JP2017058888A - Vehicle control device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately calculate a final preceding vehicle probability by reducing a risk of erroneous determination of a preceding vehicle to follow.SOLUTION: A vehicle control device optimistically estimates a probability that another vehicle is a preceding vehicle when: a first index value e1, obtained by normalizing an absolute value |A-B| where A is a difference between a preceding vehicle speed at a current time and the same at a past time first minute time before the current time and B is the difference between a communication vehicle speed at the current time and the same at the past time the first minute time before the current time, is equal to or less than a second index value e2, obtained by normalizing the absolute value |C-D| where C is the difference between the preceding vehicle speed at the current time and the same at the past time second minute time before the current time and D is the difference between the communication vehicle speed offset time to the current time and the same at the past time the second minute time before the offset time to the current time; and a first comparison value obtained by normalizing the absolute value of the difference between variations of the preceding vehicle speed and the communication vehicle speed during the first minute time from a first time is equal to or less than a second comparison value obtained by normalizing the absolute value of the difference between the variation of the communication vehicle speed during the second minute time from a second time and the variation of the preceding vehicle speed during the second minute time after the predetermined time from the second time.SELECTED DRAWING: Figure 7

Description

  The present invention is information that travels immediately before the own vehicle and is transmitted by wireless communication from other vehicles that travel in the vicinity of the own vehicle and perform wireless communication (inter-vehicle communication) with the own vehicle. The present invention relates to a vehicle control apparatus that specifies a communication follow-up target vehicle to be followed while using the vehicle and causes the vehicle to follow the communication follow-up target vehicle.

Conventionally, so-called CACC control (Cooperative Adaptive Cruise Control) is known.
CACC control refers to information related to acceleration of a preceding vehicle while wireless communication (inter-vehicle communication) is performed with a preceding vehicle in which the vehicle (hereinafter sometimes referred to as “own vehicle”) is located immediately before itself. This is control for obtaining and coordinating the vehicle with the preceding vehicle using the information.
When a plurality of other vehicles are traveling around the vehicle, in order for the vehicle to execute the CACC control while communicating with the “preceding vehicle traveling immediately before the vehicle”, the vehicle must It is necessary to identify the preceding vehicle from among a plurality of other vehicles that are communicating, and to recognize the identified preceding vehicle as a communication follow-up target vehicle.

Patent document 1 is disclosing the vehicle control apparatus which can identify a communication follow-up object vehicle.
The own vehicle equipped with the vehicle control device of Patent Document 1 includes own vehicle speed detecting means for detecting the own vehicle speed, which is the speed of the own vehicle.
Furthermore, the own vehicle is provided with a radar sensor and wireless communication means.
The own vehicle receives the detection wave emitted from the radar sensor and reflected backward by the preceding vehicle. And a vehicle control apparatus calculates the relative speed with respect to the own vehicle of a preceding vehicle based on the received detection wave.
Further, the vehicle control device for the host vehicle calculates the preceding vehicle speed, which is the speed of the preceding vehicle, based on the host vehicle speed and the relative speed.
Furthermore, the own vehicle performs wireless communication with other vehicles located around the own vehicle using wireless communication means. The other vehicle, which is a communication vehicle that performs wireless communication with the own vehicle, transmits the communication vehicle speed, which is its own vehicle speed, to the own vehicle by wireless communication. Therefore, the own vehicle can acquire the communication vehicle speed which is the speed of the other vehicle (communication vehicle).

  Based on the acquired preceding vehicle speed and communication vehicle speed, the vehicle control device of the host vehicle calculates the similarity between the communication vehicle (other vehicle) and the preceding vehicle. Specifically, a plurality of similarities are calculated using the maximum speed mean square error of the preceding vehicle speed and the communication vehicle speed, the maximum acceleration fluctuation amount difference, the speed correlation coefficient, and the like. Each similarity is represented by a numerical value having a size between 0 and 1.

And a vehicle control apparatus calculates | requires the product of each similarity as a last precedence vehicle probability. The final preceding vehicle probability is represented by a numerical value having a size between 0 and 1.
The closer the final preceding vehicle probability is to 1, the higher the probability that the communication vehicle (other vehicle) is the preceding vehicle. The closer the final preceding vehicle probability is to 0, the lower the probability that the communication vehicle is the preceding vehicle.
Therefore, when the final preceding vehicle probability is close to 1, the vehicle control device recognizes that the communication vehicle with which the vehicle is performing wireless communication is the preceding vehicle, and uses this preceding vehicle (communication vehicle) as the communication follow-up target vehicle. Identify. And a vehicle control apparatus carries out CACC control of the own vehicle, performing radio | wireless communication with this communication follow-up object vehicle (preceding vehicle).

JP 2013-228804 A

By the way, when the preceding vehicle is performing CACC control on the preceding preceding vehicle located immediately before the preceding vehicle, the time-series data (speed change history) of the speed between the preceding preceding vehicle and the preceding vehicle is obtained. May be very similar.
For this reason, in such a case, Patent Document 1 may misidentify the preceding preceding vehicle as a communication follow-up target vehicle even though the preceding preceding vehicle is not originally a communication follow-up target vehicle for the own vehicle.

  The present invention has been made to address the above-described problems. That is, one of the objects of the present invention is to reduce the possibility of erroneously determining that a preceding vehicle that is a non-communication vehicle positioned immediately after a preceding preceding vehicle that is a communication vehicle is a communication follow-up target vehicle, An object of the present invention is to provide a vehicle control device capable of specifying a communication follow-up target vehicle with higher accuracy.

In order to achieve the above object, a vehicle travel control device of the present invention comprises:
Own vehicle speed detecting means (42) for detecting the own vehicle speed which is the speed of the own vehicle (10);
Relative speed acquisition means (60, 61) for acquiring the relative speed of the preceding vehicle (11) that travels immediately before the host vehicle and the host vehicle should follow the host vehicle;
Preceding vehicle speed calculation means (20) for calculating a preceding vehicle speed, which is the speed of the preceding vehicle, based on the host vehicle speed and the relative speed;
Communication vehicle speed acquisition means for acquiring other vehicle communication information including communication vehicle speed that is the speed of the other vehicle from the other vehicle by performing wireless communication with the other vehicles (11 to 13) existing around the host vehicle ( 80, 81),
Similarity calculation means (20) for calculating the similarity between the other vehicle and the preceding vehicle based on the preceding vehicle speed and the communication vehicle speed;
Precedence of calculating the probability that the other vehicle is the preceding vehicle to be higher as the similarity increases, and determining whether the other vehicle is the preceding vehicle based on the calculated probability Vehicle determination means (20);
By controlling the acceleration of the host vehicle based on the other vehicle communication information acquired by wireless communication with the other vehicle determined to be the preceding vehicle, the host vehicle is followed by the other vehicle. Traveling control means (20) for traveling;
In a vehicle control device comprising:
The preceding vehicle specifying means is
The difference (A) (dVfr) between the present value of the preceding vehicle speed and the past value before the first minute time from the present of the preceding vehicle speed, the present value of the communication vehicle speed, and the present of the communication vehicle speed A first index value (e1) obtained by normalizing the difference (B) (dVc) from the previous value before the first minute time and the absolute value (| AB) of the difference;
The difference (C) (dVfr ′) between the present value of the preceding vehicle speed and the past value of the preceding vehicle speed before the second minute time, the past value of the communication vehicle speed before the offset time, and the Normalize the absolute value (| C−D |) of the difference between the difference (D) (dVc ′) and the past value before the second minute time from the time before the offset time from the current communication vehicle speed. The second index value (e2)
To calculate
When the first index value is less than or equal to the second index value, the probability is calculated to be higher than when the first index value is greater than the second index value.

The difference (A) between the present value of the preceding vehicle speed and the past value before the first minute time from the present value of the preceding vehicle speed, and the past value before the first minute time from the present value of the communication vehicle speed and the present value of the communicating vehicle speed. The first index value (e1) obtained by normalizing the difference (B) and the absolute value (| A-B |) of the difference between the current value of the preceding vehicle speed and the current value of the preceding vehicle speed is the second minute time. The difference (C) between the previous past value and the past value before the offset time from the current communication vehicle speed and the past value before the second minute time from the time before the offset time from the current communication vehicle speed ( D) and the second index value (e2) obtained by normalizing the absolute value of the difference (| C−D |) is different from the other vehicle, which is the communication vehicle, is the first preceding vehicle and the preceding vehicle is first. It is empirically known that there is a high possibility of following the vehicle. Therefore, if the probability that another vehicle that is a communication vehicle is a preceding vehicle is not calculated while reflecting this empirical fact, the obtained probability may be an inaccurate value.
However, the preceding vehicle specifying means of the present invention is configured to calculate the probability higher when the first index value is less than or equal to the second index value compared to when the first index value is greater than the second index value. ing.
Therefore, the preceding vehicle specifying means of the present invention reduces the possibility that the preceding vehicle located immediately after the preceding preceding vehicle that is the communication vehicle is erroneously determined to be the communicating vehicle, while the other vehicle that is the communicating vehicle is the preceding vehicle. A certain probability can be calculated with high accuracy.

