RADARAPPARAT

Abstract

Radar device for acquiring information about a target, the radar device comprising: a direction derivation part (72) that derives a peak angle indicating an angle of the target, based on received signals received by multiple antennas (51); a memory (65) that stores internal data indicating the target, the internal data being derived based on the peak angle; a reliability determination part (75) that determines the reliability of a first peak angle, which is the peak angle relevant for the internal data, based on a second peak angle, which is another peak angle derived simultaneously with the first peak angle by the direction derivation part (72); and a data erasure part (76) that, based on the reliability of the first peak angle determined by the reliability determination part (75),excludes the internal data relevant for the first peak angle from the processing objects, wherein the reliability determination part (75) is configured to determine, in a case where the number of angles is two or three, whether the reliability of the first peak angle is high based on a lateral position difference; and to determine that the reliability of the first peak angle is high in response to the determination that the lateral position difference is equal to or less than a normal vehicle width, wherein the number of angles represents the number of peak angles derived simultaneously by the direction derivation part (72), and wherein the lateral position difference is determined by a difference between a first lateral position of the target relative to the radar device based on the first peak angle and a second lateral position of the target relative to the radar device.which is based on the second peak angle, is shown...

Description

GENERAL STATE OF THE ARTFIELD OF THE INVENTIONThe invention relates to a technology for acquiring information about a target.DESCRIPTION OF THE PRIOR ARTConventionally, in a vehicle control system following another preceding vehicle, a vehicle control system that reduces collision with an obstacle or another system, a radar device that acquires information on a target present around a vehicle is used.The radar device transmits a transmission wave and receives a reflection wave reflected by a target such as another vehicle, and then acquires information on the target such as a distance and a relative speed on the basis of a reception signal. A present radar apparatus receives a reflection wave from a target through a plurality of antennas, and estimates the angle of the target (the direction toward the vehicle) based on the phases of the reception signals received by the plurality of antennas. As an angle estimation system that estimates an angle of a target, ESPRIT, MUSIC, and PRISM are well known.In the case that there are multiple targets at substantially the same distance, the angle estimation system estimates the angle of each of the multiple targets as above. However, in each of the angle estimation systems, a possible division number that limits the number of targets whose angles are correctly estimated. For example, the possible division number in ESPRIT is "3". In the case where a radar device receives the reflection waves reflected by the targets present at substantially the same distance and the number of which exceeds the possible division number, the radar device cannot correctly estimate the angles of the targets, and the reliability of the estimated angles is lowered.The phenomenon in which the reliability of the angles estimated by a radar device is lowered may occur in some edge environments where a vehicle is moving regardless of the number of actual targets. For example, when a vehicle travels in a tunnel, even if only an actual target is present, a radar device receives the reflection waves from multiple directions. Consequently, the radar device is in a state as if the radar device receives the reflection waves from the targets that are present at substantially the same distance and the number of which exceeds the possible division number. As a result, the radar apparatus cannot correctly estimate the angles, and the reliability of the angles estimated by the radar apparatus is loweredIn such a case that the reliability of the angles estimated by a radar device is lowered, a vehicle control system may perform erroneous control based on improper information about a target.From WO 2013 / 175558 A1 and post-published EP 2853915 A1 a radar device is known, wherein an azimuth of a target is determined on the basis of the phase of signals received at a plurality of antennas. US 2012 / 0249359 A1 shows a calculation for estimating angles of a target object in a radar device. DE 102012206511 A1 discloses a radar apparatus for detecting an azimuth of a target, wherein a virtual azimuth and a virtual power are determined for each of first and second frequency components. US 2010 / 0090900 A1 relates to an angle measurement for received signal waves, wherein a reliability determination is carried out.SUMMARY OF THE INVENTIONAccording to one aspect of the invention, a radar apparatus according to claim 1 acquires information on a target. The radar apparatus includes deriving means that derives an apex angle showing an angle of the target based on reception signals received by a plurality of antennas, storage means that stores internal data showing the target, the internal data being derived based on the apex angle, determination means that determines a reliability of a first apex angle, which is the apex angle relevant to the internal data, based on a second apex angle, which is another apex angle derived simultaneously with the first apex angle by the deriving means, and exclusion means that excludes, based on the reliability of the first apex angle determined by the determination means, the internal data relevant to the first apex angle from the processing objects.Further, the present invention provides a method according to claim 6.The exclusion means excludes, based on the reliability of the tip angle, the relevant ones of the processing objects for the tip angle. Consequently, it is possible to exclude the internal data in which the reliability of the apex angle is low.According to another aspect of the invention, the exclusion means excludes, based on results, the determination means determines the reliability of the first apex angle relevant to the internal data showing a same target a plurality of times, the internal data from the processing objects.It is possible to exclude the internal data in which the reliability of the apex angle is kept low.According to another aspect of the invention, the radar apparatus further includes output means that outputs, to a vehicle control apparatus that controls a vehicle, the information on the target shown by the internal data that is not excluded from the processing objects by the exclusion means.It is possible to prevent the vehicle controller from erroneously controlling the vehicle based on the information about the target shown by the internal data in which the reliability of the tip angle is low.Therefore, an object of the invention is to exclude internal data showing a target from the processing objects in the case where the reliability of a tip angle is low.These and other objects, features, aspects and advantages of the invention will become more apparent from the following detailed description of the invention when taken in conjunction with the accompanying drawings.BRIEF DESCRIPTION OF THE DRAWINGSFIG. 1 shows a configuration of a vehicle control system. FIG. 2 shows a configuration of a radar device. FIG. 3 shows a relationship between a transmission wave and a reception wave. FIG. 4 shows an example of frequency spectra. FIG. 5 shows an example of tip angles. FIG. 6 shows a flow of recognition information acquisition processing. FIG. 7 shows a case example where the radar apparatus is used. FIG. 8 shows a flow of angular reliability determination processing. FIG. 9 shows reliability conditions. FIG. 10 shows operation patterns for handling a value of a low angle reliability counter. FIG. 11 shows an example of a change in a value of the low angle reliability counter. FIG. 12 shows an example of a change of a value of a full extrapolation counter. FIG. 13 shows a transition example of a inferred longitudinal distance and a inferred lateral position. FIG. 14 shows another transition example of the derived longitudinal distance and the derived lateral position.DESCRIPTION OF THE EMBODIMENTSHereinafter, some embodiments of the invention will be described based on the attached drawings.1. CONFIGURATIONFIG. 1 shows a schematic configuration of a vehicle control system 10 of the embodiment. The vehicle control system 10 is installed in a vehicle such as an automobile. Hereinafter, the vehicle in which the vehicle control system 10 is installed will be referred to as "own vehicle". As shown in FIG. 1, the vehicle control system 10 includes a radar device 1 and a vehicle controller 2.The radar device 1 acquires information on a target present around the own vehicle (hereinafter, referred to as "target recognition information"). The radar device 1 of the embodiment acquires information on a vehicle such as another vehicle present in front of the own vehicle by using a FMCW (Frequency Modulated Continuous Wave). The radar device 1 acquires target information including a distance (m) to the target in the traveling direction of the own vehicle, a relative speed (km / h) of the target to the own vehicle, and a distance (m) to the target in the right and left direction of the own vehicle (hereinafter referred to as "lateral position"), and outputs the acquired target recognition information to the vehicle controller 2 thereafter.The vehicle control device 2 connected to a brake, a throttle, and others of the own vehicle control elements controls the behavior of the own vehicle on the basis of the target recognition information output from the radar device 1. The vehicle control device 2, while maintaining a following distance between the own vehicle and another vehicle traveling in front of the own vehicle, controls the behavior of the own vehicle to follow the other vehicle. This allows the vehicle control system 10 of the embodiment to function as an ACC (Adaptive Cruise Control) system.FIG. 2 shows a configuration of the radar apparatus 1. the radar apparatus 1 mainly includes a transmitting part 4, a receiver 5, and a signal processor 6.The transmitting part 4 includes a transmitter 41 and a signal generator 42. The signal generator 42 generates a modulation signal whose voltage changes in a triangular waveform, and provides the signal to the transmitter 41. The transmitter 41 generates a transmission signal that changes in frequency as time passes by performing frequency modulation of a continuous wave signal based on the signal generated by the signal generator 42, and thereafter outputs the generated signal to a transmission antenna 40.The transmission antenna 40 outputs, on the basis of the transmission signal output by the transmitter 41, a transmission wave TW outside the own vehicle. The transmission wave TW to be output by the transmission antenna 40 corresponds to an FMCW that periodically changes in frequency up and down. The