JPWO2005091015A1 - Radar - Google Patents

Abstract

A beam is scanned over a detection range of a predetermined azimuth direction and a predetermined distance range, information on a plurality of targets obtained for each scan is compared, and the target detected for the same target is detected. The tracking reliability of the target is increased, and the tracking reliability is decreased when the tracking reliability cannot be detected. When the tracking reliability exceeds the first threshold (S1), the target is detected as a target. When it falls below the second threshold, it is detected as not being the target. As a result, when performing signal processing for target tracking, when the threshold value is set to a small value, incorrect information such as whether there is a target that does not actually exist is output. When the threshold value is set large, the problem that the determination of the target to enter the radar detection range is delayed is solved.

Description

  The present invention relates to a radar that detects a target and tracks the target.

  For example, in-vehicle radar not only detects the distance and relative speed of multiple targets within the detection range, but also tracks the positions of the necessary targets in order to ensure the safety of the vehicle and other vehicles. It is necessary to set a target target (hereinafter simply referred to as “target”) and track the target. This target tracking function is basically a function that continues tracking a target by repeating the process of extracting the target being tracked from a plurality of targets detected at each measurement cycle. It notifies the driver and performs control according to the presence of the target. For example, when the target is in a dangerous relationship with the own vehicle, the engine and the brake are automatically controlled as necessary.

  Conventionally, the probability that a target is detected at the same position for each measurement period is obtained, and when the probability exceeds a predetermined threshold value, it is assumed that the target actually exists. For example, in Patent Document 1, when a target is successfully detected at a target position for N times out of M detection operations, the target is regarded as a target being tracked. And in this patent document 1, in order to improve the detection performance in the vicinity of the maximum detection distance of the radar, the threshold value of N, that is, the value of N is set according to the distance to the target and the relative speed of the target. I try to change it.

Further, in Patent Document 2, the position of a target (detected target) detected by measurement is compared with the position of a target (stored target) that has already been stored, and the position of the detected target and the stored target When the position is determined to correspond, the value increases, and when it is determined that the position does not correspond, the “confidence level” decreases. And when there is a long-term loss that the certainty level falls below the predetermined threshold, the data relating to the target is deleted from the storage means, and the certainty level does not fall below the predetermined threshold. For loss of sight, the data related to the target is not deleted.
JP 11-38119 A Japanese Patent No. 3065821

  However, by lowering the threshold value N shown in Patent Document 1, it becomes easier to output an erroneous tracking result due to noise or a pairing error of a stationary object. That is, if N is decreased, the possibility that the detected target is the target being tracked decreases, but the risk of a collision or the like can be reduced by determining that the target is the target being tracked. However, there is no actual target, but if noise or the like is detected as a target (target target to be tracked), unnecessary braking is often applied to the vehicle by automatic control. Comfort is impaired. Further, this Patent Document 1 corresponds to processing related to target tracking in the vicinity of the maximum detection distance, and for example, it is not possible to quickly respond to prevention of collision with another vehicle trying to enter the vicinity of the own vehicle. For example, when another vehicle enters the vicinity of the maximum detection distance, if the threshold value N is set small, the entry can be detected quickly. However, even if the target goes out of the radar detection range, it tries to determine the presence or absence of the target with a small threshold N, so the target does not actually exist within the radar detection range. In some cases, information may be output. Conversely, if the threshold value N is increased, it can be quickly determined that the target has gone out of the radar detection range, but the determination of other vehicles entering the radar detection range is delayed, and safety is likely to be impaired. Problems arise.

  The certainty factor shown in Patent Document 2 can be one of the determination factors for controlling the vehicle, but is not a measure for measuring the reliability of signal processing for target tracking. This is apparent from Patent Document 2 in which, for example, an increase in the certainty factor is increased for an object approaching high speed. In this way, it is not always possible to have the same scale, so it is not possible to immediately and accurately respond to sudden changes in the situation.

  An object of the present invention is to provide a radar that solves the above-described problems in performing signal processing for target tracking and that can perform target tracking suitable for various situations.