In the above description, in order to help the understanding of the invention, the reference numerals used in the embodiments are attached to the configuration of the invention corresponding to the embodiments in parentheses, but each constituent element of the invention is represented by the reference numerals. It is not limited to the embodiments specified.
Other objects, other features, and attendant advantages of the present invention will be readily understood from the description of each embodiment of the present invention described with reference to the following drawings.

It is a figure which shows the own vehicle and other vehicle which mount the vehicle control apparatus which concerns on one Embodiment of this invention. It is a flowchart which shows the process for the vehicle control apparatus of the own vehicle to acquire the communication vehicle speed, the preceding vehicle speed, and the relative speed of the own vehicle and the preceding vehicle. It is a flowchart showing the process for a vehicle control apparatus to specify a communication vehicle. It is a flowchart showing the process for a vehicle control apparatus to calculate speed fluctuation amount SHn. It is a graph which shows the relationship between (DELTA) Vfr and the correction coefficient k1. (A) and (b) are graphs showing changes of the communication vehicle speed Vcb, the preceding vehicle speed Vfr, and the relative speed Vr over time. It is a flowchart showing the process for a vehicle control apparatus to calculate non-preceding preceding vehicle degree apn. It is a graph which shows the change with time progress of communication vehicle speed Vcb and preceding vehicle speed Vfr. It is an example of the look-up table (map) utilized in order that a vehicle control apparatus calculates | requires the correction coefficient fn. It is a flowchart for a vehicle control apparatus to perform the averaging process of a mean square error. It is a flowchart showing the process performed before a vehicle control apparatus performs the averaging process of a speed correlation function. It is a graph showing the change with time progress of the threshold value Pth.

  Hereinafter, a vehicle control device according to an embodiment of the present invention will be described with reference to the drawings. First, main terms used in the present specification, drawings, and claims will be described.

・ Own vehicle: Own vehicle (vehicle of interest)
・ Other vehicles: Vehicles other than the own vehicle ・ Previous vehicles: Other vehicles and other vehicles traveling immediately before the own vehicle captured by a sensor (own vehicle radar sensor, that is, relative information acquisition means) included in the own vehicle Communication information: Information related to other vehicles acquired by the subject vehicle through wireless communication (inter-vehicle communication) between the own vehicle and other vehicles. Communication vehicle: Other vehicle / communication follow-up that transmits other vehicle communication information. Target vehicle: By performing wireless communication with the own vehicle, the own vehicle should control the acceleration of the own vehicle based on other vehicle communication information obtained by the wireless communication, and therefore the own vehicle should follow Leading vehicle

  As will be described later, it can be said that the vehicle control device according to the embodiment of the present invention includes a device that specifies a communication follow-up target vehicle from among other vehicles (that is, a communication follow-up target vehicle specification device). Further, the other vehicle will be described as including a vehicle control device similar to the “vehicle control device mounted on the host vehicle”.

(Constitution)
As shown in FIG. 1, the vehicle control device according to the embodiment of the present invention is mounted on a host vehicle 10.

  The own vehicle 10 includes a vehicle control ECU 20, an engine control ECU 30, a brake control ECU 40, a steering control ECU 50, a sensor ECU 60, a GPS device 70, and a radio control ECU 80. These ECUs are capable of exchanging data (communicable) via a communication / sensor system CAN (Controller Area Network) 101. The ECU is an abbreviation for an electric control unit and is an electronic control circuit having a microcomputer including a CPU, a ROM, a RAM, an interface, and the like as main components. The CPU implements various functions to be described later by executing instructions stored in a memory (ROM).

  The vehicle control ECU 20 is connected to a “plurality of vehicle control sensors 21” and a CACC switch 22 other than the sensors described later, and receives signals from these sensors 21 and the switch 22.

  The CACC switch 22 is an ON-OFF switch that is operated by a passenger of the host vehicle 10. The CACC switch 22 outputs a CACC request signal when its position is set to the ON position.

  The engine control ECU 30 is connected to an accelerator operation amount sensor 31 and a plurality of other engine control sensors (not shown), and receives detection signals from these sensors.

  The accelerator operation amount sensor 31 detects an operation amount (hereinafter referred to as “accelerator operation amount”) AP of an accelerator pedal 91 as an accelerator operation element, and outputs a signal representing the accelerator operation amount AP.

  The engine control ECU 30 is connected to an engine actuator 32 such as a throttle valve actuator and a fuel injection valve. The engine control ECU 30 drives the engine actuator 32 to change the torque generated by an engine (not shown) and adjust the acceleration of the host vehicle 10.

  The brake control ECU 40 is connected to a brake operation amount sensor 41, a vehicle speed sensor 42, and a plurality of other brake control sensors (not shown), and receives detection signals from these sensors.

The brake operation amount sensor 41 detects an operation amount (hereinafter referred to as “brake operation amount”) BP of a brake pedal 93 as a brake operation element, and outputs a signal representing the brake operation amount BP.
The vehicle speed sensor 42 detects the speed of the host vehicle (own vehicle speed) Vj and outputs a signal representing the host vehicle speed Vj.

  The brake control ECU 40 is connected to a brake actuator 43 including a hydraulic control device. The brake actuator 43 is arranged in a hydraulic circuit (both not shown) between a master cylinder that pressurizes hydraulic oil by the depression force of the brake pedal 93 and a friction brake device including a well-known wheel cylinder provided on each wheel. Established. The brake actuator 43 adjusts the hydraulic pressure supplied to the wheel cylinder. The brake control ECU 40 generates a braking force on each wheel by driving the brake actuator 43 to adjust the acceleration (negative acceleration, that is, deceleration) of the vehicle 10.

  The steering control ECU 50 is connected to a steering angle sensor 51 that detects the steering angle α of the steering wheel of the host vehicle 10 and a plurality of other steering control sensors (not shown), and receives detection signals from these sensors. It has become.

  The steering control ECU 50 is connected to a steering actuator 52 which is a motor of an electric power steering device (not shown), and drives the steering actuator 52.

  The sensor ECU 60 is connected to the own vehicle radar sensor 61. The own vehicle radar sensor 61 is a well-known millimeter wave radar sensor. The own vehicle radar sensor 61 transmits a millimeter wave in front of the own vehicle 10 in accordance with an instruction from the sensor ECU 60. The millimeter wave is reflected by the preceding vehicle 11. The own vehicle radar sensor 61 receives this reflected wave.

  The sensor ECU 60 is based on the phase difference between the millimeter wave transmitted from the vehicle radar sensor 61 and the received reflected wave, the attenuation level of the reflected wave, the time from when the millimeter wave is transmitted until the reflected wave is received, and the like. The relative speed Vr, the inter-vehicle distance Dr, the relative azimuth θp, and the like are calculated and acquired every predetermined time. The sensor ECU 60 stores (stores) the relative speed Vr, the inter-vehicle distance Dr, the relative azimuth θp, and the like in its RAM in time series. The “information (data) including the relative speed Vr, the inter-vehicle distance Dr, the relative orientation θp, etc.” acquired by the own vehicle radar sensor 61 and the sensor ECU 60 is also referred to as “own vehicle sensor information”.

The vehicle control ECU 20 acquires the preceding vehicle speed Vfr of the preceding vehicle 11 by adding the host vehicle speed Vj and the relative speed Vr. In other words, the relative speed Vr is a difference (= Vfr−Vj) between the own vehicle speed Vj of the own vehicle 10 and the preceding vehicle speed Vfr of the preceding vehicle 11.
The inter-vehicle distance Dr is a distance between the host vehicle 10 and the preceding vehicle 11.
The relative azimuth θp is an angle (relative azimuth) of the traveling direction of the preceding vehicle 11 with respect to the traveling direction of the own vehicle 10 with respect to the position of the own vehicle 10.

  The GPS device 70 is well known, and acquires a position (own vehicle position) where the host vehicle 10 is traveling based on a GPS signal transmitted from a GPS satellite every time a predetermined time elapses, and acquires the acquired position in its RAM. Are stored in chronological order. The position of the host vehicle 10 is specified by longitude X and latitude Y.

  The wireless control ECU 80 is connected to a wireless antenna 81 for performing wireless communication (inter-vehicle communication) with other vehicles. The radio control ECU 80 transmits other vehicle communication information, information on other vehicles (hereinafter also referred to as “other vehicle communication information”) transmitted from other vehicles (other vehicles 11 to 13 in FIG. 1). It is received every time a predetermined time elapses together with an ID for identifying another vehicle (another vehicle ID). The radio control ECU 80 stores information received by inter-vehicle communication in the RAM for each other vehicle ID and in time series.

The other vehicle communication information includes the following information indicating the driving state of the other vehicle (that is, the communication vehicle).
(A) Vehicle speed (communication vehicle speed) Vcb of the communication vehicle acquired by the brake control ECU 40 of the communication vehicle.
(B) The position Pc of the communication vehicle acquired by the GPS device 70 of the communication vehicle.

