Speed ​​detection device and method

Abstract

We propose a speed detection device and method that can accurately detect the speed of a moving object in real time. [Solution] A continuous sinusoidal wave signal is emitted toward a target object, the reflected wave of the sinusoidal wave signal at the target is received, the frequency components of the sinusoidal wave signal that have been Doppler-shifted according to the relative velocity of the moving object and the target are extracted from the received reflected wave and output as a Doppler-shifted signal, and within a predetermined time period, a plurality of Doppler-shifted signals of second time lengths are generated by cutting out a portion of the Doppler-shifted signal of a predetermined first time length corresponding to the time period at multiple different timings, the Doppler velocity for each of the first and second time length Doppler-shifted signals is calculated based on each of the first and second time length Doppler-shifted signals, and the velocity of the moving object in the current time period is determined based on the calculated Doppler velocities for each of the first and second time length Doppler-shifted signals.

Description

Technical Field

[0001] The present invention relates to a speed detection device and method, and is suitable for application to a vehicle speed detection device that detects the speed of a vehicle such as a railway vehicle.

Background Art

[0002] In the railway business, safe operation of vehicles and accurate train operation are important. For safe operation of vehicles, speed monitoring of the vehicle is essential, and in particular, monitoring of the vehicle speed at stops on the platform and in scenarios such as curves is required. Also in urban areas, the train schedule is becoming more and more dense, and speed monitoring of the vehicle is also required for accurate train operation. In addition, in tunnels and subways, etc., since GPS (Global Positioning System) information cannot be obtained, accurate vehicle speed information is required for vehicle position detection.

[0003] Conventionally, for speed detection of railway vehicles, a method of measuring the rotation speed of a wheel using a tachogenerator attached to the wheel and calculating the speed from the rotation speed of the wheel has been widely used. However, in this method, there is a problem that accurate speed may not be detected due to wheel slip. Therefore, in recent years, as a method for avoiding the influence of wheel slip and detecting accurate speed, research and development of a speed detection method using millimeter waves has been promoted.

[0004] As such a speed detection method using millimeter waves, a method of irradiating a sine wave of millimeter waves from a speed detection device installed on a vehicle to a rail and detecting the speed from the Doppler shift amount of the sine wave reflected from the rail is disclosed in Patent Document 1.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Patent Document 2

[0006] However, with this method, the amount of Doppler shift changes depending on the irradiation angle of the millimeter waves emitted from the speed detection device onto the rails. Therefore, in order to accurately detect the vehicle speed, it is necessary to irradiate the rails with millimeter waves at a known, designed irradiation angle and within a narrow angular range.

[0007] In this regard, for example, Patent Document 2 discloses an antenna that generates a narrow-angle beam capable of irradiating a narrow angular range, in which a patch antenna is formed on a substrate and a lens antenna is placed on top of it.

[0008] In this case, a large lens antenna is required to generate a narrow-angle beam in such an antenna. However, when applying this method to railway vehicles, it is difficult to install a large antenna because the space at the bottom of the vehicle directly above the rails is limited. For this reason, an antenna with a beam width is actually installed, but when a beam width is used, there is a problem in that a speed detection error occurs corresponding to the beam width.

[0009] As a method to suppress such velocity detection errors, for example, Patent Document 3 discloses a method of taking a moving average when performing a Fourier transform on the Doppler shift amount. However, this method has the problem that a delay equal to the time length required to calculate the moving average occurs in velocity detection.

[0010] This invention has been made in consideration of the above points, and aims to propose a velocity detection device and method that can accurately detect the velocity of a moving object in real time by suppressing velocity detection errors caused by fluctuations in Doppler shift due to the influence of the antenna beam width. [Means for solving the problem]

[0011] To solve the above problems, the present invention provides a speed detection device for detecting the speed of a moving object, comprising: an antenna unit mounted on the moving object that emits a continuous sinusoidal wave signal toward a target object, receives the reflected wave of the sinusoidal wave signal at the object object, extracts the frequency components of the sinusoidal wave signal that have been Doppler-shifted according to the relative speed of the moving object and the object object from the received reflected wave object and outputs them as a Doppler-shifted signal; and a plurality of second timings obtained by cutting out a portion of the Doppler-shifted signal of a predetermined first time length corresponding to a predetermined time period within a predetermined time period. The system includes a cutting unit that generates a Doppler shift signal of a given time length, a Fourier transform unit that performs a Fourier transform on each of the first and second time length Doppler shift signals, a Doppler velocity detection unit that calculates the Doppler velocity for each of the first and second time length Doppler shift signals based on the Fourier transformed signals, and a velocity determination unit that determines the velocity of the moving object in the current time period based on the calculated Doppler velocity for each of the first and second time length Doppler shift signals.

