Wall shape measuring device

Abstract

The present invention provides a wall shape measuring device (11, 12), which is equipped with a wall detection judgment unit (S50), a wall shape calculation unit (S230), an extrapolation unit (S170), an extrapolation prohibition unit (S160) and an extrapolation times setting unit (S740, S750). The wall detection judgment unit repeatedly judges whether a wall-shaped object is detected. The wall shape calculation unit uses a driving trajectory and a plurality of past wall distance values ​​to calculate a plurality of wall shape values. When a wall-shaped object is not detected, the extrapolation unit extrapolates the wall distance value. When the number of extrapolations is greater than a prohibition judgment value, the extrapolation prohibition unit prohibits extrapolation. When the judgment result of the wall detection judgment unit before executing the extrapolation is correct, the extrapolation times setting unit sets the prohibition judgment value to a proper judgment value, and when the judgment result of the wall detection judgment unit before executing the extrapolation is incorrect, the extrapolation times setting unit sets the prohibition judgment value to an improper judgment value that is less than the proper judgment value.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This international application claims the priority based on Japanese Patent Application No. 2019-170243 filed with the Japan Patent Office on September 19, 2019, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present disclosure relates to a wall shape measuring device which measures the shape of a wall provided along a road. Background Art

[0004] Patent Document 1 describes that wall distance information indicating the distance to a wall existing on the side of the vehicle is acquired from a radar device mounted on the vehicle, and the wall shape is calculated based on the wall distance information and the vehicle's travel trajectory.

[0005] Patent Document 1: Japanese Patent Application Publication No. 2016-85567

[0006] However, as a result of detailed research, the inventor discovered the following problem: In the technology described in Patent Document 1, when there is a situation where another vehicle is between a vehicle equipped with a radar device (hereinafter referred to as this vehicle) and a wall, and the side of the other vehicle is mistakenly identified as a wall, the shape of the wall is calculated to be closer to this vehicle than the actual position of the wall, resulting in miscalculation of the wall shape. Summary of the invention

[0007] The present disclosure suppresses the calculation of the wall shape to be closer to the host vehicle than the actual position of the wall.

[0008] One aspect of the present disclosure is a wall shape measuring device mounted on a vehicle, the wall shape measuring device comprising a wall detection and determination unit, a wall distance acquisition unit, a wall shape calculation unit, an extrapolation unit, an extrapolation prohibition unit, a correctness determination unit, and an extrapolation number setting unit.

[0009] The wall detection determination unit is configured to repeatedly determine whether a wall-like object provided along a road on which the vehicle is traveling is detected, using a reception signal obtained by receiving radar waves transmitted to and reflected in a predetermined detection range around the vehicle.

[0010] The wall distance acquisition unit is configured to acquire a wall distance value indicating a distance to the wall-like object using the received signal when the wall detection determination unit determines that a wall-like object has been detected.

[0011] The wall shape calculation unit is configured to calculate a plurality of wall shape values ​​indicating positions of a plurality of locations on the surface of the wall-shaped object using the travel trajectory of the vehicle and a plurality of past wall distance values ​​repeatedly acquired by the wall distance acquisition unit.

[0012] The extrapolation unit is configured to extrapolate the wall distance value when the wall detection determination unit determines that the wall-shaped object is not detected.

[0013] The extrapolation prohibition unit is configured to prohibit the extrapolation performed by the extrapolation unit when the number of extrapolation times by which the extrapolation unit continuously performs extrapolation is equal to or greater than a preset prohibition determination value.

[0014] The correctness determination unit is configured to determine whether a determination result of the wall detection determination unit before the extrapolation unit performs the extrapolation is correct.

[0015] The extrapolation times setting unit is configured to set the prohibited judgment value to a pre-set appropriate judgment value when the correctness judgment unit determines that the judgment result of the wall detection judgment unit is correct, and to set the prohibited judgment value to an inappropriate judgment value when the correctness judgment unit determines that the judgment result of the wall detection judgment unit is incorrect, and the inappropriate judgment value is set to be smaller than the appropriate judgment value.

[0016] The wall shape measuring device of the present disclosure configured in this way extrapolates the wall distance value when it is determined that the wall-like object is not detected. Therefore, for example, in the wall shape measuring device of the present disclosure, if the situation changes from a situation where there is no other vehicle between the vehicle and the wall-like object to a situation where there is another vehicle between the vehicle and the wall-like object, it is determined that the wall-like object cannot be detected, and the wall distance value after the wall-like object cannot be detected is calculated by extrapolation based on the wall distance value before the wall-like object cannot be detected. As a result, the wall shape measuring device of the present disclosure can prevent the wall shape from being calculated as being closer to the vehicle than the actual position of the wall-like object.

[0017] In addition, if there is another vehicle on the side of the vehicle when the vehicle is stopped, the wall shape measuring device of the present invention may detect the other vehicle as a wall-shaped object. Furthermore, if the vehicle starts before other vehicles, there is no other vehicle on the side of the vehicle. If the wall is far away or does not exist, the wall shape measuring device of the present invention may extrapolate the distance between the side of the vehicle and other vehicles when the vehicle is stopped as the wall distance value.

[0018] In view of this, the wall shape measuring device of the present invention reduces the prohibition judgment value when judging that the judgment result of the wall detection judgment unit before the extrapolation unit performs extrapolation is incorrect, compared with the case where the judgment result of the wall detection judgment unit before the extrapolation unit performs extrapolation is correct. Thus, the wall shape measuring device of the present invention can prevent the situation where the wall shape is calculated to be closer to the vehicle than the actual position of the wall-shaped object continues. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a block diagram showing the structure of the vehicle-mounted system.

[0020] Figure 2 This is a block diagram showing the structure of a radar device.

[0021] Figure 3 is a flow chart showing the first part of the wall calculation process.

[0022] Figure 4 is a flow chart showing the second part of the wall calculation process.

[0023] Figure 5 This is a flowchart showing the constant setting process.

[0024] Figure 6 : is a flowchart showing the wall shape value calculation process.

[0025] Figure 7 This is a diagram showing a method for calculating a wall shape value.

[0026] Figure 8 : is a flowchart showing the extrapolation frequency setting process.

[0027] Fig. 9 This is a graph showing the difference in calculation results depending on whether or not the extrapolation frequency setting process is performed.

[0028] Fig.10 This is a graph showing the difference in detection results depending on whether or not the extrapolation frequency setting process is performed.

[0029] Fig.11 This is a graph showing the calculation results of the wall distance when passing through a toll booth. DETAILED DESCRIPTION

[0030] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings.

[0031] like Figure 1As shown, the vehicle-mounted system 1 of this embodiment includes a driving assistance ECU 2, a radar system 3, and an alarm device 4. ECU is the abbreviation of Electronic Control Unit. In this embodiment, the vehicle-mounted system 1 is mounted on a vehicle that is a four-wheeled vehicle. Hereinafter, the vehicle equipped with the vehicle-mounted system 1 is also referred to as the vehicle itself.

[0032] The radar system 3 includes a plurality of radar devices 11, 12, 13, 14, ... In the present embodiment, the radar device 11 is disposed on the rear right side of the vehicle, the radar device 12 is disposed on the rear left side of the vehicle, the radar device 13 is disposed on the front right side of the vehicle, and the radar device 14 is disposed on the front left side of the vehicle.

[0033] The radar devices 11 and 12 are configured so that the detection ranges of the radar devices 11 and 12 include the rear along the moving direction of the vehicle and the lateral direction perpendicular to the moving direction. The radar device 11 acquires information about objects existing on the right side and the right rear side of the vehicle. The radar device 12 acquires information about objects existing on the left side and the left rear side of the vehicle.

