JP5530543B2 - Evaluation program and evaluation device for automatic brake system - Google Patents

Description

The present invention relates to an evaluation program and an evaluation apparatus for an automatic brake system.

  The automatic brake system is mounted on a vehicle and is for avoiding a collision with an object on a traveling road, and automatically activates a brake when an obstacle is detected ahead. Such an automatic brake system includes a detection unit that detects an object on the road and an operation unit that activates a brake when the detection unit detects the object (for example, Patent Document 1).

JP 2007-062604 A

  By the way, there are a stereo camera equipped with a CCD, a detection unit using a millimeter wave radar / ultrasonic radar, etc. as a detection unit adopted in the automatic brake system. In which the braking force is constant and the braking force is appropriately determined by the speed of the vehicle and the distance from the vehicle to the object.

  On the other hand, one of the performances required for an automatic brake system is a collision avoidance capability. When evaluating the collision avoidance ability, a method based on the performance of the component parts such as the detection unit and the operation unit described above can be considered. However, although such an evaluation method can be applied to the same type of automatic brake systems, it cannot be applied to different types of automatic brake systems.

  Therefore, a method for uniformly evaluating the performance of the automatic brake system is required regardless of the method, but a method for evaluating the performance of such an automatic brake system has not been established.

Therefore, in view of such a situation, the present invention intends to provide an evaluation program and an evaluation device for an automatic brake system.

The present invention is an automatic brake system evaluation program executed by a computer having at least a CPU, and the CPU is configured to determine whether a vehicle equipped with the automatic brake system collides with an obstacle by the operation of the automatic brake system. A reading unit that reads the result of the test of whether or not from the storage unit, a regression analysis execution unit that performs linear regression analysis using a one-dimensional model based on the result of the test read by the reading unit, and the linear regression analysis And an evaluation unit that evaluates the performance of the automatic braking system based on the result. The explanatory variable in the linear regression analysis is for the obstacle of the vehicle before the automatic braking system is operated. a relative speed vi, objective variable in the linear regression analysis the vehicle is the obstacle Is the relative velocity vc for the obstacle upon impact, the test results are included with the relative velocity vc regarding the vehicle of the said relative speed vi relative speed vi is the evaluation unit, the linear As the vi intercept of the regression line obtained by regression analysis increases, the performance of the automatic brake system is evaluated to be high, and the automatic brake system detects the obstacle, the obstacle sensor, A determination unit that performs a determination process for determining whether or not the obstacle is present in the traveling direction of the vehicle based on a sensing signal from an object sensor; and when the determination process determines that the obstacle is present, the vehicle And a controller for operating the brake .

The present invention is an automatic brake system evaluation program executed by a computer having at least a CPU, and the CPU is configured to determine whether a vehicle equipped with the automatic brake system collides with an obstacle by the operation of the automatic brake system. A reading unit that reads the result of the test of whether or not from the storage unit, a regression analysis execution unit that performs linear regression analysis using a one-dimensional model based on the result of the test read by the reading unit, and the linear regression analysis And an evaluation unit that evaluates the performance of the automatic braking system based on the result. The explanatory variable in the linear regression analysis is for the obstacle of the vehicle before the automatic braking system is operated. a relative speed vi, objective variable in the linear regression analysis the vehicle is the obstacle Is the relative velocity vc for the obstacle upon impact, the test results are included with the relative velocity vc regarding the vehicle of the said relative speed vi relative speed vi is the evaluation unit, the linear As the slope of the regression line obtained by the regression analysis becomes smaller, the performance of the automatic brake system is evaluated to be high, and the automatic brake system detects the obstacle, the obstacle sensor for detecting the obstacle, and the obstacle A determination unit that determines whether or not the obstacle is present in the traveling direction of the vehicle based on a sensing signal from the sensor; and when it is determined that the obstacle is present in the determination process, And a control unit that operates the brake .

The present invention is an automatic brake system evaluation program executed by a computer having at least a CPU, and the CPU is configured to check whether a vehicle equipped with the automatic brake system collides with an obstacle by the operation of the automatic brake system. a reading section for reading the result of whether the test from the storage means, wherein the reading unit is the test result of reading is, the vehicle and the collision has occurred events that did not collide to the obstacle, the vehicle is the obstacle And a light collision occurrence event that has a small adverse effect due to the collision of the vehicle and a heavy collision occurrence event in which the vehicle collides with the obstacle and the adverse effect due to the vehicle collision is greater than the light collision occurrence event. It is determined which of these is based on a predetermined threshold, and the collision has not occurred based on the result of the determination. A selection unit for assigning an identifier for identifying the event, the light collision occurrence event, and the heavy collision occurrence event to the result of the test, and performing logistic regression analysis based on the result of the test to which the identifier is assigned It functions as a component that constitutes a regression analysis execution unit and an evaluation unit that evaluates the performance of the automatic brake system based on the result of the logistic regression analysis. The relative speed vi of the vehicle to the obstacle before operation, and the objective variable in the logistic regression analysis is selected from the non-collision event, the light collision event, and the heavy collision event The occurrence probability of an event, and the regression analysis execution unit performs an integration interval determined by the relative speed vi. The integrated value S0 of the occurrence probability of the heavy collision occurrence event, the integration value S1 of the occurrence probability of the light collision occurrence event, and the integration value S2 of the occurrence probability of the non-collision occurrence event are obtained by the logistic regression analysis. As a result, the evaluation unit determines that the performance of the automatic brake system is higher as the integral value S0 becomes smaller or as the sum of the integral value S1 and the integral value S2 becomes larger. It is characterized by evaluating.

