Living body detecting device and living body detecting method

Abstract

A living body detecting device (1) includes a controller and a radar module. The controller executes sampling processing at measurement intervals (for example, 100 milliseconds). Each sampling processing includes storing in a memory, as a measurement data set, data indicating a transition in the reception strength of a reflected wave, from when the radar module transmits an impulse signal as an exploratory wave until a predetermined duration has elapsed. The controller generates, from a plurality of measurement data sets, an extracted data set obtained by extracting detected values of the reception strength at specific observation timings (in other words, extraction points). The controller analyzes the extracted data set to determine whether or not an occupant is present in a vehicle.

Description

Living body detection device and living body detection method CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application is based on Patent Application No. 2024-018013 filed in Japan on February 8, 2024, the contents of which are incorporated by reference in their entirety.

[0002] The present disclosure relates to a technique for detecting the presence of a human being in a specific space.

[0003] Patent Document 1 describes a technology for detecting passengers in a vehicle using a frequency modulated continuous wave (FM-CW) radar. Specifically, the distance to a reflector is detected by frequency analysis of the differential signal between the beat signal and the received signal.

[0004] Japanese Patent Application Laid-Open No. 2017-181225

[0005] In Patent Document 1, the detected object is determined to be a passenger. However, in an environment with many reflecting objects, such as the inside of a vehicle (a so-called multipath environment), the search wave transmitted by the radar is reflected by various objects and returns to the radar. In other words, the radar device may detect objects other than humans. The configuration disclosed in Patent Document 1 may erroneously detect structures such as seats as humans.

[0006] One of the objects of the present disclosure is to provide a technology that can determine whether or not a living body is present in a target area even in a multipath environment.

[0007] The living body detection device disclosed herein is a living body detection device comprising a transmitting unit configured to transmit radio waves in a predetermined frequency band as search waves, a receiving unit configured to receive radio waves, and a judgment unit that judges whether a living body is present in a predetermined detection target area based on the signal received by the receiving unit, wherein the judgment unit is configured to perform a sampling process at predetermined measurement intervals, which includes intermittently transmitting search waves using the transmitting unit and acquiring the reception strength of the radio waves at an observation timing, which is a timing when a predetermined time has elapsed since the search wave was transmitted, store the observed value of the reception strength at the observation timing in memory each time the sampling process is performed, generate an intensity dataset, which is a dataset in which multiple observation values ​​at the observation timing are arranged in the order in which they were acquired, and determine whether a living body is present in the detection target area by analyzing the intensity dataset.

[0008] In addition, the living organism detection method included in the present disclosure includes performing a sampling process at a predetermined measurement interval, which includes intermittently transmitting a probe wave using a transmitting unit and acquiring the reception strength of the probe wave at an observation timing, which is a timing when a predetermined time has elapsed since the probe wave was transmitted; storing the observed value of the reception strength at the observation timing in memory each time the sampling process is performed; generating an intensity dataset, which is a dataset in which multiple observation values ​​at the observation timing are arranged in the order in which they were acquired; and determining whether a living organism is present in the detection area by analyzing the intensity dataset.

[0009] The received strength of a probe wave reflected by a living organism and returning may fluctuate rhythmically in response to the organism's breathing. This characteristic is expected to be true not only for waves directly reflected from the organism, but also for indirect waves, which are signals reflected by surrounding structures after or before being reflected by the organism. Therefore, if a living organism is present in the detection target area, the received strength observed at a specific timing, which is a predetermined time after the probe wave is transmitted, is expected to fluctuate rhythmically in response to breathing over the long term. This disclosure was created based on this idea. The intensity dataset used by the above-described living organism detection device / living organism detection method to determine the presence or absence of a living organism (so-called presence / absence determination) is data in which the received strength at the observation timing is arranged at the intervals between sampling processes. The intensity dataset represents the transition of the received strength at the observation timing over a relatively long period of time. In other words, if a living organism is present in the detection target area, the intensity dataset may exhibit a waveform that oscillates with a period corresponding to breathing. Since the intensity dataset used in the above-described device / method is data indicating the presence or absence of a living organism, analyzing this data makes it possible to determine whether a living organism is present in the detection target area. Furthermore, the above determination method is not limited to direct waves, and it is also possible to determine the presence or absence of a living body using indirect waves, which means that it is possible to determine the presence or absence of a living body in a multipath environment.

[0010] Note that the symbols in parentheses in the claims indicate a correspondence with the specific means described in the embodiments described below as one aspect, and do not limit the technical scope of the present disclosure.

[0011] 1 is a diagram for explaining an example of the mounting position of a living body detection device in a vehicle. FIG. 2 is a block diagram showing the configuration of a living body detection device. FIG. 3 is a diagram for explaining the types of signals that can be received by a radar module. FIG. 4 is a diagram showing changes in reception strength. FIG. 5 is a diagram for explaining extracted data sets. FIG. 6 is a graph showing that reception strength at a certain sampling point varies in the long term in conjunction with respiration. FIG. 7 is a diagram showing a usable area in one measurement data set. FIG. 8 is a flowchart of living body detection processing. FIG. 9 is a flowchart of measurement processing. FIG. 10 is a flowchart of presence / absence determination processing. FIG. 11 is a diagram for explaining a method of determining amplitude values. FIG. 12 is a diagram for explaining a plurality of extracted data sets generated from a plurality of measurement data sets. FIG. 13 is a flowchart of presence / absence determination processing using a plurality of extracted data sets. FIG. 14 is a diagram showing the relationship between kurtosis and peak frequency. FIG. 15 is a diagram showing another configuration example of a living body detection device.

[0012] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. The present disclosure is not limited to the following embodiments. The configurations disclosed below may be modified in various ways without departing from the spirit of the present disclosure. Various modified examples may be appropriately combined as long as no technical contradictions arise. The present disclosure also includes configurations that are not explicitly stated and are formed by combining multiple modified examples. In the following description, components having the same function may be given the same reference numerals, and specific descriptions thereof may be omitted. Furthermore, components having the same function may be given the same or similar names, and specific descriptions thereof may be omitted. When only a portion of a configuration is mentioned, descriptions given elsewhere may apply to other parts.

[0013] <Introduction> Child abandonment detection will be added to EuroNCAP as an in-car occupant sensing app from 2023, and will be eligible for additional points. In particular, it has been decided that from 2025 onwards, only apps that directly detect abandoned children will be evaluated. EuroNCAP is an abbreviation for European New Car Assessment Program.

[0014] For this reason, there is a growing demand for technology that can detect biological information such as breathing and body movement throughout the vehicle and determine the presence or absence of children. Since one of the evaluation conditions is to detect whether a child is covered with a blanket or in a child seat, radio waves are considered a promising method. This is because it is difficult to detect children in these situations using cameras, etc.

[0015] The living body detection device 1 according to the present embodiment described below detects whether or not a living body is present inside a vehicle using impulse waves, as described below. The living body detection device 1 may be applied to, for example, detecting whether a child has been left behind. Children may include infants, toddlers, and elementary school students. Living bodies that can be detected by the living body detection device 1 are not limited to children and may include adults. Furthermore, the living bodies that can be detected may be animals such as dogs and cats. The living bodies that can be detected may be animals that breathe using lungs.

[0016] <Living Body Detector> The living body detector 1 of this embodiment is mounted in the vehicle interior as shown in FIG. 1 to determine whether an occupant is present inside the vehicle. The detection target area of ​​the living body detector 1 in this embodiment is the vehicle interior. For example, the living body detector 1 of this embodiment may be mounted in a vehicle HV primarily for the purpose of detecting whether a child is present in a child car seat installed in the right rear seat. In accordance with this purpose, the living body detector 1 may be mounted in an area of ​​the vehicle interior ceiling near the overhead console so that it can transmit radio waves as a search wave toward the right rear seat. The vicinity of the overhead console may be interpreted as an area of ​​the ceiling within 0.4 m from the front edge of the ceiling. The vicinity of the overhead console may include the top edge of the windshield or the vicinity of the rearview mirror. In describing the mounting location of the living body detector 1, the vicinity of a certain component may be interpreted as an area within 0.4 m from the component.