  (1) The present invention detects object information including at least target position information by repeatedly transmitting electromagnetic waves to a predetermined detection range and receiving reflected waves from the target at predetermined observation timings. The target information detection means and the target information detected at a plurality of different observation timings by the target information detection means are compared, and when target information originating from the same target is detected, the target information Means for increasing the tracking reliability, lowering the tracking reliability when no target information derived from the same target is detected, means for determining the first and second threshold values, and the tracking reliability And a means for determining the target as a target when the first threshold is exceeded, and determining that the target is not a target when the second threshold is exceeded. .

  (2) In (1), the first and second threshold values are determined according to the received signal status such as the received signal strength and noise strength.

  (3) In (1), the first and second threshold values are determined according to the relative movement situation such as the relative position and relative speed of the target.

  (4) In (1), the radar is an on-vehicle radar, and includes means for detecting the state of the road on which the vehicle travels, and the first and second threshold values are determined according to the road environment.

  (5) In (1), the radar is an on-vehicle radar, and includes means for detecting vehicle conditions such as vehicle running speed, acceleration / deceleration, steering angle, yaw rate, and tilt angle. First and second threshold values are determined according to vehicle conditions.

  (6) In (1) to (5), the threshold value setting means increases or decreases the first and second threshold values in the same direction.

  (7) In (1) to (5), the threshold value setting means increases or decreases the first and second threshold values in different directions.

  (1) According to the present invention, the first threshold value for determining that there is a target to be tracked and the second threshold value for determining that it is no longer necessary to target the target being tracked. By defining the threshold value, the degree of freedom of signal processing for target tracking is increased, and signal processing for target tracking according to the purpose is enabled. For example, the smaller the first threshold value (hereinafter referred to as S1) is, the shorter the time required for determining the detected target as a target can be shortened when an other vehicle enters the vicinity of the host vehicle. Can be supported. For example, if the second threshold value (hereinafter referred to as S2) is set large, for example, when the target is far away from the vehicle, the tracking reliability is lower than S2 earlier. The other cars that travel far away can be quickly released from the target. As a result, information that does not hinder driving is quickly deleted, and the burden on the driver can be reduced. That is, the signal from the radar does not frequently perform automatic control of the engine and the brake by the function of adaptive cruise control (hereinafter referred to as ACC), and the driving comfort is increased.

  (2) By determining the first and second threshold values according to the received signal conditions such as received signal strength and noise strength, for example, the stronger the received signal strength, the lower the possibility that it is noise Therefore, by setting S1 to be small and S2 to be large, the target can be determined early, and the target release determination can also be accelerated. Conversely, when the received signal strength is weak, increasing S1 and decreasing S2 reduces the probability of noise being erroneously detected as a target, and the target once targeted is reliable even if the signal strength is low. Tracking can be continued.

  (3) By determining the first and second threshold values in accordance with the relative motion status such as the target relative position and relative speed, optimization according to the target motion status can be achieved. For example, when this radar is an on-vehicle radar, when S1 is small when the relative speed in the approaching direction of the other vehicle is high, and S1 is large when the relative speed is low, the target vehicle is to track other vehicles. Judge early. Further, when the relative speed in the approaching direction is low, there is a time margin, and by increasing S1 accordingly, the probability that a target that does not exist or a target that does not require tracking will be erroneously detected as a target target is increased. Can be small.

  (4) Optimizing signal processing for target tracking according to road conditions by detecting the conditions of the road on which the vehicle travels and determining the first and second threshold values according to the road conditions Can be realized.

  (5) By detecting vehicle conditions such as the vehicle running speed, acceleration / deceleration, steering angle, yaw rate, and tilt angle, and determining the first and second threshold values according to the vehicle conditions, It is possible to optimize signal processing for target tracking according to the target.

  (6) By increasing or decreasing the first and second threshold values in the same direction, the first and second threshold values can be determined according to the weight of safety and comfort. For example, if both the first and second threshold values are reduced, the setting omission as a target to be tracked is reduced, and safety can be improved. Conversely, if both the first and second threshold values are increased, the probability of tracking an unnecessary target as a target is reduced, and comfort is improved without unnecessary braking by ACC.

