ENVIRONMENTAL SURVEILLANCE RADAR DEVICE

Abstract

Environmental monitoring radar device (20a, 20b) mounted on a vehicle (70) to monitor an object in the vicinity of the vehicle (70), comprising: a transmitting unit (21, 22) designed to emit a combination of transmit signals modulated using several modulation modes as a radar wave; a receiving unit (21, 23) designed to generate a received signal from a wave reflected by an object, the object reflecting the radar wave emitted by the transmitting unit (21, 22) as the reflected wave; a spectrum generating unit (21) designed to generate a frequency spectrum for the respective modulation modes based on the received signal generated by the receiving unit (21, 22);an azimuth calculation unit (21) designed to extract, for the respective modulation modes, one or more peaks corresponding to the object from the corresponding frequency spectrum generated by the spectrum generation unit (21), and to subject the extracted peak(s) to azimuth expansion in order to calculate an azimuth of the object accordingly; an environment determination unit (21) designed to determine, for the respective modulation modes, on the basis of a degree of randomness of the corresponding frequency spectrum generated by the spectrum generation unit (21), whether an environment of the vehicle (70) is a complex environment with a reduction in the computational accuracy of the azimuth for the corresponding modulation mode;and a position calculation unit (21) designed, when the environment of the vehicle (70) is determined for at least one modulation mode from the modulation modes as the complex environment, to remove at least one azimuth corresponding to the at least one modulation mode from the azimuths calculated for each of the modulation modes in order to obtain at least one target azimuth which is at least one of the remaining azimuths excluding the removed azimuth; and to calculate a position of the object based on the at least one target azimuth.

Description

Technical field The present invention relates to a radar device that monitors the surroundings of a vehicle. State of the art Radar devices used to monitor objects in the vicinity of vehicles employ multiple modulation modes in combination to improve object detection accuracy. For example, the vehicle radar described in JP 2004-340755A transmits a combination of a radar wave modulated by the FMCW modulation mode and a radar wave modulated by the CW modulation mode. The vehicle radar uses the results of the CW modulation to determine the presence or absence of overlapping peaks in a frequency spectrum calculated using the FMCW modulation mode. In the case of overlapping peaks, the vehicle radar uses an azimuth calculated from a signal obtained using the CW modulation mode.Without overlapping peaks, the vehicle radar device uses an azimuth calculated from a signal obtained through the FMCW modulation mode. A radar system mounted on a vehicle may exhibit reduced accuracy in calculating the azimuth of a monitored target due to the vehicle's surroundings. For example, in FMCW modulation mode, in an environment where another vehicle is traveling near a continuous roadside object or similar, the radar system may detect an overlap of the roadside object's peak with the monitor's peak as a single peak in the frequency spectrum. If the azimuth difference between the roadside object and the monitor, where the peaks appear at the same position, is equal to or greater than the radar system's azimuth resolution, the two will be detected as separate objects, and their individual azimuths will be calculated.However, if the azimuth difference between the roadside object and the monitored target is smaller than the radar's azimuth resolution, the two are detected as a single object, and an azimuth is calculated that lies between them. When the radar recognizes the detected object as a monitored target, the azimuth of the monitored target becomes erroneous, shifting the actual position of the other vehicle towards the roadside object, thus reducing the computational accuracy. This, in turn, reduces the accuracy of the object's position calculated using the azimuth. Furthermore, when using CW modulation mode, for example, in an environment where another vehicle is driving alongside the target vehicle, a large number of peaks with a wide variety of frequencies may appear in the frequency spectrum. These peaks are based on reflected waves from the target vehicle's wheels. This occurs because the wheels have different velocity components. Therefore, it is difficult to extract peaks corresponding to the other vehicle's speed from the frequency spectrum. Even if the peaks can be extracted, the peak azimuths and the azimuths of the reflections from the wheels cannot be precisely separated at the time of azimuth development, leading to a reduction in the computational accuracy of the target vehicle's azimuth.This leads to a reduction in the computational accuracy of the object's position, which is calculated using the azimuth. Summary of the invention The vehicle radar system