[0017] Of course, the installation location of the living body detection device 1 may be changed as appropriate. The living body detection device 1 may be attached to the center of the ceiling, the back of the front seat, an A-pillar, a B-pillar, or the like. The living body detection device 1 is not limited to being attached to the ceiling, but may also be attached to a pillar, the back of the seat, a headrest, or an instrument panel. In the present disclosure, a seat that is the target of determining (detecting) whether or not an occupant is present is also referred to as a target seat. The living body detection device 1 may be attached to a position corresponding to the seat that is the target of detection. If the living body detection device 1 has directionality, the living body detection device 1 may be attached in an orientation in which the center of its directionality faces the target seat.

[0018] 2, the living body detection device 1 includes a controller 2, a radar module 3, and a communication interface 4. The living body detection device 1 is also connected to a body ECU 5 via wired or wireless communication so as to be able to communicate with the body ECU 5. ECU is an abbreviation for Electronic Control Unit, and refers to an electronic control device.

[0019] The controller 2 is hardware that performs processing to detect living organisms inside the vehicle using the radar module 3. The controller 2 corresponds to a determination unit. The controller 2 may be a computer. The controller 2 may include a processor 21, a memory 22, a storage 23, and an input / output circuit 24. The processor 21 is an arithmetic core that performs arithmetic processing based on data received from the radar module 3 and the body ECU 5. The processor 21 may be a central processing unit (CPU) or a graphics processing unit (GPU). The controller 2 may include an FPGA (field-programmable gate array) instead of or in addition to the processor 21. The memory 22 is a rewritable volatile storage medium. The memory 22 may be a random access memory (RAM). The memory 22 may be configured to temporarily store data received from other devices, calculation results by the processor 21, programs, etc.

[0020] The storage 23 is a rewritable non-volatile memory. The storage 23 may be realized by at least one type of non-transitory tangible storage medium, such as a semiconductor memory, a magnetic medium, or an optical medium. The storage 23 may include multiple types of storage media, such as a read-only memory (ROM) and a flash memory. The storage 23 stores a living body detection program executed by the processor 21. The living body detection program may be a program for realizing at least some of the functions of the controller 2. Execution of the living body detection program by the processor 21 corresponds to execution of a living body detection method. The input / output circuit 24 is a circuit module that allows the processor 21 to communicate with other devices. In FIG. 2, the input / output circuit 24 is indicated by IO (Input / Output). The functions of the controller 2 will be described separately below.

[0021] The radar module 3 is a device that transmits radio waves in a predetermined frequency band as search waves and receives reflected waves that are radio waves reflected by an object. The radar module 3 of this embodiment is configured to transmit and receive impulse waves used in UWB (Ultra Wide Band) communication as search waves. Such a radar module 3 may also be referred to as an IR (Impulse Radio) UWB radar. The impulse waves used in UWB communication may be signals with an extremely short pulse width (e.g., 2 ns) and a bandwidth of 500 MHz (strictly speaking, 499.2 MHz) or greater. Using impulse-shaped radio waves as search waves can improve spatial resolution. Hereinafter, the frequency band used by the radar module 3 as search waves will also be referred to as the operating frequency band.

[0022] The communication interface 4 is a circuit module that enables the controller 2 to communicate with other in-vehicle devices such as the body ECU 5. The communication interface 4 may include a PHY chip or the like that complies with the communication standard of the in-vehicle network. The communication interface 4 transmits data received from the body ECU 5 to the controller 2. The communication interface 4 also transmits data received from the controller 2 to the body ECU 5. The communication interface 4 may include a wireless communication module for wireless communication with a mobile device (e.g., a smartphone) owned by the user of the vehicle Hv. The wireless communication module may be a circuit for communication via Bluetooth (registered trademark) Low Energy or Wi-Fi (registered trademark).

[0023] The body ECU 5 is an ECU that controls body-related devices such as a door lock motor, a meter display, headlights, and a horn. The body ECU 5 is configured to receive signals indicating the state of the vehicle hybrid vehicle and signals indicating user operations on the vehicle hybrid vehicle from on-board sensors. The state of the vehicle hybrid vehicle includes the open / closed state of the doors, the open / closed state of the windows, the shift position, the vehicle power supply state, and the seating state of the driver's seat. The user operations on the vehicle hybrid vehicle may include opening / closing the doors, pressing the start switch, changing the shift position, and pressing the lock sensor. The start switch is a switch for turning the vehicle power supply on and off. The lock sensor may be a touch sensor or a push switch provided on the outer door handle for locking / unlocking the vehicle hybrid vehicle.

[0024] The body ECU 5 detects a predetermined confirmation event based on a signal from an on-board sensor. The body ECU 5 may transmit a confirmation request signal to the living body detection device 1 based on the detection of the confirmation event. The confirmation request signal is a message requesting confirmation of whether or not there is an occupant in the vehicle. In a specific use case, the confirmation request signal may be interpreted as a signal requesting confirmation of whether or not a child has been left in the vehicle. The confirmation event may be that the vehicle Hv has been locked. The confirmation event may also be that the vehicle power supply has been turned off, that the driver's seat occupant has exited the vehicle, or that the driver's seat door has been closed. The driver's seat occupant's exit may be detected based on a signal from a seat occupancy sensor in the driver's seat.

[0025] The controller 2 of the living body detection device 1 may be configured to start a living body detection process, which will be described later, based on receiving a confirmation request signal from the body ECU 5. The confirmation request signal may function as a signal to start the living body detection device 1. The living body detection device 1 may be configured to start up and start the living body detection process based on receiving the confirmation request signal from the body ECU 5.

[0026] The connection between the living body detection device 1 and the body ECU 5 is an optional element and may be omitted. The living body detection device 1 may be directly connected to a predetermined in-vehicle sensor or an in-vehicle network without going through the body ECU 5. The living body detection device 1 may be configured to, for example, acquire information that the vehicle Hv has been locked from a predetermined in-vehicle sensor via wired or wireless communication and start the living body detection process. The living body detection device 1 may be configured to execute the sampling process described below multiple times within 10 seconds after the vehicle Hv is locked. Furthermore, the device to which the living body detection device 1 connects may be an ECU other than the body ECU 5, or may be a portable device (e.g., a smartphone) carried by the user.

[0027] <Radar Module> The radar module 3 includes a clock generator 31, a transmission circuit 32, a transmission antenna 33, a reception antenna 34, and a reception circuit 35. The clock generator 31 is a circuit element that outputs a clock signal for transmission processing and reception processing. The clock generator 31 may be a crystal oscillation circuit including a crystal resonator. The clock generator 31 may also be a VCO (Voltage Controlled Oscillator) that uses a crystal resonator as a resonator, i.e., a VCXO (Voltage Controlled Xtal Oscillator). The clock generator 31 inputs a clock signal to the transmission circuit 32 and the reception circuit 35.

[0028] The transmission circuit 32 is a circuit that generates an impulse signal and outputs it to the transmission antenna 33. The transmission circuit 32 corresponds to a transmission unit. The transmission unit may include the transmission antenna 33. The transmission circuit 32 may include a pulse generator, a bandpass filter, an amplifier, etc. The pulse generator is a circuit that generates an impulse signal from a clock signal using an SRD (Step Recovery Diode) or the like. The impulse signal output by the pulse generator may include not only frequency components used in UWB but also other frequency components. The bandpass filter included in the transmission circuit 32 may be a filter that passes only UWB band components from the signal output by the pulse generator. Additionally, the transmission circuit 32 may include a modulation circuit that generates a pulse sequence signal according to transmission data input from the controller 2. The transmission circuit 32 generates an impulse signal based on instructions from the controller 2 and outputs it to the transmission antenna 33.