transmission wave TW output by the transmission antenna 40 toward the front of the own vehicle becomes a reflection wave RW when the transmission wave TW is reflected by a target such as another vehicle.The receiver 5 includes a plurality of reception antennas 51 constituting an array antenna and a plurality of individual receivers 52 each connected to each of the plurality of reception antennas 51. In the embodiment, the receiver 5 includes, for example, four of the reception antennas 51 and four of the individual receivers 52. Each of the four individual receivers 52 corresponds to each of the four reception antennas 51. Each of the individual receivers 52 processes the received signal received by the corresponding one of the receiving antennas 51.Each of the individual receivers 52 includes a mixer 53 and an A / D converter (analog-to-digital converter) 54. The reception signal detected based on the reflection wave RW received by each of the reception antennas 51 is passed to the mixer 53 after being amplified by a low self-noise amplifier (not shown in FIG. 2 ). The transmission signal is transmitted from the transmitter 41 of the transmission part 4 to the mixer 53, and the mixer 53 performs mixing of the reception signal and the transmission signal. This generates a beat signal showing a beat frequency corresponding to the difference between the frequency of the transmission signal and the frequency of the reception signal. The beat signal generated by the mixer 53 is output to the signal processor 6 after being converted into a digital signal by the A / D converter 54.The signal processor 6 has a microcomputer including a CPU and a memory 65. The signal processor 6 stores various data to be subjected to calculation in the memory 65, which is a storage device. The memory 65 is, for example, a RAM. The signal processor 6 includes a transmission controller 61, a Fourier transformer 62, and a data processor 7, which are functions implemented by software in the microcomputer. The transmission controller 61 controls the signal generator 42 of the transmitting part 4.The Fourier transformer 62 performs Fast Fourier Transform (FFT) on the beat signal output by each of the plurality of individual receivers 52. This allows the Fourier transformer 62 to transform the beat signal of the reception signal received by each of the plurality of reception antennas 51 into a frequency spectrum corresponding to the data of a frequency range. The frequency spectrum obtained by the Fourier transformer 62 is passed to the data processor 7.The data processor 7 performs recognition information acquisition processing to derive target recognition information (a longitudinal distance, a relative speed, and a lateral position) about the target present in front of the own vehicle, based on the frequency spectrum of each of the plurality of reception antennas 51. The data processor 7 outputs the derived target recognition information to the vehicle controller 2. The data processor 7 receives information from various sensors, such as a vehicle speed sensor 81 and a steering sensor 82, installed on the own vehicle. The data processor 7 may use various data such as a speed of the own vehicle transmitted from the vehicle speed sensor 81 and an angle of the own vehicle transmitted from the steering sensor 82 for processing.FIG. 2 shows, as principal functional parts of the data processor 7, a peak extraction part 71, a direction derivation part 72, a recognition information derivation part 73, a recognition information output part 74, a reliability determination part 75, and a data deletion part 76.2. ACQUISITION OF DESTINATION DETECTION INFORMATIONThe next explanation deals with how (in principle) the radar apparatus 1 acquires the target recognition information (a longitudinal distance, a relative speed, and a lateral position). FIG. 3 shows the relationship between a transmission wave TW and a reflection wave RW. In order to simplify explanation, a reflection wave RW reflected by an ideal target is shown in FIG. 3. In FIG. 3, the transmission wave TW is shown with a solid line, and the reflection wave RW is shown with a dotted line. In the upper figure of FIG. 3, the horizontal axis represents time and the vertical axis represents frequency.As shown in FIG. 3, the transmission wave TW is a continuous wave that periodically changes in frequency up and down from a certain center frequency. The frequency of the transmission wave TW varies linearly with time.Hereinafter, the section where the frequency of the transmission wave TW increases is referred to as "up section", while the section where the frequency decreases is referred to as "down section". Also, the center frequency is referred to as fo, the width of the change in the frequency of the transmission wave TW is referred to as ΔF, and the inverse of an up-down cycle of the frequency of the transmission wave TW is referred to as fm.Since the transmission wave TW becomes the reflection wave RW when the transmission wave TW is reflected by a target, the reflection wave RW is also, like the transmission wave TW, a continuous wave that periodically changes in frequency up and down from a certain center frequency. However, the reflection wave RW is delayed from the transmission wave TW by a time T. The time T, that is, the delay time, is relative to a distance (longitudinal distance) R of a target to the own vehicle and is represented by Equation 1 below by using the speed of light (electric wave speed) c.A Doppler effect corresponding to a relative speed V of the target to the own vehicle causes a frequency shift by a value of a frequency fd with respect to the transmission wave TW.As above, the reflection wave RW is delayed from the transmission wave TW by a longitudinal distance, and its frequency is shifted by a relative speed. Consequently, as shown in the lower figure of FIG. 3, the value of the beat frequency (the difference of the frequencies of the transmission wave TW and the reflection wave RW) in the up section of the beat signal generated by the mixer 53 is different from that in the down section. Hereinafter, the beat frequency in the up section is referred to as fup, and the beat frequency in the down section is referred to as fdn.In the case where a relative speed of a target is "0" (in the case without a frequency shift caused by the Doppler effect), the beat frequency is referred to as fr. The beat frequency fr is represented by Equation 2 below.The value of the frequency fr is relative to the above-described time T, that is, a delay time. Thus, a longitudinal distance R of the target is obtained based on Equation 3 below by using the frequency fr.A frequency fd by which a frequency is shifted due to the Doppler effect is represented by Equation 4 below.A relative speed V of the target is obtained based on Equation 5 below by using the frequency fd.In the above description, a longitudinal distance and a relative speed of an ideal target are calculated. In reality, the radar device 1 simultaneously receives reflection waves RW from a plurality of targets present in front of the own vehicle. Thus, with respect to the reception signal received by one of the reception antennas 51, the frequency spectrum into which the Fourier transformer 62 transformed the beat signal closes information corresponding to the plurality of targets.Fig. 4 shows an example of such frequency spectra. The upper figure of Fig. 4 shows the frequency spectrum in the up section, while the lower figure of Fig. 4 shows the frequency spectrum in the down section. In FIG. 4, each of the horizontal axes represents the frequency, and each of the vertical axes represents the power of a signal.The frequency spectrum of the up section shown in the upper figure of Fig. 4 has peaks Pu at three points of a frequency fup1, a frequency fup2 and a frequency fup3, respectively. The frequency spectrum of the down section shown in the lower figure of FIG. 4 has peaks Pd at three points of a frequency fdn1, a frequency fdn2and a frequency fdn3, respectively.Without considering a relative velocity, the frequency at the point where the frequency spectrum has a peak is relative to a longitudinal distance of a target. In one example, the frequency spectrum of the up section shows that there are three targets at the longitudinal distances corresponding to the three frequencies fup 1, fup 2, and fup 3, respectively, at which the three peaks Pu are present.Thus, with respect to the two frequency spectra of the up section and the down section, the peak decoupling part 71 of the data processor 7 (see FIG. 2 ) couples out the frequencies corresponding to the peaks Pu and the peaks Pd each having a power higher than a prescribed threshold. Hereinafter, the frequency extracted in the above manner is referred to as "peak frequency".The frequency spectra of both the up section and the down section as shown in FIG. 4 are obtained based on the reception signal received by one of the reception antennas 51. Thus, the Fourier transformer 62 derives the two frequency spectra of the up section and the down section as shown in FIG. 4 based on each of the reception signals received by the four reception antennas 51.Since each of the four reception antennas 51 receives the reflection wave RW reflected by the same target, the output peak frequencies in the frequency spectra received by the four reception antennas 51 are the same. However, since the four receiving antennas 51 are disposed at different positions, the reflection waves RW received by the four receiving antennas 51 deviate in phase. Consequently, the reception signals showing the same peak frequency corresponding to the four reception antennas 51 differ in phase information.In the case that there are multiple targets at substantially the same longitudinal distance, the signal of a peak frequency in the frequency spectrum includes information about multiple targets. Thus, the direction deriving part 72 of the data processor 7 (see FIG. 2 ) divides the information on a plurality of targets relevant to the signal based on the signal of the one peak frequency, and estimates respective angles of the plurality of targets through direction calculation processing.The direction deriving part 72 captures the reception signals having the same peak frequency from all the frequency spectra of the four reception antennas 51, and estimates the angles of the targets based on the phase information of the reception signals by using a well-known angle estimation system. In the embodiment, the direction deriving part 72 estimates an angle of a target by using ESPRIT (Estimation of Signal Parameters via Rotational Invariant Techniques) as an angle estimation system. This allows the direction deriving part 72 to derive multiple angles based on the signal of a peak frequency. Further, the direction deriving part 72 separates the signal of each of the plurality of angles and derives the power thereof based on the signal of the one peak frequency.FIG. 5 shows the angles conceptually estimated by the direction deriving part 72 as an angle spectrum. In FIG. 5, the horizontal axis represents the angle (degree), and the vertical axis represents the power of a signal. Each of the peaks Pa in the angle spectrum shown in FIG. 5 shows an