The automatic brake system includes an obstacle sensor for detecting the obstacle, and a determination process for determining whether or not the obstacle exists in the traveling direction of the vehicle based on a sensing signal from the obstacle sensor. And a control unit that operates a brake of the vehicle when it is determined that the obstacle is present in the determination process. The predetermined threshold is preferably a value of the relative speed vc set based on the probability of occurrence of a phenomenon induced by the obstacle . Furthermore, the phenomenon is preferably a simplified disorder scale.

Wherein the predetermined threshold value is preferably a value of the relative velocity vc, which is set on the basis of head injury indicator of phenomena induced by the obstacle.

The CPU preferably functions to include a threshold setting unit that sets the threshold.

  An automatic brake system evaluation apparatus according to the present invention includes the above-described automatic brake system evaluation program.

  According to the above means, the performance of the automatic brake system can be uniformly evaluated regardless of the method.

It is explanatory drawing which shows the outline | summary of a mode that operation check of an automatic brake system is performed. It is explanatory drawing which shows the outline | summary of the evaluation apparatus of an automatic brake system, a vehicle, and an automatic brake system. It is a flowchart of the operation confirmation of an automatic brake system. It is a block diagram which shows the outline | summary of the evaluation apparatus of an automatic brake system. It is a graph showing the outline | summary of the experimental data obtained by operation confirmation, and the regression line obtained by the regression analysis of this experimental data, a horizontal axis is the initial speed vi, and a vertical axis | shaft is the collision speed vc. It is a graph showing the outline | summary of the regression line obtained by the regression analysis of the experimental data obtained by operation confirmation, a horizontal axis is the initial speed vi, and a vertical axis | shaft is the collision dummy variable Cdv. It is a graph showing the outline | summary of the regression line obtained by the regression analysis of the experimental data obtained by operation confirmation, a horizontal axis is the initial speed vi, and a vertical axis | shaft is the collision dummy variable Cdv. 5 is a graph showing a distribution function F (vi) representing the number of deaths when a vehicle having an initial speed vi collides with a person for each initial speed vi, the horizontal axis is the initial speed vi, and the vertical axis is the number of deaths. . (A) shows an outline of Expressions 7 to 9, and the vertical axis is a graph of the occurrence probability P, and the horizontal axis is the initial speed vi. (B) shows the integration and breakdown of each occurrence probability P shown in (A), with the vertical axis representing the occurrence probability P and the horizontal axis representing the initial velocity vi. (A) shows an outline of Expressions 7 to 9, and the vertical axis is a graph of the occurrence probability P, and the horizontal axis is the initial speed vi. (B) shows the integration and breakdown of each occurrence probability P shown in (A), with the vertical axis representing the occurrence probability P and the horizontal axis representing the initial velocity vi.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

  As shown in FIG. 1, the automatic brake system 10 automatically brakes the vehicle 20 by detecting an obstacle 30 and is mounted on the vehicle 20. Here, the obstacle 30 includes a person, an animal, a plant, and other objects (guardrail, stone, etc.).

  As shown in FIG. 2, the vehicle 20 includes an engine 21 mounted on a vehicle body (not shown), an accelerator pedal 22 for operating the output amount of the engine 21, a tire 23 driven by the output of the engine 21, A drive shaft 24 that transmits the output of the engine 21 to the tire 23, a brake mechanism 25 that suppresses the rotation of the tire 23, and a handle 27 that adjusts the direction of the tire 23 are provided. The brake mechanism 25 includes a brake member such as a disc brake and a drum brake, and a brake pedal for operating the brake member.

  As shown in FIGS. 1 and 2, the automatic brake system 10 detects a collision between the distance measuring sensor 11 that measures the distance L <b> 1 between the vehicle 20 and the obstacle 30, and the obstacle 30 provided in the front portion of the vehicle 20. A collision sensor 12, a speed sensor 13 that detects the speed of the vehicle 20, a control unit 14 that controls each unit, and a storage device 15 that stores predetermined values. As the distance measuring sensor 11, any of a radar, a camera (monocular camera or stereo camera), a laser, a combination of a radar and a camera, a combination of a radar, a camera, and a laser may be used.

  The control unit 14 brakes the vehicle 20 based on the sensing signal from the distance measuring sensor 11. More specifically, the control unit 14 determines the presence or absence of the obstacle 30 based on the sensing signal from the distance measuring sensor 11. Further, when it is determined that the obstacle 30 is present, the control unit 14 calculates TTC (Time to Collation) based on the distance L1 to the obstacle 30 and the relative speed v with respect to the obstacle 30. Here, TTC is given by L1 / v. Further, the control unit 14 controls the brake mechanism 25 based on the calculated TTC value. Note that the control unit 14 may operate the handle 27.

  Examples of the storage device 15 include a hard disk device incorporating a magnetic disk.

  Next, an outline of operation confirmation (hereinafter referred to as operation confirmation) of the automatic brake system 10 will be described.

  The vehicle 20 travels on a predetermined travel path so that the relative speed of the vehicle 20 with respect to the obstacle 30 becomes a constant value (S10 in FIG. 3). At this time, the steering operation may be performed by either the manual or the control unit 14. Hereinafter, the relative speed of the vehicle 20 before the automatic brake system 10 operates is referred to as an initial speed vi.

  The distance measuring sensor 11 measures the distance L1 to the obstacle 30. The control unit 14 continuously calculates the relative speed v of the vehicle 20 with respect to the obstacle 30, and the collision sensor 12 continuously detects the presence or absence of a collision (S20 in FIG. 3).

  Based on the sensing signal from the distance measuring sensor 11, the control unit 14 determines whether or not the obstacle 30 is present and determines the attribute of the obstacle 30, that is, whether the obstacle 30 is a person or an object. The attribute determination is performed by determination processing (S30 in FIG. 3). In the determination process, when it is determined that the obstacle 30 is present, the distance L1 (see FIG. 1) from the vehicle 20 to the obstacle 30 and the relative speed v with respect to the obstacle 30 are calculated, and TTC (Time to Collation). When the calculated TTC is equal to or less than a preset threshold value or less than the threshold value, the control unit 14 operates the brake mechanism 25. Thereby, braking of the vehicle 20 in the running state is performed by the automatic brake system 10 (S40 in FIG. 3). On the other hand, if it is determined in the determination process that there is no obstacle 30, the determination process is repeated. That is, the determination process is continuously performed until it is determined that the obstacle 30 is present.