  (7) By increasing / decreasing the first and second threshold values in different directions, when the relationship of the first threshold value <the second threshold value is satisfied, the target tracking start timing is advanced and the target value is As a result, it is possible to track only a target that needs to be continuously tracked. On the contrary, when the relationship of the first threshold value> the second threshold value is satisfied, it is possible to reduce the number of erroneous detections and reliably track the target once tracking has been started.

It is a block diagram which shows the structure on the hardware of a radar. It is a functional block diagram regarding the signal processing of the radar. It is the figure which plotted the position in a certain timing of the detected several target. It is a figure shown about the influence of the setting of the 1st and 2nd threshold value. It is a flowchart regarding the tracking process of a target. It is a flowchart regarding the setting process of a threshold value. It is a figure which shows the example of a setting of the threshold value according to the condition of the road where the own vehicle is drive | working. It is a figure which shows the example of a setting of the threshold value according to the vehicle condition of the own vehicle. It is a figure which shows the example of a setting of the threshold value according to the relative motion condition of the detected target. It is a figure which shows the example of a setting of the threshold value according to a received signal condition and other conditions.

Explanation of symbols

4-Antenna 30-Tracking processing control unit SA-Detection range

A radar according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing the configuration of the entire system including a vehicle-mounted radar and various units connected thereto. In FIG. 1, a portion indicated by 20 is a radar front end, and includes a control circuit 1, a millimeter wave circuit 2, a scan unit 3, an antenna 4, and the like. Here, the millimeter wave circuit 2 modulates the oscillation frequency with a modulation signal supplied from the control circuit 1 as described later, and outputs the transmission signal to the antenna 4 via the scan unit 3. The received signal is given to the control circuit 1 as an intermediate frequency signal (IF signal). The scan unit 3 scans the direction of the beam of the antenna 4 over a predetermined range, for example, by mechanical reciprocation.

  The control circuit 1 gives a modulation signal to the millimeter wave circuit 2 and obtains the distance to the target and the relative speed based on the IF signal from the millimeter wave circuit 2. Further, the control circuit 1 outputs a control signal to the scan unit 3 and directs the beam of the antenna 4 in a predetermined direction to scan in the azimuth direction of the detection range to obtain the azimuth of the target. Thereby, the relative motion state of the target is detected. Further, the control circuit 1 detects a received signal condition such as received signal strength and noise strength.

  The mode switch 10 is a switch that switches between controlling the ACC in the collision prevention mode and controlling the ACC in the follow-up running mode. The tracking process control unit 30 reads the state of the mode switch 10 and sets the first and second threshold values as will be described later. In addition, the tracking processing control unit 30 inputs signals from the vehicle speed sensor 11, the car navigation (car navigation device) 12, the ETC (Electronic Toll Collection System) 13, and other various sensors 14 to detect the road on which the vehicle travels. Detects environment and vehicle conditions. Then, the target target information is given to the ACC controller 15.

  The ACC controller 15 performs automatic cruise control based on the vehicle speed obtained by the vehicle speed sensor 11 and the relative position and relative speed given from the control circuit 1. For example, control data is given to the engine control unit 16 and the brake control unit 17 so that the distance between the vehicle and the preceding vehicle is always kept constant. Also, control data for avoiding a collision with a target ahead such as a preceding vehicle is given.

  The engine control unit 16 and the brake control unit 17 perform engine control and brake control based on the control data given from the ACC controller 15.

  FIG. 2 is a block diagram of signal processing of the tracking process control unit 30 shown in FIG. Here, the information input means 35 is a signal processing unit for inputting information from the mode switch 10, the vehicle speed sensor 11, the car navigation system 12, the ETC 13, and the various sensors 14 shown in FIG. 1.

  The tracking processing unit 31 inputs the target feature information such as the distance, relative speed, and azimuth of each target input from the radar front end 20, and determines whether or not it is the target being tracked, that is, during tracking. It is determined whether or not the signal is generated due to the same target as the target.

  The tracking reliability calculation unit 32 obtains “tracking reliability” that is an index of reliability that can be regarded as being correctly tracked based on the result of the tracking processing unit 31.

  The threshold value setting unit 33 sets a threshold value by a method described later based on various information input via the information input means 35.