[0034] The radar devices 13 and 14 are arranged so that the detection ranges of the radar devices 13 and 14 include the front along the moving direction of the vehicle and the lateral direction orthogonal to the moving direction. The radar device 13 acquires information about objects existing on the right side and the right front side of the vehicle. The radar device 14 acquires information about objects existing on the left side and the left front side of the vehicle.

[0035] In addition, various detection methods are known as target detection methods used in radar devices, such as FMCW method and dual-frequency CW method. The radar devices 11, 12, 13, 14, ... of this embodiment are configured as so-called "millimeter wave radars" of the FMCW method.

[0036] The plurality of radar devices 11 , 12 , 13 , 14 , . . . constituting the radar system 3 basically have the same structure and function.

[0037] like Figure 2 As shown, radar devices 11, 12, 13, 14...repeatedly transmit radar waves and receive reflected waves every time a processing cycle passes, and based on the received signal Sr, perform processing to detect the distance D to the target object that reflected the radar wave, the relative speed Vr of the target object, and the azimuth θ of the target object.

[0038] The radar devices 11 , 12 , 13 , 14 . . . generate driving assistance information for assisting the driver in driving the vehicle based on the detected observation values ​​(D, Vr, θ), and output the driving assistance information to the driving assistance ECU 2 .

[0039] The driving assistance ECU 2 performs various processes for assisting the driver in driving the vehicle based on the driving assistance information input from the radar devices 11, 12, 13, 14, etc. The processes related to driving assistance may also include, for example, a process of issuing an alert to the driver of an approaching object, and a process of avoiding a collision with an approaching object by controlling the braking system and the steering system, or performing vehicle control for automatically changing lanes.

[0040] The radar devices 11 , 12 , 13 , 14 , . . . include a transmission circuit 20 , a distributor 30 , a transmission antenna 40 , a reception antenna 50 , a reception circuit 60 , a processing unit 70 , an output unit 80 , and a network interface (hereinafter, network I / F) 90 .

[0041] The transmission circuit 20 is a circuit for supplying a transmission signal Ss to the transmission antenna 40. The transmission circuit 20 outputs a high-frequency signal in the millimeter wave band to the distributor 30 located upstream of the transmission antenna 40. Specifically, the transmission circuit 20 alternately repeats the uplink modulation interval and the downlink modulation interval with a modulation period Tm, and outputs the high-frequency signal generated in each modulation interval to the distributor 30. In addition, the transmission circuit 20 generates a high-frequency signal that is frequency-modulated in an increasing manner in the uplink modulation interval, and generates a high-frequency signal that is frequency-modulated in a decreasing manner in the downlink modulation interval.

[0042] The distributor 30 distributes the high frequency signal power input from the transmission circuit 20 to the transmission signal Ss and the local signal L.

[0043] The transmission antenna 40 radiates a radar wave having a frequency corresponding to the transmission signal Ss based on the transmission signal Ss supplied from the distributor 30 .

[0044] The receiving antenna 50 is an antenna for receiving radar waves reflected by a target object, that is, reflected waves. The receiving antenna 50 is configured as a linear array antenna having a plurality of antenna elements 51 arranged in a row. A reception signal Sr of reflected waves from each antenna element 51 is input to a receiving circuit 60 .

[0045] The receiving circuit 60 processes the receiving signal Sr input from each antenna element 51 constituting the receiving antenna 50, generates and outputs a beat signal BT for each antenna element 51. Specifically, the receiving circuit 60 generates and outputs a beat signal BT for each antenna element 51 by mixing the receiving signal Sr input from the antenna element 51 and the local signal L input from the distributor 30 using the mixer 61 for each antenna element 51.

[0046] Hereinafter, the beat signal BT generated by mixing the received signal Sr and the transmitted signal during the period of transmitting the radar wave in the uplink modulation interval is referred to as the uplink beat signal. In addition, the beat signal BT generated by mixing the received signal Sr and the transmitted signal during the period of transmitting the radar wave in the downlink modulation interval is referred to as the downlink beat signal.

[0047] The process of outputting the beat signal BT includes a process of amplifying the received signal Sr, a process of removing unnecessary signal components from the beat signal BT, and a process of converting the beat signal BT into digital data. In this way, the receiving circuit 60 converts the generated beat signal BT of each antenna element 51 into digital data and outputs it. The output beat signal BT of each antenna element 51 is input to the processing unit 70.

[0048] The processing unit 70 is an electronic control device centered on a microcomputer having a CPU 71, a ROM 72, and a RAM 73. The various functions of the microcomputer are realized by the CPU 71 executing a program stored in a non-transitional physical recording medium. In this example, the ROM 72 is equivalent to a non-transitional physical recording medium storing the program. In addition, by executing the program, a method corresponding to the program is executed. In addition, a part or all of the functions executed by the CPU 71 may be constituted in hardware by one or more ICs, etc. In addition, the number of microcomputers constituting the processing unit 70 may be one or more.

[0049] In addition, the processing unit 70 may include a coprocessor that performs fast Fourier transform (hereinafter, FFT) processing, etc. FFT is an abbreviation of Fast Fourier Transform.

[0050] Then, the radar devices 11, 12, 13, 14, etc. calculate the above-mentioned distance D and relative speed Vr based on the frequency fbu (hereinafter, the uplink beat frequency fbu) of the uplink beat signal and the frequency fbd (hereinafter, the downlink beat frequency fbd) of the downlink beat signal by using equations (1) and (2). In equations (1) and (2), c is the speed of light, Δf is the frequency variation amplitude of the transmission signal, and f0 is the center frequency of the transmission signal.

[0051] [Formula 1]

[0052]

[0053]

[0054] The network I / F 90 is a communication interface for the processing unit 70 to perform data communication with other control systems 5 via the network 6. In the present embodiment, the network 6 is a CAN known as an in-vehicle network. CAN is an abbreviation of Controller Area Network. CAN is a registered trademark.

[0055] The other control system 5 includes a plurality of vehicle-mounted devices (for example, ECUs and sensors) other than the driving support ECU 2 .

[0056] Various odometer information is periodically sent from other control systems 5 to the network 6. The processing unit 70 acquires the odometer information via the network 6. The acquired odometer information includes the speed (hereinafter, vehicle speed) Vs of the vehicle, the yaw rate ω of the vehicle, the steering angle α of the vehicle, the turning radius Rs of the vehicle, and the like.

[0057] The other control system 5 calculates the vehicle speed Vs, the yaw rate ω, and the steering angle α based on the detection signals from the vehicle speed sensor, the yaw rate sensor, and the steering angle sensor (not shown), and further calculates the turning radius Rs by a predetermined calculation method based on the steering angle α. In addition, the processing unit 70 may calculate the turning radius Rs inside the radar devices 11, 12, 13, 14, ... based on the steering angle α acquired via the network 6, instead of acquiring the turning radius Rs via the network 6.

[0058] In the vehicle-mounted system 1 configured as described above, the CPU 71 of the radar devices 11 and 12 executes a wall calculation process and an extrapolation frequency setting process.

[0059] First, the procedure of the wall calculation process executed by the CPU 71 of the radar device 11 will be described. The wall calculation process is a process that is repeatedly executed every time the above-mentioned processing cycle passes during the operation of the processing unit 70.

[0060] If wall calculation is performed, Figure 3 As shown, first in S10, CPU71 obtains odometer information from other control systems 5 via network 6. The odometer information obtained in S10 includes at least vehicle speed Vs, yaw rate ω, steering angle α, and turning radius Rs. CPU71 stores the odometer information obtained in S10 from the first N cycles of the above-mentioned processing cycle to the current odometer information in RAM73. N is an integer greater than 2. In addition, N is determined in consideration of the distance detected in the vehicle (for example, 100m behind) and the vehicle speed of the vehicle.