angle estimated by the direction deriving part 72. Hereinafter, the angle estimated by the direction deriving part 72 is referred to as "tip angle".In an angle estimation system, a possible division number, that is, the number of targets whose information can be separated based on the signal of a peak frequency (that is, the number of angles to be estimated of the targets present at substantially the same longitudinal distance) is limited. The possible division number is "3" in the ESPRIT used by the direction deriving part 72 in the embodiment. That is, the direction deriving part 72 estimates three peak angles. Thereafter, the direction deriving part 72, as actual results, selects, among the three estimated peak angles, the peak angles whose signal powers exceed a prescribed threshold value. That is, the direction deriving part 72 derives at most three peak angles based on the signal of a peak frequency. Each of the at most three peak angles simultaneously derived by the direction deriving part 72 shows the angle of each of at most three targets present at substantially the same longitudinal distance (at the longitudinal distance relative to the target peak frequency). The direction deriving part 72 derives the peak angles by the above processing with respect to each peak frequency in the frequency spectra of the up section and the down section, respectively.By the above processing, the data processor 7 derives section data corresponding to each of the plurality of destinations. Each of the section data includes parameter values of the above-described peak frequency, the peak angle, and the power of a signal. The data processor 7 derives the section data of the up section and the down section, respectively.The recognition information deriving part 73 of the data processor 7 (see FIG. 2 ) relates, by pairing processing, the section data of the up section and the section data of the down section derived by the above processing. The recognition information deriving part 73 relates the section data of the up section and the section data of the down section based on the parameter values (peak frequency, peak angle, and power of a signal). The recognition information deriving part 73 relates two sets of the section data having similar parameters to relate the two of the section data relating to the same destination. This allows the recognition information deriving part 73 to derive the target data that is internal data showing each of the plurality of targets present in front of the own vehicle. The target data is also called "pair data" because the two of the section data are acquired as a pair.The recognition information deriving part 73 is capable of deriving the target recognition information (longitudinal distance, relative speed, and lateral position) of the target relative to the target data by using the parameter values of the two section data of the upward section and the downward section included in the target data (pair data).The detection information deriving part 73 uses a peak frequency of the up section as the above-described frequency fup and a peak frequency of the down section as the above-described frequency fdn. The data processor 7 is capable of obtaining the longitudinal distance R of the target by using Equation 2 and Equation 3 described above, and is capable of obtaining the relative speed V of the target by using Equation 4 and Equation 5 described above.In addition, in Equation 6 below, the recognition information deriving part 73 acquires an angle θ of the target by using an apex angle θup of the upward section and an apex angle θdn of the downward section. Thereafter, the recognition information deriving part 73 is capable of obtaining the lateral position of the target by the calculation of trigonometric functions based on the angle θ and the longitudinal distance R of the target.3. PROCESSING FOR RECOGNITION INFORMATION ACQUISITIONThe following description deals with the entire flow of the recognition information acquisition processing performed by the data processor 7. FIG. 6 shows the flow of the recognition information acquisition processing by the data processor 7. the data processor 7 repeats the recognition information acquisition processing shown in FIG. 6 in a subsequent time and in a prescribed cycle (for example, a cycle of 1 / 20 second). At the start of the detection information acquisition processing shown in FIG. 6, the frequency spectra of both the up section and the down section with respect to each of the four reception antennas 51 are passed from the Fourier transformer 62 to the data processor 7.First, the peak decoupling part 71 couples out a peak frequency based on the frequency spectra (step S 11). The peak decoupling part 71 couples out, as a peak frequency in the frequency spectra, a frequency that forms a peak having the power exceeding a prescribed threshold value.Next, with respect to the peak frequency extracted by the peak extraction part 71, the direction extraction part 72 estimates the angle of the target with respect to the signal of the peak frequency by the direction calculation processing based on ESPRIT. This allows the direction deriving part 72 to derive the apex angle and the power of the signal for each of the plurality of targets present at substantially the same longitudinal distance (step S 12). The direction deriving part 72 derives at most three peak angles from the signal of a peak frequency.By the above processing, the data processor 7 derives the section data corresponding to each of the plurality of targets present in front of the own vehicle. The data processor 7 derives, in both the up section and the down section, the section data having parameter values of the peak frequency, the peak angle and the power of the signal.Next, the recognition information deriving part 73 relates the section data of the up section and the section data of the down section by the pairing processing (step S 13). The recognition information deriving part 73, at the start of this pairing processing, relates the two section data that are continuous in time to the target data derived by the recognition information acquisition processing in the past. This pairing processing in consideration of past target data is called "history processing".In the history processing, the recognition information deriving part 73 estimates, based on past target data, the parameter values (peak frequency and peak angle) of the section data that can be obtained by the current recognition information acquisition processing of the target relevant to the past target data. In the above manner, the recognition information deriving part 73 derives non-real section data (hereinafter referred to as "prognostic section data") including estimated parameters. The recognition information deriving part 73 derives the predictive section data of both the up section and the down section.Thereafter, the recognition information deriving part 73 selects, among a plurality of section data of the up section derived by the current recognition information acquisition processing, one of the section data including the parameter values similar to those of the prognostic section data. In the same manner, the recognition information deriving part 73 selects, among a plurality of section data of the down section derived by the current recognition information acquisition processing, one of the section data including the parameter values similar to those of the prognostic section data.Thereafter, the recognition information deriving part 73 relates the section data of the up section and the section data of the down section that have been selected on the basis of the prognostic section data. In the case where there is only one section data similar to the prognostic section data that is one of the up section and the down section while there is no section data similar to that of the other section, the recognition information deriving part 73 relates the present section data and the prognostic section data of the section in which there is no section data. The processing using prognostic section data as the section data of the one of the outward section and the downward section is referred to as "one-side extrapolation".The recognition information deriving part 73 performs the processing as above with respect to a prescribed number of sets (for example, 20 sets) of the target data derived by the recognition information acquisition processing in the past. Next, the recognition information deriving part 73 relates the section data of the up section and the section data of the down section that have not been selected in the processing. The recognition information deriving part 73 relates two section data having similar parameters, for example, by calculation by using a Mahalanobis distance.The recognition information deriving part 73, in the case that the section data of the up section and the section data of the down section are related by the processing above, derives the target data (pair data) based on the two related section data. The recognition information deriving part 73 derives, by the above-described calculation, with respect to each of the derived target data, the longitudinal distance of the target, the relative speed of the target, and the lateral position of the target as the target recognition information.The recognition information deriving part 73 stores the derived target data as a processing object in the memory 65 of the signal processor 6.Each of the target data has different kinds of parameters. The parameters of the target data include "existence counter", "moving object flag", "preceding vehicle flag", "one-side extrapolation flag", and "full extrapolation flag".The "existence counter" represents the certainty of the presence of the target shown by the target data. The higher the certainty of the presence of the target, the larger the value of the existence counter. The minimum value of the existence counter is, for example, "0", while the maximum value is, for example, "35".The "moving object flag" represents whether or not the target shown by the target data is a moving object. The "preceding vehicle flag" represents whether the target shown by the target data moves toward the same direction as the own vehicle.The "one-side extrapolation flag" represents whether "one-side extrapolation" is performed on the target data in the present recognition information acquisition processing. The "full extrapolation flag" represents whether "full extrapolation" (which will be described in detail later) is performed on the target data in the present recognition information acquisition processing.Next, by performing continuity determination processing, the recognition information deriving part 73 determines the temporal continuity between the target data derived by the current recognition information acquisition processing and the target data derived by the recognition information acquisition processing in the past (step S 14).The recognition information deriving part 73 estimates the target detection information (longitudinal distance, relative speed, and lateral position) of the target relevant to the past target data to be obtained by the present recognition information detection processing, based on the past target data. This