vehicle 2 is obtained by symbolizing and integrating the host vehicle behavior value that is a value related to the behavior of the host vehicle 2 and the host vehicle differential value that is a differential value thereof. The driving state including one side is shown. The driving scene of the preceding vehicle 3 is recognized in the same manner as the driving scene of the host vehicle 2. Therefore, the driving scene of the preceding vehicle 3 also represents a traveling state including at least one of the vehicle speed and the traveling direction. The method of acceleration, deceleration, and steering operation differ depending on the driver. Therefore, the driving characteristics of the driver are reflected in the driving scene.

  The change in the vehicle speed changes depending on the acceleration / deceleration performance of the vehicle, and the change in the traveling direction of the vehicle changes depending on the steering performance of the vehicle. For these reasons, the driving scene is affected by both the driving characteristics of the driver and the characteristics of the vehicle related to the behavior.

  When this driving scene is compared between the host vehicle 2 and the preceding vehicle 3, the driving scene tends to match when the driving capability of the driver is high than when the driving capability of the driver is low. In addition, the higher the ability related to the behavior of the host vehicle 2 that is the subsequent vehicle, the higher the possibility that the driving scene of the host vehicle 2 can follow the preceding vehicle 3 at an early stage. On the contrary, when the own vehicle 2 has a low capability regarding the behavior of the own vehicle 2 such as a truck, it takes time for the driving scene of the own vehicle 2 to follow the preceding vehicle 3. Therefore, the coincidence degree M of the driving scene reflects the high driving ability of the driver and the high ability of the vehicle related to the behavior.

  Since the traveling ability evaluation unit 44 evaluates the traveling ability of the host vehicle 2 based on the degree of coincidence M of the driving scene, the traveling ability can be evaluated with high accuracy.

  In the present embodiment, as shown in FIG. 4, the driving scene is represented by a driving word, and the driving word can clearly detect switching. Thereby, the coincidence degree M can be calculated based on the time difference Δt at which the driving word is switched. When the driving capability of the driver of the host vehicle 2 is low, when the driving capability of the driver of the host vehicle 2 is low, such as the time until the host vehicle 2 is decelerated after the preceding vehicle 3 is decelerated, The time until the behavior of the host vehicle 2 changes following the preceding vehicle 3 tends to increase.

  In addition, when the ability of the own vehicle 2 is low, such as when the acceleration / deceleration capability of the own vehicle 2 is low, the time required for a certain degree of acceleration / deceleration tends to increase, the vehicle 2 follows the preceding vehicle 3 as well. The time until the behavior of the host vehicle 2 changes tends to be longer.

  From these things, it can be considered that the time difference Δt well represents the running ability. The degree of coincidence M is calculated based on this time difference Δt. As a result, the running ability can be evaluated particularly accurately.

  Further, in the present embodiment, the degree of coincidence M is determined in consideration of the degree of coincidence s of contents. As a result, the running ability can be further accurately evaluated.

  In the present embodiment, the degree of coincidence M is calculated based on the time difference Δt, and the time at which the driving scene is switched is earlier in the host vehicle 2 that is the rear vehicle than in the preceding vehicle 3. In this case, the degree of coincidence M is calculated with the time difference Δt as 0. Thereby, it can suppress that the driving capability of the own vehicle 2 which carried out the advance driving | operation is evaluated low.

  In the present embodiment, when there is no corresponding driving scene, the coincidence degree M is set to a predetermined value θ that can be determined without comparing driving scenes. Thereby, the influence which the matching degree M determined when there is no corresponding driving scene has on the driving ability can be reduced.

  In addition, as shown in FIG. 12, the driving support unit 46 of the present embodiment facilitates the execution of the support execution condition as the driving ability decreases, and thus provides driving assistance in an appropriate situation according to the driving ability. Can be executed.

Second Embodiment
Next, a second embodiment will be described. In the following description of the second embodiment, elements having the same reference numerals as those used so far are the same as elements having the same reference numerals in the previous embodiments unless otherwise specified. Further, when only a part of the configuration is described, the above-described embodiment can be applied to the other parts of the configuration.

  In the first embodiment described above, the segment traveling state is statistically determined by processing, and the driving word generated by further statistically processing the driving symbol representing the segment traveling state is used as the driving scene. On the other hand, in the second embodiment, each driving scene shown in FIG. 13 is set. Specifically, in FIG. 13, stop, acceleration, deceleration, steady travel, right / left turn, curve travel, and others are set as driving scenes. In FIG. 13, driving scene conditions for recognizing each driving scene are also set. In FIG. 13, the speed fluctuation is a speed fluctuation in a preset time, for example, about several seconds to 10 seconds.