uses the results of a capture using the CW modulation mode to determine the presence or absence of overlapping peaks in the frequency spectrum calculated using the FMCW modulation mode. It was thus discovered that the vehicle radar system has a problem in that, if the vehicle's surroundings are characterized by reduced azimuth accuracy in the CW modulation mode, the presence or absence of overlapping peaks may be incorrectly determined.In particular, it was found that the vehicle radar device has a problem in that the azimuth calculated from a signal obtained using the CW modulation mode is used, even though the azimuth calculated from a signal obtained using the FMCW modulation mode has a higher accuracy. Accordingly, it was found that the vehicle radar device has a problem in that the computational accuracy of the object's position can decrease. The object of the present invention is to provide an environmental monitoring radar device that calculates the position of an object with high accuracy. This object is achieved by an environmental monitoring radar device with the features of claim 1. The dependent claims are directed to advantageous embodiments of the invention. One aspect of the present invention relates to an environmental monitoring radar device mounted on a vehicle to monitor an object in the vehicle's vicinity, comprising a transmitter, a receiver, a spectrum generator, an azimuth calculation unit, an environment determination unit, and a position calculation unit. The transmitter emits a combination of signals, modulated using several modulation modes, as a radar wave. The receiver generates a received signal from a wave reflected by an object. The object reflects the radar wave emitted by the transmitter as a reflected wave. The spectrum generator produces a frequency spectrum based on the received signal generated by the receiver for the respective modulation modes.The azimuth calculation unit extracts one or more peaks corresponding to the object from the relevant frequency spectrum generated by the spectrum generation unit and subjects these one or more extracted peaks to azimuth development or generation to calculate the object's azimuth for the respective modulation modes. Based on the degree of randomness of the relevant frequency spectrum generated by the spectrum generation unit, the environment determination unit determines whether the vehicle's environment is complex and thus reduces the computational accuracy of the azimuth for the relevant modulation mode.The position calculation unit, when the environment of the vehicle is determined from the modulation modes as the complex environment for at least one modulation mode, removes at least one azimuth corresponding to the at least one modulation mode from the azimuths calculated for each of the modulation modes in order to obtain at least one target azimuth which is at least one of the remaining azimuths except for the removed azimuth, and calculates a position of the object based on the at least one target azimuth. According to one aspect of the present invention, the combination of the transmitted signals, modulated by the modulation modes, is emitted as the radar wave. The frequency spectrum is then generated based on the received signal for the respective modulation modes. Furthermore, the peak corresponding to the object is extracted from the frequency spectrum, the extracted peak is subjected to azimuth development, and the azimuth of the object is calculated. For each modulation mode, the vehicle's environment, in which the azimuth calculation accuracy decreases, differs. Therefore, even if the azimuth calculation accuracy decreases in one of several modulation modes, the azimuth can still be calculated accurately in other modulation modes. Thus, based on the degree of randomness of the frequency spectrum generated for each modulation mode, it is determined whether the vehicle's environment is complex, resulting in a decrease in azimuth calculation accuracy for that specific modulation mode. The object's position is then calculated using the azimuth determined in the modulation mode other than the one for which the environment is considered complex. Therefore, the object's position can be calculated accurately.Furthermore, it is possible to accurately calculate the trajectory of the object as a monitoring target, which is determined based on the object's position information, for use in various control systems. The reference numerals in parentheses in the claims only indicate a correspondence to specific devices in an embodiment described below as a mode and do not limit the technical scope of the present invention. Brief description of the drawings Fig. 1 is a block diagram representing a configuration of a vehicle-integrated system according to the present embodiment. Fig. 2 is a diagram representing waveforms or functions of radar waves emitted by a radar device according to the present embodiment. Fig. 3 is a diagram representing a paired mapping or paired matching between a peak of a frequency spectrum waveform for a 2FCW mode and peaks