allows the recognition information deriving part 73 to derive non-real target data (hereinafter referred to as "target prognostic data") including estimated target recognition information.The recognition information deriving part 73 selects, among the current target data, a set of target data similar to those of the derived prognostic target data in the target recognition information. The recognition information deriving part 73 determines that the selected one set of the target data has continuity with the past target data, and displays the same target as the past target data.The recognition information deriving part 73 determines the continuity with respect to each of the past target data stored in the memory 65. In the case where there is no current target data similar to the prognostic target in the determination processing, the prognostic target data is used as the current target data having continuity with the past target data. The processing for using the predictive target data as the current target data is called "full extrapolation".The recognition information deriving part 73 determines that the target data whose continuity with the past target data cannot be determined shows a new target.Next, the recognition information deriving part 73 switches the value of the existence counter in the current target data (step S 15). With respect to the target data showing a new target, the recognition information deriving part 73 sets, for example, "8" as an initial value for the existence counter. With respect to the target data showing a target that is not new, the recognition information deriving part 73 switches the value of the existence counter according to the result of the current continuity determination while continuing to use the value of the existence counter of the past target data.The recognition information deriving part 73 adds "+4" to the value of the existence counter of the target data when it is normally determined that the target data has continuity with the past target data. However, the value of the existence counter by the operation never exceeds the maximum value (for example, "35").The recognition information deriving part 73 maintains the value of the existence counter of the target data on which "one-side extrapolation" is performed even when it is determined that the target data has continuity with the past target data. Then, the recognition information deriving part 73 sets the flag for one-side extrapolation of the target data to "on".The recognition information deriving part 73 adds "-2" to the value of the existence counter of the target data on which "full extrapolation" is performed. Then, the recognition information deriving part 73 sets the flag for fully extrapolating the target data to "on". Further, the recognition information deriving part 73 deletes from the memory 65 the target data whose value of the existence counter is equal to or less than "0" by the operation on the existence counter. The target data in which the value of the existence counter is equal to or less than "0" as above can be seen when "full extrapolation" is repeatedly performed in sequence. It is highly likely that the target relevant to such target data has disappeared in front of the own vehicle. Thus, the recognition information deriving part 73 deletes the target data from the memory 65 to exclude the target data from the processing objects.Next, the recognition information deriving part 73 performs filtering processing on the target data showing a target that is not new, so as to smooth the target recognition information (longitudinal distance, relative speed, and lateral position) of the target data in a time axis direction (step S 16). Specifically, the recognition information deriving part 73 derives new target recognition information of the target data by applying weighted averaging to the target recognition information of the target data derived as current values at this time and the target recognition information of the target prognostic data used in the continuity determination processing. The weight for the target recognition information of the target data derived at this time is, for example, "0.25", while the weight for the target recognition information of the target prognostic data is, for example, "0.75". The target recognition information of the target data (pair data) derived as current values may be abnormal values due to the influence of noise or other reasons in some cases. However, the filtering processing prevents the target recognition information from being abnormal values. The target data after the filtering processing may be called "filter data".Next, the recognition information deriving part 73 sets a moving object flag and a preceding vehicle flag for each target data after performing a moving object determination processing (step S 17).The recognition information deriving part 73 first derives an absolute speed and a moving direction of the target shown by the target data, based on the relative speed of the target data and the speed of the own vehicle acquired from the vehicle speed sensor 81.When an absolute speed of the target shown by the target data is equal to or above a prescribed speed (for example, 1 km / h), the recognition information deriving part 73 determines that the target is a moving object, and sets the moving object flag to "on". When an absolute speed of the target shown by the target data is less than a prescribed speed (for example, 1 km / h), the recognition information deriving part 73 determines that the target is a moving object, and sets the moving object flag to "off",When the moving direction of the target shown by the target data is the same as that of the own vehicle and further an absolute speed is equal to or above a prescribed speed (for example, 18 km / h), the recognition information deriving part 73 sets the preceding vehicle flag to "on". When the target shown by the target data does not satisfy these conditions, the recognition information deriving part 73 sets the preceding vehicle flag to "off".Next, the reliability determination part 75 and the data deletion part 76 perform processing for angle reliability determination (step S 18). In the angle reliability determination processing, the reliability determination part 75 determines the reliability of tip angles of each of the two section data with respect to the target data after the filtering processing. Thereafter, the data deleting part 76, under a prescribed condition, deletes the target data from the memory 65 to exclude the target data from the processing objects on the basis of the reliability of the apex angles detected. The processing for angle reliability determination will be described in detail later.Next, the recognition information deriving part 73 selects a prescribed number (for example, 20 sets) of target data to be subjected to the history processing (pairing processing at step S 13) in the next recognition information acquisition processing (step S 19). The recognition information deriving part 73 preferably selects, in consideration of the longitudinal distance and the lateral position of the target data, the target data showing the target that travels in the same lane as the own vehicle and that is closer to the own vehicle. The recognition information deriving part 73 recognizes the shape of the traveling lane on the basis of the steering angle of the own vehicle acquired by the steering sensor 82, and determines whether the target is moving in the same traveling lane as the own vehicle.Next, the recognition information deriving part 73 performs grouping processing to form a group with all the target data relevant to the same item among the entire target data (step S 20). In the case where a transmission wave TW is reflected by a vehicle traveling in front of the own vehicle, the transmission wave TW is normally reflected at each of a plurality of reflection points of the vehicle. Since a plurality of reflection waves RW reflected at the plurality of reflection points of the same vehicle respectively reach the radar device 1, the detection information deriving part 73 derives the target data relevant to each of the plurality of reflection points. Since the target shown by the plurality of target data is a vehicle, the recognition information deriving part 73 forms a group with such target data. The recognition information deriving part 73 forms, for example, a group including the plurality of target data having substantially the same relative speed and having similar longitudinal distances and lateral positions. As the target recognition information of the target data after grouping, for example, an averaged value of the target recognition information of the plurality of target data subjected to grouping may be selected.Next, the recognition information output part 74 (see FIG. 2 ) outputs the target recognition information (longitudinal distance, relative speed, and lateral position) of the target data derived in the above manner and not excluded from the processing objects to the vehicle control device 2 (step S 21). In the case where there are many target data, the recognition information output part 74 selects a prescribed number (for example, 8 sets) of the target data and outputs the target recognition information including only the selected target data. The recognition information output part 74 preferably selects, in consideration of the longitudinal distance and the lateral position of the target data, the target data that shows the target that travels in the same lane as the own vehicle and that is closer to the own vehicle. The recognition information output part 74 recognizes the shape of the traveling lane on the basis of the steering angle of the own vehicle acquired by the steering sensor 82, and determines whether the target is moving in the same traveling lane as the own vehicle.4. PROCESSING FOR ANGLE RELIABILITY DETERMINATIONThe next description deals with the angular reliability determination processing (step S 18 of FIG. 6 ) performed by the reliability determination part 75 and the data deletion part 76. The description starts with the problem occurring in the case where the processing for angle reliability determination is not performed based on FIG. 7.As described above, the possible division number is "3" in the ESPRIT used in the embodiment. That is, the direction deriving part 72 derives, based on the signal of a peak frequency, at most three peak angles as the angles of the targets disposed at substantially the same longitudinal distance. Consequently, when receiving the reflection waves from four or more targets disposed at substantially the same distance and the number of which exceeds the possible division number, the radar apparatus 1 cannot correctly estimate the tip angles of these targets, and the reliability of the tip angles is lowered.The phenomenon in which the reliability of the apex angles is lowered can be observed in some cases of surroundings of a moving vehicle regardless of the number of actual targets. In the case where an own vehicle 9 travels on a road 100 within a tunnel, the phenomenon in which the reliability of the apex angles is lowered can be observed.In FIG. 7, another vehicle (hereinafter referred to as "preceding vehicle") 