[0012] Furthermore, the present invention relates to a speed detection method performed in a speed detection device for detecting the speed of a moving object, comprising: a first step of emitting a continuous sinusoidal wave signal toward a target object, receiving the reflected wave of the sinusoidal wave signal at the object, and extracting the frequency components of the sinusoidal wave signal that have been Doppler-shifted according to the relative speed of the moving object and the object from the received reflected wave and outputting them as a Doppler-shifted signal; and a plurality of second time intervals obtained by cutting out a portion of the Doppler-shifted signal of a predetermined first time length corresponding to the time period at a plurality of different timings within a predetermined time period. The system includes a second step of generating a Doppler shift signal of a given time length, a third step of performing a Fourier transform on each of the first and second time length Doppler shift signals, a fourth step of calculating the Doppler velocity for each of the first and second time length Doppler shift signals based on the Fourier-transformed signals, and a fifth step of determining the velocity of the moving object in the current time period based on the calculated Doppler velocity for each of the first and second time length Doppler shift signals.

[0013] According to the velocity detection device and method of the present invention, the Doppler velocity calculated based on each Doppler shift signal of the second time length is reflected in the global Doppler velocity calculated based on the Doppler shift signal of the first time length to calculate the final Doppler velocity. [Effects of the Invention]

[0014] According to the present invention, a speed detection device and method can be realized that can accurately detect the speed of a moving object in real time. [Brief explanation of the drawing]

[0015] [Figure 1] (A) and (B) are schematic diagrams illustrating the vehicle speed detection principle according to this embodiment. [Figure 2] This is a waveform diagram used to explain fading. [Figure 3]This is a diagram for explaining update cycle data and update cycle partial data. [Figure 4] This is a block diagram showing the configuration of a vehicle speed detection device according to the present embodiment. [Figure 5] (A) to (C) are waveform diagrams for explaining experimental results using the vehicle speed detection device according to the present embodiment.

Embodiments for Carrying Out the Invention

[0016] The following describes a detailed embodiment of the present invention with reference to the drawings.

[0017] (1) Principle As shown in FIG. 1(A), consider a vehicle speed detection device 3 that irradiates millimeter waves toward a target 1 such as the ground or a rail, and detects the speed of the vehicle 2 based on the reflected wave. Here, it is assumed that the vehicle speed detection device 3 is installed at the bottom of the vehicle 2 such as a railway vehicle.

[0018] The Doppler speed Vdpr detected by this vehicle speed detection device 3 can be expressed by the following formula when the angle formed by the vehicle speed Vcar and the irradiation direction is θ.

Equation

[0019] Here, when the transmitting antenna and the receiving antenna for transmitting and receiving millimeter waves in this vehicle speed detection device 3 are separate bodies, the calculation of the Doppler speed Vdpr becomes complicated because the irradiation angle of the millimeter wave at the target 1 and the receiving angle of the millimeter wave at the receiving antenna are different. Therefore, this problem can be solved by coupling the transmitting part and the receiving part of the millimeter wave with a circulator circuit and sharing the transmitting antenna and the receiving antenna.

[0020] Furthermore, if the angular range of the millimeter waves irradiating target 1 is very narrow, like a pencil beam, then in (1), "θ" can be treated as a constant, and the Doppler velocity Vdpr is uniquely determined. However, as mentioned above, a huge antenna is required to generate a narrow-angle pencil beam.

[0021] However, considering the size of the antenna that can typically be installed on the bottom of vehicle 2, it is difficult to generate a narrow-angle pencil beam. In reality, the antenna will have a certain beam width, and therefore the range of millimeter waves illuminating target 1 will also have a certain degree of spread.