of a frequency spectrum waveform for an FMCW mode. Fig. 4 is a diagram describing a situation in which a roadside object is present in the vicinity of a vehicle. Fig. 5 is a diagram representing azimuth spectrum waveforms of the 2FCW mode and the FMCW mode in the presence of a roadside object. Fig. 6 is a diagram representing a frequency spectrum waveform of the FMCW mode in a clear environment.Figure 7 is a diagram showing a frequency spectrum waveform of the FMCW mode in a complex environment. Figure 8 is a diagram showing specifications that define azimuth information used depending on the situation. Figure 9 is a flowchart of a process procedure for calculating the position of an object. Figure 10 is a diagram showing the trajectory of another vehicle when azimuth information is used in a modulation mode in a clear environment with a roadside object present. Figure 11 is a diagram showing the trajectory of another vehicle when an average of azimuth information in 2FCW mode and azimuth information in FMCW mode with a roadside object present is used. Description of the embodiments Exemplary embodiments for carrying out the present invention are described below with reference to the drawings. 1. Configuration First, a vehicle-integrated system 100 according to the present embodiment is described with reference to Fig. 1. The vehicle-integrated system 100 is a system mounted on a vehicle that includes a radar system 10, a driver assistance ECU 30, a warning device 40, and a control ECU group 50. The radar system 10 includes radar devices 20a and 20b. Radar device 20a is a right rear radar device installed on the right side of a rear section of the vehicle. Radar device 20b is a left rear radar device installed on the left side of a rear section of the vehicle. Radar device 20a and radar device 20b have essentially the same configuration and functions. In the following description, radar device 20a and radar device 20b are collectively referred to as radar device 20. The radar system 10 preferably includes at least one radar device 20. That is, the radar system 10 can also include only one radar device 20 or three or more radar devices 20. In the present embodiment, radar device 20 corresponds to an environmental monitoring radar device. The radar device 20 is a millimeter-wave radar that monitors the surroundings of a vehicle 70 by repeatedly transmitting and receiving radar waves. The radar device 20 comprises a signal processing unit 21, a transmitting antenna unit 22, and a receiving antenna unit 23. The signal processing unit 21 generates a combination of transmitted signals, which are modulated by several modulation modes, and causes the transmitting antenna unit 22 to emit a transmitted wave as a radar wave based on the generated combination of transmitted signals. As shown in Fig. 2, in the present embodiment the transmitted signal modulated by the FMCW mode and the transmitted signal modulated by the 2FCW mode are combined into one set. The transmitting antenna unit 22 repeatedly transmits a radar wave based on this single set of transmitted signals in a predetermined cycle. FMCW is an abbreviation for frequency-modulated continuous wave, and 2FCW is an abbreviation for dual-frequency continuous wave. The receiving antenna unit 23 has N antennas arranged in a line in a vehicle width direction to receive waves reflected by an object reflecting the transmitted wave. N is an integer equal to or greater than two. The signal processing unit 21 generates a received signal from the received wave received by the respective N antennas contained in the receiving antenna unit 23 and produces a beat signal for each antenna. The beat signal is a frequency difference signal that exhibits a frequency difference between the transmitted signal and the received signal. The signal processing unit 21 also performs a frequency analysis process, such as an FFT, on the generated beat signal to produce a frequency spectrum. The signal processing unit 21 generates the frequency spectrum from the beat signal for the respective modulation modes. In the present embodiment, the signal processing unit 21 generates a frequency spectrum Sp_up for each of the antennas from the frequency rise portion of the FMCW mode of the beat signal and generates a frequency spectrum Sp_dn for each of the antennas from the frequency fall portion of the FMCW mode of the beat signal. The signal processing unit 21 then extracts an azimuth θ and power information for the respective peaks of the frequency spectra Sp_up and Sp_dn. In particular, the signal processing unit 21 performs an arrival direction estimation process in each of the frequency spectra Sp_up and Sp_dn using an algorithm such as "multiple signal classification" (hereinafter referred to as MUSIC) for N peak frequency components of the same frequency collected by the respective antennas to extract the azimuths θ. The signal processing unit 21 uses the extracted azimuths θ and the power information to perform a pair mapping or pair matching between the peak frequencies of the frequency spectrum Sp_up