JP7693768B2 - Radar device, signal processing device, and signal processing method - Google Patents

Description

Embodiments of the present invention relate to a radar device, a signal processing device, and a signal processing method.

For example, radar devices are known that are installed on aircraft to detect approaching targets. In this type of radar device, it has been proposed to suppress clutter using a pattern classifier that applies machine learning, for example.

JP 2021-015079 A

When a target is far away or the RCS (Radar Cross Section) is small, the S/N (Signal to Noise ratio) of the received signal becomes small. Increasing the aperture area of the radar device's antenna or increasing the transmission power is effective for increasing the S/N, but this increases the scale of the hardware. In addition, the S/N can also be increased by increasing the number of pulses used for coherent integration, but this may increase the update rate or limit the detection area. Furthermore, in the case of targets moving at high speeds, the detection performance is significantly reduced. Therefore, it is necessary to improve the detection performance of targets moving at high speeds without increasing the hardware scale of the radar antenna by using a signal processing method using machine learning that is different from the conventional method. However, in order to handle targets of various speeds, it is necessary to train a pattern classifier that can handle various speeds, which increases the scale of machine learning.

The objective is to provide a radar device, a signal processing device, and a signal processing method that reduce noise and have high performance in detecting moving targets.

According to an embodiment, a radar device includes an antenna unit, a transceiver unit, and a signal processor. The transceiver unit transmits radar pulses from the antenna unit and receives radio waves arriving at the antenna unit to generate a received signal. The signal processor processes the received signal. The signal processor includes a pulse compression processor, an integration processor, a first sensitivity enhancement processor, a second sensitivity enhancement processor, and a detection processor. The pulse compression processor pulse compresses the received signal to generate a compressed pulse signal. The integration processor pulse integrates the compressed pulse signal to generate an integrated pulse signal. The first sensitivity enhancement processor reduces noise in the compressed pulse signal to generate a first high sensitivity signal. The second sensitivity enhancement processor reduces noise in the first high sensitivity signal to generate a second high sensitivity signal. The detection processor detects a target from the second high sensitivity signal. The second sensitivity enhancement processor generates the second high sensitivity signal from the integrated pulse signal and the first high sensitivity signal.

FIG. 1 is a functional block diagram showing an example of a radar device according to each embodiment. FIG. 2 is a functional block diagram showing an example of a signal processing unit according to the first embodiment. FIG. 3 is a block diagram showing an example of the configuration of the sensitivity enhancement processing section 122 and the sensitivity enhancement processing section 123. As shown in FIG. FIG. 4 is a flowchart illustrating an example of a target detection process by the signal processing unit in the first embodiment. FIG. 5 is a diagram for explaining the operation of the sensitivity enhancement processors 122 and 123. As shown in FIG. FIG. 6 is a functional block diagram showing an example of a signal processing unit according to the second embodiment. FIG. 7 is a functional block diagram showing an example of a signal processing unit according to the third embodiment. FIG. 8 is a flowchart showing an example of a target detection process by the signal processing unit in the third embodiment. FIG. 9 is a functional block diagram showing an example of a signal processing unit according to the fourth embodiment. FIG. 10 is a functional block diagram showing an example of a signal processing unit according to the fifth embodiment.

The embodiments will be described below with reference to the drawings. FIG. 1 is a functional block diagram showing an example of a radar device according to each embodiment. In the embodiments, a pulse radar is assumed. As shown in FIG. 1, the radar device 1 includes an antenna unit 10, a transmitting/receiving unit 11, and a signal processing unit 12. The radar device 1 may also include a display unit 13 for visually displaying the output of the signal processing unit 12 to present it to a user. Here, the user is, for example, a watchman who monitors a target using the radar device 1. The target is, for example, an object that moves through the air at a relatively high speed, such as a flying object or aircraft. The target may or may not have a person on board. Furthermore, the target is not necessarily limited to an object that moves through the air.

The antenna unit 10 is, for example, an array antenna having multiple antenna elements arranged in a regular pattern. The antenna unit 10 transmits radio waves (radar waves) into space and receives radio waves including reflected waves (received signals) from targets.

The transmitter/receiver unit 11 generates radar pulses and sends them to the antenna unit 10. The transmitter/receiver unit 11 also generates a received signal by performing processes such as amplification and analog/digital (A/D) conversion on the radio waves arriving at the antenna unit 10. The transmitter/receiver unit 11 sends the generated received signal to the signal processing unit 12.