  The threshold value may be one or plural. When there are a plurality of thresholds, for example, the following may be performed. First, the brake mechanism 25 operates when the calculated TTC is less than the larger threshold value. When the calculated TTC is between the larger threshold value and the smaller threshold value, that is, when the time until the vehicle 20 having the relative speed v reaches the obstacle 30 is relatively long, the brake mechanism 25 Operates with a relatively small braking force BK1. When the calculated TTC is smaller than the smaller threshold value, that is, when the time until the vehicle 20 having the relative speed v reaches the obstacle 30 is relatively short, the brake mechanism 25 is larger than the braking force BK1. Operates with braking force BK2. Here, the braking force of the brake mechanism 25 may be, for example, a brake pedal depression amount.

  Thereafter, a speed determination process is performed to determine whether or not the vehicle 20 is stopped, that is, whether or not the speed of the vehicle read from the speed sensor 13 is “0” (S50 in FIG. 3). When the vehicle speed read from the speed sensor 13 is “0”, the process proceeds to S60 in FIG. On the other hand, if the vehicle speed read from the speed sensor 13 is not "0", the process proceeds to S40 in FIG.

  Next, the control part 14 performs the collision determination process which determines the presence or absence of a collision based on the detection signal from the collision sensor 12 (S60 in FIG. 3). When it is determined that there is a collision with the obstacle 30, the control unit 14 sets a value indicating that a collision has occurred (for example, “1”) as the collision dummy variable Cdv (vi) for the initial speed vi. Is stored in the storage device 15 (S70 in FIG. 3), and the value of the relative speed v of the vehicle 20 when the collision is detected is set as the collision speed vc (vi) of the vehicle 20 at the initial speed vi. (S80 in FIG. 3). On the other hand, when it is determined that there is no collision with the obstacle 30, the control unit 14 sets a value indicating that no collision has occurred as the collision dummy variable Cdv (vi) for the initial speed vi (for example, “0”). Is stored in the storage device 15 (S75 in FIG. 3). When it is determined that there is no collision, the control unit 14 may set the value of the collision speed vc (vi) of the vehicle 20 at the initial speed vi to “0” or “blank”. Here, in the first evaluation method (described later) of the automatic brake system, when data in which no collision has occurred is included in the regression analysis target, the collision speed vc (vi) of the vehicle 20 at the initial speed vi is determined. When the value is “0” and data in which no collision has occurred is not included in the regression analysis target, the value of the collision speed vc (vi) of the vehicle 20 at the initial speed vi is preferably “blank”. .

  Thus, the storage device 15 stores the operation confirmation result including the result identifier for identifying other operation confirmation results and the experimental data associated with the result identifier. The experimental data includes an initial speed vi, a collision dummy variable Cdv (vi) at the initial speed vi, and a collision speed vc (vi) at the initial speed vi.

  Next, when a sufficient number of operation confirmation data is obtained in S90 of FIG. 3, the operation confirmation is terminated, and when a sufficient number of operation confirmation data is not obtained, FIG. Returning to S10, a new operation check is performed. When a new operation check is performed, a reset operation may be performed before a new operation check depending on the type of the automatic brake system 10. As this reset operation, for example, there is a restart of the engine.

  As shown in FIG. 4, an automatic brake system evaluation device (hereinafter referred to as an evaluation device) 40 evaluates the automatic brake system 10 based on the operation confirmation data obtained as described above. CPU 41, RAM 42, ROM 43, input device 44, display device 45, input / output interface 46, and bus 47.

  The CPU 41 is a so-called central processing unit, and implements various functions of the evaluation device 40 by executing various programs. The RAM 42 is a so-called RAM (Random Access Memory), and is used as a work area for the CPU 41. The ROM 43 is a so-called ROM (read only memory), and stores a basic OS executed by the CPU 41 and various programs (for example, an evaluation program for an automatic brake system).

  The input device 44 is an input key, a keyboard, or a mouse, and inputs various information. The display device 45 is a display and displays various operation states. The input / output interface 46 inputs and outputs power and control signals for operating the storage device 15. The bus 47 is a wiring that performs communication by integrally connecting the CPU 41, the RAM 42, the ROM 43, the input device 44, the display device 45, the input / output interface 46, and the like.

  When the basic OS and various programs stored in the ROM 43 are executed by the CPU 41, as shown in FIG. 2, the evaluation apparatus 40 includes a regression analysis execution unit 40A that performs regression analysis based on the test results, It functions as an evaluation unit 40B that evaluates the automatic brake system 10 based on the result. Furthermore, the evaluation device 40 also functions as a selection unit 40C that divides the operation confirmation result in which a collision has occurred into a plurality of groups according to a predetermined condition and a setting unit 40D for setting a predetermined condition. Details of the sorting unit 40C and the setting unit 40D will be described later.

  Hereinafter, the evaluation method of the automatic brake system performed by the evaluation device 40 will be described.

  First, the evaluation method of the first automatic brake system will be described.