[0061] Furthermore, in S20, the CPU 71 calculates the driving trajectory of the vehicle from the first N cycles of the above-mentioned processing cycle to the current one. Specifically, the CPU 71 uses the odometer information acquired in S10 (i.e., the acquired values ​​in each processing cycle up to the first N cycles) to calculate the estimated value of the vehicle position in each processing cycle from the previous cycle to the first N cycles based on the current position (hereinafter referred to as the estimated position of the vehicle). Then, the CPU 71 uses the line connecting the current position and the calculated estimated position of the vehicle in each cycle as the driving trajectory of the vehicle. In addition, since the technology for calculating the driving trajectory of the vehicle based on the odometer information is well known, the detailed description thereof is omitted.

[0062] In addition, due to various factors such as detection errors of the vehicle speed sensor and yaw rate sensor and noise, the odometer information obtained in S10 contains errors. Therefore, in S20, CPU71 also calculates the estimated existence range of the estimated position of the vehicle taking into account the error of the odometer information for each estimated position of the vehicle in the past processing cycles up to the first N cycles. The estimated existence range can be expressed as an error variance based on the estimated position of the vehicle. And, by projecting the error variance into the lane width direction (i.e., the direction perpendicular to the direction of travel), the existence probability of the estimated position of the vehicle in the lane width direction can be expressed as a prescribed probability distribution centered on the estimated position of the vehicle.

[0063] In this embodiment, the error variance of the estimated position of the vehicle caused by the error factor of the odometer information is modeled as a normal distribution. That is, the existence probability of the estimated position of the vehicle calculated using the odometer information becomes the peak with the highest probability in the normal distribution, and the farther away from the estimated position of the vehicle in the lane width direction, the less the existence probability according to the normal distribution.

[0064] Then, in S30, the CPU 71 calculates the adjacent lane probability map. Specifically, the CPU 71 specifies the position of the adjacent lane (specifically, the positions of both ends in the lane width direction) for each estimated position of the vehicle up to the previous N cycles. Then, the error variance on the estimated position of the vehicle is projected to the position of the specified adjacent lane to calculate the probability distribution of the adjacent lane.

[0065] Specifically, the CPU 71 specifies the positions of two lane dividing lines (i.e., lane dividing lines on both sides of the adjacent lanes) that distinguish the adjacent lanes, i.e., the estimated value of the position (hereinafter, the inner dividing line) of the lane dividing line on the side close to the vehicle (hereinafter, the inner dividing line), i.e., the inner estimated dividing position, and the estimated value of the position (hereinafter, the outer dividing line) of the lane dividing line on the side away from the vehicle (hereinafter, the outer dividing line), i.e., the outer estimated dividing position. Then, the CPU 71 sets the probability distribution of the lane width direction of each dividing position by projecting the error variance of the estimated position of the vehicle onto each estimated dividing position as it is.

[0066] That is, for the inner partition position, the inner inferred partition position becomes the peak of the normal distribution, and as the distance from the inner inferred partition position increases, the probability of the inner partition position decreasing according to the normal distribution. For the outer partition position, the outer inferred partition position also becomes the peak of the normal distribution, and as the distance from the outer inferred partition position increases, the probability of the outer partition position decreasing according to the normal distribution.

[0067] Then, the CPU 71 calculates the adjacent lane probability map by connecting the points with equal probability of existence of the inner distinguishing position in each processing cycle up to the first N cycles (for example, points with predetermined probabilities P1, P2, and P3), and connecting the points with equal probability of existence of the outer distinguishing position (for example, points with predetermined probabilities P1, P2, and P3 as the inner side). The adjacent lane probability map is a map indicating an area with a probability of P1, P2, and P3 of the existence of an adjacent lane. In other words, it can also be said to be a map indicating that when an object exists in the area represented by the adjacent lane probability map, the probability of the object existing in the adjacent lane is P1, P2, and P3.

[0068] In addition, in the present embodiment, the probabilities P1, P2, and P3 are 100%, 70%, and 30%. Furthermore, the mapping of the region representing the probability P1 is formed by a line connecting the points of 1σ (i.e., 68.3%). The mapping of the region representing the probability P2 is formed by a line connecting the points of 2σ (i.e., 95.5%). The mapping of the region representing the probability P3 is formed by a line connecting the points of 3σ (i.e., 99.7%).

[0069] Next, in S40, the CPU 71 determines whether the wall on the right side can be detected. Specifically, the CPU 71 follows the same object mark detected in each cycle based on the distance D, relative speed Vr and azimuth θ from the previous N cycles to the currently detected object mark, thereby determining whether there is another vehicle running parallel to the right side of the vehicle (hereinafter referred to as the right parallel vehicle).

[0070] Then, when there is a vehicle running parallel to the right, the CPU 71 determines that the situation in which the wall on the right is not detectable, and when there is no vehicle running parallel to the right, the CPU 71 determines that the situation in which the wall on the right is detectable.

[0071] Here, when it is determined that the wall on the right side can be detected, in S50, the CPU 71 determines whether the wall on the right side is detected. Specifically, in order to detect the wall on the right side, first, the CPU 71 extracts multiple peaks (in this embodiment, three peaks from high to low power) in the peak extraction frequency range for each power spectrum of the uplink beat signal and the downlink beat signal, and performs azimuth expansion processing (for example, MUSIC, DBF, CAPON, etc.).

[0072] The power spectrum is calculated by performing frequency analysis processing (e.g., FFT processing) on ​​each of the uplink beat signal and the downlink beat signal. In the power spectrum, the power of the beat signal is represented for each frequency bin. The frequency bin is a frequency range that becomes a unit scale of the power spectrum. A power spectrum is obtained for each of the plurality of antenna elements 51. The power spectrum that is the object of extracting the peak in S50 is an average power spectrum obtained by averaging the power spectra obtained from each antenna element 51.

[0073] Furthermore, the CPU 71 further extracts the azimuth peak from the azimuth expansion spectrum, and performs pairing matching taking into account the characteristics of the wall at the multiple azimuth peaks in the upper and lower directions. Then, when the pairing matching is established, the CPU 71 determines that the wall on the right side is detected, and when the pairing matching is not established, the CPU 71 determines that the wall on the right side cannot be detected.

[0074] Furthermore, when the following first pairing matching determination condition, second pairing matching determination condition, and third pairing matching determination condition are all satisfied, the CPU 71 determines that pairing matching is satisfied.

[0075] The first pairing matching determination condition is that the uplink frequency and the downlink frequency are almost identical. The first pairing matching determination condition is a condition set based on the fact that the wall on the front side has no relative speed with respect to the vehicle.

[0076] The second pairing matching judgment condition is that the upward direction and the downward direction are almost the same. The second pairing matching judgment condition is a condition set based on that both the upward direction and the downward direction detect the wall in the positive side direction.

[0077] The third pairing matching determination condition is that the uplink power and the downlink power are almost the same. The third pairing matching determination condition is a condition set based on that the power of the received signal Sr reflected from the same target object is almost the same.

[0078] Here, when it is determined that the wall on the right side is detected, in S60 , the CPU 71 acquires the distance D calculated from the peak pair for which pair matching is established as the instantaneous wall distance value, and stores the acquired value in the instantaneous wall distance value WD provided in the RAM 73 .

[0079] Then, in S70 , the CPU 71 increments (ie, adds 1) the continuous detection counter C3 provided in the RAM 73 , and moves to S210 .