(C) The vehicle control ECU 20 of the communication vehicle when the vehicle control device of the communication vehicle is not executing any of “coordinated follow-up running control (CACC) and inter-vehicle distance control (ACC: Adaptive Cruise Control)” described later. Is a required acceleration Gc that is an estimated acceleration of the communication vehicle calculated based on the “accelerator operation amount AP and brake operation amount BP” of the communication vehicle.
(D) When the vehicle control device of a communication vehicle is executing any one of the “coordinated follow-up running control and the inter-vehicle distance control”, the calculation is performed to perform the control (request to the communication vehicle) Required acceleration Gc which is acceleration.
(E) The actual acceleration Ga (= dVcb / dt) of the communication vehicle obtained by time-differentiating the vehicle speed (other vehicle speed) Vcb of the communication vehicle by the vehicle control ECU 20 of the communication vehicle.

  The radio control ECU 80 transmits (transmits) the other vehicle communication information about the host vehicle 10 to the outside for a subsequent vehicle (a vehicle traveling behind the host vehicle 10) every time a predetermined time elapses. It is like that.

(Basic traveling control of own vehicle 10)
Next, a basic travel control method of the host vehicle 10 by the vehicle control ECU 20 will be described.

  When the CACC switch 22 of the host vehicle 10 is in the off position, the travel mode of the host vehicle 10 is a normal travel mode (non-automatic operation mode). In the normal travel mode, the engine control ECU 30 controls the engine actuator 32 according to the accelerator pedal operation amount AP, and the brake control ECU 40 controls the brake actuator 43 based on the brake operation amount BP. Therefore, the host vehicle 10 travels according to the driving operation of the driver.

  When the occupant of the host vehicle 10 changes the position of the CACC switch 22 to the ON position, the traveling mode of the host vehicle 10 becomes the automatic driving mode. In the automatic operation mode, when the vehicle control ECU 20 of the host vehicle 10 specifies the preceding vehicle 11 as a communication follow-up target vehicle (communication preceding vehicle), the vehicle travels following the preceding vehicle 11 in cooperation.

  More specifically, the own vehicle sensor 61 and the sensor ECU 60 acquire the own vehicle sensor information (such as “relative speed Vr and inter-vehicle distance Dr” between the own vehicle 10 and the preceding vehicle 11) at a predetermined period, The vehicle sensor information is transmitted to the vehicle control ECU 20.

  Further, the vehicle control ECU 20 acquires (calculates) the host vehicle speed Vj of the host vehicle 10 based on the signal from the vehicle speed sensor 42, and the inter-vehicle time between the host vehicle 10 and the preceding vehicle 11 (= inter-vehicle distance Dr / own vehicle speed Vj) Tj. Is calculated.

  Further, the vehicle control ECU 20 requests an FB request based on the target inter-vehicle time Tj and the predetermined target inter-vehicle time Ttgt as a target acceleration that is an acceleration for making the difference between the calculated inter-vehicle time Tj and the predetermined target inter-vehicle time Ttgt zero. Calculated as G (feedback required acceleration). The target inter-vehicle time Ttgt is a value obtained by dividing the target inter-vehicle distance Drtgt by the own vehicle speed Vj. Therefore, the vehicle control ECU 20 calculates the FB request G for making the difference between the inter-vehicle distance Dr and the target inter-vehicle distance Drtgt zero. More specifically, since acceleration is required if the inter-vehicle distance L is longer than the target inter-vehicle distance Ltgt, the FB request G is increased by a predetermined positive value. Conversely, if the inter-vehicle distance L is shorter than the target inter-vehicle distance Ltgt, deceleration is necessary, so the FB request G is decreased by a predetermined positive value. The target inter-vehicle distance Drtgt may be a constant value, or may be a variable value selected by the occupant of the host vehicle 10 with a setting switch (not shown).

  Further, the vehicle control ECU 20 acquires the “required acceleration Gc from the communication vehicle and the actual acceleration Ga of the communication vehicle” received using the wireless control ECU 80 and the wireless antenna 81, and requests the FF based on the required acceleration Gc and the actual acceleration Ga. G (feed forward required acceleration) is calculated. More specifically, in this example, the vehicle control ECU 20 determines the FF request G based on the sum of a value obtained by subjecting the requested acceleration Gc to a high-pass filter process and a value obtained by subjecting the actual acceleration Ga to a low-pass filter process. Is calculated.

  Then, the vehicle control ECU 16 calculates the sum of the FB request G and the FF request G as a target acceleration (that is, the CACC request G) that is finally required for the host vehicle 10. The vehicle control ECU 16 may calculate the weighted average value of the FF request G and the FB request G as the target acceleration Gtgt.

  Further, the vehicle control ECU 20 transmits a request command signal based on the calculated CACC request G to the engine control ECU 30 and the brake control ECU 40 via the CAN 101. When receiving the request command signal, the engine control ECU 30 and the brake control ECU 40 control the engine actuator 32 and the brake actuator 43, respectively, according to the request command signal. As a result, the acceleration of the host vehicle 10 is controlled to match the CACC request G. The above is the basic operation in CACC control.

  Further, when the vehicle control ECU 20 does not identify the preceding vehicle 11 as a communication follow-up target vehicle (communication vehicle), the vehicle control ECU 20 determines the own vehicle 10 based on ACC control (Adaptive Cruise Control) that is control based only on the FB request G. To control.

(Identification of vehicles subject to communication follow-up)
As described above, the own vehicle 10 can execute the CACC control in which the preceding vehicle 11 is the communication follow-up target vehicle only when the preceding vehicle 11 is a communication vehicle. In other words, when the communication vehicle is a vehicle other than the preceding vehicle 11, the host vehicle 10 cannot execute CACC control. Therefore, the own vehicle 10 determines whether the communication vehicle is the preceding vehicle 11 by the following method.

  The CPU (hereinafter simply referred to as CPU) of the vehicle control ECU 20 of the host vehicle 10 identifies the preceding vehicle 11 from among a plurality of other vehicles according to the flowcharts of FIGS.

  First, the processing of the flowchart of FIG. 2 will be described. When the CACC switch 22 of the host vehicle 10 is in the ON position, the CPU of the host vehicle 10 repeatedly executes the routine shown in the flowchart of FIG. 2 every predetermined time (for example, several milliseconds).

  In step 201, the host vehicle 10 uses the radio control ECU 80 and the radio antenna 81 to acquire other vehicle communication information from the other vehicles 11 to 13 which are communication vehicles. The other vehicle communication information includes the communication vehicle speed Vcb of each of the other vehicles 11-13.

  Subsequently, in Step 202, the host vehicle 10 acquires host vehicle sensor information using the host vehicle sensor 61 and the sensor ECU 60. The own vehicle sensor information includes the relative speed Vr of the preceding vehicle 11, the inter-vehicle distance Dr between the own vehicle 10 and the preceding vehicle 11, and the like. Further, in step 202, the CPU calculates the preceding vehicle speed Vfr by adding the own vehicle speed Vj and the relative speed Vr.

  Subsequently, the host vehicle 10 stores the communication vehicle speed Vcb, the relative speed Vr, and the preceding vehicle speed Vfr in the buffer memory of the RAM of the vehicle control ECU 20 in step 203.

Further, the CPU proceeds to step 204 to correct the communication vehicle speed Vcb stored in the buffer memory. Specifically, the speed bias (Vb (t)) is calculated using the following formula, and the speed bias (Vb (t)) is subtracted from the communication vehicle speed Vcb stored in the buffer memory. The speed bias is a steady error between the vehicle speed Vj calculated based on the signal from the vehicle speed sensor 42 and the actual (true) vehicle speed. This error is caused by a deviation from a design value such as the tire diameter of the host vehicle 10.

Vb (t) = k * {Vcb (t-dt) -Vfr (t-dt)} + (1-k) * Vb (t-dt)
k: filter coefficient dt: calculation cycle t: current time

Hereinafter, the communication vehicle speed corrected in this manner is referred to as communication vehicle speed (communication vehicle speed after bias correction) Vc.

Next, the flowchart of FIG. 3 will be described. When the CACC switch 22 of the host vehicle 10 is in the ON position, the CPU of the host vehicle 10 repeatedly executes the routine shown in the flowchart of FIG. 3 every predetermined time (for example, several milliseconds).
The detailed description of this flowchart will be described later. First, an outline of the basic technical idea of this flowchart will be described.

In steps 306, 309, and 312, based on the communication vehicle speed Vc and the preceding vehicle speed Vfr acquired during the processing of the flowchart of FIG. 2, the CPU calculates the similarity g1n based on the mean value H2E of the mean square error and the speed average deviation degree. The similarity g2n based on the average value He1 of e1 and the similarity g3n based on the average value Hcf of the velocity correlation coefficient coef are calculated.
These similarities g1n, g2n, and g3n are all numerical values representing the similarity between the preceding vehicle 11 and the communication vehicle, that is, the possibility that the communication vehicle is the preceding vehicle 11. Note that n is a number from 1 to 3. That is, three similarities are calculated for each of the similarities g1n, g2n, and g3n. The magnitudes of the similarities g1n, g2n, and g3n are all 0 or more and 1 or less.