  The threshold determination unit 34 determines whether or not the tracking reliability calculated by the tracking reliability calculation unit 32 exceeds the first threshold set by the threshold setting unit 33. When the target is determined as a target and falls below the second threshold, it is determined that the target is no longer the target. And the relative position and relative motion information about these several targets are output. The ACC controller 15 shown in FIG. 1 controls the engine control unit 16 or the brake control unit 17 based on this information.

  FIG. 3 is an example in which the positions at a certain timing of a plurality of targets detected by the radar front end 20 shown in FIG. 2 are plotted. In FIG. 3, O represents the radar antenna position (the position of the vehicle), and SA represents the radar detection range. Thus, by repeatedly scanning over the left and right θ in front of the vehicle centering on the radar position O, detection is performed over the angle θ in the azimuth direction and the distance L in the distance direction. C1 and C2 are other vehicles existing in front of the vehicle. In addition, the positions of many targets existing in the range surrounded by LW and RW are obtained by detecting points on the side walls provided on the left and right sides of the road as targets. When such side walls LW and RW are detected as targets, since these targets are stationary objects, they approach the host vehicle direction at the host vehicle speed.

  FIG. 4 shows the influence of the setting of the first and second threshold values. When the tracking reliability exceeds the first threshold value S1, the target is set as a target to be tracked. However, when the threshold value S1 is set small, tracking of the target is accelerated while the relative speed of the side wall or the like is set. The possibility of detecting it erroneously and tracking it as a moving object increases. On the other hand, when the threshold value S1 is increased, the possibility of tracking a target such as a side wall is reduced, but tracking of the target that should be tracked tends to be delayed. For example, in FIG. 4A, LW and RW are left and right side walls, C3 is a target position of a vehicle traveling in front of the own vehicle, and C2 is a target position of a vehicle traveling on the right side of the own vehicle. Is shown. As shown in this figure, even if a certain part of the side wall is detected as a target CW due to a pairing error or the like at a certain timing (one scanning period), such a false detection occurs every time scanning is performed. However, the tracking reliability is not high because the continuity is poor. Therefore, erroneous tracking can be prevented by appropriately increasing S1. On the other hand, if the first threshold value S1 is reduced, erroneous tracking cannot be prevented, but if a new target C2 enters the detection range SA in front of the vehicle, The tracking start of the target C2 can be accelerated.

  FIG. 4B shows the influence of the setting of the second threshold value S2. Here, N indicates a range of a target that is likely to be erroneously detected due to the influence of fixed pattern noise such as a clock of a radar switching power supply circuit or a signal processing circuit or a driving signal of a scanning mechanism. In such an area where the influence of the fixed pattern noise is large, detection of the position and relative speed of the target becomes inaccurate, and tracking of the target tends to fail and the tracking reliability tends to decrease. C4 'is the position of the target detected under the influence of the noise. Further, C3 indicates the target position of the vehicle traveling in front of the host vehicle, and C5 indicates the target position of the vehicle that is going down to the right rear of the host vehicle.

In such a case, the second threshold value S2 is set small for the range N that is likely to be erroneously detected. This can reduce the tendency of the target to be canceled even though it actually exists.
Note that the target of the vehicle (C5) that has moved outward from the detection range SA is not set (target release) regardless of the tracking reliability.

Next, processing contents of the tracking processing control unit 30 shown in FIGS. 1 and 2 will be described with reference to FIGS.
FIG. 5 is a flowchart regarding target tracking processing. First, target information such as the target's relative position, relative speed, and received signal strength profile in the azimuth direction, and the target model being tracked (target position, moving speed, scattering cross section, etc.) It is determined whether or not it is possible to associate the position of the target with the information capable of predicting the speed (ST1). If a new target is detected without being associated with the target currently being tracked, an initial value is set for the tracking reliability RC for that target, and a model for that target is created. (ST2->ST3-> ST4).

  If the association between the new target information in the current scan and the information of the model of the target being tracked succeeds in step ST1, the tracking reliability RC for the target is increased (ST1 → ST2 → ST5). ). For example, the tracking reliability RC takes an integer value, and is incremented by 1 in this step ST5. Thereafter, the target model information of the target is updated (ST6). For example, information such as position, relative velocity, and scattering cross section is updated to the latest value. Subsequently, when the tracking reliability RC exceeds the first threshold value S1, the valid flag of the target is set (ST7 → ST8). That is, the target is set as one of the targets.