The signal processing unit 12 is, for example, a computer equipped with a memory and a processor. The signal processing unit 12 performs various signal processing on the received signal from the transmission/reception unit 11. This signal processing includes, for example, signal processing such as clutter suppression, radar video generation, and target detection. The signal processing unit 12 sends information on the detected targets to the display unit 13.

The display unit 13 visually presents the target to the user based on the target information input from the signal processing unit 12.

[First embodiment]
2 is a functional block diagram showing an example of a signal processing unit according to the first embodiment. The signal processing unit 12 includes a pulse compression processing unit 121, a sensitivity enhancement processing unit 122, a sensitivity enhancement processing unit 123, and a detection processing unit 124.

The pulse compression processing unit 121 performs pulse compression on the received signal from the transmission/reception unit 11 to generate a compressed pulse signal. The pulse compression processing unit 121 outputs the compressed pulse signal to the sensitivity enhancement processing unit 122. The compressed pulse signal is three-dimensional data including, for example, distance, azimuth, and elevation angle as axes. The compressed pulse signal may be four- or more-dimensional data including other elements such as Doppler.

The high sensitivity processing unit 122 performs a first high sensitivity processing on the compressed pulse signal from the pulse compression processing unit 121 to generate a first high sensitivity signal. The high sensitivity processing unit 122 outputs the first high sensitivity signal to the high sensitivity processing unit 123. The first high sensitivity signal is a signal in which signals other than the target signal component in the compressed pulse signal have been reduced as noise. As will be described later, the high sensitivity processing unit 122 performs processing to reduce noise in the compressed pulse signal using a trained model. Here, the compressed pulse signal may be input one pulse at a time to the high sensitivity processing unit 122, or multiple compressed pulse signals, for example, in CPI (Coherent Processing Interval) units, may be input together.

The high sensitivity processing unit 123 performs a second high sensitivity processing on the first high sensitivity signal from the high sensitivity processing unit 122 to generate a second high sensitivity signal. The high sensitivity processing unit 123 outputs the second high sensitivity signal to the detection processing unit 124. The second high sensitivity signal is a signal in which signals other than the target signal component in the first high sensitivity signal have been reduced as noise. As will be explained later, the high sensitivity processing unit 123 performs processing to reduce noise in the first high sensitivity signal using a trained model, similar to the high sensitivity processing unit 122.

The detection processing unit 124 detects the target from the second high-sensitivity signal from the high-sensitivity processing unit 123. In the second high-sensitivity signal output from the high-sensitivity processing unit 123, the levels of signal components other than the target are reduced. Therefore, the detection processing unit 124 can extract only the target.

Figure 3 is a block diagram showing an example of the configuration of the high sensitivity processing unit 122 and the high sensitivity processing unit 123. The high sensitivity processing unit 122 and the high sensitivity processing unit 123 are, for example, computers having a physical form as hardware. The high sensitivity processing unit 122 and the high sensitivity processing unit 123 may be configured by a single computer, or may be configured by separate computers. In the following, the high sensitivity processing unit 122 and the high sensitivity processing unit 123 are described as being configured by a single computer. The computer that constitutes the high sensitivity processing unit 122 and the high sensitivity processing unit 123 includes a processor 21 and a storage unit 24. Furthermore, the computer includes a ROM 22, a RAM 23, and a communication unit 25.

The processor 21 is, for example, a computing device such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). The processor 21 loads a program stored in the ROM 22 or the storage unit 24 into the RAM 23, and executes the program loaded into the RAM 23 to operate as the sensitivity enhancement processing unit 122 and the sensitivity enhancement processing unit 123.

ROM 22 is a non-volatile memory such as a flash memory, and operates as a program memory. ROM 22 stores programs executed by processor 21 and control data used by processor 21.

RAM 23 is a volatile memory such as a DRAM, and operates as a working memory that temporarily stores data. Programs are loaded into RAM 23. RAM 23 also temporarily stores data to be processed by processor 21.