As described above, in the conventional example, the final preceding vehicle probability is obtained by multiplying the obtained similarities g1n, g2n, and g3n. That is, the final preceding vehicle probability is obtained without considering the amount of change per predetermined time of the communication vehicle speed Vc and the preceding vehicle speed Vfr.
However, the final preceding vehicle probability calculated in this way is that the communication vehicles other than the preceding vehicle are traveling at substantially the same constant speed, particularly when the preceding vehicle is traveling at a substantially constant speed. Then, the degree of similarity in speed between the communication vehicle other than the preceding vehicle and the preceding vehicle is increased, and thus the reliability of the final preceding vehicle probability is decreased. That is, there is a high possibility that a communication vehicle other than the preceding vehicle is erroneously recognized as the preceding vehicle.

Therefore, in the present embodiment, in steps 301 to 303 in the flowchart of FIG. 3, the correction coefficient (correction coefficient of the preceding vehicle probability) fn is calculated based on the fluctuation amounts of the communication vehicle speed Vc, the relative speed Vr, and the preceding vehicle speed Vfr. . More specifically, in steps 301 and 302, the speed change amount SHn of the own vehicle 10, the preceding vehicle 11 and the communication vehicle (hereinafter simply referred to as “speed change amount SHn”) and the non-preceding preceding vehicle degree of the communication vehicle ( A priori) apn (hereinafter simply referred to as “non-preceding preceding vehicle degree”) is calculated. Based on the speed fluctuation amount SHn and the non-preceding preceding vehicle degree apn, a correction coefficient fn is calculated in step 303. Similarly to the similarities g1n, g2n, and g3n, three types of correction coefficients fn (f1, f2, and f3) are calculated.
The correction coefficient fn is a numerical value calculated in consideration of the influence of the fluctuation amount of the communication vehicle speed Vc, the relative speed Vr, and the preceding vehicle speed Vfr on the final preceding vehicle probability Pn, and the magnitudes thereof are all 0 or more and 1 or less. The correction coefficient fn is basically calculated so as to approach “1” as the speed fluctuation amount SHn increases and to approach “1” as the non-preceding preceding vehicle degree apn increases.

  In step 313, the CPU calculates the final preceding vehicle probability Pn (= P1, P2, P3) by multiplying the similarities g1n, g2n, g3n and the same value of n of the correction coefficient fn. Further, when the largest value of the determined final preceding vehicle probabilities Pn exceeds a predetermined threshold value Pth, the CPU determines that “the similarity g1n, g2n, g3n and the correction coefficient that are the basis of the final leading vehicle probability Pn” It is determined that the communication vehicle that has transmitted other vehicle communication information including the original data of fn (communication vehicle speed Vcb and the like) is the preceding vehicle 11 ". The final preceding vehicle probability Pn is a numerical value of 0 or more and 1 or less. The closer to “1”, the higher the probability that the communication vehicle is the preceding vehicle 11, and the closer to “0” the “communication vehicle”. Represents a low probability that the vehicle is a preceding vehicle 11 ”.

Thus, the reliability of the final preceding vehicle probability Pn calculated by the CPU in consideration of not only the similarities g1n, g2n, g3n but also the correction coefficient fn is high. In other words, the final leading vehicle probability Pn more accurately represents the probability that the communication vehicle is the leading vehicle 11 than the final leading vehicle probability calculated based only on the similarities g1n, g2n, and g3n.
Accordingly, when the CPU determines that the communication vehicle is the preceding vehicle 11 based on the final preceding vehicle probability Pn and the threshold value Pth obtained using the correction coefficient fn, the preceding vehicle 11 is actually a communication vehicle. Certain probabilities increase.

  Next, the processing content of the flowchart of FIG. 3 will be specifically described.

In step 301, the CPU calculates a speed fluctuation amount SHn.
The process of step 301 is performed according to the flowchart of FIG.
When the CACC switch 22 of the host vehicle 10 is in the ON position, the CPU of the host vehicle 10 repeatedly executes the routine shown in the flowchart of FIG. 4 every predetermined time (for example, several milliseconds).

  First, in step 401, the CPU sets all preceding vehicle speeds Vfr and relative speeds Vr stored in the buffer memory between the time when the automatic operation mode is entered and the present time, and the target communication vehicle (ie, the target communication vehicle). Each “maximum value Vfrmax, Vrmax and Vcmax” and “minimum value Vfrmin, Vrmin and Vcmin} are selected from the speeds Vc.

  Further, the CPU proceeds to step 402, and based on the maximum values Vfrmax, Vrmax and Vcmax and the minimum values Vfrmin, Vrmin and Vcmin, the preceding vehicle speed fluctuation amount ΔVfr = Vfrmax−Vfrmin, the communication vehicle speed fluctuation amount ΔVc = Vcmax. -Vcmin and relative velocity fluctuation amount ΔVr = Vrmax−Vrmin are calculated.

Further, the CPU proceeds to step 403 to determine whether or not ΔVfr × k1 (ΔVfr) ≦ ΔVc is established. As will be described later, the value k1 (ΔVfr) is a correction coefficient, and is a value obtained based on the fluctuation amount ΔVfr of the preceding vehicle speed according to the lookup table shown in FIG.
If the target communication vehicle is a preceding preceding vehicle (not shown) located immediately before the preceding vehicle 11 and the preceding vehicle 11 that is a non-communication vehicle is performing ACC control to follow the preceding preceding vehicle, It is known as an empirical fact that ΔVc of the preceding vehicle (communication vehicle) tends to be larger than ΔVfr. This is because, for example, the preceding vehicle starts to accelerate after the preceding preceding vehicle has accelerated.
That is, when ΔVfr × k1 (ΔVfr) ≦ ΔVc is established, it can be estimated that the possibility that the communication vehicle is the preceding vehicle is higher than the possibility that the communication vehicle is the preceding vehicle 11.

However, it is empirically known that when ΔVfr is large, the CPU is less likely to misidentify a preceding vehicle that is a non-communication vehicle as a communication vehicle. Therefore, in step 403, ΔVfr and ΔVc are not directly compared, but a correction coefficient k1 determined according to the magnitude of ΔVfr is multiplied by ΔVfr, and the magnitude relationship between this multiplied value and ΔVc is compared.
As shown in FIG. 5, when ΔVfr is a value between 0 and ΔVfr1, which is a predetermined size, k1 = 1. However, when ΔVfr exceeds ΔVfr1, k1 gradually increases, and when ΔVfr exceeds ΔVfr2, which is a predetermined magnitude, k1 becomes a constant value greater than 1.
Therefore, in the flowchart of FIG. 4, when ΔVfr exceeds ΔVfr1 that is a predetermined magnitude, there is a higher possibility that the communication vehicle is determined to be the preceding vehicle 11 than when it does not exceed ΔVfr1.

  If the CPU determines YES in step 403, in other words, if it is determined that “the communication vehicle is likely to be the preceding vehicle” based on the values of ΔVfr and ΔVc, the CPU proceeds to step 404. Then, the CPU calculates ΔVcinv = ΔVc−ΔVfr × k1 (ΔVfr), which is a value indicating the degree of “the communication vehicle seems to be the preceding vehicle”.

As shown in step 414 described later, the flowchart of FIG. 4 finally calculates the speed fluctuation amount SHn. Further, as will be described later, the value of the correction coefficient fn increases as the value of the speed fluctuation amount SHn increases. In other words, the greater the value of the speed fluctuation amount SHn, the higher the possibility that the CPU determines that “the communication vehicle is the preceding vehicle 11”.
ΔVcinv (when selected as ΔVinv in step 413 described later) is used in step 414 to reduce the value of the speed fluctuation amount SHn. In other words, ΔVcinv is used to reduce the possibility that the CPU determines that the communication vehicle is the preceding vehicle 11. Therefore, the larger the value of ΔVcinv obtained in step 404, the lower the possibility that the CPU will determine that “the communication vehicle is the preceding vehicle 11”.

On the other hand, if the CPU determines NO in step 403, in other words, if it is determined that the communication vehicle cannot be said to be the preceding vehicle, the CPU proceeds to step 405 and sets ΔVcinv to “0 (zero)”. Set.
As is clear from the above description, when ΔVcinv is set to 0 (and ΔVcinv is selected as ΔVinv in step 413), the CPU calculates a larger value of the speed fluctuation amount SHn in step 414. That is, there is a high possibility that the CPU determines that the communication vehicle is the preceding vehicle 11.

The CPU that has finished the processing of step 404 or 405 proceeds to step 406.
In step 406, the CPU sets dVrmax / dt> 0 (if Vrmax during the current step 406 processing is greater than Vrmax during the previous step 406 processing, dVrmax / dt> 0. Conversely, during the current step 406 processing. DVrmax / dt> 0 is not satisfied if Vrmax is equal to or lower than Vrmax at the time of the previous step 406.), it is determined whether all of dVfr / dt> 0 and dVc / dt <0 are satisfied. .
When the preceding vehicle 11 accelerates and the speed of the host vehicle 10 hardly changes, dVrmax / dt> 0 and dVfr / dt> 0 are established. On the other hand, dVc / dt <0 is established when the communication vehicle is decelerating. That is, when dVrmax / dt> 0, dVfr / dt> 0, and dVc / dt <0 are all established, the sign of acceleration of the preceding vehicle 11 (plus and minus) and the sign of acceleration of the communication vehicle are different from each other. .
Therefore, when all of dVrmax / dt> 0, dVfr / dt> 0, and dVc / dt <0 are satisfied, it can be estimated that “the communication vehicle is not like the preceding vehicle 11”.