[0080] In addition, in S40, when it is determined that the wall on the right side is not detectable (that is, when it is determined that there is a right parallel vehicle), Figure 4 As shown, in S80, the CPU 71 determines whether there is a right parallel vehicle in an adjacent lane on the right side of the vehicle. In addition, the CPU 71 determines whether there is a right parallel vehicle in an adjacent lane based on the distance D and the azimuth θ of the object mark corresponding to the right parallel vehicle and the position of the adjacent lane specified in S30, or the driving track of the vehicle detected in the past.

[0081] Here, when there is a right parallel-traveling vehicle in an adjacent lane on the right side of the vehicle, in S90 , the CPU 71 stores a preset second lane determination value N1_2nd in the first threshold value Vth1 provided in the RAM 73 , and moves to S110 .

[0082] On the other hand, when there is no right parallel vehicle in an adjacent lane on the right side of the vehicle, in S100, CPU71 stores the third lane judgment value N1_3rd in the first threshold value Vth1 and moves to S110, where the third lane judgment value N1_3rd is set to be greater than the second lane judgment value N1_2nd.

[0083] Then, if the process moves to S110, the CPU 71 determines whether the value of the undetectable wall counter C1 set in the RAM 73 is less than the first threshold value Vth1. Here, when the value of the undetectable wall counter C1 is less than the first threshold value Vth1, the CPU 71 maintains the value stored in the instantaneous wall distance value WD in S120 and moves to S140. That is, the CPU 71 maintains the instantaneous wall distance value WD at the value in the previous processing cycle.

[0084] On the other hand, when the value of the undetectable wall counter C1 is greater than the first threshold value Vth1, the CPU 71 stores a preset initial value in the instantaneous wall distance value WD in S130 and moves to S140. In the present embodiment, the initial value is set to a value corresponding to 15 m, for example.

[0085] Then, when the process moves to S140, the CPU 71 increments the undetectable wall counter C1. Furthermore, in S150, the CPU 71 resets the continuous detection counter C3 (that is, sets it to 0), and moves to S210.

[0086] In addition, in S50, when it is determined that the wall on the right side is not detected, in S160, the CPU 71 determines whether the value of the wall loss counter C2 set in the RAM 73 is less than the second threshold value Vth2 set in the RAM 73. Here, when the value of the wall loss counter C2 is less than the second threshold value Vth2, in S170, the CPU 71 holds the value stored in the wall distance instantaneous value WD and moves to S190. That is, the CPU 71 holds the wall distance instantaneous value WD at the value in the previous processing cycle.

[0087] On the other hand, when the value of the wall loss counter C2 is equal to or greater than the second threshold value Vth2, the CPU 71 stores the initial value in the wall distance instantaneous value WD in S180, and moves to S190.

[0088] Then, when the process moves to S190, the CPU 71 increments the wall loss counter C2. Furthermore, in S200, the CPU 71 resets the continuous detection counter C3 and moves to S210.

[0089] Then, if we move to S210, Figure 3 As shown, the CPU 71 executes a constant setting process.

[0090] Next, the procedure of the constant setting process will be described.

[0091] If constant setting processing is executed, Figure 5 As shown, first in S310, the CPU 71 calculates the variance of the wall distance detected in a cycle corresponding to the most recent preset high reliability judgment number JC1, and stores the calculated value in the variance V1 set in the RAM 73. In this embodiment, the cycle corresponding to the high reliability judgment number JC1 is 20 cycles.

[0092] Then, in S320, the CPU 71 determines whether the value of the continuous detection counter C3 is greater than the high reliability determination number JC1 and whether the variance V1 is less than a preset variance determination value JV1. In the present embodiment, the variance determination value JV1 is set to, for example, 0.1 m / s. 2 The purpose of using variance in this process is to reduce the amount of calculation, and it is preferred to use standard deviation.

[0093] Here, when the value of the continuous detection counter C3 is greater than the high reliability judgment number JC1 and the variance V1 is less than the variance judgment value JV1, in S330, the CPU 71 stores the pre-set high reliability constant NN_h in the filter constant FT set in the RAM 73 and ends the right constant setting process. In the present embodiment, the high reliability constant NN_h is set to a value equivalent to 0.9, for example.

[0094] On the other hand, when the value of the continuous detection counter C3 is less than the high reliability judgment number JC1, or the variance V1 is greater than the variance judgment value JV1, in S340, the CPU 71 calculates the variance of the wall distance detected in a cycle corresponding to the nearest preset medium reliability judgment number JC2, and stores the calculated value in the variance V2 set in the RAM 73. In the present embodiment, the cycle corresponding to the medium reliability judgment number JC2 is 10 cycles.

[0095] Then, in S350 , the CPU 71 determines whether the value of the continuous detection counter C3 is equal to or greater than the medium reliability determination number JC2 and whether the variance V2 is smaller than the variance determination value JV1 .

[0096] Here, when the value of the continuous detection counter C3 is greater than the medium reliability judgment number JC2 and the variance V2 is less than the variance judgment value JV1, in S360, the CPU 71 stores the pre-set medium reliability constant NN_m in the filter constant FT and ends the right constant setting process. In this embodiment, the medium reliability constant NN_m is set to a value equivalent to 0.7, for example.

[0097] On the other hand, when the value of the continuous detection counter C3 is less than the medium reliability judgment number JC2, or the variance V2 is greater than the variance judgment value JV1, in S370, the CPU 71 calculates the variance of the wall distance detected in a cycle corresponding to the most recent preset low reliability judgment number JC3, and stores the calculated value in the variance V3 set in the RAM 73. In the present embodiment, the cycle corresponding to the low reliability judgment number JC3 is 5 cycles.

[0098] Then, in S380 , the CPU 71 determines whether the value of the continuous detection counter C3 is equal to or greater than the low reliability determination number JC3 and whether the variance V3 is smaller than the variance determination value JV1 .

[0099] Here, when the value of the continuous detection counter C3 is greater than the low reliability judgment number JC3 and the variance V3 is less than the variance judgment value JV1, in S390, the CPU 71 stores the pre-set low reliability constant NN_1 in the filter constant FT and ends the constant setting process. In this embodiment, the low reliability constant NN_1 is set to a value equivalent to 0.5, for example.

[0100] On the other hand, when the value of the continuous detection counter C3 is less than the low reliability judgment number JC3, or the variance V3 is greater than the variance judgment value JV1, in S400, the CPU 71 stores the pre-set normal reliability constant NN_n in the filter constant FT and ends the right constant setting process. In the present embodiment, the normal reliability constant NN_n is set to a value equivalent to 0.3, for example.

[0101] Furthermore, if the constant setting process is completed, Figure 3 As shown, in S220, the CPU 71 calculates the wall filter value. Specifically, the CPU 71 first stores the values ​​of the wall filter values ​​FV(N), FV(N-1), ..., FV(1) in the wall filter values ​​FV(N-1), FV(N-2), ..., FV(0) respectively set in the RAM 73. Next, the CPU 71 stores the value calculated by the right side of the formula (3) in the wall filter value FV(N).

[0102] FV(N)=WD×FT+FV(N-1)×(1-FT)…(3)

[0103] Furthermore, in S230 , the CPU 71 executes wall shape value calculation processing.

[0104] Next, the procedure of the wall shape value calculation process will be described.

[0105] If the wall shape value calculation process is performed, then Figure 6 As shown, first in S510, the CPU 71 determines whether the value of the undetectable wall counter C1 is less than a preset third threshold value Vth3. The third threshold value Vth3 is set to be greater than the first threshold value Vth1.