and the peak frequencies of the frequency spectrum Sp_dn for the same object. Then, for each object, the signal processing unit 21 calculates a relative velocity Vr of the object with respect to the own vehicle 70 and a distance R from the own vehicle 70 to the object from the pair-mapped peak frequencies of the frequency spectra Sp_up and Sp_dn. For the FMCW mode portions of the beat signal, the signal processing unit 21 can use the object's azimuth θ extracted from either the frequency rise or frequency fall portion of the beat signal as the FMCW mode azimuth θ. The signal processing unit 21 can also use the average of the object's azimuths θ extracted from the frequency rise and frequency fall portions as the FMCW mode azimuth θ. The signal processing unit 21 also generates a frequency spectrum Sp_cw for the respective antennas from the 2FCW mode portion of the beat signal. For the 2FCW mode portion of the beat signal, the signal processing unit 21 generates a frequency spectrum from the respective beat signals at the two transmit frequencies for each antenna and creates the frequency spectrum Sp_cw by summing the two generated frequency spectra. The signal processing unit 21 then extracts the azimuth θ and the power information at each of the peaks of the frequency spectrum Sp_cw. The azimuth θ can be determined by performing the arrival direction estimation process using an algorithm such as MUSIC. The signal processing unit 21 uses the extracted azimuths θ and the power information to calculate the relative velocity Vr of the object with respect to the own vehicle 70 and the distance R from the own vehicle 70 to the object from the peak frequencies of the frequency spectrum Sp_cw. Specifically, the signal processing unit 21 calculates the distance R, the azimuth θ, and the relative velocity Vr of the object from the portion of the beat signal obtained by the FMCW mode, and calculates the distance R, the azimuth θ, and the relative velocity Vr of the object from the portion of the beat signal obtained by the 2FCW mode. As shown in Fig. 3, the signal processing unit 21 then uses the calculated relative velocity Vr, the azimuth θ and the power information of a given object to perform a pair mapping or pair matching between the peak frequency of the frequency spectrum Sp_cw and the pair of peak frequencies of the frequency spectra Sp_up and Sp_dn corresponding to the same object. The signal processing unit 21 then generates object information based on the frequency spectra and outputs this information to the driving assistance ECU 30. The object information includes the object's position P, calculated from its distance R and azimuth θ, and its relative velocity Vr. The object's azimuth θ, used for calculating its position P, will be described in more detail later. In this embodiment, the transmitting antenna unit 22 and the signal processing unit 21 correspond to the transmitting unit, and the receiving antenna unit 23 and the signal processing unit 21 correspond to the receiving unit. The signal processing unit 21 implements the functions of a spectrum generation unit, an azimuth calculation unit, an environment determination unit, and a position calculation unit. The control ECU group 50 contains several ECUs that are mounted in the vehicle itself and do not include the driver assistance ECU 30, and is connected to a network 6. The driver assistance ECU 30 obtains object information from the respective radar devices 20 regarding the object detected by each radar device 20 and exchanges data with the control ECU group 50 via network 6. If an object is present in the vicinity of the vehicle that could collide with it, the driver assistance ECU 30 issues a warning output command to the warning device 40. The warning device 40 is an indicator located on a door mirror or inside the vehicle, a loudspeaker inside the vehicle, a display device inside the vehicle, or the like. The warning device 40 emits a sound to warn or attract attention, or displays a warning message in accordance with the warning output command from the driver assistance ECU 30. 2. Azimuth calculation accuracy The computational accuracy of the object's azimuth θ can decrease depending on the environment of the vehicle 70. The environment in which the computational accuracy of the object's azimuth θ decreases varies depending on the modulation modes. For example, in FMCW mode, the computational accuracy of the object's azimuth θ decreases if a roadside object 200, a highly reflective object, is present in the vicinity of the vehicle 70, as shown in Fig. 4. The roadside object 200 could be a guardrail, a noise barrier, or similar. For FMCW mode, the frequency of the beat signal depends on the distance R and the relative velocity Vr of the object. If a continuous object with high reflectivity, such as the roadside object 200, is present in the vicinity of the vehicle 70, a large number of peaks appear over a wide range of frequencies in the frequency spectra Sp_up and Sp_dn, as shown in Fig. 7. Accordingly, the peaks of the roadside object 200 appear at the