91 travels toward the same direction as the own vehicle 9 on an adjacent lane in front of the own vehicle 9. Consequently, the radar device 1 receives a reflection wave RWb and a reflection wave RWc reflected by the inner wall of the tunnel in addition to a reflection wave RWa immediately incident from the preceding vehicle 91. That is, as shown in FIG. 7, the radar device 1 is in a state in which the radar device 1 receives the reflection waves RWb and RWc that are the same as the reflection waves RW reflected by a target Tgb and a target Tgc that are not actually present. The distances of the target Tgb and the target Tgc are substantially the same as the distance of an actual target Tga of the preceding vehicle 91.FIG. 7 shows only two reflection waves RWb and RWc as the reflection wave RW reflected by the inner wall of the tunnel. However, the reflection wave RW reflected by the preceding vehicle 91 is actually reflected at various parts such as a ceiling, a floor, and an inner wall. Thus, the radar device 1 is in a state as if the radar device 1 receives the reflection wave RW reflected by four or more targets present at substantially the same distance.In this case, due to the effects of these reflection waves RW incident from different directions, the direction deriving part 72 cannot correctly derive the tip angle of the actual target Tga, which reduces the reliability of the tip angle. In this case, for example, the radar device 1 may erroneously recognize the target Tga of the preceding vehicle 91 as a target traveling on the same lane as the own vehicle 9.Thus, in the radar apparatus 1 of the embodiment, the reliability determination part 75 determines the reliability of the peak angle derived by the direction derivation part 72 in the angle reliability determination processing. Thereafter, the data deleting part 76 deletes the target data derived based on the low-reliability apex angle from the memory 65 under a prescribed condition to exclude the target data from the processing objects. The recognition information output part 74 outputs, to the vehicle control device 2, the target recognition information of the target data that is not excluded from the processing objects by the data deletion part 76. This prevents the vehicle control device 2 from performing erroneous control based on the target recognition information relevant to the target low-reliability tip angle data.FIG. 8 shows a detailed flow of the angular reliability determination processing (step S 18 of FIG. 6 ). The angle reliability determination processing is performed on all the target data stored in the memory 65 after filtering (step S 16 of FIG. 6 ). The angle reliability determination processing is a part of the recognition information acquisition processing shown in FIG. 6. Consequently, the processing for angle reliability determination is repeated in a subsequent time and a prescribed cycle (for example, a cycle of 1 / 20 second).The angle reliability determination processing uses various parameters set for each of the target data, respectively. The parameters of the target data include "low reliability up flags", "low reliability down flags", "low angle reliability counters", "low target reliability flags", and "full extrapolation counters".Each of the "upward low reliability flag" and the "downward low reliability flag" represents whether or not the apex angle of each of the two section data of the upward section and the downward section included in the target data is low. In the case that the reliability of the tip angle of the upward section is low, the low reliability flag upward is "on". In the case that the reliability of the tip angle of the down section is low, the low reliability flag is "on" downward.The "low angle reliability counter" represents the low reliability of the peak angle itself relevant to the target data, and is operated in accordance with the low reliability flag upward and the low reliability flag downward (which will be described in detail later). The lower the reliability of the peak angle relevant to the target data, the larger the value of the low angle reliability counter. With respect to the low angle reliability counter, the initial value is, for example, "0", the minimum value is, for example, "0", and the maximum value is, for example, "15".The "target low reliability flag" represents whether or not the reliability of the target shown by the target data is low, and is set according to the angle low reliability counter (which will be described in detail later).The "full extrapolation counter" represents the frequency at which "full extrapolation" is performed on the target data. The higher the frequency at which "full extrapolation" is performed on the target data, the larger the value of the full extrapolation counter. With respect to the full extrapolation counter, the initial value is, for example, "0", the minimum value is, for example, "0", and the maximum value is, for example, "10".Each of the "upward low reliability flag" and the "downward low reliability flag" is set each time in the recognition information acquisition processing, and represents the reliability of the apex angle derived in the direction calculation processing (step S 12 of FIG. 6 ) of this recognition information acquisition processing. On the other hand, with respect to the "low angle reliability counter", the "low target reliability flag", and the "full extrapolation counter", the values of the past target data determined in the continuity determination processing (step S 14 of FIG. 6 ), in which the continuity is determined, are continuously adopted as the values of the current target data. The initial value of each of the low reliability flag upward, the low reliability flag downward, and the low target reliability flag is "off".The next description deals with the flow of the processing for angle reliability determination based on FIG. 8, first, the reliability determination part 75 selects, among all the target data stored in the memory 65, a set of the target data as "object target data" which are processing objects (step S 31).Next, the reliability determination part 75 determines whether the object target data satisfies all the requirements from (a1) to (a4) described below (step S 32).(a1) The moving object flag of the object target data is "on".(a2) The preceding vehicle flag of the object target data is "on".(a3) The longitudinal distance of the object target data is equal to or greater than 30 (m).(a4) The flag for total extrapolation of the object target data is "off".In the case that the object target data does not satisfy any of the requirements of (a1) to (a4) (No at step S 32), the processing proceeds to step S 37 without determination by the reliability determination part 75 regarding the reliability of the tip angles of the object target data. In the case where the object target data does not satisfy any of the requirements of (a1) to (a3), the reliability of the apex angles is of little importance. In the case that the object target data does not satisfy the requirement of (a4), "complete extrapolation has been performed on the object target data, and there is no tip angle to be determined. In this case, the processing proceeds to step S 37 while both the low reliability flag is up and the low reliability flag is down "off" as it is.In the case where the object target data satisfies all of the requirements of (a1) to (a4) (Yes in step S 32), the reliability determination part 75 determines the reliability of the peak angles of the object target data. The reliability determination part 75 determines the reliability of the tip angle of both the upward section and the downward section with respect to the object target data.First, the reliability determination part 75 determines the reliability with respect to the tip angle of the upward section of the object target data (step S 33). In the case that the tip angle of the upward section does not satisfy a prescribed reliability condition (No at step S 33), the reliability determination part 75 determines that the reliability of the tip angle of the upward section is low. In this case, the reliability determination part 75 sets the low reliability flag upward to "on", which represents that the reliability of the tip angle of the upward section is low (step S 34). In the case where there is no tip angle of the upward section due to "one-side extrapolation", the reliability determination part 75 sets the low reliability flag upward to "off".Next, the reliability determination part 75 determines the reliability with respect to the tip angle of the downward section of the object target data (step S 35). In the case that the tip angle of the down section does not satisfy a prescribed reliability condition (No at step S 35), the reliability determination part 75 determines that the reliability of the tip angle of the down section is low. In this case, the reliability determination part 75 sets the low reliability flag downward to "on", which represents that the reliability of the tip angle of the downward section is low (step S 36). In the case where there is no tip angle of the down section due to "one-side extrapolation", the reliability determination part 75 sets the low reliability flag down to "off".FIG. 9 is a table showing the reliability conditions to be used when the reliability of the tip angle of the object target data is determined. The reliability determination part 75 uses the reliability conditions shown in FIG. 9 for each peak angle of the upward section and the downward section. In the case where the tip angle to be determined satisfies the reliability conditions shown in FIG. 9, the reliability determination part 75 determines that the reliability of the tip angle is high. Hereinafter, the tip angle to be determined by the reliability determination part 75 is referred to as "object tip angle".As shown in FIG. 9, the reliability determination part 75 determines "number of angles", "lateral position difference (m)", "ST ratio (dB)", "angle difference (degree)", and "angle power difference (dB)" as reliability conditions.The value of the "number of angles" represents the number of the tip angles (including the object tip angle) derived simultaneously by the direction deriving part 72 in the direction calculation processing for deriving the object tip angle (step S 12 of FIG. 6 ). That is, the value of the "number of angles" represents the number of the tip angles (including the object tip angle) into which the information is divided from the signal of a tip frequency that is the derivation source of the object tip angle. The case where the "number of angles" is "1" corresponds to the case where the direction deriving part 72 derives only the object tip angle. The case where the "number of angles" is "2" corresponds to the case where the direction deriving part 72 derives the object tip angle and another tip angle simultaneously. Further, the case that the "number of angles" is "3" corresponds to the case that the direction deriving part 72 derives the object tip angle and two other tip angles simultaneously.Hereinafter, the other tip angle derived simultaneously with the object tip angle by the direction deriving part 72 will be referred to as "other tip angle". In