[0106] Here, when the value of the undetectable wall counter C1 is smaller than the third threshold value Vth3, in S520, the CPU 71 determines whether the value of the wall loss counter C2 is smaller than a preset fourth threshold value Vth4. The fourth threshold value Vth4 is set to be larger than the second threshold value Vth2.

[0107] Here, when the value of the wall loss counter C2 is smaller than the fourth threshold value Vth4, in S530, the CPU 71 calculates the wall shape value based on the wall filter value FV and ends the wall shape value calculation process.

[0108] Specifically, if Figure 7 As shown, CPU 71 sets each of the estimated positions CP(N), CP(N-1), CP(N-2) ..., CP(1), CP(0) of the vehicle constituting the driving trajectory TR of the vehicle from the current to the previous N cycles as a starting point, and sets the position away from the wall filter value FV(N), FV(N-1), FV(N-2), ..., FV(1), FV(0) to the right along the lateral direction orthogonal to the advancing direction as the wall position WP(N), WP(N-1), WP(N-2), ..., WP(1), WP(0). The wall positions WP(N) to WP(0) are wall shape values.

[0109] In addition, if Figure 6 As shown, in S520, when the value of the wall loss counter C2 is greater than the fourth threshold value Vth4, the CPU 71 moves to S540. In addition, in S510, when the value of the undetectable wall counter C1 is greater than the third threshold value Vth3, the CPU 71 moves to S540.

[0110] Then, if the process moves to S540, the CPU 71 calculates the wall shape value based on the initial value and ends the right wall shape value calculation process. Specifically, the CPU 71 first stores the initial value in the wall filter values ​​FV(N), FV(N-1), FV(N-2), ..., FV(1), and FV(0). Then, the CPU 71 sets each of the vehicle estimated positions CP(N), CP(N-1), CP(N-2), ..., CP(1), and CP(0) constituting the vehicle's driving trajectory TR from the previous N cycles to the present as a starting point, and sets the position that is away from the wall filter values ​​FV(N), FV(N-1), FV(N-2), ..., FV(1), and FV(0) to the right in the lateral direction perpendicular to the advancing direction (i.e., the position that is away from the initial value to the right) as the wall position WP(N), WP(N-1), WP(N-2), ..., WP(1), and WP(0).

[0111] Then, if the wall shape value calculation process is completed, Figure 3 As shown, in S240, CPU71 calculates the existence probability of the target on the adjacent lane. Specifically, CPU71 adds the wall shape value calculated in S230 to the adjacent lane probability map calculated in S30 to calculate the existence probability distribution of the target. That is, CPU71 basically calculates the existence probability of the target based on the adjacent lane probability map. However, when there is a wall, CPU71 sets the existence probability of the adjacent lane to 0 for the area outside the wall in the adjacent lane probability map, and sets the existence probability of the target to 0 for the area. Alternatively, CPU71 can also delete the object mark detected outside the wall.

[0112] Further, in S250, the CPU 71 determines whether there is a target in the adjacent lane. Specifically, the CPU 71 determines whether the target exists in the adjacent lane based on the target existence probability calculated in S240 and the position information of the detected target. The target's adjacent lane existence probability is not calculated only by the value of the current cycle, but is calculated based on the filter value obtained by multiplying the past adjacent lane existence probability by the forgetting coefficient. In addition, this formula is the same as the above formula (3).

[0113] For example, when the filtered value of the object existence probability is 70% or more, the CPU 71 determines that the object exists in the adjacent lane. On the contrary, when the filtered value of the object existence probability is less than 70%, the CPU 71 determines that the object does not exist in the adjacent lane.

[0114] Next, in S260, CPU71 extracts the alarm object. Specifically, when it is determined in S250 that the target exists in the adjacent lane, CPU71 confirms the positional relationship between the target and the vehicle. Moreover, when the positional relationship between the two satisfies the prescribed conditions, CPU71 extracts the target as the alarm object. The prescribed conditions are, for example, at least one of the following first extraction conditions and second extraction conditions. The first extraction condition is that the distance between the vehicle and the target (hereinafter, relative distance) is less than the prescribed distance. The second extraction condition is that the arrival time to the vehicle inferred from the relative speed detected by the radar device is less than the prescribed time.

[0115] Then, in S270, the CPU 71 performs the alarm output process and ends the wall calculation process. Specifically, when the alarm object is extracted in S260, the CPU 71 outputs an alarm instruction to the driving assistance ECU 2, and the alarm instruction indicates the output of an alarm indicating the presence of the alarm object. If the driving assistance ECU 2 obtains the alarm instruction, it causes the alarm device 4 to output an alarm of the content indicated by the alarm instruction.

[0116] The wall calculation process executed by the CPU 71 of the radar device 12 is the same as the wall calculation process executed by the CPU 71 of the radar device 11 except that the left wall is used as the target point. Therefore, the description of the steps of the wall calculation process executed by the CPU 71 of the radar device 12 is omitted.

[0117] Next, a description will be given of the procedure of the extrapolation frequency setting process executed by the CPU 71 of the radar device 11. The extrapolation frequency setting process is a process that is repeatedly executed every time the above-mentioned processing cycle passes during the operation of the processing unit 70.

[0118] If the extrapolation times setting process is executed, Figure 8As shown, first in S650 , the CPU 71 checks the powers at the wall positions WP(N), WP(N-2), . . . , WP(1) acquired in the past processing cycle.

[0119] Specifically, the CPU 71 first calculates the distance D_j between the vehicle and the wall position WP(j), and the orientation θ_j of the wall position WP(j) relative to the vehicle based on the current position of the vehicle and the wall positions WP(N) to WP(1) acquired from the previous N cycles to the previous cycle. j is an integer from 1 to N. In addition, the CPU 71 calculates the relative speed Vr_j of the wall at the wall position WP(j) based on the driving speed and orientation θ_j of the vehicle.

[0120] Furthermore, the CPU 71 substitutes the distance D_j and the relative speed Vr_j into the equations (1) and (2) to calculate the up check frequency fcku_j and the down check frequency fckd_j corresponding to the wall position WP(j).

[0121] Then, the CPU 71 sets the confirmation frequency range centered on the uplink confirmation frequency fcku_j as the uplink confirmation frequency range CFu_j, and sets the confirmation frequency range centered on the downlink confirmation frequency fckd_j as the downlink confirmation frequency range CFd_j.

[0122] Furthermore, the CPU 71 performs azimuth expansion processing in the uplink confirmation frequency range CFu_j on the power spectrum of the uplink beat signal. Then, the CPU 71 confirms the power at the azimuth θ_j for the azimuth expansion spectrum obtained by the azimuth expansion processing.

[0123] Similarly, the CPU 71 performs azimuth expansion processing in the downlink confirmation frequency range CFd_j on the power spectrum of the downlink beat signal. Then, the CPU 71 confirms the power at the azimuth θ_j for the azimuth spread spectrum obtained by the azimuth expansion processing.

[0124] Then, in S660, CPU71 determines whether the power of at least one azimuth among the azimuths θ_1 to θ_N for the azimuth spread spectrum of the uplink beat signal is greater than the preset second wall detection judgment power, and whether the power of at least one azimuth among the azimuths θ_1 to θ_N for the azimuth spread spectrum of the downlink beat signal is greater than the second wall detection judgment power.

[0125] Here, when the azimuth spread spectrum of the uplink beat signal has a power above the second wall detection judgment power in at least one of the azimuths θ_1 to θ_N, and when the azimuth spread spectrum of the downlink beat signal has a power above the second wall detection judgment power in at least one of the azimuths θ_1 to θ_N, in S670, CPU71 reduces (i.e., subtracts 1) the wall non-continuation counter C4 set in RAM73 and moves to S690.