  In 2nd Embodiment, the own vehicle scene recognition part 41 and the preceding vehicle scene recognition part 42 recognize that it is a corresponding driving scene, when the driving scene conditions shown in FIG. 13 are satisfied.

  In the second embodiment, the degree of coincidence calculation unit 43 calculates the degree of coincidence M using Equation 1 as in the first embodiment. However, the degree of coincidence s of contents is determined using the coincidence degree table shown in FIG. The value of each coincidence s constituting the coincidence table of FIG. 14 is determined in advance.

  The driving support unit 46 of the first embodiment determines whether or not to perform driving support using the relationship shown in FIG. On the other hand, in 2nd Embodiment, the driving assistance part 46 determines a driving assistance level with the table | surface shown in FIG. And the driving assistance determined according to the determined driving assistance level is performed. The driving support level Lv0 means that the driving support is not executed, and the higher the number, the higher the driving support is executed. For example, the driving support levels Lv1 to Lv4 are only notifications, but are changed step by step to notifications that mean that the danger is high. At the driving support level Lv5, the behavior of the vehicle is automatically controlled. In the table shown in FIG. 15, the lower the traveling ability, the lower the predicted collision value for executing driving assistance. Thereby, driving support is performed at an early timing, so that driving ability is low.

<Third Embodiment>
In the third embodiment, the coincidence degree table shown in FIG. 16 is used instead of the coincidence degree table shown in FIG. In FIG. 16, the coincidence s when the driving scene of the preceding vehicle 3 is acceleration is higher than that in the coincidence table shown in FIG.

  As a result, even if the driving scene is the same acceleration and steady running (that is, speed maintenance), the preceding vehicle 3 is accelerated and the own vehicle is more accelerated than the preceding vehicle 3 is running steady and the own vehicle 2 is accelerating. The degree of coincidence s when 2 is steady running is higher.

  Thereby, in the scene where the own vehicle 2 does not need to accelerate excessively, it can suppress that driving ability will be evaluated low.

  As mentioned above, although embodiment of this invention was described, this invention is not limited to the above-mentioned embodiment, The following modification is also contained in the technical scope of this invention, Furthermore, the summary other than the following is also included. Various modifications can be made without departing from the scope.

<Modification 1>
For example, in the first embodiment and the second embodiment, the host vehicle 2 is a subsequent vehicle, and the traveling ability of the host vehicle 2 is evaluated. However, the present invention is not limited to this, and the host vehicle 2 may be a preceding vehicle, and the driving capability evaluation apparatus 1 mounted on the host vehicle 2 may evaluate the driving capability of the vehicle immediately following the host vehicle 2. In this case, the immediately following vehicle becomes the comparison vehicle and the target vehicle in the claims.

  When evaluating the driving ability of the immediately following vehicle, the behavior-related value of the immediately following vehicle is acquired from the immediately following vehicle by wireless communication by the communication unit 30. Further, a rear sensor for detecting the behavior of the rear vehicle may be provided at the rear part of the host vehicle 2, and a part or all of the behavior related values of the immediately following vehicle may be detected by the rear sensor.

  In the case of evaluating the driving ability of the immediately following vehicle, when the immediately following vehicle changes to another vehicle, the driving ability evaluated so far is discarded and the driving ability is evaluated from the beginning. This is because the traveling ability of the nearest succeeding vehicle is different for each specific vehicle corresponding to the nearest succeeding vehicle. On the other hand, when the host vehicle 2 is a succeeding vehicle and the traveling capability of the host vehicle 2 is evaluated, it is not necessary to evaluate the traveling capability from the beginning even if the preceding vehicle changes.

<Modification 2>
The driving scene may be recognized by the comparison vehicle, and the recognized driving scene may be transmitted to the surroundings wirelessly. In this case, the communication unit 30 receives the driving scene transmitted from the comparison vehicle, and the preceding vehicle scene recognition unit 42 acquires the driving scene transmitted from the comparison vehicle received by the communication unit 30.

<Modification 3>
In the previous embodiment, the degree of coincidence M is calculated from the time difference Δt and the degree of coincidence s of the contents, but the degree of coincidence M may be determined from only one of them.

<Modification 4>
In the second embodiment, seven types of driving scenes are set, but only a deceleration scene and other two types of driving scenes may be set. In addition to these two types of driving scenes, three to six types of driving scenes may be set by adding different driving scenes. Further, eight or more types of driving scenes may be set.

<Modification 5>
In the first embodiment, each driving word is set as one driving scene. However, a row of driving words, that is, a plurality of driving words may be recognized as one driving scene.