  For the tracking target model that cannot be matched with the new target information obtained by the current scan among the plurality of tracking target models in step ST1, the tracking reliability RC of the target is decreased (ST1 → ST2 → ST9). For example, RC is decreased by 1. When the tracking reliability RC falls below the second threshold value S2, the valid flag of the target is reset (ST10 → ST11). That is, the target is removed from the target.

  As described above, the tracking reliability RC is increased or decreased for a plurality of detected targets, and a plurality of tracking targets are compared by comparing the first and second threshold values S1 and S2 with the tracking reliability RC. Are extracted as targets that exceed the first threshold value S1. Also, the tracking target that falls below the second threshold value is released from the target.

  FIG. 6 shows the threshold value setting process. In the example shown in (A), the first and second threshold values S1 and S2 are determined for each target according to the received signal strength and the strength of a signal (noise) that can be regarded as noise for target detection. In the example of (B), first and second threshold values S1 and S2 are determined for each target in accordance with the relative motion status of the target. In the example shown in (C), the first and second threshold values S1 and S2 are determined according to road conditions detected based on car navigation, ETC, inclination angle, steering angle, yaw rate, and the like. In the example shown in (D), the first and second threshold values S1 and S2 are determined according to own vehicle information such as the vehicle speed of the own vehicle and the health condition of the driver. In the example shown in (E), the first and second thresholds are set according to the state of a mode switch that switches between, for example, a collision prevention mode that places importance on safety or a follow-up running mode that automatically runs following a preceding vehicle. Values S1 and S2 are set.

Next, specific setting examples of the first and second threshold values S1 and S2 will be described with reference to FIGS.
FIG. 7 shows a threshold value setting method according to the condition of the road on which the vehicle is traveling.
FIG. 7A shows an example of setting a threshold value according to the planar shape of the road on which the vehicle travels. When the radius of curvature of the road is larger than a predetermined value, that is, while the vehicle is traveling on a curve, the beam is irradiated obliquely from behind the preceding vehicle, so that the radar scattering cross section (hereinafter referred to as RCS) is reduced, and the scanning cycle. In some cases, the reflected signal cannot be detected every time, and the reflection position in the same target changes, so that the probability of erroneous detection processed as if it is a different target is also increased. Therefore, when the radius of curvature of the road is smaller than a predetermined value, S2 is set smaller than the average value. Or, the smaller the curvature radius, the smaller S2. This makes it possible to reliably capture the target once targeted. In addition, in a curve with a small curvature radius, the probability of erroneous detection of a stationary object as a moving object increases. Therefore, when the radius of curvature is smaller than a predetermined value, S1 is set larger than the average value. Alternatively, S1 is increased as the radius of curvature is smaller. This can reduce the probability of erroneous detection.

  On the contrary, when traveling on a road with a large radius of curvature, the signal strength of other vehicles traveling ahead is high and the relative position is also stable, so there are few false detections. Therefore, when the radius of curvature is larger than a predetermined value, S1 is made smaller than the average value. Alternatively, S1 is decreased as the curvature radius increases. As a result, a target that falls within the detection range can be processed as a target at an early stage. Further, when the radius of curvature is larger than a predetermined value, S2 is made larger than the average value. Alternatively, S2 is increased as the radius of curvature increases. As a result, a target that has gone out of the detection range can be removed from the target at an early stage, unnecessary tracking can be suppressed, and processing capacity can be allocated to necessary signal processing.

  Such a planar shape of the road can be detected based on the steering angle of the own vehicle, the output of the yaw rate sensor, or the map information from the car navigation system.

