Radar device and signal processing method

Abstract

To continuously manage a track even if a target is lost.SOLUTION: According to an embodiment, a radar device includes an antenna for receiving a radio wave, and an excitation reception part for generating a transmission signal and receiving the radio wave. The radar device further includes a beam forming unit, a detection processing unit, and a track management unit. The beam forming unit forms a reception beam to obtain reception data of the radio wave. The detection processing unit detects a target from received data to obtain observation data. A correlation processing unit calculates a speed and a track of the target from time-series observation data. The track management unit manages a track of the target. Also, the track management unit calculates a stop prediction point of the target whose stop is predicted from a speed change. Further, the track management unit continuously holds the track of a lost target by associating with the stop prediction point of the target.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] FIELD An embodiment of the present invention relates to a radar device and a signal processing method. [Background technology]

[0002] Radar equipment receives and analyzes echo signals from transmitted radar waves to obtain target information. Frequency filters, such as the Moving Target Indicator (MTI), are a common method of signal processing. This process suppresses (eliminates) a band similar to the transmission frequency from the echo signal, extracting only the components with Doppler shifted frequency. This eliminates unnecessary wave components such as ground clutter and sea clutter, making it possible to extract only the echo signal from the target.

[0003] Pulse Doppler processing is also a common method. This is a process that forms a filter bank and integrates the echo signal to improve (extend) the S / N ratio of the target. By ignoring the output of the zero bank, which contains a lot of clutter from fixed targets (such as ground clutter), only moving targets can be extracted. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] International Publication No. WO2017 / 138119 Summary of the Invention [Problem to be solved by the invention]

[0005] Incidentally, an unmanned aerial vehicle (UAV) is an object that moves through the air either by radio control or autonomously, and so-called drones fall into this category. Because these types of targets move at low speeds, the Doppler frequency shift is small, and they can be erased along with clutter during signal processing. For the same reason, drones that stop in the air cannot be detected by radar. The same thing happens when a helicopter or other aircraft hover in the air or a vehicle moving on the ground stops.

[0006] While it's one thing if the target is moving away or if it's still in sight, if an approaching target suddenly becomes unseen (known as a loss of sight), we are exposed to a major threat. Drones are particularly small, making it difficult to detect them from a distance, even with large radars. Even if a drone is lost, it is desirable to continue managing its flight path so that we are not exposed to a sudden threat.

[0007] Therefore, an object of the present invention is to provide a radar device and a signal processing method that are capable of continuously managing the track even if the radar is lost. [Means for solving the problem]

[0008] According to an embodiment, a radar device includes an antenna for receiving radio waves and an excitation / reception unit for generating a transmission signal and receiving the radio waves. The radar device further includes a beam forming unit, a detection processing unit, and a track management unit. The beam forming unit forms a reception beam to obtain reception data of the radio waves. The detection processing unit detects a target from the reception data and obtains observation data. The correlation processing unit calculates the speed and track of the target from the time-series observation data. The track management unit manages the track of the target. Furthermore, the track management unit calculates the predicted stopping point of the target, whose stopping is predicted based on changes in speed. Furthermore, the track management unit continuously stores the track of a lost target in association with the predicted stopping point of the target. [Brief explanation of the drawings]

[0009] [Figure 1]FIG. 1 is a functional block diagram showing an example of a radar device according to an embodiment. [Figure 2] FIG. 2 is a flowchart illustrating an example of a processing procedure of the radar device according to the embodiment. [Figure 3] FIG. 3 is a diagram showing an example of a target's flight path displayed on the display unit 60. As shown in FIG. [Figure 4] FIG. 4 is a diagram for explaining an example of the concept of the quality of the process of tracking a target. [Figure 5] FIG. 5 is a diagram illustrating the track management according to the embodiment in comparison with existing techniques. [Figure 6] FIG. 6 is a diagram for explaining an example of target determination by the rotor system determination unit 22. In FIG. [Figure 7] FIG. 7 is a diagram for explaining an example of target determination by the rotor system determination unit 22. In FIG. DETAILED DESCRIPTION OF THE INVENTION

[0010] <Summary> In recent years, the targets that must be captured by radar have become more diverse, ranging from those that move relatively monotonously at long distances to those that move with high maneuverability at close ranges. In most cases, targets do not suddenly appear within the radar's coverage area, but approach from outside the coverage area, managed as a wake. Loss of track occurs when a target whose track was being managed slows down or stops due to hovering, etc., and is lost.