<Modification 6>
In the first embodiment, the driving support level may be determined using the table shown in FIG. In the second embodiment, it may be determined based on FIGS. 11 and 12 whether or not driving assistance is executed.

1: Driving ability evaluation device 2: Own vehicle 3: Leading vehicle 10: Forward sensor 14: Word transition database 20: Own vehicle behavior sensor 30: communication unit 40: control unit 41: own vehicle scene recognition unit 42: preceding vehicle scene recognition Unit 43: coincidence calculation unit 44: driving ability evaluation unit 45: collision prediction value calculation unit 46: driving support unit 100: driving ability evaluation device 110: forward sensor 120: vehicle behavior sensor 130: communication unit 140: control unit 411: Own vehicle differential value calculation unit 412: Symbolization unit 413: Symbol integration unit 421: Preceding vehicle differential value calculation unit 422: Symbolization unit 423: Symbol combination unit 431: Driving scene storage unit 432: Detection unit 433: Time difference calculation unit 434 : Content comparison unit 435: Calculation unit 441: Running capacity storage unit 4 42: Driving ability update unit

Claims (13)

A driving capability evaluation device for evaluating the driving capability of a vehicle,
A host vehicle scene recognition unit (41) for recognizing a driving scene representing a driving state including at least one of a vehicle speed and a traveling direction with respect to the host vehicle, which is a vehicle equipped with the driving ability evaluation device;
A comparison vehicle scene recognition unit (42) for recognizing the driving scene of the comparison vehicle that is one of the preceding vehicle and the immediately following vehicle for the host vehicle;
A degree of coincidence calculation unit (43) for calculating a degree of coincidence between the driving scene of the host vehicle recognized by the host vehicle scene recognition unit and the driving scene of the comparison vehicle recognized by the comparison vehicle scene recognition unit;
A driving capability evaluation unit comprising a driving capability evaluation unit (44) that sequentially evaluates the driving capability of the target vehicle that is one of the host vehicle and the comparison vehicle based on the matching score calculated by the matching score calculation unit. apparatus. In claim 1,
The driving scene is determined from a plurality of types of driving states,
The coincidence calculation unit is a driving ability evaluation device that calculates the coincidence based on a time difference between a time when the driving scene of the host vehicle is switched and a time when the driving scene of the comparative vehicle is switched. In claim 1 or 2,
The driving scene is determined from a plurality of types of driving states,
The degree of coincidence calculation unit is a driving ability evaluation device that calculates the degree of coincidence based on the degree of coincidence between the driving scene of the host vehicle and the content of the driving scene of the comparison vehicle. In any one of Claims 1-3,
When the inter-vehicle distance between the host vehicle and the comparative vehicle is equal to or less than a preset threshold distance, the travel capability evaluation unit changes the travel capability based on the degree of coincidence as the inter-vehicle distance is shorter. Driving ability evaluation device that increases In claim 2,
The coincidence calculation unit
The driving scene of the host vehicle recognized by the host vehicle scene recognition unit is sequentially acquired, the driving scene of the comparison vehicle recognized by the comparison vehicle scene recognition unit is sequentially acquired, and the degree of coincidence is calculated. The degree of coincidence is calculated between the driving scenes that are not,
And the time when the driving scene of the front vehicle, which is a vehicle traveling on the front side of the host vehicle and the comparison vehicle, and the rear side, which is a vehicle traveling on the rear side of the host vehicle and the comparison vehicle, are The degree of coincidence is calculated based on the time difference from the time when the driving scene of the vehicle is switched,
The travel capability evaluation apparatus according to claim 1, wherein when the driving scene is switched, the time difference is set to 0 and the degree of coincidence is calculated when the rear vehicle is earlier than the front vehicle. In claim 5,
The coincidence calculation unit
The driving scene of the rear vehicle following the driving scene of the rear vehicle that calculates the degree of coincidence with the driving scene of the front vehicle before the driving scene of the front vehicle that calculates the degree of matching is switched. When is switched,
For the driving scene of the rear vehicle that is switched before the driving scene of the front vehicle for which the degree of coincidence is switched and that does not calculate the degree of coincidence with the driving scene of the front vehicle, the front side A travel capability evaluation device having the degree of coincidence that can be determined without comparing with the driving scene of the vehicle. In any one of Claims 1-4,
The coincidence calculation unit
The driving scene of the front vehicle, which is a vehicle that travels in front of the host vehicle and the comparison vehicle, is the driving scene relating to acceleration, and the rear side, which is a vehicle that travels in the rear of the host vehicle and the comparison vehicle. The degree of coincidence when the driving scene of the vehicle is the driving scene related to speed maintenance,