[0126] On the other hand, when the azimuth spread spectrum of the uplink beat signal and all azimuth powers in the azimuths θ_1 to θ_N are less than the second wall detection judgment power, or when the azimuth spread spectrum of the downlink beat signal and all azimuth powers in the azimuths θ_1 to θ_N are less than the second wall detection judgment power, in S680, CPU71 increases the wall non-continuation counter C4 and moves to S690.

[0127] When the process proceeds to S690 , the CPU 71 determines whether the value of the wall absence continuation counter C4 is smaller than a fourth threshold value Vth4 set in advance. The fourth threshold value Vth4 is set to 30, for example.

[0128] Here, when the value of the wall absence continuation counter C4 is smaller than the fourth threshold value Vth4, in S700, the CPU 71 clears the second wall absence flag F2 set in the RAM 73 and moves to S730.

[0129] On the other hand, when the value of the wall absence continuation counter C4 is equal to or greater than the fourth threshold value Vth4, the CPU 71 sets the second wall absence flag F2 in S710 and moves to S730.

[0130] Then, if moving to S730, CPU71 judges whether the second wall no mark F2 is cleared. Here, when the second wall no mark F2 is set, CPU71 moves to S750.

[0131] On the other hand, when the second wall no flag F2 is cleared, the CPU 71 moves to S740.

[0132] When the process proceeds to S740 , the CPU 71 stores the preset first extrapolation count N2_h in the right threshold value Vth2 , and ends the extrapolation count setting process.

[0133] In addition, when the process moves to S750, the CPU 71 stores the second extrapolation number N2_1 in the second threshold value Vth2, and ends the extrapolation number setting process, wherein the second extrapolation number N2_1 is set to be smaller than the first extrapolation number N2_h.

[0134] The extrapolation number setting process executed by the CPU 71 of the radar device 12 is the same as the extrapolation number setting process executed by the CPU 71 of the radar device 11, except that the left wall is used as the target point. Therefore, the description of the steps of the extrapolation number setting process executed by the CPU 71 of the radar device 12 is omitted.

[0135] Fig. 9 This is a graph showing the calculation results of the wall distance from the state where the vehicle is stopped to the state where the vehicle starts to move while being aligned with other vehicles in the adjacent lane. The horizontal axis of the graph is the number of processing cycles described above, and the vertical axis is the wall filter value.

[0136] Fig. 9 The line WD1 in represents the calculation result obtained by executing the extrapolation frequency setting process. The line WD1 represents the wall filter value calculated by setting the instantaneous value of the wall distance to the initial value (ie, 15 m) when the state in which the wall cannot be detected continues for a predetermined period. Fig. 9 The line WD2 in FIG. 1 represents the calculation result obtained without executing the extrapolation frequency setting process. The line WD2 represents the wall filter value calculated by keeping the instantaneous value of the wall distance at the previous value.

[0137] Fig.10 This is a graph showing the result of the position of other vehicles detected by the vehicle when the vehicle starts to move from a stopped state to a state in which the vehicle is aligned transversely with other vehicles in the adjacent lane, and then the other vehicles in the adjacent lane approach from behind the vehicle. The horizontal axis of the graph is the horizontal position with the vehicle as the starting point, and the vertical axis is the vertical position with the vehicle as the starting point.

[0138] Fig.10 The line CP1 in the figure is based on Fig. 9 The track of another vehicle is the result of detecting another vehicle in the adjacent lane of the vehicle using the wall filter value represented by the line WD1 in FIG. The line CP1 shows that another vehicle at least 90 m away can be detected by executing the extrapolation frequency setting process.

[0139] Fig.10 The line CP2 in the figure is based on Fig. 9The wall filter value represented by line WD2 in the figure is used to detect other vehicles in the adjacent lanes of the vehicle, and the result is the trajectory of other vehicles. Line CP2 shows that the number of continuous extrapolation times is inappropriately large, so that other vehicles cannot be detected before other vehicles approach to 35m. That is, when the vehicle is parked, the side of the vehicle parked on the front side is judged as a wall. After the vehicle starts to move, since there is no vehicle on the front side of the vehicle, it is judged that the wall is lost, and it becomes an extrapolation mode. Then, as long as the wall can be detected, the extrapolation is continued in an environment where the wall is too far away or does not exist. However, since the number of extrapolation times is inappropriately large, the wall trajectory is continuously calculated, so that there is always a wall on the front side of the vehicle, and it is mistakenly judged that the approaching vehicle is detected on the outside of the wall (that is, it is judged as a mirror ghost), resulting in a delay in the detection of the approaching vehicle.

[0140] Fig.11 The graph GR is a graph showing the calculation result of the wall distance when the vehicle passes through the toll booth TB with the ETC lane as shown by arrows L1, L2, L3, L4, and L5. ETC is a registered trademark. The horizontal axis of the graph is the number of processing cycles mentioned above, and the vertical axis is the wall filter value. In addition, the graph GR shows the calculation result of the wall distance on the right side starting from before entering the ETC lane.

[0141] The triangular point cloud in the graph GR represents the calculation results obtained without executing the extrapolation number setting process and the constant setting process. The diamond point cloud in the graph GR represents the calculation results obtained by executing the extrapolation number setting process but not executing the constant setting process. The quadrilateral point cloud in the graph GR represents the calculation results obtained by executing the extrapolation number setting process and the constant setting process.

[0142] As shown by the diamond point cloud and the quadrilateral point cloud in the region surrounded by the ellipse EL in the graph GR, by executing the constant setting process, the followability of the wall filter value when the wall distance changes suddenly is improved.

[0143] The radar devices 11 and 12 configured in this manner use reception signals Sr obtained by receiving radar waves transmitted to and reflected in a detection range set in advance around the vehicle to repeatedly determine whether a wall installed along the road on which the vehicle is traveling is detected.

[0144] When it is determined that a wall is detected, the radar devices 11 and 12 use the received signal Sr to acquire an instantaneous value of the wall distance indicating the distance to the wall.

[0145] The radar devices 11 and 12 calculate a plurality of wall shape values ​​indicating the positions of a plurality of locations on the surface of the wall, using the travel trajectory of the vehicle and a plurality of repeatedly acquired past instantaneous values ​​of the wall distance.

[0146] When it is determined that a wall is not detected, the radar devices 11 and 12 extrapolate the instantaneous value of the wall distance.

[0147] When the number of extrapolation times of consecutively executing extrapolation is equal to or greater than a preset second threshold value Vth2, the radar devices 11 and 12 prohibit the extrapolation.

[0148] The radar devices 11 and 12 determine whether the determination result of the wall detection before performing the extrapolation is correct.

[0149] When it is determined that the result of the wall detection is correct, the radar devices 11 and 12 set the second threshold value Vth2 to a pre-set first extrapolation number N2_h. When it is determined that the result of the wall detection is incorrect, the radar devices 11 and 12 set the second threshold value Vth2 to a second extrapolation number N2_l, which is set to be less than the first extrapolation number N2_h.

[0150] In addition, the radar devices 11 and 12 generate multiple azimuth spread spectrums representing the relationship between the power of the beat signal and the azimuth of the wall by performing azimuth spread processing on the power spectrum at the beat signal frequency corresponding to each of the multiple wall positions WP(1) to WP(N) calculated, wherein the power spectrum represents the relationship between the power of the beat signal and the frequency of the beat signal. Then, the radar devices 11 and 12 determine whether the judgment result of the wall detection is correct by judging whether the power in the azimuths θ_1 to θ_N corresponding to each of the multiple wall positions WP(1) to WP(N) satisfies a pre-set azimuth power existence condition for each of the generated multiple azimuth spread spectrums, wherein the azimuth power existence condition indicates the existence of power. The azimuth power existence condition is the judgment condition of S660.