[0011] Track management is an important process in radar signal processing. Generally, when a target is lost, it is immediately moved to a memory track, and if this state continues for a certain period of time, it is removed from the track management targets. While this may be acceptable for targets at a distance, if this occurs with a close-range target, there is nothing that can be done if the target remains in the same location, which is extremely dangerous. In this embodiment, a technology that can deal with such a situation will be described.

[0012] <Configuration> 1 is a functional block diagram showing an example of a radar device according to an embodiment. The radar device 10 according to the embodiment includes an antenna 40 for receiving incoming radio waves, an excitation / reception unit 50, a signal processing unit 20, a display unit 60 for displaying various information, and a radar control unit 30 for overall control of these components. The platform of the radar device 10 is not limited to a ground-mounted type, and may be any type, such as a vehicle-mounted type, an aircraft-mounted type, a ship-mounted type, or a satellite-mounted type.

[0013] 1, the excitation receiver 50 receives radio waves received by the antenna 40 and amplifies the signals using a low-noise amplifier or the like to obtain received signals. The received signals are converted from analog to digital and then passed to the signal processor 20.

[0014] The signal processing unit 20 detects targets from the data of the received signal and generates a radar video signal, which is displayed on the display unit 60. Here, the signal processing unit 20 includes a beam forming unit 11, a detection processing unit 12, a correlation processing unit 13, a micro-Doppler extraction unit 21, a rotor system determination unit 22, and a track management unit 23. These are functions that are realized by the processor executing programs installed in a computer equipped with a processor and memory.

[0015] The beam forming unit 11 simultaneously forms a plurality of receiving beams by, for example, DBF (Digital Beam Forming) processing, and obtains reception data of the received radio waves. The detection processing unit 12 detects targets from the received data and obtains observation data. The correlation processor 13 calculates the speed and track of the target from the time-series observation data. The speed of the target, together with the position of the target, is the basis of the tracking process. The existing tracking process will be described.

[0016] In the tracking process, a smoothed position is calculated using the average value of the position data, etc. In the time series observation, the current predicted position is determined from the previous smoothed position and the previous velocity vector. Then, a correlation gate of a predetermined size and shape is opened, centered on the current predicted position. If multiple target position data are observed in the current observation, for example, the data closest to the current predicted position in the correlation gate is taken as the observed position (filtering process), and the smoothed position is determined. The next predicted position can be determined from this smoothed position and the velocity vector at that time. The target's trajectory can also be calculated through this process.

[0017] The track management unit 23 manages the track of the target. In this embodiment, the track management unit 23 monitors the trend of changes in the target's speed over time, and identifies a target that is predicted to stop because its speed approaches 0. Then, from the current position and speed of the target, it calculates the position where the speed is predicted to become 0, i.e., the predicted stopping point.

[0018] A stopped target is often lost because its Doppler shift cannot be detected. The track management unit 23 flags the lost target and continuously stores the track of this target in association with its predicted stopping point. A target whose track has been lost may be remaining in the same location as before, so the detection processing unit 12 improves the detection sensitivity in a region of a predetermined size (ROI (Region of Interest)) that includes the predicted stopping point of the lost target.

[0019] Sensitivity can be improved by increasing the S / N ratio in the area near the predicted stopping point. For example, this can be done by increasing the number of pulse integrations in that area or by lengthening the transmit pulse width. This makes it possible to detect weak Doppler shifts caused by, for example, the rotors of a drone.

[0020] The micro-Doppler extraction unit 21 acquires observation data from the detection processing unit 12 and extracts micro-Doppler from the above-mentioned region of interest from this observation data. Incidentally, in recent years, research into identification using micro-Doppler has been active, particularly targeting rotary-wing platforms (helicopters, multi-rotor drones). This is a technology that determines whether an object is a rotary-wing target by examining echoes (called micro-Doppler signatures) with a unique frequency shift from rotors that are different from the target itself.

[0021] Based on the extracted micro-Doppler, the rotor system determination unit 22 creates image information by mapping the Doppler shift in the region of interest onto a two-dimensional image. This image information is known as, for example, an RDM (Range Doppler Map) and shows patterns that can be distinguished according to the type, characteristics, attributes, etc. of the target. The rotor system determination unit 22 then determines the target based on this image information. For example, AI (Artificial Intelligence) technology can be applied to determine the target, as will be described in detail later with reference to Figures 6 and 7.

