JP2012149951A - Radar device - Google Patents

Abstract

【Task】
In a weather radar apparatus, the FFT method and the PPP method, which are typical Doppler velocity calculation methods, have contradictory properties. Combine both methods to provide more accurate Doppler velocity observation data.
【Solution】
The Doppler velocity and velocity width by the FFT method and the PPP method are calculated simultaneously. Compare the Doppler speed threshold values of both methods, and if both are larger than the threshold value, the Doppler speed by the PPP method that does not suffer the deterioration of the frequency resolution due to the influence of the finite term truncation error is adopted. On the other hand, when any one of the Doppler velocities by both methods falls below the threshold value, a Doppler velocity calculation result by a method with a small calculated velocity width is output. This method provides more accurate Doppler velocity observation data.
[Selection] Figure 1

Description

  The present invention intermittently emits a pulse wave into a space at regular intervals, receives a reflected wave from an object that affects the weather, such as a cloud accompanying rain, and measures the frequency of the reflected wave, thereby measuring a target Doppler. It is an invention related to a weather radar device for calculating speed.

  Methods for estimating a target Doppler velocity to be detected by a weather radar apparatus are roughly classified into an FFT method (Fast Fourier Transform) in which estimation processing is performed in the frequency domain and a PPP method (Pulse Pair Processing) in the time domain. . (Non-Patent Document 1)

  The FFT method aggregates N signal sequences of reflected waves from rain or the like obtained by pulses emitted into space at a constant interval Ts [seconds], and Fourier-transforms this to obtain a first moment of a periodogram, that is, The Doppler velocity of the reflected wave is obtained from the most dominant frequency component, and the spectral width centered on the Doppler velocity is output as the velocity width as an index representing the accuracy of the Doppler velocity.

  The PPP method aggregates N reflected waves from meteorological echoes by pulses emitted into the space at a constant interval Ts [seconds], and declinates the cross-correlation coefficient of adjacent signal sequences, that is, N × Ts [seconds]. The Doppler velocity of the meteorological echo is output from the average phase rotation amount between pulses in, and the velocity width is output as an index representing the accuracy of the Doppler velocity calculated from the standard deviation of the instantaneous phase rotation amount.

Since both the FFT method and the PPP method are signal processing that handles discrete signals obtained by sampling the space with a pulse emission period Ts [seconds], the wind measurement range of the Doppler velocity is limited by the sampling theorem. The wind measurement range Vmax of the Doppler velocity is given by (Equation 1), where λ [m] is the carrier wavelength of a pulse emitted into the space.
(Equation 1)
Vmax = −λ × Ts / 4

Edited by Masahito Ishihara "Meteorological Research Note No.200" Japan Meteorological Society December 26, 2001 First edition issued p24-26 Edited by Masahito Ishihara "Meteorological Research Note No.200" Japan Meteorological Society December 26, 2001 First edition issued p30-32

  The FFT method multiplies the time domain signal before processing by a window function in order to suppress the occurrence of side lobes due to the finite length truncation error of the Fourier transform processing. The multiplication of the window function is effective for suppressing the side lobes, but causes degradation of the frequency resolution of the periodogram obtained as a result of the FFT processing, that is, deterioration of the Doppler velocity calculation accuracy.

  However, since the FFT method has frequency selectivity, when a reflected wave from a fixed target and a reflected wave from a weather echo such as rainfall are superimposed, the former has a main component in DC and the latter has a main component in AC. Thus, only the AC component, that is, the weather echo can be extracted.

  On the other hand, since the PPP method does not go through the window function processing, there is no deterioration in the frequency resolution, that is, the Doppler velocity calculation accuracy in the process of the processing seen in the FFT method. However, because it is time domain processing, it does not have frequency selectivity, and when a reflected wave from a fixed target and a reflected wave from a weather echo such as rainfall are superimposed, a fixed target having a large reflection cross-section to the weather echo The Doppler velocity calculation result shifts to the direct current side, and the velocity width is expanded to deteriorate the Doppler velocity calculation accuracy.

  Therefore, the FFT method and the PPP method, which are Doppler velocity calculation methods for weather radar, have contradictory properties.

  It is an object of the present invention to provide Doppler velocity observation data with higher accuracy by combining both methods having contradictory properties.

