JP6498425B2 - measuring device - Google Patents

Description

  The present invention relates to a measuring apparatus that calculates a frequency spectrum for each AC signal by performing Fourier transform on instantaneous values obtained by sampling a plurality of AC signals.

  As this type of measuring apparatus, a measuring apparatus disclosed by the applicant of the present application in Patent Document 1 below is known. In this measuring apparatus, a plurality of alternating current signals having the same frequency are sampled so that the number of instantaneous values for each period is the same, and the instantaneous values obtained by sampling are subjected to Fourier transform, thereby obtaining each alternating current signal. The frequency spectrum of the signal is calculated.

  Incidentally, there is a demand for calculating the frequency spectrum of each AC signal even when the frequencies of the plurality of AC signals are not all the same and the frequencies of some AC signals are different from the frequencies of other AC signals. In this case, by applying the configuration of the measurement apparatus disclosed in Patent Document 1 below, sampling is performed so that the number of instantaneous values for each cycle is the same for a plurality of AC signals having different frequencies. Then, it is conceivable that the frequency spectrum of each AC signal is calculated by Fourier transforming the instantaneous value obtained by sampling.

JP 2012-154664 A (page 6, FIG. 1)

  However, the AC signals having different frequencies are sampled so that the number of instantaneous values for one period is the same (predetermined number), and the obtained instantaneous values are Fourier transformed. A measuring apparatus that employs a configuration for calculating the frequency spectrum of an AC signal has the following problems to be improved. That is, in this type of measuring apparatus, within one cycle, processing for obtaining one cycle of the instantaneous value of the AC signal within this cycle, and the instantaneous value of the AC signal within this cycle acquired during the immediately preceding cycle In parallel, a process of calculating a frequency spectrum by Fourier transforming the above is executed.

  Therefore, this measuring apparatus shortens one cycle for an AC signal having a high frequency, and therefore, from the start of execution of Fourier transform to a predetermined number of instantaneous values until the completion of output of the calculated frequency spectrum. There is a problem to be improved that the processing may not be completed within one cycle. Further, in this measuring apparatus, since an AC signal having a low frequency is long for one cycle, the process from the start of the Fourier transform to the completion of the output of the frequency spectrum is completed at an early stage within one cycle. There is also a problem to be improved that the time from the completion of the output of the frequency spectrum to the end of the one cycle becomes a dead time.

  The present invention has been made to improve such a problem, and can reliably calculate and output a frequency spectrum regardless of the frequency of an input AC signal, and can avoid generation of dead time. The main purpose is to provide a measuring device.

In order to achieve the above object, the measuring apparatus according to claim 1 is configured such that a plurality of AC signals are input and sampled so that the instantaneous value of each of the AC signals is a power of 2. A plurality of AC signals obtained by Fourier transforming the instantaneous values of the plurality of AC signals output from the sampling unit, which are obtained and output so that the number corresponds to each frequency. A processing unit that performs a spectrum calculation process for calculating a frequency spectrum of the AC signal, and the sampling unit calculates the number of the one period as the frequency range including the frequency of the AC signal becomes a lower frequency range. a measuring apparatus for increasing the sampling unit, support a plurality of alternating signals at a common sampling clock A sampling process for obtaining the instantaneous values by pulling, a frequency detection process for detecting the frequencies of the plurality of AC signals based on the determined instantaneous values, and the AC signals based on the detected frequencies The range identification process for identifying the frequency range in which the frequency is included, and the instantaneous values for the plurality of AC signals obtained by the sampling process correspond to the frequency range identified by the range identification process. A thinning process for thinning out the number corresponding to the period is executed .

According to a second aspect of the present invention, the plurality of AC signals are input and sampled, and the instantaneous values of the plurality of AC signals are a power of 2 for one cycle , and the measurement apparatus according to each frequency. The frequency spectrum for the plurality of AC signals is calculated by Fourier-transforming the instantaneous values for the plurality of AC signals output from the sampling unit and the sampling units that are acquired and output so as to be the same number A processing unit that executes a spectrum calculation process, wherein the sampling unit is a measuring device that increases the number of the one period as the frequency range including the frequency of the AC signal becomes a lower frequency range. Te, the processing unit, in the spectrum calculation processing includes said frequency to a higher the frequency ranges Serial preferentially calculation of the frequency spectrum of the AC signal to perform.

The measuring apparatus according to claim 3 inputs and samples a plurality of AC signals, so that each instantaneous value of each of the plurality of AC signals is a power of 2 for one cycle and corresponds to each frequency. The frequency spectrum for the plurality of AC signals is calculated by Fourier-transforming the instantaneous values for the plurality of AC signals output from the sampling unit and the sampling units that are acquired and output so as to be the same number A processing unit that executes a spectrum calculation process, wherein the sampling unit is a measuring device that increases the number of the one period as the frequency range including the frequency of the AC signal becomes a lower frequency range. Te, wherein the plurality of alternating signals are grouped into one or more groups comprising a plurality of said AC signal, Serial processor, when calculating the frequency spectrum for all of the AC signal included in the group has been completed, perform operations that combine the frequency spectrum more.