If the CPU determines NO in step 406, the CPU proceeds to step 407 and determines whether or not all of dVrmax / dt <0, dVfr / dt <0, and dVc / dt> 0 are satisfied.
When the preceding vehicle 11 decelerates and the speed of the host vehicle 10 hardly changes, dVrmax / dt <0 and dVfr / dt <0 are established. On the other hand, dVc / dt> 0 is established when the communication vehicle is accelerating. That is, when dVrmax / dt <0, dVfr / dt <0, and dVc / dt> 0 are all established, the sign of the acceleration of the preceding vehicle 11 and the sign of the acceleration of the communication vehicle are different from each other.
Therefore, even when dVrmax / dt <0, dVfr / dt <0, and dVc / dt> 0 are all established, it can be estimated that “the communication vehicle is not like the preceding vehicle 11”.

  FIGS. 6A and 6B show changes of the preceding vehicle speed Vfr, the relative speed Vr, and the communication vehicle speed Vc assumed in steps 406 and 407 with time. As shown in the figure, the changes in the preceding vehicle speed Vfr, the relative speed Vr, and the communication vehicle speed Vc from time 0 to time ta in FIGS. 6A and 6B are the same. On the other hand, when the time ta elapses, the “preceding vehicle speed Vfr and relative speed Vr” and the “communication vehicle speed Vc” change in different ways.

  At times before time ta in the graphs of FIGS. 6A and 6B, dVrmax / dt> 0, dVfr / dt> 0, and dVc / dt> 0. That is, when the CPU performs the process of step 406 at a time before time ta, NO is determined in step 406 and the process proceeds to step 407. Further, when the CPU performs the process of step 407 at a time before time ta in FIGS. 6A and 6B, NO is determined in step 407 and the process proceeds to step 409.

On the other hand, in the graph of FIG. 6A, when time ta elapses, dVrmax / dt> 0, dVfr / dt> 0, and dVc / dt <0. Accordingly, when the preceding vehicle speed Vfr, the relative speed Vr, and the communication vehicle speed Vc change as shown in FIG. 6A, when the CPU performs the process of step 406 at the time when the time ta has elapsed, the CPU determines YES. Then, the process proceeds to step 408.
In the graph of FIG. 6B, when time ta elapses, dVrmax / dt <0, dVfr / dt <0, and dVc / dt> 0. Therefore, when the preceding vehicle speed Vfr, relative speed Vr, and communication vehicle speed Vc change as shown in FIG. 6B, when the CPU performs the process of step 407 at the time when the time ta has elapsed, the CPU determines YES. Then, the process proceeds to step 408.

When the CPU proceeds to step 409, the CPU initializes the “invalid counter” (ie, sets it to “0”).
The invalid counter represents the degree of “the target communication vehicle is not likely to be the preceding vehicle 11” calculated based on the determination results of Steps 406 and 407. The larger the value, the “the target vehicle. The degree to which the communication vehicle is not like the preceding vehicle 11 increases.
Further, in step 409, the CPU sets the current ΔVr as ΔVrbase. That is, ΔVr at the current time before time ta in the graphs of FIGS. 6A and 6B is set as ΔVrbase.
As will be described later, this ΔVrbase is a value for obtaining ΔVrinv in step 411.

On the other hand, when the CPU proceeds to step 408, “1” is added to the “invalid counter”.
Further, in step 408, the CPU holds the previous value which is ΔVrbase obtained in step 409 or step 408 during the previous flowchart processing of FIG. 4 as ΔVrbase. Here, for example, when the processing time of the current step 408 is the time tb in the graphs of FIGS. 6A and 6B and the previous processing time is the time ta, the previous value of ΔVrbase is ΔVr at the time ta. Become. For example, when the current processing time of step 408 is time tc in the graphs of FIGS. 6A and 6B and the previous processing time is time tb, the previous value of ΔVrbase is ΔVrbase at time tb, that is, ΔVr at time ta. That is, ΔVrbase set in step 408 becomes ΔVr last acquired in step 409 before time ta.

The CPU that has undergone step 408 or 409 proceeds to step 410, and determines whether or not the number of invalid counters exceeds a predetermined threshold inv.
If the CPU determines that the threshold inv is exceeded, that is, if it is determined that the possibility that “the communication vehicle is not likely to be the preceding vehicle 11” can be estimated at a certain level, the CPU determines YES in step 410 and performs step Go to 411.

In step 411, the CPU calculates a value obtained by subtracting ΔVrbase obtained in step 409 or 408 from ΔVr at the present time as ΔVrinv.
This ΔVrinv is used to reduce the value of the speed fluctuation amount SHn in step 414, similarly to ΔVcinv (when it is selected as ΔVinv in step 413 described later). Therefore, as the value of ΔVrinv increases, the CPU finally calculates a smaller value of the speed fluctuation amount SHn in step 414.

For example, when the CPU performs steps 406 and 407 before time ta in FIGS. 6A and 6B, the CPU proceeds to step 410 after step 409, and further determines NO in step 410. .
As described above, in this case, the signs of the accelerations of the preceding vehicle speed Vfr and the communication vehicle speed Vc are the same. That is, since there is a possibility that the communication vehicle is actually the preceding vehicle 11, it is not necessary to reduce the value of the speed fluctuation amount SHn in step 414.
Therefore, in this case, the CPU proceeds to step 412 to set ΔVrinv to 0.

On the other hand, when the CPU performs steps 406 and 407 after the time ta in FIGS. 6A and 6B has elapsed, the CPU proceeds to step 410 after step 408. In this case, the CPU may determine YES in step 410. If the CPU determines YES, the CPU proceeds to step 411.
As described above, in this case, the signs of the accelerations of the preceding vehicle speed Vfr and the communication vehicle speed Vc are different from each other. That is, the possibility that “the communication vehicle is not likely to be the preceding vehicle 11” can be estimated at a certain level. Therefore, in step 414, the value of the speed fluctuation amount SHn needs to be reduced.
Therefore, in this case, ΔVrinv is not set to zero. That is, the CPU calculates ΔVrinv = ΔVr−ΔVrbase in step 411.

However, if the current ΔVr acquired after the time ta has passed is directly used as ΔVrbase in step 411, ΔVrinv becomes zero. In other words, the CPU can easily determine that the communication vehicle is the preceding vehicle 11 even though the possibility that the communication vehicle is not likely to be the preceding vehicle 11 can be estimated at a certain level. Therefore, in this case, the current ΔVr should not be used as ΔVrbase.
On the other hand, in the situation before time ta, the possibility that “the communication vehicle seems to be the preceding vehicle 11” can be estimated to some extent (that is, it cannot be said that the communication vehicle is not likely to be the preceding vehicle 11). ΔVr is calculated by subtracting ΔVr obtained in step S5 as ΔVrbase from current ΔVr. When ΔVrinv is calculated in this manner, the possibility that it is determined that “the communication vehicle is the preceding vehicle 11” in step 411 is higher than when ΔVrbase = 0. However, since this ΔVr is ΔVr in a time zone in which the possibility that “the communication vehicle seems to be the preceding vehicle 11” can be estimated to some extent, the determination result of whether or not the communication vehicle calculated using this ΔVrinv is the preceding vehicle 11 is It is considered that there is little risk of a major departure from
Therefore, in step 411 in this case, Vrinv = ΔVr−ΔVrbase is acquired after setting ΔVr last acquired in step 409 in the time zone before time ta to ΔVrbbase.

The CPU that has finished the processing of step 411 or step 412 proceeds to step 413, where ΔVcinv in the immediately preceding step 404, ΔVrinv in the immediately preceding step 411, and ΔVinv in the previous processing in step 413 are the most numerical values. Is selected as ΔVinv.
This ΔVinv is used in step 414 to reduce the value of the speed fluctuation amount SHn.
The reason for selecting the largest numerical value in step 413 is to make the speed fluctuation amount SHn obtained in step 414 as small as possible by setting ΔVinv to a large value. In other words, the possibility that the CPU erroneously determines that “the communication vehicle is the preceding vehicle 11” is reduced.

Further, the CPU proceeds to step 414 to calculate a value obtained by subtracting ΔVrinv from the minimum value among ΔVfr, ΔVr, and ΔVc as the speed fluctuation amount SHn. Accordingly, the larger the values of ΔVfr, ΔVr, and ΔVc, the larger the speed fluctuation amount SHn, and the larger the value of ΔVrinv, the smaller the speed fluctuation amount SHn.
The reason why the smallest numerical value is selected from ΔVfr, ΔVr, and ΔVc in step 414 is to make the finally calculated speed fluctuation amount SHn as small as possible. In other words, the possibility that the CPU erroneously determines that “the communication vehicle is the preceding vehicle 11” is reduced.