[0029] The transmitting antenna 33 is an antenna for transmitting radio waves in the operating frequency band. The transmitting antenna 33 is connected to the transmitting circuit 32. The transmitting antenna 33 emits a signal input from the transmitting circuit 32 as a radio wave. The receiving antenna 34 is an antenna for receiving radio waves in the operating frequency band. The receiving antenna 34 is connected to the receiving circuit 35. The receiving antenna 34 inputs a received signal to the receiving circuit 35. Note that in other embodiments, the transmitting antenna 33 and the receiving antenna 34 may be integrated into one antenna. The radar module 3 may be configured to use one antenna for both transmission and reception.

[0030] The receiving circuit 35 is a circuit that performs predetermined signal processing on the signal received by the receiving antenna 34. Hereinafter, the signal input from the receiving antenna 34 will also be referred to as the received signal. The receiving circuit 35 corresponds to the receiving unit. The receiving unit may include the receiving antenna 34. The receiving circuit 35 includes an intensity detection circuit 351 that detects the intensity of the received signal (i.e., the reception intensity). The reception intensity may be measured in power (e.g., dBm). The reception intensity may also be expressed as a dimensionless index ranging from 0 to a predetermined value (e.g., 3000 or 5000). The receiving circuit 35 may have a function to detect the reception of an impulse signal based on the level (intensity) of the received signal.

[0031] The receiving circuit 35 may become active from the point in time when the transmitting circuit 32 transmits an impulse signal. The active state of the receiving circuit 35 refers to a state in which the strength of the received signal is being measured, and therefore a state in which the reception of an impulse signal can be detected. The receiving circuit 35 may acquire the timing of transmission of the impulse signal by the transmitting circuit 32 from the transmitting circuit 32 or the controller 2. The receiving circuit 35 outputs data indicating the reception strength at each time to the controller 2, with the time when the transmitting circuit 32 transmits the impulse signal as the origin (T=0). In the present disclosure, a collection of data indicating the detected values ​​of the reception strength at each time from the start to the end of detection of the reception strength is referred to as a measurement dataset. The measurement dataset may also be referred to as a measurement result file. Furthermore, the measurement dataset is a collection of data indicating a short-term trend in the reception strength, compared to the extraction dataset described below. Therefore, the measurement dataset may also be referred to as a short-term trend dataset. One measurement dataset is generated in response to one impulse transmission.

[0032] The receiving circuit 35 may be configured to detect the received signal strength every 0.2 nanoseconds. The time bin width for measuring the received signal strength is not limited to 0.2 nanoseconds, and may be other values ​​such as 0.1 nanoseconds or 0.3 nanoseconds. The time bin width may be referred to as a sampling interval. Each time bin may be referred to as a sampling point, observation point, or observation timing. The receiving circuit 35 is configured to measure the received signal strength until a predetermined time (e.g., 10 nanoseconds) has elapsed since the transmission of the impulse. One measurement data set may include a plurality of detection values, e.g., 50, 100, or 200. The number of detection values ​​(i.e., samples) included in one measurement data set may be designed as appropriate.

[0033] The detection duration (DT), which is the time for which detection of the reception intensity continues, may be 10 nanoseconds, 13 nanoseconds, 15 nanoseconds, or 20 nanoseconds. The detection duration may be set to twice the time required for the impulse signal to travel from the front end to the rear end of the vehicle interior. The detection duration may be set to a value obtained by dividing twice the length of the cabin in the longitudinal direction (e.g., 4 m) by the speed of light. The elapsed time since the impulse signal was transmitted may be measured based on a clock signal input from the clock generator 31. In the present disclosure, a series of processes from when the radar module 3 transmits the search wave to when measurement data is generated is referred to as a measurement process.

[0034] Additionally, the receiving circuit 35 may include a circuit (a so-called demodulation circuit) that demodulates the received signal and restores the data before modulation. The receiving circuit 35 may include an amplifier, a detector, an AD (Analog-Digital) converter, etc. The radar module 3 may also include a calculation unit that calculates the distance to the reflecting object (hereinafter referred to as the distance measurement value) from the time from when the transmitting circuit 32 transmits the impulse signal to when the receiving circuit 35 first receives the impulse signal.

[0035] <Detection Principle> Here, we will explain the concept behind the method by which the living body detection device 1 detects an occupant left inside the vehicle. First, when an occupant is inside the vehicle, the search wave transmitted by the radar module 3 is reflected by the occupant's chest BD as shown in FIG. 3 and can be received by the radar module 3. The position of the occupant's chest can fluctuate by approximately 1 mm to 10 mm as they breathe. The breathing frequency is said to be approximately 30 times per minute for young children and approximately 25 times per minute for elementary school children. In other words, the chest position can fluctuate in a cycle of approximately several seconds.

[0036] If the radar module 3 can extract only the reflected waves from the occupant (particularly the chest BD) from the received signal, the distance from the radar module 3 to the chest BD can be obtained as a distance measurement value. Such a distance measurement value may fluctuate with a period corresponding to breathing. Therefore, the living body detection device 1 can determine the presence or absence of a living body from the viewpoint of whether the distance measurement value fluctuates with a period corresponding to the breathing frequency. For example, if it is assumed that the first pulse observed after impulse transmission is a reflected wave from the chest BD, the living body detection device 1 can determine the presence or absence of a living body from the time-series data of the distance measurement value. Here, the first pulse refers to the first impulse signal detected after the receiving circuit 35 starts receiving.

[0037] However, there are multiple vehicle components inside the vehicle that reflect the search waves, such as door panels, pillars, and ceiling panels. Therefore, the radar module 3 may receive not only the reflected waves from the chest BD but also signals reflected by the vehicle components. If the radar module 3 is installed in a location far from the chest BD, such as on the ceiling, there is a high possibility that reflected waves from other vehicle components will reach the radar module 3 before the reflected waves from the chest BD. In an environment where the radar module 3 may receive reflected waves from vehicle components before the reflected waves from the chest BD, the reflected waves from the chest BD will be mixed in with the reflected waves from the vehicle components, making it difficult to estimate the distance to the chest BD. As a result, it is difficult to determine the presence or absence of a living body using the above method.

[0038] On the other hand, the received strength of the signal reflected by the chest BD is expected to fluctuate in conjunction with respiration over the long term, as shown in FIG. 4 . Long-term here refers to observation over a period of several seconds or more (more preferably, five seconds or more). For example, the strength of the direct wave is expected to fluctuate in conjunction with respiration. Here, direct wave refers to a signal that travels directly back and forth between the radar module 3 and the chest BD. Furthermore, the strength of not only direct waves but also indirect waves is expected to fluctuate in conjunction with respiration over the long term. In the present disclosure, indirect waves are signals that are reflected by the chest BD and have been reflected by vehicle components on the outbound, inbound, or both routes. The search waves transmitted by the radar module 3 reach the chest BD via various routes. The search waves that reach the chest BD may travel back to the radar module 3 via various routes.

[0039] The solid arrow R1 in FIG. 3 represents a direct wave. The dashed arrows R2a, R2b, and R2c represent indirect waves. Specifically, arrow R2a represents a signal that is transmitted from the radar module 3 and reflected directly by the chest BD, then reflected by a vehicle component and returned to the radar module 3. In other words, arrow R2a represents an indirect wave that is reflected on the return path. Arrow R2b represents a signal that is reflected by a vehicle component, reaches the chest BD, and then reflected by the chest BD and returned to the radar module 3. Arrow R2b represents an indirect wave that is reflected on the outgoing path. Arrow R2c represents an indirect wave that is reflected on both the outgoing and return paths. Arrow R3 represents a wave reflected by a nearby reflecting member 6, which is a vehicle component located near the radar module 3, or in other words, the living body detection device 1. The nearby reflecting member 6 may be a metal body such as a ceiling panel, B-pillar, or seat frame.