the case where the "number of angles" is "2" or "3", the reliability determination part 75 determines the reliability of the object tip angle based on the other tip angle(s). Specifically, the reliability determination part 75 determines whether the object tip angle is affected by the reflection wave from the other tip angle(s) when the object tip angle is derived, based on the relationship between the object tip angle and the other tip angle(s), to determine the reliability of the object tip angle.The value of the "lateral position difference" represents the difference between the lateral position based on the object tip angle and the lateral position based on the other tip angle. These lateral positions are obtained by calculation through the use of trigonometric functions based on the longitudinal distance of the object target data.The value of the "ST ratio" represents the ratio of the power of the signal (signal) of the object tip angle to the threshold value (threshold) (see FIG. 5 ) used by the direction deriving part 72 to derive the object tip angle. The larger the power of the signal of the object tip angle, the larger the ST ratio.The value of the "angle difference" represents the difference between the object tip angle and the other tip angle corresponding to ΔD in FIG. 5.The value of the "angular power difference" represents the difference between the powers of the object tip angle signal and the signal of the other tip angle. Specifically, the "angle power difference" is obtained in a manner in which the power of the signal of the other tip angle is subtracted from the power of the signal of the object tip angle, corresponding to ΔS in FIG. 5.Next, the reliability condition for each "number of angles" will be described. In the case where the "number of angles" is "1", only the "ST ratio" needs to be determined. In the case where the "ST ratio" ≥ threshold value Aa, the reliability determination part 75 determines that the reliability of the object tip angle is high. The threshold Aa is 0 (dB), for example.In the case that the "number of angles" is "2", a first "lateral position difference" needs to be determined. In the case where the "lateral position difference" is equal to or less than 2.5 meters, that is, a normal vehicle width, long, the reliability determination part 75 determines that the reliability of the object tip angle is high. In this case, assuming that the target of the object tip angle and the target of the other tip angle are the same vehicle, the reliability determination part 75 determines that the reliability of the object tip angle is high.In the case where the "lateral position difference" exceeds 2.5 meters, "ST ratio", "angular difference", and "angular power difference" must be determined. In the case where any one of the following conditions (b 1) and (b 2) is satisfied, the reliability determination part 75 determines that the reliability of the object tip angle is high.(b1) "ST ratio" ≥ threshold Ab and "angle difference" ≥ threshold B.(b2) "ST ratio" ≥ threshold Ab and "angular power difference" ≥ threshold Cb.Under these conditions, the threshold Ab is larger than the threshold Aa, and is, for example, 15 (dB). The threshold B is, for example, 8 (degrees). The threshold Cb is, for example, 4 (dB).When the condition (b1) is satisfied, the power of the signal of the object tip angle is relatively large, and the object tip angle and the other tip angle are sufficiently distant from each other. Consequently, based on the determination that the object tip angle can be derived from the other tip angle without large effects of the reflection wave, the reliability determination part 75 determines that the reliability of the object tip angle is high.When the condition (b2) is satisfied, the power of the signal of the object tip angle is relatively large, and the power of the signal of the object tip angle is sufficiently large compared with the power of the signal of the other tip angle. Consequently, based on the determination that the object tip angle can be derived from the other tip angle without large effects of the reflection wave, the reliability determination part 75 determines that the reliability of the object tip angle is high.In the case that the "number of angles" is "3", a first "lateral position difference" needs to be determined. In the case where the "lateral position difference" with respect to each of the other two tip angles is equal to or less than 2.5 meters, that is, a normal vehicle width, is long, the reliability determination part 75 determines that the reliability of the object tip angle is high.In the case that the "lateral position difference" with respect to the other two tip angles exceeds 2.5 meters, "ST ratio", "angle difference", and "angle power difference" must be determined. In the case where any one of the following conditions (c 1) and (c 2) is satisfied, the reliability determination part 75 determines that the reliability of the object tip angle is high. In this case, regarding "angle difference" and "angle power difference", both of the other two tip angles must satisfy the conditions.(c1) "ST ratio" ≥ threshold value Ac and "angle difference" ≥ threshold value B.(c2) "ST ratio" ≥ threshold Ac and "angular power difference" ≥ threshold Cc.Under these conditions, the threshold Ac is larger than the threshold Ab, and is 20 (dB), for example. The threshold B is, for example, 8 (degrees). The threshold Cc is larger than the threshold Cb, and is, for example, 8 (dB). In the case where the "number of angles" is "3", the reliability of the object tip angle tends to be lower compared to the case where the "number of angles" is "2". Consequently, as compared with the case that the "number of angles" is "2", larger values than a part of the threshold values are used for the reliability conditions so as to make the conditions more severe to determine that the reliability of the object tip angle is high.In the case that the condition (c1) is satisfied, the power of the signal of the object tip angle is relatively large, and each of the other two tip angles is sufficiently far from the object tip angle. Thus, based on the determination that the object tip angle can be derived from each of the other two tip angles without large effects of the reflection wave, the reliability determination part 75 determines that the reliability of the object tip angle is high.In the case that the condition (c2) is satisfied, the power of the signal of the object tip angle is relatively large, and the power of the signal of the object tip angle is sufficiently large compared with the power of each signal of the other two tip angles. Consequently, based on the determination that the object tip angle can be derived from each of the other two tip angles without large effects of the reflection wave, the reliability determination part 75 determines that the reliability of the object tip angle is high.After determining the reliability of the peak angles of both the upward section and the downward section of the object target data as above, the reliability determination part 75 switches the value of the low angle reliability counter (hereinafter, a symbol C 1 is used herein) of the object target data (step S 37 of FIG. 8 ).In addition, the reliability determination part 75 switches the value of the target low reliability flag (hereinafter, a symbol FL is used herein) according to the value of the switched low angle reliability counter C 1. In the case where the target low reliability flag FL is "off" and the value of the angle low reliability counter C 1 is equal to or greater than an on-state threshold value, the reliability determination part 75 sets the target low reliability flag FL to "on". In the case where the target low reliability flag FL is "on" and the value of the angle low reliability counter C 1 is equal to or less than a prescribed off-state threshold value, the reliability determination part 75 sets the target low reliability flag FL to "off",FIG. 10 is a table showing the operation patterns in which the reliability determination part 75 switches the low angle reliability counter C 1 of the object target data. The reliability determination part 75 switches the low angle reliability counter C 1 of the object target data in one of the patterns (p 1) to (p 8) shown in FIG. 10. Each of the patterns includes the on-state threshold value for use in the case that the low target reliability flag FL is set to "on".The patterns (p 1) to (p 4) correspond to the case that the longitudinal distance of the object target data is less than 70 meters, and the patterns (p 5) to (p 8) correspond to the case that the longitudinal distance of the object target data is equal to or greater than 70 meters. The reliability of the target shown by the object target data, which has a longer longitudinal distance, tends to be lower. Consequently, the low target reliability flag FL of each of the patterns (p5) to (p8) is more easily set to "on" than in the case of each of the patterns (p1) to (p4).The case where both the low reliability flag upward and the low reliability flag downward of the object target data is "on" corresponds to the first pattern (p 1) or the fifth pattern (p 5) in accordance with the longitudinal distance. In this case, the reliability of the tip angle of both the upward section and the downward section is low.Consequently, the reliability determination part 75 in the first pattern (p 1) shifts "+1" to the value of the low angle reliability counter C 1. In the case where the target low reliability flag FL is "off" and the value of the angle low reliability counter C 1 is equal to or greater than the on-state threshold value "8", the reliability determination part 75 sets the target low reliability flag FL to "on". In the fifth pattern (p5), the reliability determination part 75 adds "+2" to the value of the low angle reliability counter C1. In the case where the target low reliability flag FL is "off" and the value of the angle low reliability counter C 1 is equal to or greater than the on-state threshold value "5", the reliability determination part 75 sets the target low reliability flag FL to "on".The case that one of the low reliability flag upward and the low reliability flag downward of the object target data is "on" corresponds to the second pattern (p 2) or the sixth pattern (p 6) in accordance with the longitudinal distance. In this case, the reliability of the tip angle of both the upward section and the downward section is low.In the second pattern (p 2), the reliability determination part 75 maintains the value of the low angle reliability counter C 1. In the sixth pattern (p2), the reliability determination part 75 adds "+1" to the value of the low angle reliability counter C1. In the case where the target low reliability flag FL is "off" and the value of the angle low reliability counter C 1 is equal to or greater than the on-state threshold value "8", the reliability determination part 75 sets the target low reliability flag FL to "on".The case where both the low reliability flag upward and the low reliability flag downward of the object target data is "off", and one of the full extrapolation flag and the one-side extrapolation flag is "on" corresponds to the third