[0022] <effect> 2 is a flowchart showing an example of a processing procedure of a radar device according to an embodiment. In FIG. 2, the radar device 10 tracks targets and manages the track of each captured target (step S1). During this process, the radar device 10 monitors changes in the target's speed and predicts the possibility of the target's track stopping (step S2). That is, if there is a target whose speed is predicted to become zero (stop) based on the trend of changes in moving speed (Yes in step S2), the radar device 10 calculates the predicted stopping point of that target, sets a flag to distinguish that target's track from others, and continues to manage it (step S3).

[0023] Next, if observation data of the flagged target is input (Yes in step S4), the radar device 10 resets (removes) the flag (step S5) and returns to normal track management processing. On the other hand, if there is a continued state in which no observation data is input (No in step S4), the radar device 10 determines that the target has stopped and continues tracking (memory track) of this target (step S6).

[0024] Furthermore, the radar device 10 determines whether or not there is observation data, and if there continues to be no input (No in step S7), the radar device 10 observes whether or not there is micro Doppler in the region of interest including the predicted stopping point (step S8). Note that if there is input of target observation data in step S7 (Yes), the flag is reset and the processing procedure returns to step S1.

[0025] If there is no micro-Doppler in the observation data from the region of interest (No in step S9), the radar device 10 resets the flag and returns to the processing procedure of step S1. On the other hand, if it is determined that there is micro-Doppler (Yes in step S9), the radar device 10 determines that the target whose track has been lost has stopped, and sets the quality (TQ) of the processing for tracking the target to 0. At this time, the track is not deleted, but is continuously stored in the memory track (step S10).

[0026] Furthermore, the radar device 10, while retaining the track, determines whether or not observation data has been input into the correlation gate (step S11). Steps S10 and S11 may be looped a predetermined number of times. If observation data is detected (Yes in step S11), the radar device 10 regards the observation data as coming from the same target as the target determined to have been stopped in step S6, and resumes tracking the track (step S12).

[0027] That is, in step S12, if a wake is generated again from the area of ​​interest, the radar device 10 takes over management of the wake by regarding it as the wake of the same target as the target whose wake was retained, and continues to manage the same wake with attribute information and the like remaining the same. At this time, an appropriate correlation process may be performed. That is, whether the target that generated a wake from the area of ​​interest and the target whose wake was retained are the same may be determined using a threshold value through correlation process.

[0028] 3 is a diagram showing an example of a target's flight path displayed on the display unit 60. The hatched (□) symbol is an icon indicating the target, and the straight line extending from this icon indicates the target's velocity vector (direction and magnitude).

[0029] FIG. 4 is a diagram illustrating an example of the concept of the quality of target tracking processing. Tracking quality (TQ) is a parameter that indicates the state in which a target is observed within the correlation gate. In other words, if a target observed in the tracking process is present within the correlation gate, the TQ value increases, and if it is not present within the correlation gate, the TQ value decreases. In FIG. 4, as the target is observed within the correlation gate from a state where TQ is minimum, TQ improves, for example, reaching a maximum value of TQ4, where it is maintained. On the other hand, each time observation within the correlation gate fails, TQ decreases and reaches a minimum state. If TQ remains at its minimum value for a while, existing technology will reject the track.

[0030] 5A and 5B are diagrams comparing track management in an embodiment with existing technologies. FIG. 5A shows track management in existing technologies. For example, suppose the speed of a target managed under track number N decreases over time (number of scans) and disappears due to hovering. In this case, with existing technologies, the track managed under track number N is deleted, and if observation data is detected again, management is resumed as a different track (track number M).

[0031] In contrast to this, in this embodiment, even if a track disappears, it is retained at the predicted stopping point with a flag set, with the same track number N, as shown in Figure 5(b). If observation data is detected again, it is regarded as the same track, and track management is resumed under track number N.

[0032] 6 and 7 are diagrams for explaining an example of target determination by the rotor system determination unit 22. As shown in Fig. 6, a set of previously observed RDMs and target information corresponding to each RDM (for example, bird, drone, plane, etc.) is repeatedly provided as training data to a neural network, such as a deep neural network (DNN), for learning. By inputting the RDM generated from current observation data into the trained model obtained in this way, it becomes possible to determine the most likely target.

[0033] <Effects> As described above, in the embodiment, in addition to the normal tracking processing (observation, smoothing, prediction), a flag is set for a track that is expected to stop. Then, in addition to the normal track management (track establishment, TQ control, track deletion), for a flagged track (expected to stop), the track is not deleted, but is maintained as is at the predicted stopping point.