[0040] As described above, indirect waves can take various paths. Different propagation paths result in different signal reception timings. Therefore, the radar module 3 can receive signals on which indirect waves are superimposed at various times after transmitting an impulse signal. For example, the radar module 3 can receive signals on which indirect waves are superimposed at almost all times after a predetermined time has elapsed since transmission. Accordingly, the radar module 3 detects reception intensities including components of indirect waves or direct waves at multiple times (time bins).

[0041] Thus, the received signal strength observed after a predetermined time has elapsed since transmission is expected to often include components derived from signals reflected by the chest BD. Therefore, the received signal strength observed at a specific timing after a predetermined time has elapsed since transmission may fluctuate over the long term with a period corresponding to respiration. For example, as shown in FIG. 5 , assume that 80 measurement data sets are obtained as a result of performing measurement processing 80 times every 100 milliseconds. From these measurement data sets, the received signal strengths P1, P2, P3, ..., P80 at points where 6 nanoseconds have elapsed since the transmission of the probe wave are extracted. The data, arranged in chronological order, is expected to have a waveform in which the strength oscillates with a period corresponding to respiration, as shown in FIG. 6 .

[0042] In the present disclosure, data obtained by extracting detection values ​​of reception strength at the same sampling point from multiple measurement datasets and arranging these detection values ​​in chronological order of acquisition time is also referred to as an extracted dataset. The extracted dataset is data in which detection values ​​of reception strength are arranged at measurement process execution intervals (e.g., 100 milliseconds). Because the extracted dataset is data showing the long-term trend of reception strength at a specific sampling point, it may be referred to as a long-term observation dataset. Furthermore, because the extracted dataset is data showing the time-dependent change in reception strength with a fixed observation point on the time axis, it may be referred to as a fixed-point observation dataset. Alternatively, the extracted dataset corresponds to an intensity dataset. The extracted dataset may also be referred to as a primary processed dataset. Each graph shown in FIG. 5 shows the reception strength at each time after transmission. The horizontal axis represents the elapsed time from transmission, and the vertical axis represents reception strength. In FIG. 5, "#1" indicates the trend of reception strength observed in the first measurement process, and "#80" indicates the trend of reception strength observed in the 80th measurement process.

[0043] Furthermore, the developers of the present disclosure specifically analyzed the extracted data set and found that the extracted data set contains many high-frequency components. Data obtained by passing the extracted data set through a band-pass filter that passes only the frequency band corresponding to respiration may exhibit characteristics corresponding to respiration more prominently than the raw extracted data set. The aforementioned FIG. 6 shows the waveform of a filtered data set, which is an extracted data set that has been passed through a band-pass filter. P1a, P2a, P3a, and P80a in FIG. 6 are samples corresponding to P1, P2, P3, and P80 before passing through the filter.

[0044] As described above, the extracted data set can be used as information for determining whether or not an occupant is present in the vehicle. The living body detection device 1 of the present disclosure was created based on this idea. As will be described later, the living body detection device 1 of this embodiment analyzes data obtained by performing BPF processing on the extracted data set, and determines that an occupant is present in the vehicle when characteristics corresponding to breathing are observed. The living body detection device 1 of the present disclosure may be understood to have been created based on the technical idea of ​​actively utilizing multipath.

[0045] Note that the signal received immediately after transmission is likely to be a wave reflected from a nearby reflecting member 6 (e.g., the ceiling surface) of the radar module 3, and does not include components reflected from the chest. Furthermore, the time region immediately after transmission is heavily influenced by signals from nearby reflecting members, resulting in high reception intensity. In time regions with high reception intensity, the influence of timing discrepancies in intensity measurement between measurement processes is significant. Timing discrepancies in intensity measurement refer to discrepancies in the start timing of measurement of reception intensity relative to the transmission timing of the impulse signal. The transmission timing and measurement start timing do not necessarily align every time. There may be a discrepancy between the transmission timing and measurement start timing. In the region immediately after transmission, minute fluctuations such as breathing may be obscured by intensity changes due to timing discrepancies, etc., and breathing may not be detected.

[0046] In view of the above, it is preferable that the extraction point (Tc) be selected from a time region where the elapsed time from the transmission of the probe wave is equal to or greater than a predetermined backoff time (Tbo), as shown in FIG. 7 . The extraction point here refers to a sampling point, among the multiple sampling points included in the measurement data set, that is used to create the extraction data set. In this disclosure, the time region where the elapsed time from the transmission of the probe wave is equal to or greater than the predetermined backoff time (Tbo), is also referred to as a usable region or a candidate region. "Tz" in FIG. 7 represents the time at which measurement of the reception strength in one measurement process ends. In other words, Tz represents the timing at which the detection duration time (DT) has elapsed since the impulse transmission.

[0047] The backoff time may be a fixed value, such as 3 nanoseconds, 4 nanoseconds, or 5 nanoseconds. The backoff time may be set to a value according to the estimated distance from the radar module 3 to the chest BD. For example, if the estimated distance from the radar module 3 to the chest BD is 0.45 m or more (a round-trip distance of 0.9 m), the backoff time may be 3 nanoseconds. The backoff time may be set to be shorter than the time required to receive a direct wave. The backoff time may be set to exclude noise regions. The noise region refers to a region where the received signal is likely to be noise. The backoff time may also be referred to as a discard time or invalid time. While FIG. 5 shows an example in which the sampling point is set 6 nanoseconds after impulse transmission, the sampling point may of course be set at other times, such as 7 nanoseconds or 8 nanoseconds. One extracted data set may be a collection of reception intensities at a common sampling point (in other words, observation timing).

[0048] <Living Body Detection Process> Here, the living body detection process will be described with reference to Figures 8, 9, and 10. As shown in Figure 8, the living body detection process may include steps S1 to S4. Step S1 is a step in which the radar module 3 executes the measurement process a predetermined number of times. The radar module 3 may execute the measurement process based on an execution command for the measurement process input from the controller 2. Step S1 corresponds to a step in which the controller 2 causes the radar module 3 to execute the measurement process a predetermined number of times. Step S1 may be started based on the occurrence of a predetermined confirmation event, such as locking the vehicle Hv. As described above, the confirmation event may be detected by an external ECU such as the body ECU 5, or may be detected by the controller 2.

[0049] The measurement interval, which is the interval between executions of the measurement process, may be a time that is sufficiently short compared to the respiratory cycle, such as 100 milliseconds or 200 milliseconds. The number of measurements (M), which is the number of times the measurement process is executed, may be 80, 100, 150, etc. Note that FIG. 5 illustrates an example in which the number of measurements is set to 80. The value obtained by multiplying the number of measurements by the measurement interval defines the data collection time. The data collection time is the time required to collect data used to determine the presence or absence of a living body. The data collection time may be interpreted as the length of the period during which the measurement process is executed periodically. The data collection time may be set to a length that allows the periodicity of the respiratory cycle to be detected. The number of measurements and the measurement interval may be set so that the data collection time is approximately 5 to 10 seconds.

[0050] The controller 2 may be configured to periodically execute the measurement process until a predetermined termination condition is met, while the minimum execution count of the measurement process is set, and may not be configured to terminate the measurement process until a predetermined termination condition is met. The termination condition may be that a predetermined time (e.g., 10 seconds) has elapsed since the door was locked.

[0051] As shown in Figure 9, one measurement process includes steps S101 to S105. Step S101 is a step in which the radar module 3 transmits an impulse signal as a search wave. Step S102 is a step in which the radar module 3 starts measuring the reception strength. Steps S101 and S102 may be executed simultaneously so that the origin (t = 0) of the measurement data and the transmission timing coincide. The terms measurement, detection, and observation may be used interchangeably.

[0052] Step S103 is a step in which the radar module 3 stores the detected values ​​of the reception intensity. The detected values ​​may also be expressed as observed values. In step S103, the acquisition and storage of the detected values ​​is performed at a sampling interval (e.g., 0.2 nanoseconds). The detected values ​​may be stored in a temporary memory built into the radar module 3. The detected values ​​may be promptly provided from the radar module 3 to the controller 2 and stored in chronological order in the memory 22 provided in the controller 2.