The memory unit 24 is a storage configured by a non-volatile memory such as an SSD (Solid State Drive). The memory unit 24 stores a program 24a executed by the processor 21, data 24b used by the processor 21, a trained model 24c, and a trained model 24d. The trained model 24c is a trained model for the high sensitivity processing unit 122. The trained model 24c is a trained model in which the weights of each layer are set so as to input a compressed pulse signal and output a first high sensitivity signal in which noise components other than the target component contained in the compressed pulse signal are reduced. In other words, the training of the trained model 24c is performed to distinguish between a target signal component and a noise signal component contained in the input compressed pulse signal, and to output a signal in which the intensity of the signal component identified as noise is reduced. Limited training may be performed so that the signal range identified as the target in the compressed pulse signal includes a margin. The trained model 24d is a trained model in which the weights of each layer are set so as to input the first high-sensitivity signal and output a second high-sensitivity signal in which noise components other than the target component contained in the first high-sensitivity signal are reduced. In other words, the trained model 24d is trained to distinguish between the target signal components and noise signal components contained in the input first high-sensitivity signal and to output a signal in which the intensity of the signal components identified as noise is reduced. Limited training may be performed so that the signal range identified as the target in the first high-sensitivity signal includes a margin. The margin of the signal range identified as the target in the first high-sensitivity signal may be narrower than the margin of the signal range identified as the target in the compressed pulse signal.

The communication unit 25 is connected to the pulse compression processing unit 121 and acquires a compressed pulse signal from the pulse compression processing unit 121. The communication unit 25 is also connected to the detection processing unit 124 and outputs a second high sensitivity signal to the detection processing unit 124.

The processor 21 also includes a learning unit 21a and a learning unit 21b. The processor 21 operates as the learning unit 21a and the learning unit 21b by executing processing according to instructions written in the program 24a.

The learning unit 21a generates a trained model by repeating machine learning using the compressed pulse signal acquired from the pulse compression processing unit 121 via the communication unit 25 as training data. The generated trained model is stored in the memory unit 24 as trained model 24c.

The learning unit 21b repeats machine learning using the first high-sensitivity signal, which is the output of the high-sensitivity processing unit 122 generated by the learning unit 21a, as training data to generate a trained model. A signal that simulates the first high-sensitivity signal may be used as the training data. The generated trained model is stored in the memory unit 24 as trained model 24d.

For example, the trained model 24c and the trained model 24d can be generated by repeatedly providing training data to a Convolutional Neural Network (CNN). This type of machine learning is called deep learning. In the first embodiment, the learning unit 21a repeats supervised machine learning using a compressed pulse signal as training data to generate the trained model 24c. The training data in this case is, for example, data of a compressed pulse signal that includes only a target. Also, in the first embodiment, the learning unit 21b repeats supervised machine learning using a first high-sensitivity signal as training data to generate the trained model 24d. The training data in this case is, for example, data of a first high-sensitivity signal that includes only a target.

Here, the learning data used in the learning to generate the trained model 24c may be created taking into account the movement of the target between the compressed pulse signals. Similarly, the learning data used in the learning to generate the trained model 24d may be created taking into account the movement of the target between the first high-sensitivity signals.

In addition, the learning data used in learning to generate the trained models 24c and 24d may be data for each target velocity assumed for each Doppler filter bank.

The high sensitivity processing unit 122 and the high sensitivity processing unit 123 may also be implemented as software. In this case, the high sensitivity processing unit 122 and the high sensitivity processing unit 123 may be configured as a program that is expanded in RAM and executed by a processor.

The pulse compression processing unit 121 and the detection processing unit 124 may also be configured as dedicated hardware configured to execute the above-mentioned operations, or may be configured as software that causes a processor to execute the above-mentioned operations.

Next, the operation of the signal processing unit 12 of the first embodiment will be described. FIG. 4 is a flowchart showing an example of a target detection process by the signal processing unit 12 in the first embodiment. In step S1, the pulse compression processing unit 121 of the signal processing unit 12 performs pulse compression on the received signal from the transmitting/receiving unit 11 to generate a compressed pulse signal. When a high S/N signal is detected in the compressed pulse signal by prior threshold processing or the like, the target signal with this high S/N may be replaced with noise.

In step S2, the sensitivity enhancement processing unit 122 generates a first sensitivity enhancement signal from the compressed pulse signal. Specifically, the sensitivity enhancement processing unit 123 inputs the compressed pulse signal to the trained model 24c to obtain the first sensitivity enhancement signal.

In step S3, the sensitivity enhancement processing unit 123 generates a second sensitivity enhancement signal from the first sensitivity enhancement signal. Specifically, the sensitivity enhancement processing unit 123 inputs the first sensitivity enhancement signal to the trained model 24d to obtain the second sensitivity enhancement signal.

In step S4, the detection processing unit 124 detects a target from the second high-sensitivity signal. For example, the detection processing unit 124 detects a signal range in the second high-sensitivity signal that has an intensity equal to or greater than a threshold as the target.