[0022] As shown in Figure 1(B), when the irradiation range R of the millimeter waves irradiated onto the target 1 has a range from θ-δ to θ+δ, the Doppler velocity Vdpr is also detected by the vehicle speed detection device 3 with an error range corresponding to this range. For example, if the irradiation angle θ of the millimeter waves on the target 1 is 45 degrees and the beam width of the antenna is ±5 degrees, the detected Doppler velocity Vdpr will have an error range of -9% to +8% because the range of θ in cosθ has a range of 40 to 50 degrees.

[0023] Furthermore, since the received waveform from the antenna is a superposition of waveforms with different Doppler shift amounts, a fading waveform like the one shown in Figure 2 is output from the antenna. This fading waveform has instantaneous amplitude drops, as shown in the area enclosed by the dashed line K1 in the figure, making it difficult to detect the received waveform at such points.

[0024] Therefore, in this embodiment, in order to avoid detection errors of the Doppler velocity Vdpr caused by the beam width of millimeter waves and undetected received waveforms due to fading, the vehicle speed detection device 3 uses all available data of the fading waveform acquired within the period in which it detects the vehicle speed Vcar (for example, 25 milliseconds, which will be referred to as the update period) (hereinafter referred to as the update period data), as well as data extracted from parts of this update period data at different timings (hereinafter referred to as update period partial data), to detect the vehicle speed Vcar.

[0025] Specifically, in this embodiment, as shown in Figure 3 for example, Fourier transforms are performed on "Data 1," which is the largest first time portion of data available for data processing, excluding the time spent storing the data in memory, etc., from all the data of the fading waveform acquired within the update cycle period, and on "Data 2" to "Data 6," which are portions of "Data 1" extracted at different timings and with a time length sufficiently shorter than the time length of "Data 1."

[0026] Furthermore, based on the Fourier transform results of "Data 1" to "Data 6", the peak frequency is f0, the speed of light is c, and the frequency of millimeter waves is f d As follows,

number

[0027] Furthermore, each Doppler velocity Vdpr' obtained in this way is weighted according to its reliability.

[0028] For example, in areas enclosed by the dashed line K1 in Figure 2, where instantaneous amplitude drops occur due to fading, signals with low received power due to antenna gain reduction at the millimeter-wave beam edge of target 1 are strongly affected. Therefore, using data from the area enclosed by the dashed line K1 will result in a large error in the detected vehicle speed Vcar.

[0029] Therefore, in this embodiment, the Doppler velocity Vdpr' derived from the Fourier transform of the update cycle data ("Data 1" in Figure 3) and the Doppler velocity Vdpr' derived from partial data ("Data 2" to "Data 6" in Figure 3) extracted from the update cycle data at different timings are weighted with a magnitude corresponding to the peak power after the Fourier transform or the signal power before the Fourier transform.

[0030] Specifically, for example, the weight of update cycle data that includes time periods when the received power is low, as enclosed by the dashed line K1 in Figure 2, is reduced, while the weight of update cycle data that includes time periods when the received power is high, as enclosed by the dashed line K2 in Figure 2, is increased. This weighting helps to suppress errors associated with the antenna beamwidth and the effects of amplitude drops in the received signal due to fading.

[0031] Furthermore, since vehicle speed Vcar changes continuously with time, tracking allows for the estimation of the vehicle speed Vcar for the next update cycle. A well-known tracking method uses a Kalman filter, which can be used to estimate the vehicle speed Vcar for the next update cycle. In addition to the Kalman filter, another method for estimating vehicle speed Vcar is to estimate the vehicle speed Vcar for the next update cycle using the average value of past vehicle speed Vcars and acceleration information.

[0032] In this embodiment, the estimated vehicle speed Vcar for the next update cycle, which is estimated by this tracking process, is also used to set the weight of the Doppler speed Vdpr' calculated based on the above-mentioned "Data 1" to "Data 6".