[0053] Step S104 is a step of determining whether the detection duration has elapsed since the start of measurement of the reception strength. Step S104 may be interpreted as a step of determining whether the time to end measurement of the reception strength for one measurement process has arrived. If the detection duration has elapsed since the start of measurement of the reception strength (YES in S104), measurement of the reception strength is ended in step S105. Step S105 may include the radar module 3 transmitting a measurement data set indicating the reception strength for each time collected from the impulse transmission to the controller 2. Step S105 may also include stopping the receiving circuit 35. Note that in another embodiment, the controller 2 may generate one measurement data set based on the detection value of the reception strength for each time provided from the radar module 3 and store it in the memory 22. The functional layout related to the generation of the measurement data set may be changed as appropriate.

[0054] When the measurement process has been executed a predetermined number of times, the living body detection process proceeds to step S2 as shown in Fig. 8. Step S2 is a step in which the controller 2 executes a presence / absence determination process. The presence / absence determination process is a process for determining whether or not there is an occupant (living body) in the vehicle. The presence / absence determination process of step S2 may include steps S201 to S209 as shown in Fig. 10. The presence / absence determination process may be started when the measurement process has been executed a predetermined number of times.

[0055] Step S201 is a step of generating an extracted dataset using the multiple measurement datasets generated in step S1. As described above, the extracted dataset may be data obtained by extracting reception intensity values ​​at specific observation timings from the multiple measurement datasets and arranging them in chronological order. The specific observation timing may be, for example, a point in time 6 nanoseconds after the transmission of the impulse wave. The observation timings used to generate the extracted dataset may be referred to as extraction points.

[0056] Step S202 is a step of filtering the extracted data set generated in step S201 to remove high-frequency components as noise. The filter used here may be a band-pass filter (BPF) that passes signals in a frequency band corresponding to respiration. Step S202 may be interpreted as a step of extracting frequency components corresponding to respiration. For example, the BPF may be a filter that passes frequencies between 0.1 Hz and 1 Hz. Of course, the filter used in step S202 may also be a low-pass filter. Step S202 may be a step of removing frequency components of 1 Hz or 2 Hz or higher. The output data of step S202 corresponds to the filtered data set. If an occupant is present in the vehicle, the waveform of the filtered data set is expected to have a waveform that oscillates with a rhythm corresponding to respiration, as shown in FIG. 6 .

[0057] Step S203 is a step of analyzing the filtered data set to obtain an amplitude value (Am) of the received signal strength as one feature. In the figure, Am represents the amplitude value. As shown in FIG. 11 , the filtered data set can be divided into multiple wave sections by multiple valley points. A valley point is a point where the received signal strength changes from decreasing to increasing. A section between two valley points is a wave section. A wave section may be set to have one peak point. A peak point is a point where the received signal strength changes from increasing to decreasing. In FIG. 11 , VP1, VP2, VP3, and VP4 represent valley points, and the area surrounded by a dashed line represents one wave section. Note that a wave section whose width is equal to or less than a predetermined value (e.g., 0.5 seconds) may represent an intensity fluctuation that does not correspond to the breathing of an occupant. A wave section whose width is less than a predetermined value (e.g., 0.5 seconds) may be ignored as noise. The controller 2 may calculate the difference between the valley point and the peak point in one wave section as the amplitude of that wave section. The calculation of the amplitude may be performed only for wave sections whose width is equal to or greater than a predetermined value. A1 to A4 in Figure 11 represent the amplitude for each wave section.

[0058] The amplitude value (Am) identified in step S203 may be, for example, the amplitude of the wave section having the highest peak point among the multiple wave sections included in the filtered data set. Furthermore, the amplitude value identified in step S203 may be the average, median, or maximum value of the amplitudes for each wave section. After identifying the amplitude value from the filtered data set, the controller 2 performs the determination process of step S204.

[0059] Step S204 is a step for determining whether the amplitude value (Am) identified in step S203 is greater than a predetermined amplitude threshold (ThA). "ThA" in the figure represents the amplitude threshold. The amplitude threshold is a parameter for distinguishing between fluctuations in reception strength due to breathing and other factors. For example, it may be set to a value corresponding to the range of fluctuations in reception strength due to breathing. If the amplitude value is greater than the amplitude threshold (YES in S204), the controller 2 determines in step S205 that there is a living occupant in the vehicle. On the other hand, if the amplitude value is equal to or less than the amplitude threshold (NO in S204), the controller 2 executes the processing from step S206 onwards.

[0060] Step S206 is a step of performing frequency analysis on the filtered data set. Step S206 may be, for example, a step of performing a fast Fourier transform (FFT) on the filtered data set. The controller 2 acquires a frequency spectrum by performing an FFT on the filtered data set. The frequency spectrum is data indicating the intensity for each frequency. The frequency spectrum corresponds to frequency characteristic data. The frequency resolution and frequency bin width in the analysis process may be, for example, 0.05 Hz or 0.1 Hz.

[0061] Step S207 is a step of identifying the peak frequency and its kurtosis as frequency characteristics from the frequency spectrum obtained in step S206. In the figure, "FP" represents the peak frequency, and "K" represents kurtosis. The peak frequency is the frequency with the greatest intensity in the frequency spectrum. Kurtosis is calculated by dividing the received intensity at the peak frequency by the average intensity of the other frequencies. Here, the other frequencies may be frequencies that are a certain value (e.g., 0.2 Hz) or more away from the peak frequency. Furthermore, the other frequencies may be frequencies located in the frequency domain before the rise and after the fall of the crest (wave) corresponding to the peak frequency. The rise and fall points of the crest corresponding to the peak frequency may be points where the intensity is 25%, 33%, or 50% of the peak.

[0062] A large kurtosis means that the intensity of the peak frequency is higher than the intensity of other frequencies. In other words, a large kurtosis means that there are many peak frequency components. If the filtered data set contains fluctuation components due to breathing, the peak frequency is expected to be located in a frequency band corresponding to the breathing frequency, and its intensity (kurtosis) will be large. After obtaining the peak frequency and its kurtosis in step S207, the process proceeds to step S208.

[0063] Step S208 is a step of determining whether an occupant is present in the vehicle from the viewpoint of peak frequency and kurtosis. For example, step S208 may be a step of determining whether the peak frequency is within a predetermined allowable range and whether the kurtosis is greater than a predetermined kurtosis threshold. In the figure, "FL" represents the lower limit of the allowable range, and "FH" represents the upper limit of the allowable range. "ThS" in the figure represents the kurtosis threshold. The lower limit of the allowable range (FL) may be 0.25 Hz or 0.3 Hz. The upper limit of the allowable range (FH) may be 0.8 Hz.

[0064] The respiratory frequency increases as the age (months) decreases. For example, the respiratory frequency of a newborn baby less than one month old is said to be about 50 breaths per minute (approximately 0.83 Hz). The respiratory frequency of a child about one year old is said to be about 30 breaths per minute (approximately 0.5 Hz). The respiratory frequency of a child seven years old or older is said to be 18 to 20 breaths per minute (approximately 0.3 Hz). The controller 2 may be configured to acquire the age of the occupant to be detected based on a user operation. The controller 2 may also store the acquired target age information in the storage 23. The controller 2 may be configured to automatically change the lower and upper limits of the allowable range according to the set target age. The controller 2 may increase the lower and upper limits of the allowable range as the registered target age decreases.

[0065] The kurtosis threshold is a parameter for determining whether or not an occupant is present in the vehicle from the viewpoint of kurtosis. A specific value of the kurtosis threshold may be designed through experiments. If the peak frequency is within the allowable range and the kurtosis is greater than the kurtosis threshold (YES in S208), the controller 2 may determine in step S205 that an occupant is present in the vehicle. Alternatively, if the peak frequency is outside the allowable range or the kurtosis is equal to or less than the kurtosis threshold (NO in S208), the controller 2 may determine in step S209 that no occupant is present in the vehicle.