FIG. 5 is a diagram for explaining the operation of the sensitivity enhancement processing units 122 and 123. In FIG. 5, the sensitivity enhancement processing unit 122 receives compressed pulse signals in a time series. The sensitivity enhancement processing unit 122 aggregates the input compressed pulse signals into a plurality of units. The compressed pulse signals of the plurality of units are, for example, M compressed pulse signals X(T), X(T-1), ..., X(T-M+1) including the compressed pulse signal X(T) at time T shown in FIG. 5. After the compressed pulse signals are aggregated into a plurality of units in this way, the sensitivity enhancement processing unit 122 performs sensitivity enhancement processing on the compressed pulse signals for each unit. (Equation 1) shows the first sensitivity enhancement signal Y1(T) which is the output of the sensitivity enhancement processing unit 122 for M compressed pulse signals from time T. D1(X(T), X(T-1), ..., X(T-M+1)) in (Equation 1) indicates that noise reduction processing is performed on the input compressed pulse signal as a high-sensitivity processing using the trained model 24c.

As described above, the sensitivity enhancement processing unit 122 may perform noise reduction processing as sensitivity enhancement processing on each compressed pulse signal, without consolidating the input compressed pulse signals into multiple units.

After the high sensitivity processing in the high sensitivity processing unit 122, the high sensitivity processing unit 123 performs high sensitivity processing, for example, on the past N first high sensitivity signals Y1(T), Y1(T-T _s1 ), ..., Y1(T-T _sN-1 ). (Equation 2) shows the second high sensitivity signal Y2(T) which is the output of the high sensitivity processing unit 123 on the past N first high sensitivity signals. D2(Y1(T), Y1(T-T _s1 ), ..., Y1(T-T _sN-1 )) in (Equation 2) shows that a noise reduction process is performed as a high sensitivity process using the trained model 24d on the input first high sensitivity signal.

As described above, according to the first embodiment, a noise reduction process is performed as a high sensitivity process on the compressed pulse signal, and a noise reduction process is further performed as a high sensitivity process on the first high sensitivity signal that has been made high sensitivity by the noise reduction process. In this way, in the first embodiment, a two-stage high sensitivity process is performed on the compressed pulse signal, so that a signal with higher sensitivity than that of the existing technology can be generated. By performing a target detection process using the high sensitivity signal, a target can be detected with a smaller number of pulse hits than that of the existing technology. Since a target can be detected with a small number of pulse hits, it is also suitable for detecting targets moving at high speed. Furthermore, the target detection process is performed by increasing the sensitivity of a low S/N signal. Therefore, there is no need to increase the S/N by increasing the aperture area or transmission power of the antenna. Therefore, the ease of installation is not impaired. In other words, in the first embodiment, when the RCS detects the same target, the same detection accuracy as that of the existing technology can be maintained even if the hardware scale is smaller than that of the existing technology.

Here, in the first embodiment, padding may be performed on the boundary portions of each compressed pulse signal. The padding is a process of expanding the boundary portions of the compressed pulse signal with a predetermined value, for example, data of 0. However, when padding is performed, there is a possibility that the performance of the high sensitivity processing for the compressed pulse signal may differ between the center portion and the boundary portion of the compressed pulse signal. Therefore, instead of expanding the boundary portion with data of 0, the boundary portion may be expanded with noise data. By expanding the boundary portion with noise data, the performance of the high sensitivity processing for the compressed pulse signal is made uniform between the center portion and the boundary portion of the compressed pulse signal. Here, the noise data may be generated, for example, by a pseudorandom number that matches the noise level of the received signal. Furthermore, the noise data may be data of a fixed pattern that is stored in advance in the ROM 22 or the like. Furthermore, the noise data may be noise data obtained by performing noise measurement.

Second Embodiment
A second embodiment will be described. Fig. 6 is a functional block diagram showing an example of a signal processing unit according to the second embodiment. The signal processing unit 12 in the second embodiment includes a pulse compression processing unit 121, a sensitivity enhancement processing unit 122, a sensitivity enhancement processing unit 123, a detection processing unit 124, and an integration processing unit 125. Here, in Fig. 6, the same elements as those in Fig. 2 are given the same reference symbols as those in Fig. 2. Descriptions of elements given the same reference symbols as those in Fig. 2 will be simplified or omitted.

In the second embodiment, the pulse compression processing unit 121 outputs a compressed pulse signal to the sensitivity enhancement processing unit 122 and the integration processing unit 125.