same positions as the peaks of the object as a monitoring target in the frequency spectra Sp_up and Sp_dn. If the azimuth difference between the object and the roadside object 200, whose peaks appear at the same positions, is smaller than the azimuth resolution of the radar device 20, the radar device 20 cannot calculate the azimuth of the roadside object 200 and the azimuth of the object separately.As a result, the object's azimuth is calculated as an intermediate azimuth between the azimuth of the street-side object 200 and the object's actual azimuth. That is, the object's azimuth is calculated as an erroneous azimuth with a shift relative to the actual azimuth in the direction of the street-side object 200. In contrast, the frequency of the beat signal in 2FCW mode depends on the relative velocity Vr of the object, but not on the distance R of the object. Therefore, even if a roadside object 200 is present in the vicinity of the vehicle 70, the peaks appear in a frequency band corresponding to the velocity obtained by projecting the vehicle's own velocity in the direction of the roadside object onto the frequency spectrum Sp_cw. Furthermore, the roadside object 200, located behind the vehicle 70, has a velocity in a direction away from the vehicle 70, i.e., a negative relative velocity. Thus, there is essentially no overlap between the frequency peak of the object to be detected approaching the vehicle 70 and the frequency peak of the roadside object 200, or similar objects not being detected.Even if the roadside object 200 is present in the vicinity of the vehicle 70, the azimuth θ of the object, which is calculated in the 2FCW mode, does not shift in the direction of the roadside object 200, and thus the computational accuracy is not reduced. On the other hand, if a vehicle is present near the vehicle 70, for example, if another vehicle is driving side-by-side with the vehicle 70, the computational accuracy of the object's azimuth θ decreases in 2FCW mode. When another vehicle is present near the vehicle 70, the received waveform picked up by the radar device 20 contains reflected waves from the other vehicle's wheels. Since the wheels have different velocity components, the frequency of the beat signal calculated from the received waveform contains these different velocity components when the received waveform includes the waves reflected by the wheels. Therefore, when another vehicle is present near the vehicle 70, a large number of peaks appear in the frequency spectrum Sp_cw over a wide range of frequencies.As a result, the computational accuracy of the object's azimuth θ, calculated in 2FCW mode, decreases. When another vehicle is 70° away from the pilot's own vehicle, the influence of reflections from the other vehicle's wheels becomes small, thus avoiding the occurrence of a large number of peaks across a wide range of frequencies in the Sp_cw frequency spectrum. As can be seen from the above and is shown in Fig. 6, for FMCW mode, if the environment of the own vehicle 70 is a clear environment in which no roadside object 200 or similar is present, the degree of randomness of the frequency spectra Sp_up and Sp_dn is relatively low, and the azimuth θ of the object can be calculated with high accuracy. As shown in Fig. 7, for FMCW mode, if the environment of the own vehicle 70 is a complex environment in which a roadside object 200 or similar is present, the degree of randomness of the frequency spectra Sp_up and Sp_dn is relatively high, and the computational accuracy of the azimuth θ of the object decreases. If the environment surrounding the vehicle 70 is clear and unobstructed, with no other vehicles nearby, the degree of randomness of the frequency spectrum Sp_cw will be relatively low for 2FCW mode, allowing the object's azimuth θ to be calculated with high accuracy. Conversely, if the environment surrounding the vehicle 70 is complex and another vehicle is nearby, the degree of randomness of the frequency spectrum Sp_cw will be relatively high for 2FCW mode, reducing the accuracy of the object's azimuth θ calculation. When the degree of randomness of the frequency spectrum is low, i.e., the normal state prevails, the mean value of the object's azimuths θ, calculated using the two modulation modes, is used in the present embodiment to calculate the object's position P, as shown in Fig. 8, for both FMCW and 2FCW modes. This improves the stability of the object's azimuth θ. If the degree of randomness of the frequency spectrum is high in either FMCW or 2FCW mode, the object's azimuth θ calculated using that modulation mode is excluded, and the object's azimuth θ calculated in the other modulation mode is used to calculate the object's position P. If the degree of randomness of the frequency spectrum is high for both FMCW and 2FCW modes, the object's azimuth θ, calculated using FMCW mode, is used to determine the object's position P. If the computational accuracy of the object's azimuth θ decreases, the azimuth between the object and a roadside object (e.g., Object 200) is generally used for FMCW mode. Conversely, if