  The Doppler speed and the speed width are calculated by both the FFT method and the PPP method. It is assumed that the Doppler velocity and velocity width calculation results by the FFT method have undergone a DC component removal process based on frequency selectivity. Comparing the Doppler velocity calculation results of both and the threshold value given in advance, if both have Doppler velocities greater than or equal to the threshold, it is determined that there is no reflected wave superposition from the fixed target, and the velocity width is the smaller of the two That is, a highly accurate Doppler velocity is output as a calculation result.

  On the other hand, if one of the values is smaller than the threshold value, it is assumed that there is a possibility of reflection wave reflection from the fixed target, and the Doppler velocity calculation by the FFT method is performed through the DC component removal processing based on the frequency selectivity. Output the result.

  The threshold used to select the system of the Doppler velocity calculation results by the FFT method and the PPP method is that the observation data is acquired in an environment where there is no meteorological echo such as rain or rain, and only a fixed target is observed, and the Doppler velocity histogram The test is performed, and the determination is made based on the mode value and variance σ of the same statistics.

  By the above method, Doppler velocity calculation results by the FFT method and the PPP method having conflicting properties are dynamically selected, and highly accurate Doppler velocity observation data is output over the entire observation region.

It is a block diagram which shows the structure of the radar apparatus of this invention. It is a conceptual diagram which shows the statistic which determines the threshold value used for the Doppler velocity calculation result selection by the FFT method and the PPP method in the radar apparatus of this invention. It is a flowchart of Doppler speed calculation result selection by FFT method and PPP method in the embodiment of the present invention.

  The most preferred embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a radar apparatus according to the present invention.

  The transmission timing timing instructing unit 1 in FIG. 1 measures the timing at which a pulse is emitted into space, and generates a transmission trigger that indicates the timing.

  The transmission unit 2 in FIG. 1 generates a pulse wave for observing the target according to the transmission trigger instruction generated by the transmission timing timing instruction unit 1, converts the frequency to a frequency to be emitted into space, and amplifies the power.

  The antenna unit 3 of FIG. 1 emits a pulse wave that has been frequency-converted by the transmission unit 2 and amplified in power, and receives a reflected wave from the target.

  The receiving unit 4 in FIG. 1 frequency-converts the reflected wave from the target received by the antenna unit 3 to an intermediate frequency and amplifies the power.

  The A / D converter 5 in FIG. 1 samples / quantizes the reflected wave from the target converted to the intermediate frequency by the receiver 4.

  The digital quadrature detection unit 6 in FIG. 1 performs quadrature detection of the reflected wave from the target sampled / quantized by the A / D conversion unit.

  The memory unit 7 in FIG. 1 temporarily stores the quadrature detection signal output from the digital quadrature detection unit 6.

  The write address counter unit 8 in FIG. 1 sequentially instructs addresses for writing the quadrature detection signal output by the digital quadrature detection unit in the memory unit 7 using the transmission trigger output from the transmission timing timing instruction unit 1 as the count start timing. To do.

  1 counts the transmission trigger output from the transmission timing counter 1.

  The process start timing instruction unit 10 in FIG. 1 counts the number of transmissions instructed by the transmission number count instruction unit 9, and instructs the process start timing every N transmission times.

  The read address counter unit 11 in FIG. 1 instructs an address for sequentially reading quadrature detection signals from the memory unit 7 with the process start timing instructed by the process start timing instruction unit 10 as a count start timing.

  The FFT processing unit 12 in FIG. 1 performs FFT processing on the quadrature detection signal originating from the N pulse waves read from the memory unit 7 according to the processing start timing instructed by the processing start timing instruction unit 10. Output a periodogram.

  The DC component removing unit 13 in FIG. 1 removes a DC component from the periodogram output from the FFT processing unit 12. The method for removing the DC component is according to Non-Patent Document 2.

  The Doppler velocity calculation unit (FFT) 14 in FIG. 1 calculates the Doppler velocity from the periodogram output from the DC component removal unit 13.

  The speed width calculation unit (FFT) 15 in FIG. 1 calculates the speed width from the periodogram output from the DC component removal unit 13.

  The PPP processing unit 16 in FIG. 1 determines the cross-correlation coefficient of the adjacent signal sequence from the N orthogonal detection signals read from the memory unit 7 according to the processing start timing indicated by the processing start timing instruction unit 10. Calculate sequentially.