The high sensitivity processing unit 122 is the same as in the first embodiment. That is, the high sensitivity processing unit 122 performs a first high sensitivity processing on the compressed pulse signal from the pulse compression processing unit 121 to generate a first high sensitivity signal. The high sensitivity processing unit 122 outputs the first high sensitivity signal to the high sensitivity processing unit 123.

The integration processing unit 125 generates an integration pulse signal by pulse integrating the compressed pulse signal from the pulse compression processing unit 121. The integration processing unit 125 outputs the integration pulse signal to the sensitivity enhancement processing unit 123. For example, coherent integration may be used for the integration processing in the integration processing unit 125.

The high sensitivity processing unit 123 performs a second high sensitivity process on the first high sensitivity signal from the high sensitivity processing unit 122 and the integral pulse signal from the integral processing unit 125 to generate a second high sensitivity signal. The high sensitivity processing unit 123 outputs the second high sensitivity signal to the detection processing unit 124. The high sensitivity processing unit 123 performs a process to reduce noise in the first high sensitivity signal using a trained model, similar to the high sensitivity processing unit 122. Note that the high sensitivity processing unit 123 in the second embodiment may also be realized by the computer shown in FIG. 3.

In the second embodiment, the trained model 24d used in the sensitivity enhancement processing unit 123 is generated by repeating supervised machine learning using the first sensitivity enhancement signal and the integral pulse signal as training data. The training data in this case is, for example, data of the first sensitivity enhancement signal and the integral pulse signal that includes only the target.

The detection processing unit 124 is the same as in the first embodiment. That is, the detection processing unit 124 detects a target from the second high-sensitivity signal from the high-sensitivity processing unit 123.

As described above, according to the second embodiment, the noise reduction process and the integration process are performed on the compressed pulse signal as high sensitivity processing, and the first high sensitivity signal and the integration pulse signal are used as inputs to perform a further noise reduction process as high sensitivity processing. Like the first high sensitivity signal, the integration pulse signal can be said to be a high sensitivity compressed pulse signal. These two types of high sensitivity signals are used as learning data to perform learning in the high sensitivity processing unit 123. In other words, the learning data can be increased more than in the first embodiment. Furthermore, by performing the high sensitivity process in the trained model 24d according to these two types of high sensitivity signals, it is expected that even higher sensitivity will be achieved than in the first embodiment.

Here, in the second embodiment, padding may be applied to the boundary portions of each compressed pulse signal, or the boundary portions may be expanded with noise data.

[Third embodiment]
A third embodiment will be described. Fig. 7 is a functional block diagram showing an example of a signal processing unit according to the third embodiment. The signal processing unit 12 in the third embodiment includes a pulse compression processing unit 121, a sensitivity enhancement processing unit 122, a detection processing unit 124, and a probability output unit 126. Here, in Fig. 7, the same elements as in Fig. 2 are given the same reference symbols as in Fig. 2. Descriptions of the elements given the same reference symbols as in Fig. 2 will be simplified or omitted.

In the third embodiment, the sensitivity enhancement processing unit 123 is replaced with a probability output unit 126. Also, in the third embodiment, the compressed pulse signal generated by the pulse compression processing unit 121 is output to the detection processing unit 124 in addition to the sensitivity enhancement processing unit 122. The sensitivity enhancement processing unit 122 and the detection processing unit 124 may receive a compressed pulse signal of one pulse each, or may receive multiple compressed pulse signals collectively, for example, in CPI units.

The probability output unit 126 performs probability output processing on the first high sensitivity signal from the high sensitivity processing unit 122 to generate a probability signal. The probability signal is a signal that indicates the distribution of target-likeliness in the first high sensitivity signal as a distribution of the probability that the corresponding signal component is the target signal component. The probability output unit 126 performs processing to generate the probability signal using a trained model. Note that the probability output unit 126 may also be realized by the computer shown in FIG. 3.

The learning unit 21b in the third embodiment repeats machine learning using the first high-sensitivity signal, which is the output of the high-sensitivity processing unit 122 generated by the learning unit 21a, as learning data to generate a trained model 24d. The teacher data in this case is, for example, target map data indicating the target signal range in the input first high-sensitivity signal and noise map data indicating the noise signal range in the input first high-sensitivity signal.