[0066] Note that the determination process of step S208 may be simplified. Step S208 may be a step of determining only whether the peak frequency is within the allowable range. Furthermore, step S208 may be a step of determining only whether the kurtosis is greater than the kurtosis threshold. The condition for determining that an occupant is present in the vehicle (hereinafter, the presence determination condition) may be changed as appropriate. The above embodiment corresponds to a configuration in which the presence determination condition is set to the peak frequency being within the allowable range and the kurtosis being greater than the kurtosis threshold. The presence determination condition may be only that the peak frequency is within the allowable range, or may be only that the kurtosis is greater than a predetermined value.

[0067] The results of the above presence / absence determination process are stored in the memory 22 or storage 23 and are referenced in step S3 of FIG. 8. The presence / absence determination result may be stored linked to the determination time. Step S3 may be executed upon completion of the presence / absence determination process of step S2. If it is determined in the presence / absence determination process of step S2 that an occupant (living body) is present, step S3 is affirmative and the controller 2 executes step S4. On the other hand, if it is not determined in the presence / absence determination process of step S2 that an occupant (living body) is present, step S3 is negative and the series of flows ends.

[0068] Step S4 is a step in which the controller 2 performs a notification process. The notification process here is a process for notifying the user of the vehicle Hv that an occupant is still inside the vehicle. The notification process may be to sound the horn in a predetermined pattern. The notification process may also be to display an image on the meter display indicating that an occupant is still inside. The notification process may be to send a message indicating that an occupant is still inside to a pre-registered contact, such as a mobile phone number or email address. The notification process may also be to send a message indicating that an occupant is still inside to a user's mobile device using wireless communication such as Bluetooth Low Energy. The notification process may be to flash a lamp attached to the vehicle Hv in a predetermined pattern. The lamp may be a headlight, a hazard lamp, or a welcome lamp. The welcome lamp may be a lamp that emits light toward the road surface near the door.

[0069] The notification process may be a combination of the above processes. For example, the notification process may sound the horn and turn on the hazard lights. Instead of sounding the horn, the controller 2 may output a warning sound of a predetermined pattern from an external speaker attached to the vehicle Hv. These processes may be performed in cooperation with the body ECU 5. By performing the notification process, the risk of an occupant such as a small child being left inside the vehicle can be reduced.

[0070] Summary of the embodiment The living body detection device 1 was created by focusing on the fact that reflected waves that hit an occupant are received by the radar module 3 while being delayed in the time direction along various paths inside the vehicle. In the living body detection device 1 of this embodiment, the radar module 3 executes measurement processing multiple times. The controller 2 then analyzes multiple measurement data sets across the board to detect the presence or absence of a living body 100. This configuration makes it possible to accurately determine the presence or absence of an occupant even in a multipath environment such as the inside of a vehicle. In other words, even in a situation where it is difficult to accurately detect the distance to the chest BD, the risk of erroneously determining the presence or absence of an occupant can be reduced.

[0071] The controller 2 also performs a multi-stage presence / absence determination. That is, the controller 2 first (in the first stage) performs the presence / absence determination using the amplitude value of the filtered data set. If the amplitude value is equal to or greater than a predetermined value, it is determined that an occupant is present, and subsequent processing such as FFT is not performed. This reduces the processing load on the controller 2. If the amplitude value is less than the predetermined value, the controller 2 performs the presence / absence determination in the second stage using frequency characteristics such as peak frequency. A configuration that performs the presence / absence determination using frequency characteristics can reduce the risk of erroneously determining that an occupant is absent when there is actually an occupant in the vehicle, even when the signal strength is weak. Furthermore, a configuration that uses at least one of peak frequency and kurtosis as a criterion for determining whether an occupant is present in the vehicle can reduce the risk of erroneously determining the presence / absence of an occupant due to noise.

[0072] Furthermore, in the time domain where indirect waves can be superimposed (i.e., the usable domain), the transition level of the reception intensity is lower than immediately after transmission. Therefore, even if there is a nanosecond-order difference in the reception start timing between multiple measurement processes, there is little risk of the detection accuracy deteriorating. According to the configuration of this embodiment, the detection accuracy of the living body 100 inside the vehicle can be improved. For example, the accuracy of detecting whether a child has been left behind can be improved.

[0073] The radar module 3 in this embodiment is a UWB radar. In recent years, there has been a growing market need for using smartphones as car keys. One such method, UWB, uses pulsed radio waves to measure the communication time between the smartphone and the car, thereby locating the smartphone and using this information to trigger the car's lock and unlock and the engine start. By using a UWB radar for a smart entry system as the radar module 3 for live body detection, the introduction cost of the live body detection device 1 can be reduced.

[0074] <Modification> In the above-described embodiment, one arbitrary observation timing is selected as an extraction point from multiple measurement data sets to generate one extraction data set. However, a measurement data set may include multiple observation timings, in other words, sampling points. As shown in FIG. 12 , the controller 2 may generate multiple extraction data sets for different observation timings and perform presence / absence determination based on the peak frequency and kurtosis information of each set. "M" in FIG. 12 represents the number of measurements that have been set. "T1" in the figure represents the first sampling point in the usable area. "T2" represents the sampling point next to T1, and "T3" represents the sampling point next to T2. "Tj" may be the last sampling point in the usable area.

[0075] The controller 2 may generate extracted data sets for all sampling points belonging to the usable area. In other words, all sampling points may be used as extracted points. The controller 2 may also generate extracted data sets for only some of the sampling points. For example, the controller 2 may use 10 sampling points from 30 or more sampling points as extracted points and generate 10 sets of intensity measurement data. As described above, generating an extracted data set for a certain sampling point corresponds to extracting (cutting out) detection values ​​for that sampling point from multiple measurement data sets and generating data in which the detected values ​​are arranged in chronological order of acquisition time. The extracted data set may be associated with the extraction point.

[0076] 13 shows a process flow for determining whether or not a presence is present using a plurality of extracted data sets. As shown in FIG. 13, the presence / absence determination process may include steps S211 to S220. Step S211 is a step for generating a plurality of extracted data sets from a plurality of measurement data sets. The plurality of extracted data sets each have a different extraction point.

[0077] Step S212 is a step of performing a filter process on each of the multiple extracted data sets generated in step S211. The processing content for each extracted data set may be the same as that in step S202. As a result of step S212, multiple filtered data sets are obtained.

[0078] Step S213 is a step of determining an amplitude value for each of the multiple filtered data sets. The method of determining the amplitude value for one filtered data set may be the same as step S203. Similar to the extracted data set, the filtered data file is associated with sampling points (i.e., extraction points). The amplitude value of a certain filtered data set indicates the amplitude of the low-frequency component at the corresponding sampling point. Therefore, the amplitude value of a certain filtered data set can be rephrased as the amplitude value at the corresponding sampling point.

[0079] Step S213 may include determining a representative amplitude value based on the plurality of amplitude values ​​obtained from the plurality of filtered data sets. The representative amplitude value may be a median, maximum, or average of the plurality of amplitude values ​​from the plurality of filtered data sets. "RA" in the figure represents the representative amplitude value.

[0080] In step S214, the controller 2 determines whether the representative amplitude value determined in step S213 exceeds the amplitude threshold. If the representative amplitude value exceeds the amplitude threshold (YES in S214), the controller 2 determines that an occupant is present in the vehicle and ends the presence determination process. On the other hand, if the representative amplitude value is equal to or less than the amplitude threshold (NO in S214), the controller 2 executes the processes from step S216 onward.

[0081] Step S216 is a step in which the controller 2 performs frequency analysis processing (for example, FFT) on each of the plurality of filtered data sets, and as a result of step S216, a plurality of frequency spectra corresponding to the plurality of sampling points are obtained.