the computational accuracy of the object's azimuth θ decreases in 2FCW mode, an azimuth unrelated to the object being detected—for example, an azimuth caused by a tire dip (tire bulge, tire edge) of the vehicle in the surrounding area—may be detected.If the computational accuracy of the object's azimuth θ decreases for both FMCW mode and 2FCW mode, the object's azimuth θ calculated using FMCW mode will be used accordingly, since there is a high probability that the azimuth θ calculated using FMCW mode is closer to the azimuth θ of the object being detected. For FMCW and 2FCW modes, the computational accuracy of the object's azimuth θ decreases when the environment surrounding the vehicle is complex, but the computational accuracies of the object's distance R and relative velocity Vr remain unaffected. Therefore, the values ​​calculated using either FMCW or 2FCW mode can be used for the object's distance R and relative velocity Vr. In the present embodiment, the distance R and relative velocity Vr calculated using FMCW mode are used. 3. Process The following describes a process procedure for calculating the object's position with reference to the flowchart in Fig. 9. This process procedure is performed by the signal processing unit 21 each time the frequency spectra Sp_up, Sp_dn, and Sp_cw of a beat signal are generated. First, the signal processing unit 21 in S10 extracts peaks from the frequency spectra Sp_up, Sp_dn, and Sp_cw, extracts power information for the respective peaks, and then extracts the azimuths θ at which reflected waves originate from peak frequency components collected by the N antennas. The signal processing unit 21 then uses the extracted azimuths θ and the power information to perform a pairwise mapping between the frequency peaks of the frequency spectra Sp_up and Sp_dn corresponding to the same object in order to calculate the relative velocity Vr and the distance R of the object. The signal processing unit 21 also calculates the relative velocity Vr and the distance R of the object from the peak frequency of the frequency spectrum Sp_cw. Then, for each object, the signal processing unit 21 in S20 performs a pair mapping between the peak frequency in the frequency spectrum Sp_cw and the pair of peak frequencies in the frequency spectra Sp_up and Sp_dn corresponding to the same (for the same) object. Then, the signal processing unit 21 in S30 determines, based on the degree of randomness of the respective frequency spectra Sp_up, Sp_dn and Sp_cw, whether the environment of the own vehicle 70 is a complex environment with a reduction in the computational accuracy of the azimuth θ of the object, for the FMCW mode and the 2FCW mode.In FMCW mode, the signal processing unit 21 determines that the environment of the vehicle 70 is a complex environment if at least the following conditions (i) and (ii) are met: (i) in a complex environment determination range, the number of peaks in the frequency spectrum Sp_up or the frequency spectrum Sp_dn is greater than a preset threshold; and (ii) in the complex environment determination range, the mean of the peak powers, obtained by averaging the power values ​​at the peaks in the frequency spectrum Sp_up or the frequency spectrum Sp_dn, is greater than a preset peak threshold. With three peaks, the mean of the peak powers would be a value obtained by averaging the three power values. The complex environment determination range is a predetermined region of a frequency spectrum, as shown in Figures 6 and 7. The complex environment determination range is preset according to the range of the distance R of the object as a monitoring target. That is, the region in which the object is present as a monitoring target constitutes the complex environment determination range. In the case of 2FCW mode, the signal processing unit 21 similarly determines that the environment of the own vehicle 70 is a complex environment if at least one of conditions (i) and (ii) is met. However, in the case of 2FCW mode, the complex environment determination range is preset according to the range of the relative velocity Vr of the object as a monitoring target. In FMCW mode, the signal processing unit 21 can determine that the environment of the own vehicle 70 is a complex environment if at least one of conditions (i), (ii), and (iii) is met: (iii) in the complex environment determination area, the average power in the frequency spectrum Sp_up or the frequency spectrum Sp_dn is greater than a preset average threshold. In 2FCW mode, the signal processing unit 21 can similarly determine that the environment of the own vehicle 70 is a complex environment if at least one of conditions (i), (ii), and (iii) is met. However, in 2FCW mode, the average power in the frequency spectrum Sp_cw is compared to the average threshold. Subsequently, the signal processing unit 21 in S40 generates object information about the object extracted in S10. First, the signal processing unit 21 determines the azimuth θ of the object, which is to be used in calculating the object's position, based on the results of the determination in S30 and the specifications shown in Fig. 8. Then, the signal processing unit 21 calculates the object's position P from the determined azimuth θ and the object's distance R, which was calculated in S10, and generates the object information containing the object's position P and its relative velocity Vr, which was also calculated in S10. The signal processing unit 21 then terminates this process. 