pattern (p 3) or the seventh pattern (p 7) in accordance with the longitudinal distance. In this case, "full extrapolation" or "one-side extrapolation" is performed on the object target data. In each of the third pattern (p 3) and the seventh pattern (p 7), the reliability determination part 75 maintains the value of the low angle reliability counter C 1.The case where both the low reliability flag upward and the low reliability flag downward of the object target data is "off", and both the full extrapolation flag and the one-side extrapolation flag is "off", corresponds to the fourth pattern (p 4) or the eighth pattern (p 8) in accordance with the longitudinal distance. In this case, the object target data is normally derived.In each of the fourth pattern (p4) and the eighth pattern (p8), the reliability determination part 75 adds "-1" to the value of the low angle reliability counter C1. In the case where the target low reliability flag FL is "on" and the value of the angle low reliability counter C 1 is equal to or less than a prescribed off-state threshold value (for example, "0"), the reliability determination part 75 sets the target low reliability flag FL to "off".FIG. 11 shows an example of changing the values of the low angle reliability counter C1 and the low target reliability flag FL switched in the above manner. At a time T0, the value of the low angle reliability counter C1 is "0", and the low target reliability flag FL is "off". FIG. 11 shows the case where the longitudinal distance of the object target data is less than 70 meters.In the period from the time point T 0 to a time point T 1, both the low reliability flag upward and the low reliability flag downward are "on". Consequently, since this case corresponds to the first pattern (p 1), the value of the low angle reliability counter C 1 increases.In the next period from the time T 1 to a time T 2, both the low reliability flag upward and the low reliability flag downward are "off", and the full extrapolation flag is "on". Consequently, since this case corresponds to the third pattern (p3), the value of the low angle reliability counter C1 is held as it is.In the next period from the time T 2 to a time T 4, both the low reliability flag upward and the low reliability flag downward are "on". Consequently, since this case corresponds to the first pattern (p 1), the value of the low angle reliability counter C 1 increases. At a time T 3 in this period, since the value of the low angle reliability counter C 1 is equal to or greater than the on-state threshold value, the low target reliability flag FL is set to "on".In the next period from the time point T 4 to a time point T 6, both the low reliability flag upward and the low reliability flag downward are "off", and both the full extrapolation flag and the one-side extrapolation flag are "off". Since this case corresponds to the fourth pattern (p4), the value of the low angle reliability counter C1 decreases. At a time T 5 in this period, although the value of the low angle reliability counter C 1 is equal to or greater than the on-state threshold value, the low target reliability flag FL is held "on" as it is.In the next period from the time T 6 to a time T 7, both the low reliability flag upward and the low reliability flag downward are "off", and the one-side extrapolation flag is "on". Since this case corresponds to the third pattern (p3), the value of the low angle reliability counter C1 is held as it is.In the next period from the time T 7 to a time T 8, both the low reliability flag is "off" upward and the low reliability flag is "off". Since this case corresponds to the fourth pattern (p4), the value of the low angle reliability counter C1 decreases. At the time T 8, since the value of the low angle reliability counter C 1 is equal to or less than the off-state threshold value "0", the low target reliability flag FL is set to "off".As above, the value of the low target reliability flag FL reflects the result when the reliability determination part 75 has determined the reliability of the apex angle relevant to the target data showing the same target a plurality of times. In the case where the reliability of the tip angle of the target data is kept low, the low target reliability flag FL is set to "on".After switching the value of the low angle reliability counter C 1 of the object target data as above, the reliability determination part 75 switches the value of the full extrapolation counter (hereinafter, a symbol C 2 is used herein) (step S 38 of FIG. 8 ). The reliability determination part 75 switches the value of the full extrapolation counter C 2 according to the full extrapolation flag and the one-side extrapolation flag of the object target data.In the case where the full extrapolation flag is "on", the reliability determination part 75 shifts "+1" to the value of the full extrapolation counter C 2. In the case where the one-side extrapolation flag is "on", the reliability determination part 75 maintains the value of the full extrapolation counter C 2. In the case where both the full extrapolation flag and the one-side extrapolation flag are "off", the reliability determination part 75 shifts "-1" to the value of the full extrapolation counter C2.FIG. 12 shows an example of the change with time of the value of the full extrapolation counter C2 switched as above. At a time T 10, the value of the full extrapolation counter C 2 is "0",In the period from the time point T 10 to a time point T 11, since the full extrapolation flag is "on", the value of the full extrapolation counter C 2 increases. In the next period from the time point T 11 to a time point T 12, since the one-side extrapolation flag is "on", the value of the full extrapolation counter C 2 is maintained as it is. In the next period from the time point T 12 to a time point T 13, since both the full extrapolation flag and the one-side extrapolation flag are "off", the value of the full extrapolation counter C 2 decreases. In the next period from the time point T 13 to a time point T 14, since the full extrapolation flag is "on", the value of the full extrapolation counter C 2 increases.After switching the value of the counter C 2 for fully extrapolating the object target data as above, the reliability determination part 75 determines whether the object target data satisfies all of the following deletion conditions (d 1) to (d 6) (step S 39 of FIG. 8 ). In order to obtain the value of the power of the peak angle signal for use in the condition (d 4), the powers of the peak angle signals derived in the past in the recognition information acquisition processing performed by a prescribed number of times (for example, 10 times) are averaged.(d1) The moving object flag of the object target data is "on".(d2) The preceding vehicle flag of the object target data is "on".(d3) The longitudinal distance of the object target data is equal to or greater than 70 (m).(d4) The power of the peak angle signal of the object target data is equal to or greater than -30 (dB).(d5) The target reliability flag FL of the object target data is "on".(d 6) The value of the total extrapolation counter C 2 of the object target data is equal to or greater than a deletion threshold value.In the case where the object target data does not satisfy any of the conditions (d 1) to (d 6) (No in step S 39), the data deleting part 76 acquires the object target data stored in the memory 65. In the case that the object target data satisfies all of the deletion conditions (d 1) to (d 6) (Yes in step S 39), the data deleting part 76 deletes the object target data from the memory 65 to exclude the target data from the processing objects (step S 40).In the example of FIG. 12, the value of the full extrapolation counter C 2 at the time 14 is equal to or greater than the deletion threshold value (for example, "5"), and the condition (d 6) is satisfied. Thus, if the object target data also satisfies the other conditions (d 1) to (d 5), the data deleting part 76 deletes the object target data from the memory 65 at the time T 14.In the case that the target reliability flag FL of the object target data is on as indicated in the condition (d 5), that is, when the reliability of the tip angle of the target data is kept low, the data deleting part 76 deletes the object target data from the memory 65. Therefore, this prevents the vehicle control device 2 from performing erroneous control based on the target recognition information on the target data in which the reliability of the tip angle is low.Each time the processing of a set of target data is completed in the above manner, the reliability determination part 75 determines whether there are unprocessed target data which are not treated as object target data in the memory 65 (step S41). In the case where such data is present (Yes in step S41), the reliability determination part 75 selects a set of the target data other than new object target data (step S31), and performs the same processing as above. This processing is repeated until the last of the target data stored in the memory 65 is processed so as to exclude all of the target data in which the reliability of the tip angle is low from the processing objects.The radar apparatus 1 is capable of, by performing the angle reliability determination processing as above, deleting the target data in which the reliability of the tip angle is low more quickly than in the normal case of deleting the target data based on the existence counter (step S 15) to exclude this target data from the processing objects.Both Figs. 13 and 14 show transition examples of the longitudinal distance and the lateral position derived with respect to the target data relevant to the same target. FIG. 13 shows the case where the angle reliability determination processing is not performed, and FIG. 14 shows the case where the angle reliability determination processing is performed. In each of the figures, the horizontal axis represents time, and the left line of the vertical axes represents a longitudinal distance, and the right line of the vertical axes represents a lateral position. In each of the figures, the dot-and-dash line represents a longitudinal distance of the target data, while the solid line represents a lateral position of the target data.The target shown by the target data in each of the figures is an actual vehicle traveling on the left lane adjacent to the own vehicle 9, like the preceding vehicle shown in FIG. 7. The radar device 1 of the embodiment understands, for example, the target data whose lateral position (absolute value) is equal to or less than 1.5 meters (the target data whose lateral position is within the range of Ws in each of the figures) as the target data of the target traveling in the same lane.As shown in FIG. 13, in the case where the angular reliability determination processing is not performed, the longitudinal distance and the lateral distance of the target data are continuously derived. A frequently performed "full extrapolation" is necessary to clear the target data based on the existence counter, while an intermittently performed "full extrapolation" is not enough to clear the target data immediately.Consequently, as shown in FIG. 13, even when "full extrapolation" is performed on the