  The regression analysis execution unit 40A reads the operation confirmation result extracted under a predetermined condition from the storage device 15 in which a plurality of operation confirmation results are stored. Next, the regression analysis execution unit 40A extracts the initial speed vi and the collision speed vc (vi) at the initial speed vi from the read operation confirmation result. Then, the regression analysis execution unit 40A performs a regression analysis of the one-dimensional model with the objective variable as the collision speed vc (vi) at the initial speed vi and the explanatory variable as the initial speed vi. Furthermore, the regression analysis execution unit 40A outputs the result of the regression analysis of the one-dimensional model, that is, the data of the regression line 50 (see FIG. 5) to the display device 45. In this way, the display device 45 displays the linear regression line 50 obtained by the regression analysis of the one-dimensional model. The regression analysis execution unit 40A may store the obtained regression line 50 data in the storage device 15. Note that the regression line 50 preferably extends not only to a range of values that can be taken by the experimental data but also to a range in which one or both of the initial velocity vi and the collision velocity vc (vi) are negative values.

In the regression line 50, when the value of the initial speed vi at which “the collision speed vc (vi) = 0”, that is, the vi intercept is a positive value, “the vehicle having the initial speed less than the vi intercept is the automatic brake system. Will stop before the obstacle. " On the other hand, when the vi intercept is 0 or a negative value, it can be predicted that the collision of the vehicle with the obstacle cannot be avoided by the automatic braking system. Therefore, it can be said that the larger the value of the vi intercept, the higher the collision avoidance capability of the automatic brake system to be evaluated. That is, the evaluation unit 40B can evaluate the automatic braking system by evaluating the vi intercept. For example, the collision avoidance capability of the automatic braking system can be ranked by evaluating the vi intercept based on the following criteria. Here, A to C ranks represent those having performance as an automatic braking system, and D ranks represent those having no performance as an automatic braking system.
Rank A: vi intercept is 5 km / hour or more.
Rank B: The vi intercept is 1 km / hour or more and less than 5 km / hour.
C rank: vi intercept is 0 km / hour or more and less than 1 km / hour.
D rank: vi intercept is less than 0 km / hour.

Further, in the range where the collision speed vc (vi) takes a positive value, the impact applied to the obstacle by the collision increases as the collision speed vc (vi) increases. Therefore, as the slope of the regression line 50 (= Δvc (vi) / Δvi) becomes smaller, the impact on the obstacle due to the collision becomes smaller. It can be said that it will be higher. In this way, the automatic brake system can be evaluated by the evaluation unit 40B evaluating the inclination of the regression line 50. For example, the collision avoidance ability of the automatic brake system can be ranked by evaluating the inclination based on the following criteria. Here, A to C ranks represent those having performance as an automatic braking system, and D ranks represent those having no performance as an automatic braking system.
Rank A: The inclination is less than 1.0.
B rank: The inclination is 1.0 or more and less than 2.0.
C rank: The inclination is 2.0 or more and less than 3.0.
D rank: The inclination is 3.0 or more.

  The evaluation unit 40B may perform a combination of evaluation based on the vi intercept and evaluation based on the slope. In addition, the evaluation unit 40B evaluates based on the area surrounded by the regression line 50, the vertical axis, and the horizontal axis, instead of combining the evaluation based on the vi intercept and the evaluation based on the slope. Evaluation of the performance of the automatic brake system may be performed.

  When the obstacle is a person, the following evaluation can be performed.

From the collision speed vc, a head injury index (HIC) for a person who is an obstacle (hereinafter referred to as a collided person) can be measured.
Here, the head injury index HIC is defined as follows using the head acceleration α (acceleration received by the head of the collisionee due to the collision).
HIC = Max {(t ₂ −t ₁ ) ⁻¹ · ∫α ^2.5 dt}

Then, an abbreviated injury scale (AIS) can be obtained from the obtained head injury index HIC and a predetermined damage risk curve. Here, the damage risk curve is `` Performance of Collision Damage Mitigation Braking System and their Effects on Human Injury in the Event of Car-to-Pedestrian accidents, Stapp Car Crash Journal '' by Yasuhiro Matsui, Yong Hun, and Koji Mizuno , Vol 55 (November 2011) ”, the one described in Mertz HJ (2000) Injury risk assessments based on dummy responses, Accidental Injury, Springer-Verlag, pp89-102 can be applied. Table 1 shows an example of the probability Q (AIS ≧ 4) that is calculated based on the head injury index HIC and that the rank of the simplified injury scale AIS is 4 or more.

This summary injury scale is ranked as follows:
1: Slight injury 2: Moderate injury 3: Severe 4: Severe 5: Drowning 6: Heaviest, practically unable to save lives

And the evaluation part 40B can rank the collision avoidance capability of an automatic brake system by evaluating the probability Q based on the following criteria. Here, A to C ranks represent those having performance as an automatic braking system, and D ranks represent those having no performance as an automatic braking system.
Rank A: The probability Q is less than 0.5%.
B rank: The probability Q is 0.5% or more and less than 5%.
C rank: probability Q is 5% or more and less than 10%.
D rank: The probability Q is 10% or more.

  Although the regression analysis of the one-dimensional model is performed in the first automatic braking system evaluation method, the present invention is not limited to this. In the first automatic braking system evaluation method, other polynomial models (for example, You may use the thing using a quadratic function, the thing using a cubic function, and the thing using an exponential function.

  Next, a method for evaluating the second automatic brake system will be described.

  The regression analysis execution unit 40A reads the operation confirmation result extracted under a predetermined condition from the storage device 15 in which a plurality of operation confirmation results are stored. Next, the regression analysis execution unit 40A extracts the experimental data included in the read operation confirmation result, that is, the initial speed vi and the collision dummy variable Cdv (vi) at the initial speed vi. Then, the regression analysis execution unit 40A performs logistic regression analysis with the objective variable as the collision dummy variable Cdv (vi) and the explanatory variable as the initial speed vi. Furthermore, the regression analysis execution unit 40A outputs the result of the logistic regression analysis, that is, the data of the regression line 60 (see FIG. 6) to the display device 45. Here, the regression line 60 represents the probability R that the vehicle having the initial speed vi will collide with the collision object. Thus, the display device 45 displays the regression line 60 obtained by the logistic regression analysis. The regression analysis execution unit 40A may store the obtained regression line 60 data in the storage device 15.