  FIG. 7B shows a threshold value setting method according to the vertical plane shape of the road on which the vehicle travels, that is, the inclination angle of the road. On the slope, the gain decreases according to the directivity of the antenna in the vertical direction, and the received signal strength decreases. Therefore, the probability that the continuity of target detection for each scanning period is lost increases. Therefore, S2 is made smaller than the average value when the inclination angle of the road on which the vehicle is traveling is inclined from a predetermined angle. Alternatively, S2 is decreased as the inclination angle increases. As a result, the target once detected as a target can be reliably captured. Also, on such a slope, the frequency at which a target newly entering the detection range cannot be detected increases, so S1 is also set smaller than the average value. Alternatively, S1 is decreased as the inclination angle increases. This makes it possible to set a target within the detection range as a target at an early stage.

  On the contrary, since the received signal strength is high and there are few false detections on flat ground, when the road inclination angle is gentler than a predetermined angle, S1 is made smaller than the average value. Alternatively, S1 is made smaller as it is gentler. As a result, on a flat ground, a target that falls within the detection range can be detected as a target at an early stage. Further, since there are few false detections on flat ground, S2 is set larger than the average value as in the case of traveling on a straight road. Alternatively, S2 is increased as the surface becomes flatter. As a result, a target that has gone out of the detection range can be removed from the target at an early stage, unnecessary tracking can be suppressed, and processing capacity can be allocated to necessary signal processing.

  FIG. 7C shows an example of threshold setting according to the type of road on which the vehicle is traveling. While driving on an expressway, the situation of the target ahead changes relatively slowly. In such a situation, it is not necessary to set the target as early as possible, and S1 is made larger than the average value in order to make it less susceptible to noise. Thereby, it becomes possible to detect the target as a target when the presence of the target becomes certain. Further, since the target can be detected stably during traveling on the highway, S2 is made smaller than the average value. By reducing S2, even if the tracking reliability of the target once captured falls to some extent, it can be captured as a target.

On the other hand, during traveling on a general road, S1 is made smaller than the average value, and S2 is made larger than the average value. As a result, a new target is recognized as a target at an early stage, and when the tracking reliability of the target starts to decrease (that is, when the existence of the target becomes suspicious), the target can be quickly removed from the target.
The type of road on which the vehicle is traveling may be detected from information from a car navigation system or ETC.

  (D) of FIG. 7 is an example of setting a threshold value according to the weather condition during traveling. In bad weather, the received signal strength decreases, and continuity for each scanning cycle may not be maintained. Therefore, S2 is made smaller than the average value when the weather is worse than a predetermined state. Or S2 is made small so that it is bad weather. Thereby, the target once detected as a target can be reliably captured. Further, since the probability of occurrence of an accident such as a slip accident becomes high under such bad weather, S1 is also made smaller than the average value. Or S1 is made small so that it is bad weather. In this way, even if the probability of erroneous detection increases, early response can be achieved by reducing S1. Such weather information may be obtained by receiving a weather sensor provided in the own vehicle or broadcast information.

FIG. 8 shows a setting method according to the vehicle situation of the own vehicle.
FIG. 8A shows a threshold setting method according to the physical condition (health state) of the driver. Based on the driver's self-declaration, driving conditions obtained from external sensors, etc., the driver's mental and physical health is judged, and when malfunctioning, both S1 and S2 are set small. Physical condition that tends to be slow to respond to dangerous conditions (e.g., injuries such as reduced limb movement ability due to injury, visual impairment such as bird's eye or narrowing of visual field, continuous driving for a long time, lack of sleep, When S1 is small in the case of a hungry state, a mental depression state, etc., a quick response to the target (response by ACC) becomes possible. In addition, by reducing S2 as well, a target with low tracking reliability can be regarded as a target, and safety is improved.

  FIG. 8B shows an example of setting a threshold value according to the loading amount in a truck or the like. If the load is large, both S1 and S2 are set small. If the load is large, the brake will become less effective. In such a case, it is possible to deal with a new target earlier by reducing S1. Also, by reducing S2, it becomes possible to handle cautiously even for targets with low tracking reliability. The load amount can be obtained from a sensor or input externally.

  FIG. 8C shows an example of setting a threshold for the traveling speed of the vehicle. Thus, when the traveling speed of the own vehicle is higher than a predetermined speed, S1 is set larger than the average threshold value. Alternatively, S1 is increased as the speed increases. This prevents sudden braking due to erroneous target detection. Further, by making S2 smaller than S1, a target once detected as a target can be reliably captured.