[0033] Specifically, the vehicle speed Vcar for the next update cycle is estimated by tracking processing, and the Doppler speed Vdpr for the next update cycle is estimated using equation (1) based on this estimated value. The closer the Doppler speed Vdpr' obtained based on the Fourier transform results of "Data 1" to "Data 6" in Figure 3 is to the estimated value of this Doppler speed Vdpr (hereinafter referred to as the estimated Doppler speed), the higher the reliability of the Doppler speed Vdpr'. Therefore, in this embodiment, the closer the Doppler speed Vdpr' obtained based on the Fourier transform results is to the estimated Doppler speed based on the vehicle speed Vcar obtained by tracking processing for "Data 1" to "Data 6", the greater the weight given to "Data 1" to "Data 6".

[0034] In this embodiment, the final Doppler velocity Vdpr is calculated by taking the average of the weighted Doppler velocities Vdpr', and based on the calculated Doppler velocity Vdpr, the following equation is used:

number

[0035] As described above, the detection error of the vehicle speed Vcar caused by fluctuations in Doppler shift due to the influence of the antenna beam width can be suppressed within the update cycle. Below, a vehicle speed detection device according to this embodiment, which applies this principle, will be described.

[0036] (2) Configuration of the vehicle speed detection device according to this embodiment Figure 4 shows a vehicle speed detection device 10 according to this embodiment. This vehicle speed detection device 10 comprises an antenna unit 11 installed on the bottom of the vehicle and a calculation unit 12 that calculates the vehicle speed Vcar based on the output from the antenna unit 11.

[0037] The antenna unit 11 includes a synthesizer 20 that generates a sine wave signal in the millimeter wave band, for example, 76 GHz. The beam of the sine wave signal generated by the synthesizer 20 is continuously emitted from the transmitting antenna 22 via an amplifier 21, illuminating the target object 1 (Figure 1(A)).

[0038] Furthermore, the antenna unit 11 receives a portion of the reflected wave obtained when the beam of this sinusoidal signal is reflected by the target 1, with the receiving antenna 23. At this time, the received signal received by the receiving antenna 23 is a sinusoidal signal that has been Doppler-shifted according to the relative speed between the target 1 and the vehicle speed detection device 10.

[0039] This received signal is down-converted by a high-frequency converter, the mixer 25, via the amplifier 24. Here, a sine wave signal generated by the synthesizer 20 is supplied as the local signal to the mixer 25. Therefore, the mixer 25 outputs a frequency component of the reflected sine wave signal received by the receiving antenna 23 that has been Doppler-shifted according to the relative speed between the vehicle 2 (Figure 1) and the target object 1 (Figure 1). Hereafter, this frequency component of the sine wave signal will be referred to as the Doppler-shifted signal.

[0040] Then, high-frequency components of this Doppler-shifted signal that are above a predetermined frequency are removed by the filter 26, and after being converted into a digital signal by the A / D (Analog / Digital) converter 27, it is transmitted to the arithmetic unit 12.

[0041] The calculation unit 12 stores the Doppler shift signal data provided by the antenna unit 11 in update cycle units for detecting the vehicle speed Vcar. The Doppler shift signal data stored at this time is the maximum available time for data processing, excluding the time spent storing the data in memory as described above. The stored update cycle unit data is then provided as update cycle data D1 to the Fourier converter 31, which is provided in conjunction with the update cycle data D1.

[0042] The update cycle data D1 is also supplied to each of the multiple data extraction units 30. Each data extraction unit 30 then extracts a portion of the update cycle data D1 at different timings and for the same duration, but significantly shorter than the update cycle data D1, as shown in "Data 2" to "Data 6" in Figure 3. Each data extraction unit 30 then outputs the resulting update cycle portion data D2 to a Fourier transformer 31 that is associated with it.

[0043] Note that each update cycle portion data D2 may have some overlap in time periods with the update cycle portion data D2 of the preceding and succeeding time periods, as shown in Figure 3. However, multiple update cycle portion data D2 are generated so that they cover the entire period from the beginning to the end of the update cycle data D1. In other words, the update cycle portion data D2 are generated so that all parts of the update cycle data D1 are distributed and included in one of the update cycle portion data D2.