  In the regression line 60, it can be said that the collision avoidance capability of the automatic brake system increases as the value of the collision dummy variable Cdv (vi) decreases at the initial speed vi within a certain range. In this way, the evaluation unit 40B can evaluate the automatic brake system based on the value taken by the collision dummy variable Cdv (vi) at the initial speed vi within a certain range.

For example, the collision avoidance capability of the brake system can be ranked based on the value taken by the collision dummy variable Cdv (vi) at an initial speed vi within a certain range. An example of determination criteria in this case is shown below.
Rank A: The maximum value of the collision dummy variable Cdv (vi) is less than 0.01 in the range where the initial speed vi is less than 60 km / hour.
Rank B: In the range where the initial speed vi is less than 60 km / hour, the maximum value of the collision dummy variable Cdv (vi) is 0.01 or more and less than 0.1.
C rank: In the range where the initial speed vi is less than 60 km / hour, the maximum value of the collision dummy variable Cdv (vi) is 0.1 or more and less than 0.2.
D rank: In the range where the initial speed vi is less than 60 km / hour, the maximum value of the collision dummy variable Cdv (vi) is 0.2 or more.

  In addition, the collision avoidance capability of the brake system may be ranked based on the value of the collision dummy variable Cdv (vi) in the range where the initial speed vi is less than 80 km / hour. At this time, the value of the collision dummy variable Cdv (vi) may be set as appropriate according to the initial speed vi.

  In addition, as shown in FIG. 7, the impact given to the obstacle by the collision becomes smaller as the value taken by the initial speed vi becomes smaller in the collision dummy variable Cdv (vi) within a certain range on the regression line 60. Therefore, it can be said that the ability to control a serious accident is enhanced for the automatic brake system to be evaluated. Thus, the automatic brake system can be evaluated based on the value taken by the initial speed vi with the collision dummy variable Cdv (vi) within a certain range.

For example, the capabilities of the automatic brake system can be ranked based on the value of the initial speed vi corresponding to a predetermined collision dummy variable Cdv (vi). An example of determination criteria in this case is shown below.
Rank A: When the collision dummy variable Cdv (vi) is 0.5, the initial speed vi is 50 km / hour.
Rank B: When the collision dummy variable Cdv (vi) is 0.5, the initial speed vi is 40 km / hour.
Rank C: When the collision dummy variable Cdv (vi) is 0.5, the initial speed vi is 30 km / hour.
D rank: When the collision dummy variable Cdv (vi) is 0.5, the initial speed vi is 20 km / hour.

  Furthermore, the regression analysis execution unit 40A has a function F (vi) (see FIG. 8) that represents the number N of deaths in a certain period for each vehicle traveling speed, a probability R (vi) that the regression line 60 represents, and an expression A “ΔN = F (1−R)” is read from the storage device 15. Next, the regression analysis execution unit 40A calculates the number of deaths ΔN (hereinafter referred to as the number of death reductions) ΔN that decreases due to the mounting of the automatic brake system based on the formula A. Then, the evaluation unit 40B determines that the ability to suppress a serious accident is high with respect to the automatic brake system as the death reduction number ΔN decreases. In this way, an automatic brake system can be evaluated. The number N of deaths may be, for example, the number of deaths in a certain year or the number of deaths in a plurality of years.

Note that the evaluation unit 40B may rank the capabilities of the automatic brake system based on, for example, the value of the death reduction number ΔN. An example of determination criteria in this case is shown below.
Rank A: Death reduction number ΔN is 500 or more.
Rank B: The number of death reductions ΔN is 100 or more and less than 500.
C rank: The number of death reductions ΔN is 50 or more and less than 100.
D rank: The number of death reductions ΔN is less than 50.

  In the above embodiment, it is preferable that the speed limit on a certain road is the initial speed vi. By doing in this way, it is possible to determine whether the ranking performed by the evaluation unit 40B is an automatic brake system suitable for the road. For example, when the initial speed vi is 40 km / hour, the automatic brake system corresponding to the A rank can be evaluated as being suitable for a general road. When the initial speed vi is 80 km / hour, the automatic brake system corresponding to the A rank can be evaluated as being suitable for the highway.

In the above embodiment, the regression analysis execution unit 40A extracts the initial speed vi and the collision speed vc (vi) at the initial speed vi from the storage device 15, but the present invention is not limited to this. For example, a sensor capable of detecting the speed of the vehicle 20 and the presence or absence of a collision with an obstacle is provided in the evaluation apparatus 40, and the initial speed vi is determined from the speed of the vehicle 20 detected by this sensor and the presence or absence of a collision with an obstacle. Then, the collision speed vc (vi) at the initial speed vi may be calculated. Examples of the sensor provided in the evaluation device 40 include the distance measuring sensor 11.

  The above-described evaluation program, evaluation apparatus, and evaluation method for the automatic brake system are not limited to the above-described embodiments, and various changes can be made without departing from the scope of the present invention. It is.

  As described above, evaluating the automatic brake system based on a predetermined probability has the following advantages.

  Considering that the aspect of actual traffic accidents is complicated, various factors must be considered in order to treat the evaluation results based on the fixed criteria as the ability to avoid actual traffic accidents. As a result, an enormous amount of calculation is required to perform an evaluation to the extent that it can be obtained as an actual traffic accident avoidance capability. On the other hand, according to the present invention, since the automatic braking system is evaluated based on a predetermined probability, it is possible to easily perform an evaluation that can be handled as an actual traffic accident avoidance capability. Thus, the evaluation method of the automatic brake system using a predetermined probability can also be applied to automobile assessment. In addition, by defining the performance of the automatic brake system with probability, overconfidence of the driver in the system can be prevented.