The driving ability evaluation device that makes the degree of coincidence higher when the driving scene of the front vehicle is the driving scene related to speed maintenance and the driving scene of the rear vehicle is the driving scene related to acceleration. In any one of Claims 1-7,
When the support execution condition is satisfied, the vehicle is provided with a driving support unit (46) for executing predetermined driving support control,
The driving support unit is a driving capability evaluation device that changes the support execution condition to a condition that is more likely to be established as the driving capability evaluated sequentially by the driving capability evaluation unit is lower. In any one of Claims 1-8,
A host vehicle behavior detection unit (20) for detecting a host vehicle behavior value which is a value related to the behavior of the host vehicle;
A host vehicle differential value calculation unit (411) that calculates a host vehicle differential value that is a differential value of the host vehicle behavior value detected by the host vehicle behavior detection unit;
Based on the own vehicle behavior value detected by the own vehicle behavior detection unit and the own vehicle differential value calculated by the own vehicle differential value calculation unit, the traveling state of the own vehicle is classified into divisional traveling states as a unit of division. A host vehicle symbolization unit (412) that generates a host vehicle symbol string by representing the segmented traveling state by the corresponding symbol.
A host vehicle segmentation unit (413) for segmenting the host vehicle symbol string generated by the host vehicle symbolization unit for each driving word meaning a predetermined traveling state;
The host vehicle scene recognizing unit recognizes the driving scene of the host vehicle from the driving word segmented by the host vehicle segmenting unit. In claim 9,
A comparative vehicle behavior sensor (10) for detecting a comparative vehicle behavior value that is a value related to the behavior of the comparative vehicle;
A comparative vehicle differential value calculation unit (421) that calculates a comparative vehicle differential value that is a differential value of the comparative vehicle behavior value detected by the comparative vehicle behavior sensor;
Based on the comparative vehicle behavior value detected by the comparative vehicle behavior sensor and the comparative vehicle differential value calculated by the comparative vehicle differential value calculation unit, the traveling state of the comparative vehicle is classified into the divided traveling state and classified. A comparison vehicle symbol string generation unit (422) that generates a comparison vehicle symbol string by representing the segmented traveling state by a corresponding symbol;
A comparison vehicle segmentation unit (423) that segments the comparison vehicle symbol string generated by the comparison vehicle symbol string generation unit for each driving word;
The comparative vehicle scene recognizing unit recognizes the driving scene of the comparative vehicle from the driving word segmented by the comparative vehicle segmenting unit. In claim 9,
A receiving unit (30) for receiving a comparative vehicle behavior value, which is a value related to the behavior of the comparative vehicle, transmitted from the comparative vehicle;
A comparative vehicle differential value calculation unit (421) that calculates a comparative vehicle differential value that is a differential value of the comparative vehicle behavior value received by the reception unit;
Based on the comparative vehicle behavior value received by the receiving unit and the comparative vehicle differential value calculated by the comparative vehicle differential value calculation unit, the traveling state of the comparative vehicle is divided into the divided traveling states, and the divided vehicle state is divided. A comparison vehicle symbol string generation unit (422) that generates a comparison vehicle symbol string by representing the segment traveling state by a corresponding symbol;
A comparison vehicle segmentation unit (423) that segments the comparison vehicle symbol string generated by the comparison vehicle symbol string generation unit for each driving word;
The comparative vehicle scene recognizing unit recognizes the driving scene of the comparative vehicle from the driving word segmented by the comparative vehicle segmenting unit. In claim 9,
A receiving unit (30) for receiving information representing the driving scene of the comparison vehicle transmitted from the comparison vehicle;
The comparative vehicle scene recognition unit is a driving ability evaluation device that recognizes the driving scene of the comparative vehicle from information representing the driving scene of the comparative vehicle received by the receiving unit. In any one of Claims 2, 5, and 6,
The own vehicle scene recognition unit recognizes two or more types of the driving scene including a deceleration scene as the driving scene of the own vehicle,
The comparative vehicle scene recognition unit recognizes two or more types of the driving scenes including the deceleration scene as the driving scenes of the comparative vehicle,
The coincidence calculation unit includes a time when the driving scene of the host vehicle recognized by the host vehicle scene recognition unit is switched from the driving scene other than the deceleration scene to the deceleration scene, and the comparison vehicle scene recognition unit A driving ability evaluation device that calculates the degree of coincidence based on a time difference between the recognized driving scene of the comparative vehicle and the time when the driving scene other than the deceleration scene is switched to the deceleration scene.

Images