Further, if the following vehicle (not shown) located immediately after the own vehicle 10 performs wireless communication with the own vehicle 10, the CACC control is performed, and ΔVr between the preceding vehicle 11 and the own vehicle 10 which are non-communication vehicles is small. In this case, the speeds of the preceding vehicle 11 and the following vehicle are almost the same. In this situation, it becomes difficult for the CPU of the host vehicle 10 to distinguish between the preceding vehicle 11 and the following vehicle. Therefore, in such a situation, there is a high risk that the CPU erroneously determines that the communication vehicle (that is, the following vehicle) is the preceding vehicle 11. On the other hand, even when the subsequent vehicle communicates with the host vehicle 10 and executes the CACC control, if the ΔVr between the host vehicle 10 and the preceding vehicle 11 is large, the speeds of the preceding vehicle 11 and the following vehicle match. hard. Therefore, when ΔVr is large, the CPU of the own vehicle 10 can easily identify the preceding vehicle 11 and the following vehicle. Therefore, in such a situation, the risk that the CPU erroneously determines that the communication vehicle (that is, the following vehicle) is the preceding vehicle 11 is low.
In the flowchart of FIG. 4, ΔVr is a calculation factor of the speed fluctuation amount SHn in step 414. Therefore, when ΔVr is small, that is, when it is difficult for the CPU to distinguish between the preceding vehicle 11 and the following vehicle, the value of the speed fluctuation amount SHn calculated in step 414 tends to be small. On the other hand, when ΔVr is large, that is, when the CPU can easily identify the preceding vehicle 11 and the following vehicle, the value of the speed fluctuation amount SHn calculated in step 414 tends to be large.
In this way, by using ΔVr as a calculation factor of the speed fluctuation amount SHn in step 414, the difficulty of distinguishing the preceding vehicle 11 and the following vehicle due to the magnitude of ΔVr is reflected in the calculated value of the speed fluctuation amount SHn. . Therefore, it is possible to reduce a possibility that the CPU erroneously determines that “the communication vehicle is not the following vehicle but the preceding vehicle 11”.

As described above, the value of the correction coefficient fn increases (closer to “1”) as the speed fluctuation amount SHn obtained in step 414 (step 301) increases. In other words, the greater the value of the speed fluctuation amount SHn, the higher the possibility that it is finally determined that “the target communication vehicle is the preceding vehicle 11”.
Conversely, the smaller the speed fluctuation amount SHn, the smaller the value of the correction coefficient fn (closer to “0”). In other words, the smaller the value of the speed fluctuation amount SHn, the lower the possibility that it is finally determined that “the target communication vehicle is the preceding vehicle 11”.

The CPU that has finished the processing of step 301 in FIG. 3 (steps 401 to 414 in FIG. 4) proceeds to step 302 and calculates the non-preceding preceding vehicle degree apn.
The process of step 302 is performed according to the flowchart of FIG.

  When the CACC switch 22 of the host vehicle 10 is in the ON position, the CPU of the host vehicle 10 repeatedly executes the routine shown in the flowchart of FIG. 7 every predetermined time (for example, several milliseconds).

In step 701, the CPU calculates e1 (first index value) = | dVc−dVfr | / min (| dVc |, | dVfr |). Further, the CPU stores the calculated e1 in the buffer memory.
Here, dVc and dVfr are defined as follows.
dVc: Change amount of the communication vehicle speed Vc from the time (first time = t−t1) before the current time t by the minute time t1 to the current time (= Vc (t) −Vc (t−t1) ), See the graph of FIG.
dVfr: Change amount of the preceding vehicle speed Vfr from the time (first time = t−t1) before the current time t by the minute time t1 to the current time (= Vfr (t) −Vfr (t−t1) ), See the graph of FIG.

This e1 indicates the degree of deviation in speed change between the preceding vehicle 11 and the communication vehicle (hereinafter simply referred to as “degree of deviation in speed change”).
If the communication vehicle is the preceding vehicle and the preceding vehicle 11 travels following the preceding vehicle, a difference may occur between the dVc of the communication vehicle speed Vc and the dVfr of the preceding vehicle speed Vfr in the same time zone. Known empirically. Accordingly, the larger the value of the speed change divergence e1, in other words, the greater the value of | dVc−dVfr |, it is estimated that “the communication vehicle seems to be the preceding vehicle (that is, the communication vehicle is not likely to be the preceding vehicle 11)”. it can. On the other hand, the smaller the value of the speed change divergence e1, in other words, the smaller | dVc-dVfr | is, the less likely that the communication vehicle does not look like the preceding vehicle (that is, the communication vehicle seems to be the preceding vehicle 11). it can. Note that min (| dVc |, | dVfr |) is a smaller value of | dVc | and | dVfr |, and is a value for normalizing the deviation degree e1 of the speed change.

Subsequently, the CPU proceeds to step 702 to calculate e2 (second index value) = | dVc′−dVfr ′ | / min (| dVc ′ |, | dVfr ′ |). Further, the CPU stores the calculated e2 in the buffer memory.
Here, dVc ′ and dVfr ′ are defined as follows.
dVc ′: a time (= second time = t−toff−t2) before the current time t and a time before the current time t by the offset time toff (= second time = t−toff−t2) The amount of change in the communication vehicle speed Vc until (t−toff) (= Vc (t−toff) −Vc (t−toff−t2), see the graph of FIG. 8).
dVfr ′: the amount of change in the preceding vehicle speed Vfr from the time (= t−t2) before the minute time t2 from the current time t to the current time t (= Vfr (t) −Vfr (t−t2)) See the graph of FIG.
The minute time t2 may be different from the minute time t1, or may be the same length. The offset time toff changes when the preceding vehicle is a communication vehicle and the preceding vehicle is a non-communication vehicle that follows the preceding vehicle by ACC control. Is a delay time (for example, 0.5 to 2 s) from when the preceding vehicle speed Vfr starts to change, which is an experimentally determined value.

This e2 also indicates the degree of deviation in speed change between the preceding vehicle 11 and the communication vehicle (hereinafter simply referred to as “degree of deviation in speed change”).
If the communication vehicle is the preceding vehicle and the preceding vehicle 11 travels following the preceding vehicle, if the communication vehicle speed Vc changes by dVc 'in a certain time zone, the preceding vehicle will pass after the offset time toff has elapsed. It is empirically known that the velocity Vfr changes by (almost) the same as dVc ′. Accordingly, the smaller the value of the speed change divergence e2, in other words, the closer | dVc′−dVfr ′ | On the other hand, it can be estimated that the greater the value of the speed change divergence e2, in other words, the greater | dVc'-dVfr '| Note that min (| dVc ′ |, | dVfr ′ |) is a smaller value of | dVc ′ | and | dVfr ′ |, and is a value for normalizing the deviation degree e2 of the speed change.

  Subsequently, in step 703, the CPU obtains the speed change deviation degree e1 and the speed change deviation degree e2 that are currently stored in the buffer memory.

Subsequently, the CPU proceeds to step 704 to compare the magnitude relationship between the deviation degree e1 of the speed change and the deviation degree e2 of the speed change calculated in the immediately preceding steps 701 and 702, respectively.
As described above, it can be estimated that the smaller the value of the speed change divergence e1, the “communication vehicle is not like the preceding vehicle”. Further, it can be inferred that “the communication vehicle is not like the preceding vehicle” as the value of the deviation e2 of the speed change is larger.
Therefore, when the deviation degree e1 of the speed change ≦ the deviation degree e2 of the speed change is established, it can be estimated that “the communication vehicle is not like the preceding vehicle”.
On the other hand, when the deviation degree e1 of the speed change ≦ the deviation degree e2 of the speed change is not established, it can be estimated that “the communication vehicle seems to be the preceding vehicle”.

  If the CPU determines YES in step 704, the CPU proceeds to step 705 to determine whether or not the communication data between the vehicle 10 and the communication vehicle has been updated.

If the CPU determines YES in step 705, the CPU proceeds to step 706 to add “1” to the non-preceding preceding vehicle degree counter.
This non-preceding preceding vehicle degree counter represents the degree of “the communication vehicle is not likely to be the preceding preceding vehicle” calculated based on the deviation degree e1 of the speed change and the deviation degree e2 of the speed change, and the numerical value can be increased. The higher the degree that “the communication car is not like the preceding car” increases.

Subsequently, the CPU proceeds to step 707 to determine whether or not the number (the number of buffers) of the speed change divergence e1 and the speed change divergence e2 stored in the buffer memory is equal to or greater than a predetermined value.
If the number of speed change divergence e1 and speed change divergence e2 is equal to or greater than a predetermined value, a sufficient number of speed change divergence e1 and speed to determine the degree of divergence between the preceding vehicle 11 and the communication vehicle. It can be determined that the change divergence e2 has already been acquired.
If it is determined NO in either step 704 or step 705, the CPU proceeds directly to step 707 without passing through step 706.

  If the CPU determines YES in step 707, the CPU proceeds to step 708 to calculate non-preceding preceding vehicle degree apn = non-preceding preceding vehicle degree count / buffer data number. That is, the CPU has a speed change divergence e1 ≦ speed change divergence in all the pairs of speed change divergence e1 and speed change divergence e2 acquired up to the current time. It is calculated whether there exists a pair that satisfies the condition of e2. As the numerical value of the non-preceding preceding vehicle degree apn increases, it can be estimated that the communication vehicle is not likely to be the preceding preceding vehicle.