The detection processing unit 124 detects a target by extracting the signal components of the target from the compressed pulse signal from the pulse compression processing unit 121 based on the probability signal from the probability output unit 126. The detection processing unit 124 generates a binary signal by binarizing the probability signal, for example, using a threshold value. The binary signal is, for example, a signal that is 1 when the probability of being the target is equal to or greater than the threshold value, and is 0 when it is less than the threshold value. The detection processing unit 124 then matches the binary signal with the compressed pulse signal by calculating the product of the binary signal and the compressed pulse signal. The detection processing unit 124 detects a signal range in which the product of the binary signal and the compressed pulse signal is not 0 as a target.

Here, the threshold value used when binarizing the probability signal may be set appropriately. For example, if the threshold value is small, the signal range detected as a target is likely to be wide. Therefore, it becomes easier to extract a target from the compressed pulse signal. On the other hand, if it is necessary to further improve the accuracy of target detection, the threshold value may be set to a large value.

Next, the operation of the signal processing unit 12 in the third embodiment will be described. FIG. 8 is a flowchart showing an example of a target detection process by the signal processing unit 12 in the third embodiment. In step S11, the pulse compression processing unit 121 of the signal processing unit 12 performs pulse compression on the received signal from the transmission/reception unit 11 to generate a compressed pulse signal.

In step S12, the sensitivity enhancement processing unit 122 generates a first sensitivity enhancement signal from the compressed pulse signal. Specifically, the sensitivity enhancement processing unit 123 inputs the compressed pulse signal to the trained model 24c to obtain the first sensitivity enhancement signal.

In step S13, the probability output unit 126 generates a probability signal from the first sensitivity-enhanced signal. Specifically, the sensitivity-enhanced processing unit 123 inputs the first sensitivity-enhanced signal to the trained model 24d to obtain a probability signal.

In step S14, the detection processing unit 124 detects a target from the compressed pulse signal and the probability signal. The detection processing unit 124 calculates, for example, the product of the compressed pulse signal and the binary signal of the probability signal. Then, the detection processing unit 124 detects a signal range in which the product is not 0 as a target.

As described above, according to the third embodiment, a noise reduction process is performed as a high sensitivity process on the compressed pulse signal, and a probability signal output process is performed to represent the distribution of target-likeness in the first high sensitivity signal that has been high sensitivity due to the noise reduction process. Then, a target is detected using the compressed pulse signal and the probability signal. In this way, in the third embodiment, a probability output process is performed on the compressed pulse signal instead of the second stage high sensitivity process. Even in this case, as in the first embodiment, a target can be detected with a smaller number of pulse hits than in the existing technology. Since a target can be detected with a small number of pulse hits, it is also suitable for detecting targets moving at high speed. Furthermore, a target detection process is performed by increasing the sensitivity of a low S/N signal. Therefore, there is no need to increase the S/N by increasing the aperture area or transmission power of the antenna. Therefore, ease of installation is not impaired. In other words, in the first embodiment, when the RCS detects the same target, the same detection accuracy as that of the existing technology can be maintained even if the hardware scale is smaller than that of the existing technology.

Here, in the third embodiment, padding may be applied to the boundary portions of each compressed pulse signal, or the boundary portions may be expanded with noise data.

[Fourth embodiment]
A fourth embodiment will be described. Fig. 9 is a functional block diagram showing an example of a signal processing unit according to the fourth embodiment. The signal processing unit 12 in the fourth embodiment includes a pulse compression processing unit 121, a sensitivity enhancement processing unit 122, a detection processing unit 124, and a probability output unit 126a. Here, in Fig. 9, the same elements as those in Fig. 2 are given the same reference symbols as those in Fig. 2. Descriptions of the elements given the same reference symbols as those in Fig. 2 will be simplified or omitted.

In the fourth embodiment, the high sensitivity processing unit 122 and the probability output unit 126a are provided between the pulse compression processing unit 121 and the detection processing unit 124.

The high sensitivity processing unit 122 in the fourth embodiment generates a first high sensitivity signal from the compressed pulse signal. The high sensitivity processing unit 122 outputs the first high sensitivity signal to the detection processing unit 124.

The probability output unit 126a performs probability output processing on the compressed pulse signal from the pulse compression processing unit 121 to generate a probability signal. The probability output unit 126a performs processing to generate a probability signal using a trained model. The probability output unit 126a may be realized by the computer shown in FIG. 3.

The learning unit 21b in the fourth embodiment repeats machine learning using the compressed pulse signal as learning data to generate a learned model 24d. In this case, the teacher data is, for example, target map data indicating the target signal range in the input compressed pulse signal and noise map data indicating the noise signal range in the input compressed pulse signal.