[0082] Step S217 is a step of identifying the peak frequency (FP) and its kurtosis (K) for each sampling point. The details of this process may be similar to those of step S207. The peak frequency and its kurtosis of a certain sampling point are determined from the frequency spectrum of the filtered data set of the sampling point. Step S217 obtains multiple frequency samples corresponding to multiple sampling points. One frequency sample may be a data set including the peak frequency and its kurtosis.

[0083] Step S218 is a step that includes calculating the standard deviation (σ) and representative kurtosis (RK) of the peak frequencies. The standard deviation may be the standard deviation of the multiple peak frequencies calculated in step S217. The representative kurtosis may be the median or average value of the multiple kurtosis values ​​calculated in step S217.

[0084] Step S218 may include a step of removing frequency samples that are likely to be noise from the plurality of frequency samples for more accurate determination. The standard deviation (σ) and representative kurtosis (RK) may be calculated for a population of some frequency samples that are likely to be affected by respiration from among all frequency samples. Step S218 may include extracting N frequency samples from a specific perspective from the plurality of frequency samples. The extracted K samples are referred to as adopted samples in this disclosure. The number of samples N may be 10, 20, or the like.

[0085] As shown in FIG. 14 , frequency samples at sampling points that are likely to be affected by breathing are expected to have high kurtosis and to be concentrated in a specific frequency band corresponding to the occupant's breathing frequency. Therefore, the controller 2 may select, as adopted samples, frequency samples with the top N kurtosis values ​​from among the multiple frequency samples obtained in step S217. The number of samples extracted, N, may be indefinite. The controller 2 may select, as adopted samples, frequency samples with kurtosis values ​​equal to or greater than a predetermined value. The horizontal axis of FIG. 14 represents frequency, and the vertical axis represents kurtosis. FIG. 14 is a graph showing the distribution of frequency samples obtained in step S217. A peak frequency whose kurtosis satisfies a predetermined adoption condition may be a peak frequency whose kurtosis is within the top N or a peak frequency whose kurtosis is equal to or greater than a predetermined value. The adoption condition may be that the kurtosis is within the top N or that the kurtosis is equal to or greater than a predetermined value.

[0086] The frequency samples obtained in step S217 may contain noise. The standard deviation of the peak frequencies calculated based on the selected samples is more suitable as a basis for determining whether an occupant is present than the standard deviation of the peak frequencies calculated based on all frequency samples. For this reason, the standard deviation (σ) of the peak frequencies calculated in step S218 may be a standard deviation based on the peak frequencies of the selected samples as a population. In another embodiment, step S218 may be a step of calculating the standard deviation (σ) of all frequency samples generated in step S217. The standard deviation may become smaller as the number of frequency samples affected by the occupant's breathing increases.

[0087] Furthermore, the representative kurtosis calculated in step S218 may be the median, maximum, or average value of the kurtosis of the adopted samples. That is, the representative kurtosis may be the median, maximum, or average value of the top N kurtosis values ​​among the multiple kurtosis values. In other embodiments, the representative kurtosis may be the median, maximum, or average value of the kurtosis of all the frequency samples obtained in step S217. The more frequency samples that are affected by the occupant's breathing, the larger the representative kurtosis value may be.

[0088] Step S219 is a step for determining whether parameters such as the standard deviation (σ) and the representative kurtosis (RK) satisfy a predetermined presence determination condition. For example, the presence determination condition may be that the standard deviation is equal to or less than a predetermined value and the representative kurtosis is greater than a predetermined value. The presence determination condition may be that the value obtained by dividing the representative kurtosis by the standard deviation (= RK / σ) is greater than a predetermined threshold. If the standard deviation (σ) and the representative kurtosis (RK) satisfy the predetermined presence determination condition (YES in S219), the controller 2 determines in step S215 that an occupant is present in the vehicle. If the standard deviation (σ) and the representative kurtosis (RK) do not satisfy the predetermined presence determination condition (NO in S219), the controller 2 determines in step S220 that an occupant is not present in the vehicle.

[0089] It should be noted that the controller 2 does not necessarily need to determine whether or not an occupant is present in the vehicle using both the standard deviation (σ) and the representative kurtosis (RK). The controller 2 may determine that an occupant is present in the vehicle when the standard deviation is equal to or less than a predetermined value. The controller 2 may also determine that an occupant is present in the vehicle when the standard deviation is equal to or less than a predetermined value and the median or average of the peak frequencies is within an acceptable range. The controller 2 may also determine that an occupant is present in the vehicle when the representative kurtosis is greater than a predetermined value. The controller 2 may also determine that an occupant is present in the vehicle when the representative kurtosis is greater than a predetermined value and the average or median of the peak frequencies is within an acceptable range. According to the above determination methods, it is possible to more accurately determine whether or not an occupant is present in the vehicle.

[0090] The above describes a pattern in which an FFT is performed on a filtered data set to obtain peak frequencies, etc., but the specific procedure for the analysis process may be changed. An FFT may be performed on an extracted data set, and data with frequencies equal to or greater than a predetermined value (e.g., 1.0 Hz) may be discarded from the frequency spectrum obtained by the FFT. In other words, peak frequencies, etc. may be identified by focusing on a specific frequency range from the frequency spectrum obtained by the FFT.

[0091] As another example of a modified system configuration, the radar module 3 may be a millimeter-wave radar instead of a UWB radar. The radar module 3 and the controller 2 may be located in different locations within the vehicle. Furthermore, multiple radar modules 3 may be installed in the vehicle HV. In other words, the biological detection device 1 may include multiple radar modules 3. For example, as shown in FIG. 15 , the biological detection device 1 may include a first radar module 3A, a second radar module 3B, and a third radar module 3C. The first radar module 3A may be a radar module 3 located near the overhead console. The second radar module 3B may be a radar module 3 located on the left / right B-pillar or on the backrest of the right / left front seat. The third radar module 3C may be a radar module 3 located in the center of the vehicle ceiling. The third radar module 3C may be a radar module 3 located a predetermined distance (e.g., 0.5 m) behind the center of the vehicle ceiling.

[0092] When multiple radar modules 3 are provided, the controller 2 may be connected to each of the multiple radar modules 3. The controller 2 may control the drive timing of each radar module 3 so that the multiple radar modules 3 do not interfere with each other. For example, the controller 2 may control the first radar module 3A and the second radar module 3B so that the start timings of their measurement processes are shifted by half the measurement interval. The multiple radar modules 3 may be configured to transmit impulse waves with a delay of 1 millisecond or more. The operation intervals between the radar modules 3 may be determined based on the detection duration and reverberation time. The reverberation time refers to the time that the power of the transmitted impulse wave remains in the detection target area (here, inside the vehicle).

[0093] Each radar module 3 may be provided with a sub-controller having some of the functions of the controller 2. The sub-controller may be configured to use the radar module 3 to transmit data indicating a provisional presence / absence determination result to the controller 2. The controller 2 may be configured to integrate the determination results of the sub-controllers and make a final determination.

[0094] The living body detection device 1 of the present disclosure may be applied to applications other than vehicles. For example, the living body detection device 1 may be used to determine whether or not a living body is present in a space with many reflective objects (more preferably, an enclosed space), such as a room, elevator, warehouse, or luggage compartment. The living body detection device 1 may also be used to determine whether or not a living body is present in a specific open area, such as a specific corridor. The detection target area may be changed as appropriate depending on the application.

[0095] <Supplementary Remarks (1)> [Technical Idea 1] A living organism detection device comprising: a transmitter (32) configured to be able to transmit radio waves in a predetermined frequency band as search waves; a receiver (35) configured to be able to receive the radio waves; and a determination unit (2) that determines whether or not a living organism is present in a predetermined detection target area based on a signal received by the receiver, wherein the determination unit is configured to: perform a sampling process at predetermined measurement intervals, the sampling process including intermittently transmitting the search waves using the transmitter and acquiring the reception strength of the radio waves at an observation timing, which is a timing when a predetermined time has elapsed since the search waves were transmitted; store in a memory, each time the sampling process is performed, an observed value of the reception strength at the observation timing; generate an intensity dataset, which is a dataset in which a plurality of observation values ​​at the observation timing are arranged in the order in which they were acquired; and determine whether or not the living organism is present in the detection target area by analyzing the intensity dataset.