  On the other hand, S1 is set to be smaller than the average threshold during low-speed traveling. Alternatively, S1 is reduced as the speed decreases. As a result, it is possible to quickly detect a target that moves in the transverse direction with respect to the irradiation direction of the radar beam, such as a passing vehicle, a crossing vehicle, a crossing person, etc., and an early response is possible. Also, by making S2 larger than S1, the target once released as a target can be released quickly, and a variety of targets can be quickly dealt with.

FIG. 9 shows an example in which the threshold value is set according to the detected relative motion status of the target.
FIG. 9A shows an example of setting a threshold according to the distance from the vehicle to the target. When the distance to the target is farther than a predetermined distance, S1 is increased, and when it is a short distance, S1 is decreased. Alternatively, S1 increases as the distance increases, and S1 decreases as the distance decreases. In this way, as the distance to the target is farther away, there is a time allowance for determining whether or not the target is the target, so that the detection error can be reduced by increasing S1 accordingly. . On the other hand, when the distance is short, S1 becomes small, and safety is more important than erroneous detection, and the collision determination can be performed quickly.

  FIG. 9B shows a threshold setting method for the relative speed of the target. When the target is approaching at high speed in the direction of the vehicle, there is no time to detect it as a target. Therefore, S1 is set smaller when approaching faster than a predetermined speed. Alternatively, S1 is reduced as the speed approaches. Thus, by making S1 small, it can be aimed at an early stage, and a collision determination can be performed quickly. On the contrary, when the relative speed is small in the approaching direction or in the direction away from the own vehicle, S1 is increased because there is a time margin in determining whether or not the target is the target. This can reduce false detection accordingly.

  FIG. 9C is an example of setting a threshold value according to the RCS of the target. Generally, in the vicinity of a target with a large RCS, the level of the received signal strength rises over a wide azimuth angle, and there is a tendency to cover the peak of the received signal strength of a target with a small RCS. That is, it becomes difficult to detect a target with a small RCS. Therefore, when a target having a large RCS exists in the vicinity, S2 is determined according to the RCS size of the target in the vicinity. For example, the larger the RCS of a nearby target, the smaller S2. Thereby, the target once detected as a target can be reliably tracked.

  In general, a virtual image due to noise or side lobe tends to appear around a target having a large RCS. Therefore, when a target having a large RCS exists in the vicinity, S1 is determined according to the RCS size of the target in the vicinity. That is, as the RCS of the nearby target is larger, S1 is increased. When the RCS of the nearby target is smaller, S1 is decreased. Thereby, the probability of erroneous detection of a virtual image around a target with a large RCS can be reduced.

  (D) of FIG. 9 is an example of setting a threshold value for the driving temper of the driver of another vehicle around the own vehicle. When the driving temper of the other vehicle driver is “impatient”, S1 is set smaller than the average value. For example, “impatience” is evaluated on the basis of the overtaking frequency of the vehicle, the inter-vehicle distance at the time of overtaking, the acceleration at the time of overtaking, and S1 is reduced when a predetermined evaluation score is exceeded. Alternatively, S1 is made smaller as it exceeds. In this way, an early response is possible when a highly dangerous vehicle is present in the vicinity.

FIG. 10 shows an example in which a threshold value is set according to the received signal status and other situations.
FIG. 10A shows an example of setting a threshold for the received signal strength. When the received signal strength is weaker than a predetermined value or weaker, S1 is set larger. Thereby, the false detection which detects noise as a target can be decreased. Further, when the received signal strength is weak or weak, S2 is set smaller. As a result, the target once detected as a target can be more reliably maintained. Conversely, for a target having a high received signal strength, since it is unlikely that it is caused by noise, S1 is reduced accordingly. Alternatively, as the received signal strength increases, the time until detection as a target can be shortened by decreasing S1. Also, S2 is set to be large for a target having a strong received signal strength. Alternatively, S2 is increased as the received signal strength increases. For such a target having a strong received signal strength, the tracking can be reliably continued, so that the target that is no longer the target can be canceled early by increasing S2.