  In the evaluation method of the second automatic brake system described above, the operation confirmation result belonging to the event where the collision occurred (collision occurrence event) and the operation confirmation result belonging to the event where the collision did not occur (non-collision event). On the other hand, after giving a different collision dummy variable Cdv for each event to which it belongs, a binomial logistic regression analysis is performed with the explanatory variable as the initial speed vi and the objective variable as the collision dummy variable Cdv. And it can evaluate from a viewpoint of the collision avoidance capability of an automatic brake system by using the result of such a binomial logistic regression analysis.

  By the way, the viewpoint of the evaluation of the automatic brake system is not limited to the viewpoint of the collision avoidance ability, but the viewpoint of the ability to avoid the high-risk collision having a great adverse effect due to the collision is also important. However, the automatic braking system cannot be evaluated from the viewpoint of the ability to avoid high-risk collisions using the results of the above-described binomial logistic regression analysis. Therefore, a method for evaluating an automatic brake system from the viewpoint of avoidance ability of high-risk collisions is also as important as a method of evaluating an automatic brake system from the viewpoint of collision avoidance ability.

  Hereinafter, a method for evaluating an automatic brake system from the viewpoint of the ability to avoid high-risk collisions will be described.

  A method for evaluating an automatic brake system from the viewpoint of the ability to avoid high-risk collisions includes a selection step, an analysis step, and an evaluation step. In the selection step, the operation confirmation results are divided into two groups, a collision occurrence event and a collision non-occurrence event, and the collision occurrence event is classified into a low-risk collision event (light collision event that is less adversely affected by the collision according to a predetermined condition. ) And high-risk collision events (heavy collision events) that have a large negative impact. In the analysis step, a multinomial logistic regression analysis is performed for these three events to obtain the occurrence probability of each event. In the evaluation step, the automatic braking system is evaluated from the viewpoint of the ability to avoid high-risk collisions from the integrated value of the occurrence probability of each event obtained.

  Details of each step will be described below.

  In the selection step, the setting unit 40D (see FIG. 2) selects the selection condition from the selection conditions stored in the storage device 15 according to the output signal from the control unit 14 or the input device 44 (see FIG. 4). . Next, the sorting unit 40C (see FIG. 2) reads the operation confirmation result from the storage device 15, and extracts the one corresponding to the collision occurrence event from the read operation confirmation result. Further, the sorting unit 40C sorts the extracted operation confirmation results into two groups of light collision events and heavy collision events according to the sorting conditions selected by the setting unit 40D. Furthermore, the sorting unit 40C assigns an identification variable Y for identifying the distribution result to each operation confirmation result belonging to the event.

For example, when the identification variable Y for identifying a heavy collision event, a light collision event, and a collision non-occurrence event is defined as follows:
Heavy collision event: Y = 0
Light collision event: Y = 1
Non-collision event: Y = 2
The sorting unit 40C sets “0” for the heavy collision event, “1” for the light collision event, and “2” for the non-collision event as the value of the variable Y. Set.

  Next, a specific example of the sorting step will be described together with the sorting conditions.

  When the selection condition is “probability Q (AIS ≧ 4) = Qy” in Table 1, the selection unit 40C refers to the data in Table 1 stored in the storage device 15 in advance, and the probability Q regarding the extracted operation confirmation result It is determined whether (AIS ≧ 4) is equal to or less than Qy. Further, the selection unit 40C sets “1” representing a light collision event to the identification variable Y for the operation confirmation result with the probability Q (AIS ≧ 4) being Qy or less, and the probability Q (AIS ≧ 4). For an operation confirmation result that is greater than Qy, “0” representing a heavy collision event is set in the identification variable Y. Note that the sorting unit 40C sets “2” representing the collision non-occurrence event in the identification variable Y for the operation confirmation result corresponding to the collision non-occurrence event.

  For example, when the threshold value of the probability Q (AIS ≧ 4), that is, Qy is “5%”, the operation confirmation result with the collision speed vc of 30 km / hour or less belongs to the light collision event in the vehicle type “sedan”. The operation confirmation result with vc of 40 km / h or more belongs to the heavy collision event. In addition, in vehicle types “light (light vehicle)” and “SUV (Sport Utility Vehicle)”, the operation confirmation result when the collision speed vc is 40 km / hour or less belongs to the light collision event, and the operation with the collision speed vc of 50 km / hour or more is performed. The confirmation result belongs to the heavy collision event.

  In addition, instead of the simplified injury scale (AIS), the “head injury index HIC” in Table 1 may be used as the selection condition. For example, when the selection condition is “HICy = 500”, in the vehicle type “sedan”, the operation confirmation result of the collision speed vc of 20 km / hour or less belongs to the light collision event, and the operation confirmation result of the collision speed vc of 30 km / hour or more. Belongs to a heavy collision event. In addition, in vehicle types “light (light vehicle)” and “SUV (Sport Utility Vehicle)”, the operation confirmation result when the collision speed vc is 30 km / hour or less belongs to the light collision event, and the operation with the collision speed vc of 40 km / hour or more is performed. The confirmation result belongs to the heavy collision event.

  Instead of the simplified injury scale (AIS) or head injury index HIC, the collision speed vc may be used as the selection condition, or the threshold value of the collision speed vc determined for each vehicle type may be used as the selection condition.

ここで、式１におけるパラメータβ_ｉ０、β_ｉ１は、最尤推定を用いて、作動確認結果から算出される。そして、回帰分析実行部４０Ａは、所定の事象の発生確率Ｐを、式３〜４から導く。

Next, in the analysis step, the regression analysis execution unit 40A reads the operation confirmation result from the storage device 15, and performs a predetermined multinomial logistic regression analysis. The explanatory variable of the multinomial logistic regression analysis performed in the analysis step is the initial velocity vi, and the objective variable is the occurrence probability P (Equations 3 to 4) expressed using the functions of Equations 1 and 2.