  The Doppler velocity calculation unit (PPP) 17 in FIG. 1 calculates the Doppler velocity based on the deviation angle of the cross-correlation coefficient output from the PPP processing unit 16.

  A Doppler velocity width calculator (PPP) 18 in FIG. 1 calculates a velocity width based on the standard deviation of the cross-correlation coefficient output from the PPP processor 16.

  The threshold value holding unit 19 in FIG. 1 holds a threshold value used for selecting a Doppler velocity calculation result by the FFT method and the PPP method.

  The threshold only needs to have the highest Doppler speed at which the fixed target is statistically distributed. (See Figure 2)

For example, the threshold value is obtained in an environment where there is no meteorological echo such as rain, rain, and only a fixed target is observed, and a Doppler velocity histogram test is performed. To (Formula 2).
(Equation 2)
Threshold = mode value + 3 × σ
The constant in (Equation 2) is arbitrary and is not limited to 3.

  In the above-described embodiment, the means for calculating the threshold is automated, and manual statistical calculation does not have to be performed.

  The Doppler speed and speed width determination unit 20 in FIG. 1 dynamically selects the Doppler speed calculation result by the FFT method and the PPP method according to the flowchart shown in FIG.

  In S1 and S2 of FIG. 3, the threshold value held by the threshold value holding unit 19 of FIG. 1 is compared with the absolute value of the Doppler velocity calculation results by the FFT method and the PPP method. If the absolute values of the Doppler velocity calculation results by the FFT method and the PPP method are both larger than the threshold value, the fixed target is not superimposed, and the smaller one of the velocity widths, that is, the Doppler velocity with high accuracy is calculated. Output as a result. (S3, S4)

  On the other hand, if one of the absolute values of the Doppler velocity calculation results by the FFT method and the PPP method is smaller than the threshold value held by the threshold value holding unit in FIG. As a result, the Doppler velocity calculation result by the FFT method is adopted. (S5)

  With the above operation, the Doppler velocity calculation result by the FFT method and the PPP method having contradictory properties is dynamically selected, and the object such as clouds or the like that affects the weather can be accurately measured over the entire observation area. Output high Doppler velocity observation data.

DESCRIPTION OF SYMBOLS 1 Transmission timing timing instruction part 2 Transmission part timing timing instruction part 3 Aerial part 4 Reception part 5 A / D conversion part 6 Quadrature detection part 7 Memory part 8 Write address counter part 9 Transmission frequency count instruction part 10 Process start timing instruction part 11 Read address counter unit 12 FFT processing unit 13 DC component removal unit 14 Doppler velocity calculation unit (FFT)
15 Speed range calculator (FFT)
16 PPP processing unit 17 Doppler velocity calculation unit (PPP)
18 Speed range calculator (PPP)
19 Threshold value holding unit 20 Doppler speed and speed width determining unit

Claims (3)

  In a meteorological radar device that analyzes a reflected wave from an object that affects the weather and calculates a Doppler velocity, a Doppler velocity 1 and a velocity width 1 from a periodogram obtained by fast Fourier transform of the input time-domain quadrature detection signal A first processing unit that calculates the Doppler velocity 2 and a velocity width 2 from the cross-correlation coefficient of the time-domain quadrature detection signal, the first processing unit and the second processing unit. A Doppler speed and a speed width determination unit that selects a Doppler speed and a speed width from the calculation result of the processing unit. The Doppler speed and the speed width determination unit are variously set to predetermined threshold values of the Doppler speed 1 and the Doppler speed 2. A weather radar apparatus characterized by selecting both in response.   The Doppler speed and speed width determination unit according to claim 1 compares the speed width 1 and the speed width 2 when both of the Doppler speed 1 and the Doppler speed 2 exceed the threshold. The Doppler speed with the smaller speed range is selected, and if either one of the Doppler speed 1 and the Doppler speed 2 or both are smaller than the threshold value, the Doppler speed 1 is selected. Weather radar device. 2. The meteorological radar apparatus according to claim 1, wherein a threshold value that contributes to selection of the Doppler velocities 1 and 2 is determined from a mode value and a variance of a Doppler velocity histogram test in data in which only a fixed target in clear weather is observed. A featured weather radar device.