[0151] In this way, when it is determined that the wall is not detected, the radar devices 11 and 12 extrapolate the instantaneous value of the wall distance. Therefore, for example, if the situation changes from a situation where there is no other vehicle between the vehicle and the wall to a situation where there is another vehicle between the vehicle and the wall, the radar devices 11 and 12 determine that the wall cannot be detected, and calculate the instantaneous value of the wall distance after the wall cannot be detected by extrapolation based on the instantaneous value of the wall distance before the wall cannot be detected. In this way, the radar devices 11 and 12 can suppress the calculation of the wall shape as being closer to the vehicle than the actual position of the wall.

[0152] In addition, if there is another vehicle on the side of the vehicle when the vehicle is stopped, the radar devices 11 and 12 may detect the other vehicle as a wall. In addition, if the vehicle starts earlier than other vehicles, there is no other vehicle on the side of the vehicle, and then in an environment where the wall is far away or does not exist and cannot be detected, the radar devices 11 and 12 extrapolate the distance between the side of the vehicle and the other vehicle when the vehicle is stopped as the instantaneous value of the wall distance.

[0153] In contrast, when the radar devices 11 and 12 determine that the result of the wall detection before the extrapolation is incorrect, they reduce the second threshold value Vth2 compared to the case where the result is correct. Thus, the radar devices 11 and 12 can prevent the situation where the wall shape is calculated to be closer to the position of the vehicle than the actual position of the wall continues.

[0154] The radar devices 11 and 12 determine whether a preset undetectable condition indicating that the wall cannot be detected continues to be satisfied. The undetectable condition is a condition that is negatively determined in S510 or S520. Then, when it is determined that the undetectable condition is satisfied, the radar devices 11 and 12 calculate a plurality of wall shape values ​​so that the distance between the driving trajectory and the plurality of wall shape values ​​becomes a preset initial value.

[0155] Thus, the radar devices 11 and 12 can suppress the wall shape from being calculated as being closer to the vehicle than the actual wall position by setting the initial value to be larger than the instantaneous value of the wall distance when another vehicle is detected as a wall on the side of the vehicle when the vehicle is stopped.

[0156] The radar devices 11 and 12 determine whether there is a parallel vehicle in the adjacent lane. The parallel vehicle in the adjacent lane is an obstacle that hinders the detection of the wall. Then, when it is determined that there is a parallel vehicle in the adjacent lane, the radar devices 11 and 12 extrapolate the instantaneous value of the wall distance. Further, when the number of times of hindered extrapolation (i.e., the value of the undetectable wall counter C1) of continuous extrapolation is greater than the first threshold value Vth1, the radar devices 11 and 12 prohibit extrapolation. Then, the radar devices 11 and 12 set the first threshold value Vth1 so that there is a positive correlation between the lane distance between the driving lane of the vehicle and the adjacent lane where the parallel vehicle is driving. In addition, "there is a positive correlation between the lane distance and the first threshold value Vth1" not only includes that the first threshold value Vth1 increases continuously with the increase of the lane distance, but also includes that the first threshold value Vth1 increases in stages with the increase of the lane distance.

[0157] As a result, the radar devices 11 and 12 can prevent the situation where the wall shape is calculated to be closer to the position of the vehicle than the actual position of the wall continues to occur as the parallel running vehicle approaches the vehicle.

[0158] The radar devices 11 and 12 calculate the wall filter value after filtering the instantaneous value of the wall distance, and use the driving trajectory and the multiple wall filter values ​​to calculate multiple wall shape values. Then, the radar devices 11 and 12 set the reliability of the instantaneous value of the wall distance obtained. Further, the radar devices 11 and 12 set the filter constant FT of the filtering process so that there is a positive correlation between the set reliability and the followability of the wall filter value relative to the instantaneous value of the wall distance. The so-called "positive correlation between reliability and followability" not only includes that the followability continuously improves with the increase of reliability, but also includes that the followability improves in stages with the increase of reliability.

[0159] In addition, in the case of continuous detection of a wall, when the variance of a plurality of instantaneous values ​​of the wall distance acquired during the continuous detection period of continuous detection of the wall is less than a preset variance judgment value JV1, the radar devices 11 and 12 set the reliability so that there is a positive correlation between the continuous detection period and the reliability. The so-called "there is a positive correlation between the continuous detection period and the reliability" includes not only that the reliability increases continuously with the increase of the continuous detection period, but also that the reliability increases in stages with the increase of the continuous detection period.

[0160] Therefore, for the radar devices 11 and 12, the higher the reliability of the instantaneous wall distance value, the better the followability of the wall filter value relative to the instantaneous wall distance value can be, and the situation where the wall shape is significantly different from the actual wall position can be suppressed.

[0161] In the above-described embodiment, the radar devices 11 and 12 correspond to the wall shape measuring device, S50 corresponds to the process as the wall detection determination unit, and the right wall and the left wall correspond to the wall-shaped object.

[0162] In addition, S60 corresponds to the processing as a wall distance acquisition unit, the wall distance instantaneous value corresponds to the wall distance value, and S230 corresponds to the processing as a wall shape calculation unit.

[0163] In addition, S170 is equivalent to the processing as the extrapolation unit, the second threshold Vth2 is equivalent to the prohibition judgment value, S160 is equivalent to the processing as the extrapolation prohibition unit, the value of the wall loss counter C2 is equivalent to the number of extrapolations, and S650~S730 are equivalent to the processing as the correctness judgment unit.

[0164] In addition, S740 and S750 correspond to the processing of the extrapolation number setting unit, the first extrapolation number N2_h corresponds to the appropriate time judgment value, and the second extrapolation number N2_1 corresponds to the inappropriate time judgment value.

[0165] In addition, S510 and S520 correspond to the processing as the undetectable determination unit, S40 corresponds to the processing as the obstacle determination unit, and S120 corresponds to the processing as the obstacle extrapolation unit.

[0166] S110 corresponds to processing as an obstruction extrapolation prohibition unit, the value of the undetectable wall counter C1 corresponds to the obstruction extrapolation count, the first threshold Vth1 corresponds to the obstruction prohibition determination value, and S80 to S100 correspond to processing as an obstruction extrapolation count setting unit.

[0167] In addition, the wall filter value FV is equivalent to the wall distance filter value, S320, S350, and S380 are equivalent to the processing as a reliability setting unit, S330, S360, S390, and S400 are equivalent to the processing as a filter constant setting unit, and the variance judgment value JV1 is equivalent to the variance judgment value.

[0168] In addition, the power of the beat signal corresponds to the power parameter, the frequency of the beat signal corresponds to the distance parameter, and the orientation of the wall relative to the vehicle corresponds to the orientation parameter.

[0169] As mentioned above, although one embodiment of the present disclosure has been described, the present disclosure is not limited to the above embodiment, and can be implemented with various modifications.

[0170] [Modification 1]

[0171] For example, in the above embodiment, the method of detecting the target object by using the FMCW method is shown, but the detection method is not limited thereto. For example, the target object may be detected by using a dual-frequency CW method or an FCM method.

[0172] [Modification 2]

[0173] In the above embodiment, a method of determining whether there are parallel vehicles by following the same object mark is shown, but the method of determining the presence of parallel vehicles is not limited to this. For example, in the dual-frequency CW method, when other vehicles are parallel near the vehicle, there is a situation where multiple peaks are generated in a wide frequency range in the power spectrum due to the reflected waves reflected by the wheels of the other vehicles. This is because the wheels have various speed components. Therefore, in the dual-frequency CW method, when multiple peaks are generated in a wide frequency range, it can be determined that there are parallel vehicles.