[0044] The Fourier transformer 31 performs a Fourier transform on the given update period data D1 or update period partial data D2, and outputs the resulting frequency domain update period data (hereinafter referred to as the Fourier-transformed update period data) D3 or frequency domain update period partial data (hereinafter referred to as the Fourier-transformed update period partial data) D4 to the Doppler velocity detection unit 32 and the level detection unit 33, which are provided in conjunction with itself.

[0045] The Doppler velocity detection unit 32 extracts the peak frequency from the Fourier transform update period data D3 or the Fourier transform update period partial data D4 provided by the Fourier transform 31, and calculates the Doppler velocity Vdpr' based on the extracted peak frequency using equation (2) above. The Doppler velocity detection unit 32 then outputs the calculated Doppler velocity Vdpr' to the weight calculation unit 35 and the multiplication circuit 36, respectively, which are provided in the weighting processing unit 34, in association with itself.

[0046] Furthermore, the level detection unit 33 detects the peak frequency power (hereinafter referred to as peak power) PV from the Fourier transform update period data D3 or the Fourier transform update period partial data D4 provided by the Fourier transform converter 31, and outputs the detected peak power PV to the weight calculation unit 35 provided in the weighting processing unit 34, associating it with itself.

[0047] Furthermore, as will be described later, each weight calculation unit 35 is given an estimated Doppler velocity for the current update cycle as the estimated Doppler velocity EVdpr by the tracking velocity estimation unit 39.

[0048] Thus, the weight calculation unit 35 calculates a weight W for the corresponding Fourier transform update period data D3 or Fourier transform update period partial data D4 based on the Doppler velocity Vdpr' provided by the corresponding Doppler velocity detection unit 32, the peak power PV provided by the corresponding level detection unit 33, and the estimated Doppler velocity EVdpr provided by the tracking velocity estimation unit 39, and outputs the calculation result to the multiplication circuit 36 ​​associated with the weight calculation unit 35.

[0049] Specifically, for example, the weight calculation unit 35, which is associated with the Fourier transform update period data D3, calculates the vehicle speed Vcar from the Doppler speed Vdpr detected by the Doppler speed detection unit 32 in the current update period using equation (3) above, and takes the calculated vehicle speed Vcar as S, and then the following equation

number

[0050] Furthermore, each weight calculation unit 35, which is associated with the updated period partial data D4 after the Fourier transform, is calculated using the following equation:

number

[0051] Incidentally, in equations (4) and (5) above, "Vdpr'-EVdpr" is the error between the estimated Doppler velocity EVdpr estimated by the tracking velocity estimation unit 39 and the corresponding Doppler velocity Vdpr' detected by the Doppler velocity detection unit 32. Therefore, in equations (4) and (5), the peak power PV is multiplied by the reciprocal of the error so that the larger the error, the smaller the weight W becomes.

[0052] The multiplication circuit 36 ​​multiplies the Doppler velocity Vdpr' provided by the corresponding Doppler velocity detection unit 32 by the weight W provided by the corresponding weight calculation unit 35, and outputs the multiplication result as the weighted detected Doppler velocity WVdpr' to the average value calculation unit 37. The average value calculation unit 37 then calculates the average of the weighted detected Doppler velocities WVdpr' provided by each multiplication circuit 36, and outputs the calculation result as the final Doppler velocity Vdrp to the vehicle speed determination unit 38.

[0053] The vehicle speed determination unit 38 then calculates the vehicle speed Vcar for the current update cycle using equation (3) above, based on the Doppler speed Vdpr given by the average value calculation unit 37, and outputs the calculated vehicle speed Vcar to the outside, as well as to the tracking speed estimation unit 39.

[0054] The tracking speed estimation unit 39 estimates the Doppler velocity in the next update cycle using an existing tracking method, for example, one that utilizes a Kalman filter. Specifically, it estimates the vehicle speed Vcar for the next cycle by inputting the vehicle speed Vcar provided by the vehicle speed determination unit 38 into the Kalman filter, and then estimates the Doppler velocity for the next cycle based on the estimated vehicle speed Vcar. The tracking speed estimation unit 39 then outputs the estimated Doppler velocity as the estimated Doppler velocity EVdpr to each weight calculation unit 35 of the weighting processing unit 34.