[0034] Additionally, when a stop is predicted and observation data cannot be acquired, the S / N ratio around the predicted stop point is improved, making it possible to detect, for example, a weak Doppler shift caused by the drone's rotor.

[0035] In existing technology, a tracking filter detects the deceleration of a track and, if it is lost, transfers to a memory track. In this embodiment, a stop from deceleration is assumed and an expected stopping point is calculated. Then, assuming that the track will continue to exist at this expected stopping point, the track is retained (the track is not deleted).

[0036] In the meantime, it is possible to improve the S / N ratio by increasing the number of beams, and add a function to search for micro-Doppler. If micro-Doppler is detected, it is determined to be a rotorcraft system, and if it is not, it can be determined to be a vehicle system (this can also be determined by altitude). Then, if the lost track begins to move again, the Doppler-displaced echo can be detected, and it can be placed on the original track.

[0037] Therefore, according to the embodiments, it is possible to provide a radar device and a signal processing method that can continuously manage the track even if the target is lost. As a result, it is possible to continuously evaluate the threat level of the target, and to eliminate damage from the threat or minimize any damage that may occur.

[0038] It should be noted that the present invention is not limited to the above-described embodiment. For example, in the flowchart of Fig. 2, whether or not the observation data has been restored is checked twice (steps S4 and S7), but this does not have to be limited to two times. The number of times this check is made can be freely set by the user as a parameter for system operation. Similarly, the upper limit of the number of times the processing loop of steps S10 and S11 is repeated can also be freely set as a parameter.

[0039] Although an embodiment has been described, this embodiment is presented as an example and is not intended to limit the scope of the invention. This novel embodiment can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. This embodiment and its modifications are included within the scope and spirit of the invention, and are also included in the inventions described in the claims and their equivalents. [Explanation of symbols]

[0040] 10... radar device, 11... beam forming unit, 12... detection processing unit, 13... correlation processing unit, 20... signal processing unit, 21... micro-Doppler extraction unit, 22... rotor system determination unit, 23... track management unit, 30... radar control unit, 40... antenna, 50... excitation receiving unit, 60... display unit.

Claims

1. An antenna that receives radio waves, an excitation receiving unit that generates a transmission signal and receives the radio wave; a beam forming unit that forms a receiving beam to obtain received data of the radio wave; a detection processing unit that detects a target from the received data and obtains observation data; a correlation processor that calculates the speed and track of the target from the time-series observation data; a track management unit that manages the track of the target, The track management unit calculating an expected stopping point of the target that is predicted to stop from the change in speed; A radar device that continuously stores the track of a lost target in association with the target's predicted stopping point.

2. The radar device according to claim 1 , wherein the detection processing unit improves detection sensitivity in a region of interest of a predetermined size that includes an expected stopping point of the lost target.

3. The radar device according to claim 2 , further comprising a micro-Doppler extraction unit that extracts micro-Doppler in the region of interest from the observation data.

4. 4. The radar device according to claim 3, further comprising a determination unit that generates image information by mapping the Doppler shift in the region of interest onto a two-dimensional image based on the micro-Doppler, and determines the target based on the image information.

5. 5. The radar device according to claim 4, wherein the determination unit inputs the created image information into a neural network that has been trained by repeatedly providing a set of image information and target information prepared in advance, and determines the target.

6. 6. The radar device according to claim 5, wherein, when the target is determined to be a rotary wing body, the track management unit regards the lost target as the rotary wing body and continues to retain the track of the lost target.

7. When a track is generated again from a predetermined size attention area including the predicted stopping point of the lost target, The track management unit 7. The radar device according to claim 1, wherein if a result of correlation processing between the retained track and the re-occurring track is equal to or greater than a predetermined value, the radar device assumes that these tracks originate from the same target and takes over management of the tracks.

8. A signal processing method by a computer for a radar device that captures and receives radio waves using an antenna, comprising: a step in which the computer forms a receiving beam and obtains received data of the radio waves; a step in which the computer detects a target from the received data and obtains observation data; a step in which the computer calculates the speed and trajectory of the target from the time-series observation data; said computer managing the track of said target; a step in which the computer calculates an expected stopping point of the target that is predicted to stop from the change in speed; and a step in which the computer continuously stores the track of the lost target in association with the predicted stopping point of the target.

Images