  FIG. 10B shows an example of setting a threshold value according to the noise intensity. In a radar that obtains a distance to a target by frequency analysis, there is a region that is affected by noise of a fixed pattern as shown in FIG. Since the frequency of erroneous detection is high at such a position, the erroneous detection can be reduced by setting S1 large. Further, by setting S2 to be small in that area, it becomes possible to more reliably track a target that is difficult to continue tracking.

  FIG. 10C shows a method for setting a threshold value inside and outside the radar detection range. Since the target information is not necessarily required for a target outside the detection range, both S1 and S2 are set higher than the upper limit value of the tracking reliability. As a result, the tracking reliability is updated outside the detection guarantee range, but is not treated as a target.

  FIG. 10D shows a threshold value setting method according to the gain that changes according to the beam direction. In this example, both S1 and S2 are set to be smaller toward the outside of the scanning range. By reducing S2 in this way, it becomes easier to maintain tracking of the target in the direction in which the gain of the antenna is reduced. Further, by making S1 small, an interrupting vehicle of another vehicle enters near the outside of the detection range, so that it is possible to deal with those targets at an early stage.

  FIG. 10E shows an example of setting a threshold value according to a time zone. When statistics of the occurrence probability of road traffic accidents are taken for each time zone, there is a certain distribution, and there are times when the accident occurrence rate is high. For example, it may be 4:00 to 6:00 in the evening, or a time zone in which the time zone in which the toll on the expressway is approaching is close. In such a time zone, both S1 and S2 are set smaller than the average value. As a result, safety-oriented operation control is possible.

  In addition, for example, when obtaining information on the accident occurrence rate at each point or section of the road statistically obtained in advance from the outside, and traveling in a point or section with a high accident occurrence rate, Both S1 and S2 may be set small. As a result, safety-oriented operation control becomes possible.

  In any of the cases shown in FIGS. 7 to 10, if the tracking reliability exceeds S1 in the relationship of S1 <S2, the target is targeted, but thereafter, the measurement timing until the tracking reliability exceeds S2 If the target cannot be detected, the target may be immediately removed from the target at that time, or the target may be regarded as the target until the tracking reliability falls below S1. .

  In the examples shown in FIGS. 6 to 10, the first and second threshold values are set for various conditions. However, these conditions are not applied independently. Instead, S1 and S2 may be determined so as to satisfy a plurality of conditions simultaneously. For example, the upper right column of each table shown in FIGS. 7 to 10 is the antecedent part, the contents of the table are the consequent part, and finally S1 and S2 are generated one by one by fuzzy inference using min-max centroid method etc. You may make it do.

Claims (7)

Target information detecting means for detecting transmission of electromagnetic waves to a predetermined detection range and reception of reflected waves from the target at predetermined observation timings to detect target information including at least target position information;
The target information detected at a plurality of different observation timings by the target information detection means is compared, and when the target information originating from the same target is detected, the tracking reliability of the target is increased. Means for lowering the tracking reliability when the target information resulting from the same target is not detected;
Means for respectively determining first and second threshold values;
Means for determining the target as a target when the tracking reliability exceeds a first threshold, and determining that the target is not a target when the tracking reliability falls below a second threshold;
A radar comprising:   2. The radar according to claim 1, wherein the threshold value setting means determines the first and second threshold values according to received signal conditions such as intensity of the received signal and noise intensity.   2. The radar according to claim 1, wherein the threshold value setting means determines the first and second threshold values according to a relative movement situation such as a relative position and a relative speed of the target being tracked.   The radar is provided in a vehicle, and includes means for detecting a situation of a road on which the vehicle travels, and the threshold value setting means sets the first and second threshold values according to the environment of the road. The radar according to claim 1.   The radar is provided in a vehicle, and includes means for detecting a vehicle condition such as a traveling speed, acceleration / deceleration, steering angle, yaw rate, and tilt angle of the vehicle, and the threshold value setting means corresponds to the vehicle condition. The radar according to claim 1, wherein the first and second threshold values are defined.   The radar according to claim 1, wherein the threshold value setting means increases or decreases the first and second threshold values in the same direction.   The radar according to claim 1, wherein the threshold value setting means increases or decreases the first and second threshold values in different directions.

Images