4. Businesses Fig. 10 represents a trajectory T determined from the position P of another vehicle 80, calculated using the object's azimuth θ, determined according to the specifications shown in Fig. 8. Fig. 11 represents a trajectory T determined from the other vehicle 80's position P, calculated using the mean of the object's azimuths θ, calculated using the FMCW and 2FCW modes. Figs. 10 and 11 indicate the latitude of a road with an x-coordinate and the rear of a vehicle in the direction of travel with a y-coordinate, and show a state in which the vehicle 70 is traveling along the y-direction on the road with the roadside object 200 on its right. That is, Fig. 10 and Fig. 11 show a state in which the environment of the own vehicle 70 is a complex environment for the FMCW mode.The other vehicle 80 travels linearly behind the first vehicle 70. The straight dashed line shows the actual trajectory of the other vehicle 80. According to Fig. 10, the azimuth θ of the object with lower computational accuracy is not used; instead, only the azimuth θ of the object with higher computational accuracy is used, so that the trajectory T corresponding to the actual behavior of the other vehicle 80 is captured. In contrast, according to Fig. 11, the mean value of the azimuth θ of the object containing an error with a displacement in the direction of the roadside object 200 and the azimuth θ of the object with higher computational accuracy is used, so that the trajectory T is captured midway between the actual trajectory of the other vehicle 80 and the roadside object 200. According to Fig. 11, the trajectory T is closer to the actual trajectory of the other vehicle 80 as it gets closer to the own vehicle 70. One factor that makes the trajectory T closer to the actual trajectory is that the azimuth difference between the other vehicle 80 and the roadside object 200, as seen from the radar device 20, increases as the other vehicle 80 approaches the own vehicle 70. In particular, as the other vehicle 80 gets closer to the own vehicle 70, the azimuth difference between the other vehicle 80 and the roadside object 200 exceeds the radar's azimuth resolution in the frequency spectra Sp_up and Sp_dn at peaks at the same positions. Accordingly, the azimuth of the other vehicle 80 and the azimuth of the roadside object 200 can be calculated separately.Other factors according to which the trajectory T comes closer to the actual trajectory are that, as the other vehicle 80 approaches the own vehicle 70, the roadside object 200 is hidden behind the other vehicle 80 and the radar device 20 no longer receives reflected waves from the roadside object 200, and that the reflection intensity of the other vehicle 80 becomes higher. 5. Beneficial effects According to the embodiment described above, the following advantageous effects can be achieved. (1) It is determined, based on the degree of randomness of the frequency spectrum generated using the FMCW mode and the 2FCW mode respectively, whether the vehicle's environment is a complex environment that reduces the computational accuracy of the object's azimuth θ for that modulation mode. Then, the object's position P is calculated with high accuracy using the object's azimuth θ calculated using a modulation mode other than the one for which the environment is determined to be complex. (2) At least one of the number of peaks in the frequency spectra Sp_up, Sp_dn, and Sp_cw, the mean values ​​of the peak powers, and the mean power of the frequency spectra can be used as an index of the degree of randomness.(3) For the two modulation modes FMCW and 2FCW, averaging the object's azimuths θ calculated for FMCW and 2FCW modes respectively improves the stability of the object's azimuth θ when the vehicle's environment is clear. This leads to an improvement in the stability of the calculated object position information. (4) The vehicle's environment, which reduces the computational accuracy of the object's azimuth θ, differs between FMCW and 2FCW modes. Therefore, even if the computational accuracy of the object's azimuth θ decreases in one mode, the object's azimuth θ can often be calculated with high accuracy in the other mode.Accordingly, the use of the object's azimuth θ, calculated using the FMCW mode, and / or the object's azimuth θ, calculated using the 2FCW mode, depending on the situation, allows the object's position P to be calculated with high accuracy. Other embodiments Above, an embodiment of the present invention has been described. However, the present invention is not limited to the embodiment above, but can be modified in various ways. (a) In the embodiment above, the modulation modes FMCW