The detection processing unit 124 detects the target by extracting the signal component of the target from the first high sensitivity signal from the high sensitivity processing unit 122 based on the probability signal from the probability output unit 126a. The detection processing unit 124 generates a binary signal by binarizing the probability signal, for example, using a threshold value. The detection processing unit 124 then matches the binary signal with the first high sensitivity signal by calculating the product of the binary signal and the first high sensitivity signal. The detection processing unit 124 detects a signal range in which the product of the binary signal and the first high sensitivity signal is not zero as the target.

Here, the threshold value used when binarizing the probability signal may be set appropriately. For example, if the threshold value is small, the signal range detected as a target is likely to be wide. Therefore, it becomes easier to extract a target from the compressed pulse signal. On the other hand, if it is necessary to further improve the detection accuracy, the threshold value may be set to a large value.

As described above, according to the fourth embodiment, even in this case where a target is detected by matching the first high-sensitivity signal with a probability signal, the target can be detected with a smaller number of pulse hits than with existing technology, as in the first embodiment. Since the target can be detected with a small number of pulse hits, it is also suitable for detecting targets moving at high speed. Furthermore, the target detection process is performed by increasing the sensitivity of a low S/N signal. Therefore, there is no need to increase the S/N by increasing the aperture area or transmission power of the antenna. Therefore, ease of installation is not compromised. In other words, in the first embodiment, when the RCS detects the same target, the same detection accuracy as with existing technology can be maintained even if the hardware scale is smaller than that of existing technology.

Here, in the fourth embodiment, the probability output unit 126a generates a probability signal from a compressed pulse signal. In contrast, the probability output unit 126a may be configured to generate a probability signal from a first high-sensitivity signal, as in the third embodiment. Also, the probability output unit 126a may be configured to generate a probability signal from an integrated pulse signal generated by the integration processing unit 125 described in the second embodiment.

The detection processing unit 124 may detect the target by matching the second high-sensitivity signal with the probability signal instead of the first high-sensitivity signal. In this case, the high-sensitivity processing unit 123 is provided between the high-sensitivity processing unit 122 and the detection processing unit 124.

Furthermore, in the fourth embodiment, padding may be applied to the boundary portions of each compressed pulse signal, or the boundary portions may be expanded with noise data.

[Fifth embodiment]
A fifth embodiment will be described. Fig. 10 is a functional block diagram showing an example of a signal processing unit according to the fifth embodiment. The signal processing unit 12 in the fifth embodiment includes a pulse compression processing unit 121, a detection processing unit 124, and a probability output unit 126a. Here, in Fig. 10, the same elements as those in Fig. 2 are given the same reference symbols as those in Fig. 2. Descriptions of the elements given the same reference symbols as those in Fig. 2 will be simplified or omitted.

In the fifth embodiment, the probability output unit 126a is provided between the pulse compression processing unit 121 and the detection processing unit 124.

The probability output unit 126a performs probability output processing on the compressed pulse signal from the pulse compression processing unit 121 to generate a probability signal, similar to the probability output unit 126a in the fourth embodiment.

The detection processing unit 124 detects a target by extracting a signal component of the target from the compressed pulse signal from the pulse compression processing unit 121 based on the probability signal from the probability output unit 126a. The detection processing unit 124 may detect a target by performing threshold processing on the probability signal from the probability output unit 126a.

Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the gist of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are included in the scope of the invention and its equivalents described in the claims.

1...radar device, 10...antenna unit, 11...transmitter/receiver unit, 12...signal processing unit, 13...display unit, 21...processor, 21a...learning unit, 21b...learning unit, 22...ROM, 23...RAM, 24...memory unit, 24a...program, 24b...data, 24c...learned model, 24d...learned model, 25...communication unit, 121...pulse compression processing unit, 122...high sensitivity processing unit, 123...high sensitivity processing unit, 124...detection processing unit, 125...integration processing unit, 126...probability output unit, 126a...probability output unit.