[0174] [Variation 3]

[0175] In the above-mentioned embodiment, a method of determining whether the judgment result of the wall detection is correct by processing S650 and S660 is shown. However, it is also possible to extract one or more peaks from the power spectrum representing the relationship between the power of the beat signal and the frequency of the beat signal, and determine whether the extracted peaks exist in the beat signal frequency corresponding to each of the multiple calculated wall positions WP(1) to WP(N), thereby determining whether the judgment result of the wall detection is correct. In addition, the above-mentioned power spectrum may include all directions that can be detected by the radar devices 11 and 12, or may only include the direction corresponding to the front side of the vehicle. Among them, the power spectrum that only includes the direction corresponding to the front side of the vehicle is easier to extract the peak corresponding to the wall position than the power spectrum that includes all directions.

[0176] The processing unit 70 and the method thereof described in the present disclosure may also be implemented by a special-purpose computer, which is provided by constituting a processor and a memory programmed to perform one or more functions embodied by a computer program. Alternatively, the processing unit 70 and the method thereof described in the present disclosure may also be implemented by a special-purpose computer provided by constituting a processor using one or more special-purpose hardware logic circuits. Alternatively, the processing unit 70 and the method thereof described in the present disclosure may also be implemented by one or more special-purpose computers, which are composed of a combination of a processor and a memory programmed to perform one or more functions, and a processor composed of one or more hardware logic circuits. In addition, a computer program may also be stored as an instruction executed by a computer in a non-transitional tangible recording medium that can be read by a computer. In the method for implementing the functions of each part included in the processing unit 70, it is not necessarily necessary to include software, and one or more hardware may be used to implement all its functions.

[0177] The multiple functions of one component in the above-mentioned embodiment can be realized by multiple components, or a function of one component can be realized by multiple components. In addition, multiple functions of multiple components can be realized by one component, or a function realized by multiple components can be realized by one component. In addition, a part of the structure of the above-mentioned embodiment can also be omitted. In addition, at least a part of the structure of the above-mentioned embodiment can also be added to or replaced with the structure of other above-mentioned embodiments.

[0178] In addition to the above-mentioned radar devices 11 and 12, the present disclosure can also be implemented in various ways such as a system that uses the radar devices 11 and 12 as components, a program for causing a computer to function as the radar devices 11 and 12, a non-transitional physical recording medium such as a semiconductor memory that records the program, a wall shape measurement method, etc.

Claims

1. A wall shape measuring device, mounted on a vehicle, comprising: a wall detection determination unit configured to repeatedly determine whether a wall-like object set along the road on which the vehicle is traveling is detected, using a reception signal obtained by receiving a radar wave transmitted to and reflected from a predetermined detection range around the vehicle; a wall distance acquisition unit configured to acquire a wall distance value indicating a distance to the wall-like object using the received signal when the wall detection and determination unit determines that the wall-like object is detected; a wall shape calculation unit configured to calculate a plurality of wall shape values ​​indicating positions of a plurality of locations on the surface of the wall-shaped object using the travel trajectory of the vehicle and the plurality of past wall distance values ​​repeatedly acquired by the wall distance acquisition unit; an extrapolation unit configured to extrapolate the wall distance value when the wall detection and determination unit determines that the wall-shaped object is not detected; an extrapolation prohibition unit configured to prohibit the extrapolation of the extrapolation unit when the number of extrapolation times by which the extrapolation unit continuously performs extrapolation is equal to or greater than a preset prohibition judgment value; A correctness judgment unit configured to judge whether a judgment result of the wall detection judgment unit before the extrapolation unit performs extrapolation is correct; and The extrapolation times setting unit is configured to set the prohibited judgment value to a pre-set appropriate judgment value when the correctness judgment unit determines that the judgment result of the wall detection judgment unit is correct, and to set the prohibited judgment value to an inappropriate judgment value when the correctness judgment unit determines that the judgment result of the wall detection judgment unit is incorrect, and the inappropriate judgment value is set to be smaller than the appropriate judgment value.

2. The wall shape measuring device according to claim 1, wherein: The apparatus comprises an undetectable determination unit configured to determine whether a preset undetectable condition indicating that the wall-like object cannot be detected continues to be satisfied. When the undetectable determination unit determines that the undetectable condition is satisfied, the wall shape calculation unit calculates the plurality of wall shape values ​​so that the distances between the driving track and the plurality of wall shape values ​​become preset initial values.

3. The wall shape measuring device according to claim 1, wherein: have: an obstacle determination unit configured to determine whether there is an obstacle that hinders the detection of the wall-like object; an obstacle extrapolation unit configured to extrapolate the wall distance value when the obstacle determination unit determines that the obstacle exists; an obstruction extrapolation prohibition unit configured to prohibit extrapolation by the obstruction extrapolation unit when the number of obstruction extrapolation times by which the obstruction extrapolation unit continuously performs extrapolation is equal to or greater than a preset obstruction prohibition determination value; and The obstruction extrapolation frequency setting unit is configured to set the obstruction prohibition determination value so as to have a positive correlation with a lane distance between a lane in which the vehicle is traveling and an adjacent lane in which the obstruction existing beside the vehicle is traveling.

4. The wall shape measuring device according to any one of claims 1 to 3, wherein: The wall shape calculation unit calculates a wall distance filter value obtained by filtering the wall distance value acquired by the wall distance acquisition unit, and further calculates a plurality of wall shape values ​​using the driving trajectory and the plurality of wall distance filter values. The above-mentioned wall shape measuring device comprises: a reliability setting unit configured to set the reliability of the wall distance value acquired by the wall distance acquisition unit; and The filter constant setting unit is configured to set the filter constant of the filtering process so that there is a positive correlation between the reliability set by the reliability setting unit and the followability of the wall distance filter value with respect to the wall distance value.

5. The wall shape measuring device according to claim 4, wherein: In the case where the above-mentioned wall-like object is continuously detected, when the variance of the multiple wall distance values ​​obtained during the continuous detection period of the above-mentioned wall-like object is less than a preset variance judgment value, the above-mentioned reliability setting unit sets the above-mentioned reliability so that there is a positive correlation between the above-mentioned reliability during the above-mentioned continuous detection period.

6. The wall shape measuring device according to any one of claims 1 to 3, wherein: The correctness judgment unit generates a plurality of azimuth spread spectrums by performing azimuth spread processing on the power spectrum under the distance parameter corresponding to each of the plurality of wall shape values ​​calculated by the wall shape calculation unit, wherein the power spectrum represents the relationship between the power parameter associated with the power of the received signal and the distance parameter associated with the distance to the wall-like object, and the azimuth spread spectrum represents the relationship between the power parameter and the azimuth parameter associated with the azimuth of the wall-like object, and whether the judgment result of the wall detection judgment unit is correct is judged by judging whether the power parameter in the azimuth parameter corresponding to each of the plurality of wall shape values ​​calculated by the wall shape calculation unit satisfies a preset azimuth power existence condition for each of the plurality of azimuth spread spectrums generated, wherein the azimuth power existence condition represents the existence of power.

7. The wall shape measuring device according to any one of claims 1 to 3, wherein: The correctness judgment unit judges whether the judgment result of the wall detection judgment unit is correct by extracting one or more peaks from the power spectrum and judging whether the extracted peaks exist at the position of the distance parameter corresponding to each of the multiple wall shape values ​​calculated by the wall shape calculation unit, wherein the power spectrum represents the relationship between the power parameter associated with the power of the received signal and the distance parameter associated with the distance to the wall-shaped object.

Images