[0096] [Technical Idea 2] Analyzing the intensity dataset includes performing frequency analysis on the intensity dataset or a filtered dataset obtained by filtering the intensity dataset to obtain frequency characteristic data indicating frequency characteristics, and obtaining a peak frequency, which is the frequency component with the strongest intensity in the intensity dataset, based on the frequency characteristic data, wherein the determination unit is configured to determine whether the living organism is present in the detection target area based on the peak frequency, in the living organism detection device described in Technical Idea 1.

[0097] [Technical Idea 3] The analyzing of the intensity data set includes calculating kurtosis, which is a parameter indicating the difference between the intensity of the peak frequency and the intensity of other frequency components, and the determination unit is configured to determine whether or not the living organism is present in the detection target area based on the peak frequency and its kurtosis, in the living organism detection device described in Technical Idea 2.

[0098] [Technical Idea 4] The living body detection device described in Technical Idea 3, wherein the judgment unit is configured to determine that the living body is present in the detection target area based on the peak frequency being within a predetermined range and the kurtosis being equal to or greater than a predetermined value.

[0099] [Technical Idea 5] The sampling process includes acquiring the reception strength of the radio wave at a plurality of different observation times after the exploration wave is transmitted, and the judgment unit is configured to: store the observation values ​​of the reception strength at the plurality of observation times in a memory each time the sampling process is performed; acquire the intensity data set for each of the observation times; acquire the peak frequency and its kurtosis for each observation time based on the intensity data set for each of the observation times; and determine whether or not the living organism is present in the detection target area based on the peak frequency, among the plurality of peak frequencies, whose kurtosis satisfies a predetermined adoption condition. This is the living organism detection device described in Technical Idea 3.

[0100] [Technical Concept 6] The biological detection device according to any one of Technical Concepts 1 to 5, wherein the search wave is an impulse-shaped radio wave.

[0101] [Technical Idea 7] A biological detection device according to any one of Technical Ideas 1 to 6, comprising a plurality of radar modules each including the transmitting unit and the receiving unit, wherein the plurality of radar modules include a first radar module (3A) and a second radar module (3B), and the first radar module and the second radar module are configured to perform sampling processing at different times.

[0102] [Technical Idea 8] A living body detection device according to any one of Technical Ideas 1 to 7, which is used in a vehicle, wherein the determination unit is configured to determine whether or not an occupant is present in the vehicle.

[0103] [Technical Idea 9] The judgment unit is configured to execute the sampling process multiple times within 10 seconds after the vehicle is locked, and is configured to determine whether or not an occupant remains in the vehicle based on reception strength data acquired within 10 seconds after the vehicle is locked, in the living body detection device described in Technical Idea 8.

[0104] <Supplementary Note (2)> The various flowcharts shown in this disclosure are all examples, and the number of steps constituting the flowcharts and the execution order of the processes can be changed as appropriate. The controls shown in each flowchart may be combined / executed in parallel to the extent that there is no contradiction. Expressions such as acquisition, determination, detection, generation, and calculation may be used interchangeably. When a device acquires certain data, it also includes the device generating the data based on a signal input from another device / sensor.

[0105] The devices, systems, and methods described herein may be implemented by a special-purpose computer having a processor programmed to execute one or more functions embodied in a computer program. The devices and methods described herein may be implemented using dedicated hardware logic circuits. The devices and methods described herein may be implemented by one or more special-purpose computers configured by combining a processor executing a computer program with one or more hardware logic circuits. The processor may be any computing core, such as a CPU, MPU, GPU, or DFP (Data Flow Processor). Some or all of the functions of the biometric detection device may be implemented as hardware. Some or all of the functions of the biometric detection device may be implemented using a system-on-chip (SoC), an integrated circuit (IC), or an FPGA.

[0106] The computer program includes instructions that are executed by a computer. The computer program may be stored in a computer-readable non-transitory tangible storage medium. The storage medium for the computer program may be a variety of media, such as a hard-disk drive (HDD), a solid-state drive (SSD), or a flash memory.

Claims

1. A living organism detection device comprising: a transmitting unit (32) configured to be able to transmit radio waves in a predetermined frequency band as search waves; a receiving unit (35) configured to be able to receive the radio waves; and a determining unit (2) that determines whether or not a living organism is present in a predetermined detection target area based on the signal received by the receiving unit, wherein the determining unit is configured to: perform a sampling process at predetermined measurement intervals, which includes intermittently transmitting the search waves using the transmitting unit and acquiring the reception strength of the radio waves at an observation timing, which is a timing when a predetermined time has elapsed since the search waves were transmitted; store in memory the observed value of the reception strength at the observation timing each time the sampling process is performed; generate an intensity dataset, which is a dataset in which multiple observation values at the observation timing are arranged in the order in which they were obtained; and determine whether or not a living organism is present in the detection target area by analyzing the intensity dataset.

2. The living body detection device of claim 1, wherein analyzing the intensity data set includes: performing frequency analysis on the intensity data set or a filtered data set obtained by filtering the intensity data set to obtain frequency characteristic data indicating frequency characteristics; and obtaining a peak frequency, which is the frequency component with the strongest intensity in the intensity data set, based on the frequency characteristic data; and wherein the determination unit is configured to determine whether or not the living body is present in the detection target area based on the peak frequency.

3. The biological detection device of claim 2, wherein analyzing the intensity data set includes calculating kurtosis, which is a parameter indicating the difference between the intensity of the peak frequency and the intensity of other frequency components, and the judgment unit is configured to judge whether or not the biological organism is present in the detection target area based on the peak frequency and its kurtosis.

4. The living body detection device of claim 3, wherein the judgment unit is configured to determine that the living body is present in the detection target area based on the peak frequency being within a predetermined range and the kurtosis being equal to or greater than a predetermined value.

5. The biological detection device of claim 3, wherein the sampling process includes acquiring the reception strength of the radio wave at a plurality of different observation times after the exploration wave is transmitted, and the judgment unit is configured to: store the observed values of the reception strength at the plurality of observation times in memory each time the sampling process is performed; acquire the intensity data set for each observation time; acquire the peak frequency and its kurtosis for each observation time based on the intensity data set for each observation time; and determine whether or not the biological organism is present in the detection target area based on the peak frequency, among the plurality of peak frequencies, whose kurtosis satisfies a predetermined adoption condition.

6. The biological detection device according to claim 1, wherein the probe wave is an impulse-shaped radio wave.

7. The biological detection device according to claim 1, comprising a plurality of radar modules each including the transmitting unit and the receiving unit, the plurality of radar modules including a first radar module (3A) and a second radar module (3B), the first radar module and the second radar module being configured to perform the sampling process at different times.

8. A living body detection device according to claim 1, which is used in a vehicle, wherein the determination unit is configured to determine whether or not an occupant is present in the vehicle.

9. The living body detection device described in claim 8, wherein the judgment unit is configured to perform the sampling process multiple times within 10 seconds after the vehicle is locked, and to determine whether or not an occupant remains in the vehicle based on reception strength data obtained within 10 seconds after the vehicle is locked.

10. A living organism detection method comprising: performing a sampling process at predetermined measurement intervals, the sampling process including intermittently transmitting a probe wave using a transmitting unit and acquiring the reception strength of the probe wave at an observation timing, which is a timing when a predetermined time has elapsed since the probe wave was transmitted; storing the observed value of the reception strength at the observation timing in memory each time the sampling process is performed; generating an intensity dataset, which is a dataset in which multiple observation values at the observation timing are arranged in the order in which they were acquired; and determining whether a living organism is present in the detection target area by analyzing the intensity dataset.

Images