Claims (12)

An antenna portion;
a transceiver unit that transmits radar pulses from the antenna unit and receives radio waves arriving at the antenna unit to generate a received signal;
A signal processing unit that processes the received signal;
Equipped with
The signal processing unit includes:
a pulse compression processing unit that performs pulse compression on the received signal to generate a compressed pulse signal;
an integration processing unit that performs pulse integration on the compressed pulse signal to generate an integrated pulse signal;
a first sensitivity enhancement processor that reduces noise in the compressed pulse signal to generate a first sensitivity enhancement signal;
a second sensitization processor that reduces noise in the first sensitized signal to generate a second sensitized signal;
a detection processing unit that detects a target from the second sensitivity-enhanced signal;
Equipped with
the second sensitivity enhancement processing unit generates the second sensitivity enhancement signal from the integral pulse signal and the first sensitivity enhancement signal;
Radar equipment. The first sensitivity enhancement processing unit includes a first learning unit that generates a first trained model by repeating machine learning using the compressed pulse signal,
The second sensitivity enhancement processing unit includes a second learning unit that generates a second trained model by repeating machine learning using the first sensitivity enhancement signal.
The radar device according to claim 1 . 3. The radar device according to claim 1 , wherein a boundary portion of the compressed pulse signal is expanded by a noise signal. An antenna portion;
a transceiver unit that transmits radar pulses from the antenna unit and receives radio waves arriving at the antenna unit to generate a received signal;
A signal processing unit that processes the received signal,
The signal processing unit includes:
a pulse compression processing unit that performs pulse compression on the received signal to generate a compressed pulse signal;
a sensitization processing unit that reduces noise in the compressed pulse signal to generate a sensitized signal;
a probability output unit that calculates a probability of the sensitized signal being a target and generates a probability signal;
a detection processing unit that detects a target from the compressed pulse signal and the probability signal. The sensitivity enhancement processing unit includes a first learning unit that generates a first trained model by repeating machine learning using the compressed pulse signal,
The radar device according to claim 4 , wherein the probability output unit includes a second learning unit configured to generate a second trained model by repeating machine learning using the sensitivity-enhanced signal. An antenna portion;
a transceiver unit that transmits radar pulses from the antenna unit and receives radio waves arriving at the antenna unit to generate a received signal;
A signal processing unit that processes the received signal,
The signal processing unit includes:
a pulse compression processing unit that performs pulse compression on the received signal to generate a compressed pulse signal;
a sensitization processing unit that reduces noise in the compressed pulse signal to generate a sensitized signal;
a probability output unit that calculates a probability of the compressed pulse signal being a target and generates a probability signal;
a detection processing unit that detects a target from the high sensitivity signal and the probability signal. The sensitivity enhancement processing unit includes a first learning unit that generates a first trained model by repeating machine learning using the compressed pulse signal,
The radar device according to claim 6 , wherein the probability output unit includes a second learning unit configured to generate a second trained model by repeating machine learning using the compressed pulse signal. The radar device according to claim 2 , 5 , or 7 , wherein the machine learning is deep learning that generates the first trained model and the second trained model by repeatedly providing training data to a convolutional neural network. In a signal processing device for processing a received signal of a radar device,
a pulse compression processing unit that performs pulse compression on the received signal to generate a compressed pulse signal;
a first sensitivity enhancement processor that reduces noise in the compressed pulse signal to generate a first sensitivity enhancement signal;
an integration processing unit that performs pulse integration on the compressed pulse signal to generate an integrated pulse signal;
a second sensitization processor that reduces noise in the first sensitized signal to generate a second sensitized signal;
a detection processing unit that detects a target from the second sensitivity-enhanced signal;
Equipped with
the second sensitivity enhancement processing unit generates the second sensitivity enhancement signal from the integral pulse signal and the first sensitivity enhancement signal;
Signal processing device. In a signal processing device for processing a received signal of a radar device,
a pulse compression processing unit that performs pulse compression on the received signal to generate a compressed pulse signal;
a sensitization processing unit that reduces noise in the compressed pulse signal to generate a sensitized signal;
a probability output unit that calculates a probability of the sensitized signal being a target and generates a probability signal;
a detection processing unit that detects a target from the compressed pulse signal and the probability signal. A signal processing method for processing a received signal of a radar device by a computer, comprising:
the computer pulse compressing the received signal to generate a compressed pulse signal;
the computer pulse-integrating the compressed pulse signal to generate an integrated pulse signal;
The computer reduces noise in the compressed pulse signal to generate a first sensitized signal;
the computer generates a second high-sensitivity signal from the integration pulse signal and the first high-sensitivity signal by reducing noise of the first high-sensitivity signal;
and detecting a target from the second sensitized signal by the computer. A signal processing method for processing a received signal of a radar device by a computer, comprising:
the computer pulse compressing the received signal to generate a compressed pulse signal;
The computer reduces noise in the compressed pulse signal to generate a sensitized signal;
The computer calculates a probability of the compressed pulse signal being a target to generate a probability signal;
and detecting a target from the sensitized signal and the probability signal by the computer.