target data, the target data is obtained as it is without deletion in the memory. Thereafter, in a period from a time point T 20 to a time point T 23, a distance equal to or less than 1.5 meters is derived as the lateral position (absolute value) of the target data, and the target data is erroneously understood as the target data showing that the target is traveling in the same lane. Therefore, the vehicle control device can perform erroneous control to follow the target based on the target recognition information of the target data.On the other hand, in the case where the angle reliability determination processing is performed, the target data at a time point 21 is deleted from the memory 65 as the target data in which the reliability of the tip angle is low, as shown in FIG. 14. In the case where the processing for angle reliability determination is performed, the target that is most likely to be erroneously understood may be immediately deleted from the memory 65 as above to exclude the target data from the processing objects. This prevents the vehicle control device 2 from performing erroneous control. Thereafter, at a time T 22, the target data of the target is derived again as the target data whose lateral position (absolute value) exceeds 1.5 meters (target data of the target traveling in another lane).As above, in the radar apparatus 1 of the embodiment, the direction deriving part 72 derives the tip angle showing the angle of the target on the basis of the reception signals received by the plurality of reception antennas 51. Thereafter, the recognition information deriving part 73 derives the target data, which is internal data showing the target, on the basis of the tip angle and stores the derived target data in the memory 65. the reliability determining part 75 determines the reliability of the object tip angle relevant to the target data, on the basis of the other tip angle(s) derived simultaneously with the object tip angle by the direction deriving part 72. Thereafter, based on the reliability of the object tip angle determined by the reliability determination part 75, the data deletion part 76 deletes the target data relevant to the object tip angle from the memory 65 to exclude the target data from the processing objects. This excludes the target data in which the reliability of the tip angle is low from the processing objects.Moreover, the data deleting part 76 excludes the target data from the processing objects on the basis of the low target reliability flag FL representing the results, and the reliability determining part 75 determines the reliability of the apex angles relevant to the target data showing the same target a plurality of times. The reliability of the tip angle used as an instantaneous value may be low due to the effects of noise or for some other reason. However, the use of the low target reliability flag FL may exclude the target data from the processing objects in which the reliability of the apex angle is kept low.5. MODIFICATIONThe embodiment of the invention has been described so far. However, the invention is not limited to the above-described embodiment and can provide various modifications. All the embodiments including the above-described embodiment and the embodiments to be described below may be combined with others as desired.In the above-described embodiment, the data deleting part 76 retrieves from the memory 65 those target data in which the reliability of the apex angle is low, in order to exclude the target data from the processing objects. However, the target data in which the reliability of the apex angle is low may be excluded from the processing objects by another method. In one example, the detection information output part 74 that obtains the target data in which the reliability of the apex angle is low without deletion from the memory 65 in a memory 65 may exclude the target data from the processing objects by not selecting the target data as the objects to be output to a vehicle controller 2.In the above-described embodiment, ESPRIT is used as an angle estimation system for direction calculation processing. However, MUSIC, PRISM or other systems may be used as an angle estimation system.Moreover, the electric power of a perpendicular component of a mode vector corresponding to a tip angle can be additionally regarded as a reliability condition for determining the reliability of the tip angle. The electric power of the perpendicular component increases in the case that a radar device receives more reflection waves RW than a possible division number in an angle estimation system. Consequently, the fact that the electric power of the perpendicular component is less than a prescribed threshold value can be set as one of the reliability conditions.In the above-described embodiment, the radar device 1 includes four reception antennas 51. However, the radar device may include five or more reception antennas or include less than four reception antennas, unless it includes only one reception antenna. Normally, the more the number of receiving antennas increases, the larger the possible division number in an angle estimation system.In the above-described embodiment, the vehicle control device 2 controls the behavior of the own vehicle to follow another vehicle. However, a vehicle controller 2 may control the behavior of the own vehicle for other purposes, for example, to reduce collision with obstacles.In the above-described embodiments, various functions are implemented by software, particularly processing based on programs. However, some or all of these functions may be implemented by hardware electrical circuits. Conversely, in the above-described embodiments, the function converted as a block may be implemented by the combination of software and hardware.While the invention has been shown and described in detail, the foregoing description is in all aspects illustrative and not restrictive. It is therefore to be understood that numerous other modifications and variations may be devised without departing from the scope of the invention.

Claims

A radar apparatus that acquires information on a target, the radar apparatus comprising: a direction deriving part (72) that derives an apex angle showing an angle of the target based on reception signals received by a plurality of antennas (51); a memory (65) that stores internal data showing the target, the internal data being derived based on the apex angle; a reliability determining part (75) that determines a reliability of a first apex angle, which is the apex angle relevant to the internal data, based on a second apex angle, which is another apex angle derived simultaneously with the first apex angle by the direction deriving part (72); and a data deleting part (76) that determines a reliability of a first apex angle, which is the apex angle relevant to the internal data, based on the reliability of the first apex angle determined by the reliability determination part (75) excluding the internal data relevant to the first apex angle from the processing objects, wherein the reliability determination part (75) is configured to determine whether the reliability of the first apex angle is high based on a lateral positional difference in a case where the number of angles is two or three; and determining that the reliability of the first apex angle is high in response to determining that the lateral position difference is equal to or less than a normal vehicle width, wherein the number of angles represents the number of apex angles derived simultaneously by the direction deriving part (72), and wherein the lateral position difference is represented by a difference between a first lateral position of the target relative to the radar device based on the first apex angle and a second lateral position of the target relative to the radar device based on the second apex angle.The radar apparatus according to claim 1, wherein the data deleting part (76) determines the reliability of the first peak angle relevant to the internal data showing a same target excluding the internal data from the processing objects a plurality of times based on results, wherein the reliability determining part (75) determines the reliability of the first peak angle.The radar apparatus according to claim 1 or 2, wherein the data deleting part (76) excludes the internal data from the processing objects by deleting the internal data from the memory (65).The radar apparatus according to any one of claims 1 to 3, wherein the reliability determination part (75) determines the reliability of the first apex angle based on a difference between a power of the first apex angle and a power of the second apex angle.The radar apparatus according to any one of claims 1 to 4, further including: a recognition information output part (74) that outputs, to a vehicle controller that controls a vehicle, the information on the target shown by the internal data that is not excluded from the processing objects by the data deletion part (76).A signal processing method for acquiring target data showing a target, the method comprising the steps of: (a) deriving an apex angle showing an angle of the target based on reception signals received by a plurality of antennas, (b) storing internal data showing the target in a memory (65), the internal data being derived based on the apex angle, (c) determining a reliability of a first apex angle, which is the apex angle relevant to the internal data, based on a second apex angle, which is another apex angle derived simultaneously with the first apex angle in the step (a), and (d) excluding, based on the reliability of the first apex angle determined in the step (c), The internal data relevant to the first apex angle from the processing objects, wherein step (c) comprises: in a case where the number of angles is two or three, determining whether the reliability of the first apex angle is high based on a lateral position difference; and determining that the reliability of the first apex angle is high, in response to determining that the lateral position difference is equal to or less than a normal vehicle width, wherein the number of angles represents the number of apex angles derived simultaneously in the deriving step (a), and wherein the lateral position difference is represented by a difference between a first lateral position of the target relative to the radar device based on the first apex angle and a second lateral position of the target relative to the radar device based on the second apex angle.The signal processing method according to claim 6, wherein the step (d) determines the reliability of the first apex angle relevant to the internal data showing a same target a plurality of times based on results, wherein the step (c) decides the internal data from the processing objects.The signal processing method according to claim 6 or 7, wherein the step (d) excludes the internal data by deleting the internal data from the memory (65) from the processing objects.The signal processing method according to any one of claims 6 to 8, further comprising the step of: (e) outputting, to a vehicle controller (2) controlling a vehicle, the information on the destination shown by the internal data not excluded from the processing objects in the step (d).

Images