On the other hand, if NO is determined in step 707, the CPU proceeds to step 709 and sets “0” as the non-preceding preceding vehicle degree apn.
The case where NO is determined in step 707 is a case where the number of speed change divergence degrees e1 and speed change divergence degrees e2 for determining the degree of divergence between the preceding vehicle 11 and the communication vehicle is insufficient.

The CPU that has completed the processing of step 302 in FIG. 3 (steps 701 to 709 in FIG. 7) subsequently proceeds to step 303, and calculates the speed fluctuation amount SHn and the non-preceding preceding vehicle degree apn obtained in steps 301 and 302. A correction coefficient fn is obtained using a lookup table (map) as an argument.
As shown in FIG. 9, three types of lookup tables (Mapf1, Mapf2, and Mapf3) for obtaining the correction coefficient fn (n = 1, 2, 3) are prepared.
That is, the table Mapf1 illustrated in FIGS. 9A and 9B is a table particularly adapted to SH1 in which the fluctuation amount of the speed fluctuation amount SHn is small.
The table Mapf2 illustrated in FIGS. 9C and 9D is a table particularly adapted to SH2 in which the fluctuation amount of the speed fluctuation amount SHn is medium.
The table Mapf3 illustrated in FIGS. 9E and 9F is a table particularly adapted to SH3 having a large fluctuation amount of the speed fluctuation amount SHn.
Each of the tables (Mapf1, Mapf2, Mapf3) shown in FIG. 9 is a table when the non-preceding preceding vehicle degree apn is 0.5 and 0.7. It is created for each value in the range (0 to 1) that the non-preceding vehicle degree apn can take. Further, “speed fluctuation amount SHn and non-preceding preceding vehicle degree apn” as arguments of each table are common to each table.
The CPU obtains three types of correction coefficients f1, f2, and f3 based on the lookup tables Mapf1, Mapf2, and Mapf3, respectively.

  In the lookup tables Mapf1, Mapf2, and Mapf3, the correction coefficient fn increases as the speed fluctuation amount SHn increases (maintains "1" after approaching "1"), and the speed fluctuation amount SHn decreases. The correction coefficient fn becomes smaller as it gets closer ("0" is maintained after approaching "0"). In the lookup tables Mapf1, Mapf2, and Mapf3, the correction coefficient fn increases as the non-preceding preceding vehicle degree apn increases, and the correction coefficient fn decreases as the non-preceding preceding vehicle degree apn decreases. Has been created. Therefore, as described above, when the speed fluctuation amount SHn and the non-preceding preceding vehicle degree apn are large, the correction coefficient fn becomes large (approaching “1”), and when the speed fluctuation amount SHn and the non-preceding preceding vehicle degree apn are small. The correction coefficient fn becomes smaller (closer to “0”).

  As described above, the correction coefficient fn is calculated using the three types of lookup tables Mapf1, Mapf2, and Mapf3 corresponding to the magnitude of the speed fluctuation amount SHn, and therefore, the magnitude of the speed fluctuation amount SHn is large, medium, or small. In addition, it is possible to accurately calculate the correction coefficient fn according to the degree of the speed fluctuation amount SHn.

  Further, the CPU performs steps 304 to 312 in parallel with the steps 301 to 303.

ｍ：バッファメモリに記憶されている通信車速度Ｖｃ及び先行車速度Ｖｆｒの数 In step 304, the CPU calculates the mean square error MS of the communication vehicle speed Vc and the preceding vehicle speed Vfr using the following equation. The calculated mean square error MS is stored in the buffer memory.

m: Number of communication vehicle speed Vc and preceding vehicle speed Vfr stored in the buffer memory

  The mean square error MS decreases as the preceding vehicle speed Vfr and the communication vehicle speed Vc continue to take close values during a period from a predetermined point in the past to the present time. That is, the mean square error MS is an index value (one of error statistics) representing the degree to which the preceding vehicle speed Vfr and the communication vehicle speed Vc are similar.

Further, the CPU proceeds to step 305 and averages all the acquired mean square errors MS.
This averaging process is performed according to the flowchart of FIG.

First, in step 1001, the CPU changes whether the vehicle located immediately before the host vehicle 10 has changed from the preceding vehicle 11 to another vehicle during the process from the previous step 305 to the current process, or the communication vehicle and the host vehicle. 10 determines whether or not the wireless communication with 10 is interrupted.
If the CPU determines YES in step 1001, the CPU proceeds to step 1002 and erases all the mean square errors MS acquired so far from the buffer memory. Then, the CPU sets both the sum (total value) and the total number of the mean square errors MS acquired so far to “0”.

  On the other hand, if the CPU determines NO in step 1001, the CPU proceeds to step 1003. The CPU determines whether or not the communication vehicle information has been updated between the previous processing in step 305 and the current processing.

  If the CPU determines YES in step 1003, the CPU proceeds to step 1004 and calculates the total sum of the mean square errors MS at the current time. Specifically, by adding the value of the mean square error MS acquired in the immediately preceding step 304 to the sum of the mean square error MS acquired in the previous step 1004 (previous value), the mean square error MS is calculated. Update the sum.

  Further, the CPU proceeds to step 1005 to calculate the total number of mean square errors MS at the present time. Specifically, the total number of mean square errors MS is updated by adding “1” to the total number of mean square errors MS acquired in the previous processing of step 1005.

  CPU which complete | finishes the process of step 1002 or 1005, or determines with NO at step 1003 progresses to step 1006. When the CPU passes through step 1002 or 1005, the CPU determines in step 1006 whether the total number of mean square errors MS acquired in step 1002 or 1004 is equal to or greater than a predetermined threshold bl (lower limit number). On the other hand, if the CPU determines NO in step 1003, it is determined in step 1006 whether or not the total number of mean square errors MS acquired in step 1002 or 1004 in the previous processing of step 305 is equal to or greater than the threshold bl. To do.

  If the CPU determines YES in step 1006, the process proceeds to step 1007, where the sum of the mean square error MS acquired in step 1002 or 1004 is divided by the total number of mean square error MS acquired in step 1002 or 1005 to obtain an average. The average value (H2E) of the square error MS is calculated.

On the other hand, if the CPU determines NO in step 1006, the CPU proceeds to step 1008 and sets the initial value (constant) stored in the buffer memory as the average value of the mean square error MS.
This initial value is a large constant. Therefore, the communication vehicle is the preceding vehicle as in the case where the vehicle located immediately before the own vehicle 10 changes from the preceding vehicle 11 to another vehicle or the number of mean square errors MS accumulated in the buffer memory is small. In order to accurately determine whether the average square error is 11, the average value of the mean square error MS does not become a small value when the data of the mean square error MS is still small. That is, by making step 1008 such processing, the risk that the CPU makes an erroneous determination that “the preceding vehicle speed Vfr and the communication vehicle speed Vc are similar” is reduced in such a case.

Further, the CPU that has finished the processing of step 305 in FIG. 3 proceeds to step 306 and uses a lookup table (map) that uses the average value (H2E) of the mean square error MS acquired in step 305 as an argument. g1n is obtained.
Similar to the lookup table for the correction coefficient fn, three types (Mapg11, Mapg12, Mapg13) specialized for each of the magnitudes (large, medium, and small) of the speed fluctuation amount SHn are also prepared for this lookup table. (Not shown).
Each of the lookup tables Mapg11, Mapg12, and Mapg13 is created such that the value of the similarity g1n (n = 1, 2, 3) increases as the average value (H2E) of the mean square error MS decreases.
Then, the CPU obtains three types of similarity g1n as g11, g12, and g13 based on the lookup tables Mapg11, Mapg12, and Mapg13. Accordingly, the similarity g1n according to the degree of the average value (H2E) can be accurately calculated regardless of whether the speed fluctuation amount SHn is large, medium, or small.

  In step 307, the CPU calculates a deviation degree e1 = | dVc−dVfr | / min (| dVc |, | dVfr |) of the speed change. Then, the CPU stores the calculated speed change deviation e1 in the buffer memory.

  Subsequently, in step 308, the CPU averages the deviation degree e1 of the speed change to obtain an average value (He1) of the deviation degree e1 of the speed change. This averaging processing method is the same as that in step 305 (the flowchart in FIG. 10).

Further, the CPU proceeds to step 309 and uses the look-up table (map) with the average value (He1) of the speed change divergence e1 acquired in step 308 as an argument, and the similarity g2n (n = 1, 2, 3) is determined.
Similar to the lookup table for the correction coefficient fn, this lookup table has three types (Mapg21, Mapg22, and Mapg23) specialized for the magnitude (large, medium, and small) of the speed fluctuation amount SHn. (Not shown).
Each of the lookup tables Mapg21, Mapg22, and Mapg23 is created such that the value of the similarity g2n increases as the average value (He1) of the divergence e1 decreases.
Then, the CPU obtains three types of similarity g2n as g21, g22, and g23 based on the lookup tables Mapg21, Mapg22, and Mapg23, respectively. Therefore, it is possible to accurately calculate the similarity g2n according to the degree of the average value (He1) regardless of whether the speed fluctuation amount SHn is large, medium, or small.