[0055] Based on this estimated Doppler velocity EVdpr, the weights of the Doppler velocity Vdpr' calculated based on the update cycle data D1 in the next update cycle, as well as the weights of the Doppler velocity Vdpr' calculated based on each update cycle partial data D2, are calculated as described above.

[0056] However, if the vehicle speed Vcar provided by the vehicle speed determination unit 38 is below a predetermined threshold speed (for example, 2 km / h), the tracking speed estimation unit 39 instructs the weight calculation unit 35 corresponding to the Fourier transform update period data D3 to set the weight W to "1", and instructs the weight calculation unit 35 corresponding to each Fourier transform update period partial data D4 to set the weight W to "0".

[0057] In addition, the tracking speed estimation unit 39 instructs the average value calculation unit 37 to output the weighted detected Doppler speed WVdpr', which is provided by the weight calculation unit 35 corresponding to the Fourier transform update period data D3, directly to the vehicle speed determination unit 38.

[0058] In this case, the weighted detected Doppler velocity WVdpr' given by the multiplication circuit 36 ​​corresponding to the Fourier transform-post-update period data D3 is output directly to the vehicle speed determination unit 38 as the final Doppler velocity Vdrp, and the vehicle speed determination unit 38 calculates the vehicle speed Vcar based on this Doppler velocity Vdrp.

[0059] In the case of low vehicle speed Vcar, the reason for providing the vehicle speed determination unit 38 with the weighted detected Doppler velocity WVdpr calculated based on the Fourier transform-based update period data D3 as the final Doppler velocity Vdrp is as follows:

[0060] When the vehicle speed Vcar is slow, the amount of Doppler shift is small. Also, the longer the data length of the update cycle data, the more the low-frequency region of the Doppler shift can be detected, making it possible to detect low vehicle speeds. Therefore, in this embodiment, when the vehicle speed Vcar determined by the vehicle speed determination unit 38 is low, the vehicle speed Vcar is calculated using only the Doppler speed Vdpr' detected based on the Fourier-transformed update cycle data D3, thereby effectively suppressing errors.

[0061] However, instead of setting the weight in the weight calculation unit 35 corresponding to each post-Fourier transform update period partial data D4 to "0", if the vehicle speed Vcar given by the vehicle speed determination unit 38 is a low speed below a predetermined threshold speed, the tracking speed estimation unit 39 may control each weight calculation unit 35 to increase the weight W calculated by equation (4) in the weight calculation unit 35 corresponding to the post-Fourier transform update period data D3, and decrease the weight W calculated by equation (5) in each weight calculation unit 35 corresponding to each post-Fourier transform update period partial data D4.

[0062] Figures 5(A) to 5(C) show an example of the detection result of the vehicle speed Vcar when the beam width is ±5 degrees. Figure 5(A) shows the result of detecting the vehicle speed Vcar using the method of this embodiment, and the standard deviation error of the detected vehicle speed Vcar was approximately 0.12 km / h.

[0063] Figure 5(B) shows the results of detecting the vehicle speed Vcar using only the update cycle data D1. Compared with the results in Figure 5(A), the detected vehicle speed Vcar fluctuates, and the standard deviation error of the detected vehicle speed Vcar was 0.31 km / h.

[0064] Furthermore, Figure 5(C) shows the results of calculating the vehicle speed Vcar based on the simple average of the Doppler velocities Vdpr' calculated from the Fourier transform results of "Data 1" to "Data 6" in Figure 3, without weighting these Doppler velocities Vdpr'. Similar to Figure 5(B), a comparison with the results in Figure 5(A) reveals that the detected vehicle speed Vcar is fluctuating. The standard deviation error of the detected vehicle speed Vcar at this time was approximately 0.28 km / h.

[0065] From the results shown in Figures 5(A) to (C) above, it was confirmed that the error in the detected vehicle speed Vcar can be suppressed by weighting the Doppler speed Vdpr' calculated based on the update cycle data D1 and the Doppler speed Vdpr' calculated based on each update cycle partial data D2 with a weight W corresponding to the reliability of the update cycle partial data D2.