and 2FCW are used. However, the present invention is not limited thereto. For example, a pulse modulation mode and the FMCW mode can be used, or the pulse modulation mode and the 2FCW mode can be used. The modulation modes can include any combination of modulation modes. The 2FCW mode can be a multi-frequency CW mode in which continuous waves of three or more transmit frequencies are transmitted sequentially. Furthermore, the modulation modes can include a combination of three or more modulation modes.In the case of using a combination of three or more modulation modes, the positions θ of the object, calculated using the two or more modulation modes in a clear environment, can be averaged and used to calculate the position P of the object, provided that the environment of the vehicle 70 is a clear environment for two or more of the modulation modes. (b) In the above embodiment, multiple functions of one component can be implemented by multiple components, or one function of one component can be implemented by multiple components. Furthermore, multiple functions of multiple components can be implemented by one component, or one function of multiple components can be implemented by one component. Some components of the above embodiment can be omitted.At least some of the components of the above embodiment can be added to or replace other components of the above embodiment. All modes included within the technical concept of the claims are embodiments of the present invention. (c) Apart from the environmental monitoring radar device described above, the present invention can be implemented in various modes, for example as a system comprising the environmental monitoring radar device as a component and an object detection method.

Claims

Environmental monitoring radar device (20a, 20b) mounted on a vehicle (70) to monitor an object in the vicinity of the vehicle (70), comprising: a transmitting unit (21, 22) designed to emit a combination of transmit signals modulated using several modulation modes as a radar wave; a receiving unit (21, 23) designed to generate a received signal from a wave reflected by an object, the object reflecting the radar wave emitted by the transmitting unit (21, 22) as the reflected wave; a spectrum generating unit (21) designed to generate a frequency spectrum for the respective modulation modes based on the received signal generated by the receiving unit (21, 22);an azimuth calculation unit (21) designed to extract, for the respective modulation modes, one or more peaks corresponding to the object from the corresponding frequency spectrum generated by the spectrum generation unit (21), and to subject the extracted peak(s) to azimuth expansion in order to calculate an azimuth of the object accordingly; an environment determination unit (21) designed to determine, for the respective modulation modes, on the basis of a degree of randomness of the corresponding frequency spectrum generated by the spectrum generation unit (21), whether an environment of the vehicle (70) is a complex environment with a reduction in the computational accuracy of the azimuth for the corresponding modulation mode;and a position calculation unit (21) designed, when the environment of the vehicle (70) is determined for at least one modulation mode from the modulation modes as the complex environment, to remove at least one azimuth corresponding to the at least one modulation mode from the azimuths calculated for each of the modulation modes in order to obtain at least one target azimuth which is at least one of the remaining azimuths excluding the removed azimuth; and to calculate a position of the object based on the at least one target azimuth. Environmental monitoring radar device (20a, 20b) according to claim 1, wherein the environmental determination unit (21) is designed to determine, for the respective modulation modes, that the environment of the vehicle (70) is the complex environment when at least one of the following conditions is met: (i) in a preset determination range, the number of one or more peaks in the frequency spectrum is greater than a preset threshold; (ii) in the determination range, an average peak power obtained by averaging power values ​​at the respective one or more peaks in the frequency spectrum is greater than a preset peak threshold; and (iii) in the determination range, an average power in the frequency spectrum is greater than a preset average threshold. Environmental monitoring radar device (20a, 20b) according to claim 1 or 2, wherein the position calculation unit (21) is designed, when there are two or more of the modulation modes for which the environment determination unit (21) does not determine that the environment of the vehicle (70) is the complex environment, to calculate the position of the object using an average of the azimuths calculated by the azimuth calculation unit (21) using the corresponding two or more of the modulation modes. Environmental monitoring radar device (20a, 20b) according to claim 3, wherein the modulation modes include an FMCW mode and a 2FCW mode.

Images