In step 310, the CPU calculates a velocity correlation coefficient coef (correlation coefficient) using the following equation. The calculated speed correlation coefficient coef is stored in the buffer memory.

  As is well known, the correlation coefficient coef is a statistic indicating a correlation (degree of similarity) between two parameters. In principle, the correlation coefficient is a real number between −1 and 1, and the closer the correlation coefficient is to 1, the more positive the two parameters are, and the higher the similarity between the two parameters is. As the correlation coefficient approaches 0, the correlation between the two parameters becomes weaker. The closer the correlation coefficient is to -1, the more negative the two parameters are. Therefore, in this example, the closer the speed correlation coefficient coef is to 1, the higher the similarity between the communication vehicle speed Vc and the preceding vehicle speed Vfr.

Further, the CPU executes the process of the flowchart in FIG. 11 in parallel with the process of step 310.
That is, in step 1101, the CPU determines whether a = sum ((Vfr (i) −ave (Vfr)) ² is equal to or less than a predetermined threshold x, and in step 1102, b = sum ((Vc ( i) −ave (Vc)) It is determined whether ² is equal to or less than a predetermined threshold y, where sum (Z) is a function that takes the sum of the variables Z, and ave (Z) is the variable Z This is an average function.

When the speed fluctuation amount of the communication vehicle speed Vc and the preceding vehicle speed Vfr is small, the CPU determines NO in step 1101 or 1102. When the speed fluctuation amount of the communication vehicle speed Vc and the preceding vehicle speed Vfr is small, an error is likely to occur in the final determination regarding the specification of the communication vehicle by the CPU. Therefore, if the CPU determines NO in step 1101 or 1102, it is better not to use the velocity correlation coefficient coef acquired in step 310 for the subsequent averaging process in step 311. Therefore, in this case, the CPU proceeds to step 1104. That is, the CPU holds the average value Hcf acquired in the previous step 311 and proceeds to step 312.
On the other hand, if the CPU determines YES in steps 1101 and 1102, the CPU proceeds to step 1103 in order to use the velocity correlation coefficient coef acquired in step 310 for the averaging process. That is, the CPU proceeds to step 311.

  In step 311, the CPU averages the velocity correlation coefficient coef to obtain an average value (Hcf) of the velocity correlation coefficient coef. This averaging processing method is the same as that in step 305 (the flowchart in FIG. 10).

Further, the CPU proceeds to step 312 to obtain the similarity g3n using a look-up table (map) using the average value (Hcf) of the velocity correlation coefficient coef acquired in step 311 as an argument.
Similar to the lookup table for the correction coefficient fn, this lookup table has three types (Mapg31, Mapg32, and Mapg33) specialized for each of the magnitudes (large, medium, and small) of the speed fluctuation amount SHn. (Not shown).
Each of the lookup tables Mapg31, Mapg32, and Mapg33 is created such that the value of the similarity g3n (n = 1, 2, 3) increases as the average value (Hcf) of the velocity correlation coefficient coef approaches 1. Yes.
Then, the CPU obtains three types of similarity g3n as g31, g32, and g33 based on the lookup tables Mapg31, Mapg32, and Mapg33, respectively. Therefore, the similarity g3n can be calculated with high accuracy regardless of whether the speed fluctuation amount SHn is large, medium, or small.

The CPU proceeds to step 313 when the processing of steps 303, 306, 309, and 312 is completed.
In step 313, the CPU calculates the final preceding vehicle probability Pn (= P1, P2, P3) by multiplying the similarities g1n, g2n, g3n and the same value of n of the correction coefficient fn.

  Subsequently, the CPU proceeds to step 314 and selects only one of the largest preceding vehicle probabilities P from P1, P2, and P3 obtained in step 313 as the final preceding vehicle probability P.

  Further, the CPU proceeds to step 315 to determine whether or not one final preceding vehicle probability P selected in step 314 exceeds a predetermined threshold value Pth.

As shown in FIG. 12, this threshold value Pth changes according to the elapsed time since the CPU started the process of the flowchart of FIG. That is, until the time te, the threshold value Pth takes a constant value of the largest value. However, the threshold value Pth gradually decreases when the time te elapses, and the minimum value is maintained after the time tf.
When the time has not passed so much since the start of the process of the flowchart of FIG. 3, the number of data such as the communication vehicle speed Vc, the relative speed Vr, and the preceding vehicle speed Vfr stored in the buffer memory is small. When the final determination is made as to whether or not the communication vehicle is the preceding vehicle 11, the CPU is likely to make an erroneous determination that the communication vehicle that is not a true communication follow-up target vehicle is the preceding vehicle 11. Therefore, the value of the threshold value Pth is increased until the elapsed time from the start of the processing of the flowchart of FIG. On the other hand, if a certain amount of time has elapsed from the start of the process of the flowchart of FIG. 3, the number of data stored in the buffer memory increases. In this case, even if the threshold Pth is slightly lowered, There is a low risk of erroneous determination that a communication vehicle that is not a communication tracking target vehicle is the preceding vehicle 11.
Therefore, the value of the threshold value Pth is changed as shown in the graph of FIG.

If the communication vehicle is only the preceding vehicle 11 among the other vehicles 11 to 13, when the CPU determines YES in step 315, the CPU specifies the preceding vehicle 11 as a communication vehicle in step 316.
On the other hand, if all of the other vehicles 11 to 13 are communication vehicles, the CPU acquires the same number of final preceding vehicle probabilities Pn as the number of communication vehicles in step 313, and further, the same number of final preceding vehicle probabilities as the communication vehicles in step 314. Get P. Further, in step 315, the CPU compares each final preceding vehicle probability P with a threshold value Pth. If the plurality of final preceding vehicle probabilities P exceeds the threshold Pth, the CPU corresponds to the final preceding vehicle probability P having the largest numerical value among the plurality of final leading vehicle probabilities P exceeding the threshold Pth in step 316. One other vehicle 11-13 is determined as a communication follow-up target vehicle. Therefore, when the final preceding vehicle probability P corresponding to the preceding vehicle 11 is the maximum and exceeds the threshold value Pth, the CPU identifies the preceding vehicle 11 as a communication follow-up target vehicle in step 316.

  As mentioned above, although this invention was demonstrated based on said each embodiment, this invention is not limited to said each embodiment, A various change is possible unless it deviates from the objective of this invention.

For example, the speed fluctuation amount SHn and the non-preceding preceding vehicle degree apn may be calculated using the communication vehicle speed Vcb instead of the communication vehicle speed Vc.
The correction coefficient k1 shown in FIG. 5 may be “1” regardless of the magnitude of ΔVfr.

  DESCRIPTION OF SYMBOLS 10 ... Own vehicle, 11-13 ... Other vehicle, 11 ... Prior vehicle, 20 ... Vehicle control ECU, 30 ... Engine control ECU, 60 ... Sensor ECU, 61 ... Vehicle radar sensor, 80 ... Radio control ECU.

Claims (1)

Own vehicle speed detecting means for detecting the own vehicle speed which is the speed of the own vehicle;
A relative speed acquisition means for acquiring a relative speed of the preceding vehicle that travels immediately before the host vehicle and the host vehicle should follow the vehicle;
Preceding vehicle speed calculation means for calculating a preceding vehicle speed, which is the speed of the preceding vehicle, based on the host vehicle speed and the relative speed;
Communication vehicle speed acquisition means for performing wireless communication with other vehicles around the host vehicle and acquiring other vehicle communication information including the communication vehicle speed that is the speed of the other vehicle from the other vehicle;
Similarity calculation means for calculating the similarity between the other vehicle and the preceding vehicle based on the preceding vehicle speed and the communication vehicle speed;
Precedence of calculating the probability that the other vehicle is the preceding vehicle to be higher as the similarity increases, and determining whether the other vehicle is the preceding vehicle based on the calculated probability Vehicle determination means;
By controlling the acceleration of the host vehicle based on the other vehicle communication information acquired by wireless communication with the other vehicle determined to be the preceding vehicle, the host vehicle is followed by the other vehicle. Traveling control means for traveling;
In a vehicle control device comprising:
The preceding vehicle specifying means is
The difference (A) between the present value of the preceding vehicle speed and the past value before the first minute time from the present of the preceding vehicle speed, and the first minute value from the present value of the communication vehicle speed and the present of the communication vehicle speed. A first index value (e1) obtained by normalizing the difference (B) from the previous value before time and the absolute value (| AB) of the difference;
The difference (C) between the present value of the preceding vehicle speed and the past value before the second minute time from the present of the preceding vehicle speed, the past value before the offset time from the present of the communication vehicle speed, and the communication vehicle speed. The second index value (e2) obtained by normalizing the difference (D) between the current time before the offset time and the difference (D) from the past value before the second minute time. When,
To calculate
A vehicle control device configured to calculate the probability higher when the first index value is less than or equal to the second index value compared to when the first index value is greater than the second index value.

Images