[0066] (3) Effects of this embodiment As described above, the vehicle speed detection device 10 of this embodiment generates update cycle partial data D2 by extracting parts of the update cycle data D1 at different timings, and calculates the Doppler velocity Vdpr' based on the Fourier transformed update cycle data D3 and the Fourier transformed update cycle partial data D4 obtained by Fourier transforming the update cycle data D1 and the update cycle partial data D2, respectively. The vehicle speed detection device 10 also weights these calculated Doppler velocities Vdpr' with a weight W of a magnitude corresponding to their reliability, calculates the final Doppler velocity Vdpr based on each weighted Doppler velocity Vdpr', and determines the vehicle speed Vcar based on that Doppler velocity Vdpr.

[0067] According to the vehicle speed detection method of this embodiment, the Doppler speed Vdpr' calculated based on update cycle partial data D2, which has a shorter time duration, is reflected in the overall Doppler speed Vdpr' within the update cycle calculated based on update cycle data D1, thereby correcting the Doppler speed Vdpr' calculated based on update cycle data D1.

[0068] In this case, the global Doppler velocity Vdpr' within the update cycle, calculated based on the update cycle data D1, is greatly influenced by the update cycle partial data D2 from reliable time periods within the update cycle, whereas the fluctuations in the Doppler shift have only a small influence on the update cycle partial data D2 from less reliable time periods within the update cycle.

[0069] Therefore, the vehicle speed detection method of this embodiment can significantly suppress speed detection errors caused by fluctuations in Doppler shift resulting from the beam width of the sinusoidal signal emitted from the antenna unit 11. Furthermore, with the vehicle speed detection method of this embodiment, the vehicle speed Vcar for each update cycle can be calculated within that update cycle, so there is no delay in detecting the vehicle speed Vcar. Thus, with this vehicle speed detection method, the vehicle speed can be detected accurately in real time.

[0070] (4) Other embodiments In the embodiments described above, the present invention was applied to a vehicle speed detection device for detecting the speed of a vehicle. However, the present invention is not limited to this and can be broadly applied to speed detection devices for detecting the speed of moving objects other than vehicles, such as elevators.

[0071] Furthermore, while the above-described embodiment mentions a case where the entire vehicle speed detection device 10 is mounted on the vehicle, the present invention is not limited to this, and the calculation unit 12 may be installed in a location other than the vehicle. In this case, the calculation unit 12 may be configured not as a single computer device, but as a distributed computing system composed of multiple computer devices.

[0072] Furthermore, although the above-described embodiment mentions a case where all update cycle partial data D2 are generated with the same time length, the present invention is not limited to this, and some or all of each update cycle partial data D2 may have different time lengths. [Industrial applicability]

[0073] The present invention can be widely applied to speed detection devices of various configurations for detecting the speed of a moving object. [Explanation of Symbols]

[0074] 10...Vehicle speed detection device, 11...Antenna unit, 12...Calculation unit, 30...Data extraction unit, 31...Fourier transformer, 32...Doppler velocity detection unit, 33...Level detection unit, 34...Weighting processing unit, 35...Weight calculation unit, 36...Multiplication circuit, 37...Average value calculation unit, 38...Vehicle speed determination unit, 39...Tracking speed estimation unit, D1...Update cycle data, D2...Update cycle partial data, D3...Update cycle data after Fourier transform, D4...Update cycle partial data after Fourier transform, PV...Peak voltage, Vcar...Vehicle speed, Vdpr, Vdpr'...Doppler velocity, W...Weight.

Claims

1. In a speed detection device for detecting the speed of a moving object, An antenna unit mounted on the mobile body transmits a continuous sinusoidal signal toward a target object, receives the reflected wave of the sinusoidal signal from the target object, extracts the frequency components of the sinusoidal signal that have been Doppler-shifted according to the relative velocity of the mobile body and the target object from the received reflected wave object, and outputs them as a Doppler-shifted signal. An extraction unit that generates a plurality of Doppler shift signals of a second time length by extracting a portion of the Doppler shift signal of a predetermined first time length corresponding to the time period within a predetermined time period at a plurality of different timings, A Fourier transform unit that performs a Fourier transform on each of the first and second time-length Doppler shift signals, A Doppler velocity detection unit calculates the Doppler velocity for each of the first and second Doppler shift signals of the time lengths based on the Fourier-transformed Doppler shift signals of the time lengths, A speed determination unit determines the speed of the moving body in the current time period based on the calculated Doppler velocity for each of the first and second time-length Doppler shift signals. A speed detection device characterized by comprising the following features.