Here, the parameters β _i0 and β _i1 in Equation 1 are calculated from the operation confirmation result using maximum likelihood estimation. Then, the regression analysis execution unit 40A derives the occurrence probability P of the predetermined event from Equations 3-4.

また、回帰分析実行部４０Ａは、重衝突事象の発生確率Ｐ（Ｙ＝０｜ｘ）、軽衝突事象の発生確率Ｐ（Ｙ＝１｜ｘ）、及び衝突未発生事象の発生確率Ｐ（Ｙ＝２｜ｘ）を、式７〜９に基づいて算出する。

さらに、回帰分析実行部４０Ａは、算出された各事象の発生確率Ｐ（Ｙ＝ｉ｜ｘ）を表示装置４５に出力する（図９（Ａ）、１０（Ａ）参照）。 In the case of this embodiment, that is, when the operation confirmation result is divided into three events (an event in which no collision has occurred, a light collision event, and a heavy collision event) and the reference event is a heavy collision event, the analysis step The objective variable of the multinomial logistic regression analysis is the occurrence probability P (Y = i | x) (where i = 0, 1, 2) expressed using the function of Formula 5 or the function of Formula 6 (Formula 7 To 9). The regression analysis execution unit 40A may calculate the parameters β _i0 and β _i1 using the maximum likelihood estimation.

The regression analysis execution unit 40A also includes the occurrence probability P (Y = 0 | x) of the heavy collision event, the occurrence probability P (Y = 1 | x) of the light collision event, and the occurrence probability P (Y of the non-collision event). = 2 | x) is calculated based on Equations 7-9.

Further, the regression analysis execution unit 40A outputs the calculated occurrence probability P (Y = i | x) of each event to the display device 45 (see FIGS. 9A and 10A).

なお、本実施形態では、積分区間の下端を「０」、上端を「７０」としているが、積分区間は、自動ブレーキシステムを評価したい初期速度ｖｉの範囲に応じて設定すればよい。 In the evaluation step, the evaluation unit 40B calculates the integrated values S ₀ , S ₁ , and S ₂ of the occurrence probability P of each event based on Expressions 7s to 9s, respectively.

In the present embodiment, the lower end of the integration interval is set to “0” and the upper end is set to “70”. However, the integration interval may be set according to the range of the initial speed vi for which the automatic braking system is to be evaluated.

Next, the evaluation unit 40B outputs the integrated values S ₀ , S ₁ , S ₂ of the occurrence probability P of each event obtained by the equations 7s to 9s to the display device 45 (FIGS. 9B, 10 (B )reference). Here, the integral values S ₀ , S ₁ , S ₂ are represented as areas of the regions W0, W1, W2 in FIGS. 9B and 10B, respectively. Furthermore, the evaluation unit 40B stores the calculated integration values S ₀ , S ₁ , S ₂ in the storage device 15, respectively.

And the evaluation part 40B evaluates an automatic brake system using these integral values. Specifically, according to the sum of the integral values S ₁ and the integrated value S ₂ is greater, or, in accordance with the sum of the integrated value S ₀ and the integral value S ₁ is decreased, high-risk collision automatic braking system Evaluate the ability to avoid Of course, the evaluation unit 40B in accordance with the integral value S ₀ is reduced, or, according to the integrated value S ₂ is greater, high evasion risk collision automatic braking system may be evaluated as high.

  In this way, it is possible to evaluate the automatic braking system from the viewpoint of the ability to avoid high-risk collisions by dividing the collision occurrence event into light collision event and heavy collision event and performing multinomial logistic analysis. . Analyzing collision occurrence events into light collision events and heavy collision events is more effective than the case of analyzing two categories of collision occurrence events and non-collision events ( For example, death accident avoidance ability) can be evaluated. In addition, since multinomial logistic regression analysis is performed as an analysis method for each event, consistency with the results of binomial logistic regression analysis for two classifications of a collision occurrence event and a collision non-occurrence event can be maintained.

  Note that the probability Q (AIS ≧ 4) is used as a condition for dividing the collision occurrence event into a plurality of groups. However, the present invention is not limited to this, and by using the probability Q (AIS ≧ 3), a simplified lifetime A scale of “severe” or higher can be evaluated as a heavy collision event.

  In the above embodiment, the operation confirmation results are classified into three groups. However, in the present invention, the operation confirmation results may be classified into four or more groups. For example, since the summary lifetime measure is ranked in six stages, the collision occurrence event may be divided into six ranks, and analysis may be performed for each rank.

  In the above embodiment, the regression analysis result for each event is used to evaluate the automatic brake system, but the present invention is not limited to this. For example, for a predetermined initial speed vi (for example, 10 km / hour, 20 km / hour, 30 km / hour,...), An operation confirmation test is performed to determine whether or not to collide with a collision object by the operation of an automatic brake system. This is performed a predetermined number of times (for example, 10 times). Then, the operation confirmation result is divided into three events (non-collision event, light collision event, and heavy collision event), and a score for each event {= number of corresponding events × predetermined score (weight)} is calculated. The sum of the scores of the events thus obtained may be obtained for each identical initial speed vi, and the automatic brake system may be evaluated based on the sum of the scores. The scoring is increased as the risk of collision is reduced (for example, the scoring of a collision-free event is “5”, the scoring of a light collision event is “3”, and the scoring of a heavy collision event is “0”). . In this case, the collision avoidance capability of the automatic brake system increases as the sum of the scores of each event at the same initial speed vi increases, and the collision avoidance capability of the automatic brake system decreases as the sum decreases. It can be said.

  The automatic brake system evaluation program, the evaluation device, and the evaluation method described above can uniformly evaluate the performance of the automatic brake system regardless of the method.