2. The first time length is, This is the maximum time length for which the Doppler shift signal can be used within a predetermined time period. Each of the Doppler shift signals of the second time length is, All portions of the Doppler shift signal of the first time length are distributed and generated to be included in any of the Doppler shift signals of the second time length. The speed detection device according to claim 1.

3. The system includes a weighting processing unit that weights the Doppler velocity calculated based on the Doppler shift signal of the first time length and the respective Doppler velocities calculated based on each of the Doppler shift signals of the second time length, according to the reliability of the calculated Doppler velocities. The speed detection device according to claim 2.

4. The system further includes a level detection unit that detects the peak voltages of the first and second time-length Doppler shift signals, respectively, which have been Fourier transformed. The aforementioned weighting processing unit, Each of the first and second time-length Doppler shift signals is weighted with a weight corresponding to the magnitude of the peak voltage of the Fourier-transformed Doppler shift signal. The speed detection device according to claim 3.

5. The system includes a Doppler velocity estimation unit that estimates the Doppler velocity based on the velocity of the moving object determined by the velocity determination unit, The aforementioned weighting processing unit, The Doppler shift signals of the first and second time lengths are weighted according to the difference between the Doppler velocity calculated by the Doppler velocity detection unit based on the Doppler velocity of each of the first and second time length Doppler shift signals, and the Doppler velocity estimated by the Doppler velocity estimation unit. The speed detection device according to claim 3 or 4, characterized in that it is as described above.

6. A speed detection method performed in a speed detection device for detecting the speed of a moving object, A first step involves transmitting a continuous sinusoidal wave signal toward a target object, receiving the reflected wave of the sinusoidal wave signal from the target, and extracting the frequency components of the sinusoidal wave signal that have been Doppler-shifted according to the relative velocity of the moving object and the target from the received reflected wave, and outputting them as a Doppler-shifted signal. A second step of generating a plurality of Doppler shift signals of a second time length by extracting a portion of the Doppler shift signal of a predetermined first time length corresponding to the time period at a plurality of different timings within a predetermined time period, A third step involves performing a Fourier transform on each of the first and second time-length Doppler shift signals, A fourth step of calculating the Doppler velocity for each of the first and second Doppler shift signals of each time length based on the Fourier-transformed Doppler shift signals of each time length, A fifth step in which the speed of the moving object in the current time period is determined based on the calculated Doppler velocity for each of the first and second time lengths of the Doppler shift signals, A speed detection method characterized by comprising the following features.

7. The first time length is, This is the maximum time length for which the Doppler shift signal can be used within a predetermined time period. Each of the Doppler shift signals of the second time length is, All portions of the Doppler shift signal of the first time length are distributed and generated to be included in any of the Doppler shift signals of the second time length. The speed detection method according to feature 6.

8. In the fourth step, the speed detection device is The Doppler velocity calculated based on the Doppler shift signal of the first time length and the Doppler velocity calculated based on each of the Doppler shift signals of the second time length are weighted according to the reliability of the calculated Doppler velocity. The speed detection method according to feature 7.

9. In the fourth step, the speed detection device is The peak voltages of the first and second time-length Doppler shift signals, which have been Fourier transformed, are detected, Each of the first and second time-length Doppler shift signals is weighted with a weight corresponding to the magnitude of the peak voltage of the Fourier-transformed Doppler shift signal. The speed detection method according to feature 7.

10. The speed detection device is Based on the determined speed of the moving object, the Doppler velocity is estimated. In the fourth step, the speed detection device is The Doppler velocities of the first and second time lengths, respectively, are weighted according to the difference between the Doppler velocities calculated based on the respective Doppler velocities of the first and second time lengths and the estimated Doppler velocities. The speed detection method according to claim 8 or 9, characterized by the features described above.

Images