DESCRIPTION OF SYMBOLS 10 Automatic brake system 11 Distance sensor 12 Collision sensor 13 Speed sensor 14 Control part 15 Memory | storage device 20 Vehicle 21 Engine 22 Accelerator pedal 23 Tire 24 Drive shaft 25 Brake mechanism 27 Handle 30 Obstacle 40 Evaluation apparatus 40A Regression analysis execution part 40B Evaluation 50, 60 regression line

Claims (9)

An evaluation program for an automatic braking system executed by a computer having at least a CPU,
The CPU
A reading unit that reads from the storage means a test result of whether or not a vehicle equipped with the automatic brake system collides with an obstacle by the operation of the automatic brake system;
A regression analysis execution unit that performs linear regression analysis using a one-dimensional model based on the result of the test read by the reading unit;
And an evaluation unit that evaluates the performance of the automatic braking system based on the result of the linear regression analysis,
The explanatory variable in the linear regression analysis is the relative speed vi of the vehicle to the obstacle before the automatic braking system is activated,
The objective variable in the linear regression analysis is a relative speed vc to the obstacle when the vehicle collides with the obstacle,
The results of the test include the relative speed vi and the relative speed vc for the vehicle at the relative speed vi;
The evaluation unit evaluates that the performance of the automatic braking system is high as the vi intercept of the regression line obtained by the linear regression analysis increases.
The automatic brake system includes:
An obstacle sensor for detecting the obstacle;
A determination unit that performs a determination process as to whether or not the obstacle exists in the traveling direction of the vehicle based on a sensing signal from the obstacle sensor;
A controller that activates a brake of the vehicle when it is determined in the determination process that the obstacle is present.
An evaluation program for an automatic brake system characterized by An evaluation program for an automatic braking system executed by a computer having at least a CPU,
The CPU
A reading unit that reads from the storage means a test result of whether or not a vehicle equipped with the automatic brake system collides with an obstacle by the operation of the automatic brake system;
A regression analysis execution unit that performs linear regression analysis using a one-dimensional model based on the result of the test read by the reading unit;
And an evaluation unit that evaluates the performance of the automatic braking system based on the result of the linear regression analysis,
The explanatory variable in the linear regression analysis is the relative speed vi of the vehicle to the obstacle before the automatic braking system is activated,
The objective variable in the linear regression analysis is a relative speed vc to the obstacle when the vehicle collides with the obstacle,
The results of the test include the relative speed vi and the relative speed vc for the vehicle at the relative speed vi;
The evaluation unit evaluates that the performance of the automatic braking system is high as the slope of the regression line obtained by the linear regression analysis decreases,
The automatic brake system includes:
An obstacle sensor for detecting the obstacle;
A determination unit that performs a determination process as to whether or not the obstacle exists in the traveling direction of the vehicle based on a sensing signal from the obstacle sensor;
A controller that activates a brake of the vehicle when it is determined in the determination process that the obstacle is present.
An evaluation program for an automatic brake system characterized by An evaluation program for an automatic brake system executed by a computer having at least a CPU,
The CPU
A reading unit that reads from the storage means a test result of whether or not a vehicle equipped with the automatic brake system collides with an obstacle by the operation of the automatic brake system;
As a result of the test read by the reading unit, a collision non-occurrence event in which the vehicle did not collide with the obstacle, and a light collision occurrence in which the vehicle collides with the obstacle and an adverse effect due to the collision of the vehicle is small. Based on a predetermined threshold, it is determined whether an event or a heavy collision occurrence event in which the vehicle collides with the obstacle and an adverse effect due to the vehicle collision is larger than the light collision occurrence event. And a selection unit that assigns an identifier for identifying the non-collision event, the light collision occurrence event, and the heavy collision occurrence event to the result of the test based on the determination result;
A regression analysis execution unit that performs logistic regression analysis based on the result of the test to which the identifier is assigned;
An evaluation unit that evaluates the performance of the automatic braking system based on the result of the logistic regression analysis,
The explanatory variable in the logistic regression analysis is the relative speed vi of the vehicle to the obstacle before the automatic braking system is activated,
The objective variable in the logistic regression analysis is an occurrence probability of an event selected from the non-collision event, the light collision event, and the heavy collision event,
The regression analysis execution unit includes an integration value S0 of the occurrence probability of the heavy collision occurrence event, an integration value S1 of the occurrence probability of the light collision occurrence event, and the non-collision occurrence in the integration interval determined by the relative speed vi. An integrated value S2 of the occurrence probability of the event is calculated as a result of the logistic regression analysis,
The evaluation unit evaluates that the performance of the automatic brake system is high as the integral value S0 decreases or as the sum of the integral value S1 and the integral value S2 increases. An evaluation program for automatic braking systems. The automatic brake system includes:
An obstacle sensor for detecting the obstacle;
A determination unit that performs a determination process as to whether or not the obstacle exists in the traveling direction of the vehicle based on a sensing signal from the obstacle sensor;
The automatic brake system evaluation program according to claim 3 , further comprising: a control unit that operates a brake of the vehicle when it is determined that the obstacle is present in the determination process. 5. The automatic brake system evaluation program according to claim 3 , wherein the predetermined threshold value is a value of a relative speed vc set based on a probability of occurrence of a phenomenon induced by the obstacle. 6. The automatic brake system evaluation program according to claim 5 , wherein the phenomenon is a simplified fault scale. 5. The automatic brake system evaluation according to claim 3 , wherein the predetermined threshold value is a value of a relative speed vc set based on a head failure index related to a phenomenon induced by the obstacle. program. The evaluation program for an automatic brake system according to any one of claims 3 to 7 , wherein the CPU functions to include a threshold setting unit that sets the threshold. Evaluation device for an automatic brake system comprising the evaluation program of the automatic braking system of any one of claims 1 to 8.

Images