  In the measuring apparatus according to claim 1, the sampling unit increases the number of instantaneous values for one cycle as the frequency range including the frequency of the input AC signal becomes a lower frequency range.

  Therefore, according to this measuring apparatus, since the number of instantaneous values for one cycle is small for an AC signal (a signal having a short cycle) whose frequency is included in a high frequency range, the FFT for this number of instantaneous values is performed. For an AC signal (a signal with a long cycle) that can be reliably completed within one cycle and the frequency is included in a low frequency range, the number of instantaneous values for one cycle increases, so there is no wasted time. While avoiding the occurrence, it is possible to reliably complete the FFT with higher accuracy within one cycle. That is, according to this measuring apparatus, it is possible to reliably execute calculation and output of the frequency spectrum regardless of the frequency of the input AC signal, and to reliably avoid generation of dead time within one cycle. be able to.

Further, in the measuring apparatus This sampling section detects a sampling process for obtaining the respective instantaneous values by sampling the plurality of alternating signals at a common sampling clock, the frequency of each ac signal based on the instantaneous value determined Frequency detection processing, range specification processing for specifying a frequency range including the frequency of each AC signal based on the detected frequency, and instantaneous value for each AC signal obtained by sampling processing are specified by range specification processing Thinning-out processing is performed for thinning out the number of one period corresponding to the frequency range. Therefore, according to this measuring apparatus, the instantaneous value for each AC signal can be automatically thinned to the number corresponding to the frequency range including the frequency of this AC signal.

In the measuring apparatus according to claim 2 , the sampling unit increases the number of instantaneous values for one cycle as the frequency range including the frequency of the input AC signal becomes a lower frequency range. Therefore, according to this measuring apparatus, since the number of instantaneous values for one cycle is small for an AC signal (a signal having a short cycle) whose frequency is included in a high frequency range, the FFT for this number of instantaneous values is performed. For an AC signal (a signal with a long cycle) that can be reliably completed within one cycle and the frequency is included in a low frequency range, the number of instantaneous values for one cycle increases, so there is no wasted time. While avoiding the occurrence, it is possible to reliably complete the FFT with higher accuracy within one cycle. That is, according to this measuring apparatus, it is possible to reliably execute calculation and output of the frequency spectrum regardless of the frequency of the input AC signal, and to reliably avoid generation of dead time within one cycle. be able to.
In this measurement apparatus, the processing unit preferentially executes the calculation of the frequency spectrum for the AC signal whose frequency is included in the higher frequency range in the spectrum calculation process. Therefore, according to this measuring apparatus, since the number of instantaneous values is large, it takes a long time to complete the FFT, and the calculation of the low frequency AC signal with the low frequency results in the calculation of the AC signal with a small number of instantaneous values. The occurrence of a situation in which the FFT calculation of the signal) is awaited can be prevented, so that it is preferable to calculate the frequency spectrum in a shorter time interval without delaying the FFT calculation for the high frequency AC signal. Can be executed.

In the measuring apparatus according to claim 3 , the sampling unit increases the number of instantaneous values for one cycle as the frequency range including the frequency of the input AC signal becomes a lower frequency range. Therefore, according to this measuring apparatus, since the number of instantaneous values for one cycle is small for an AC signal (a signal having a short cycle) whose frequency is included in a high frequency range, the FFT for this number of instantaneous values is performed. For an AC signal (a signal with a long cycle) that can be reliably completed within one cycle and the frequency is included in a low frequency range, the number of instantaneous values for one cycle increases, so there is no wasted time. While avoiding the occurrence, it is possible to reliably complete the FFT with higher accuracy within one cycle. That is, according to this measuring apparatus, it is possible to reliably execute calculation and output of the frequency spectrum regardless of the frequency of the input AC signal, and to reliably avoid generation of dead time within one cycle. be able to.
Moreover, in this measuring apparatus, when the calculation of the frequency spectrum about all the alternating current signals contained in a group is completed, the calculation which combined two or more of these frequency spectra is performed. Therefore, according to this measuring apparatus, while the plurality of AC signals input to the sampling unit include, for example, three phase voltage signals for one three-phase circuit, these phase voltage signals are Even when irrelevant alternating current signals are included, by grouping the three phase voltage signals as the alternating current signals into the same group, in the state where the frequency spectrum for each phase voltage signal is aligned, The calculation combining the frequency spectrum can be executed reliably.

1 is a configuration diagram of a measuring device 1. FIG. FIG. 6 is a state transition diagram for explaining the operation of the measuring apparatus 1. FIG. 5 is a procedure diagram for processing for each input channel CH for explaining the operation of the measuring apparatus 1. FIG. 6 is another procedure diagram for processing for each input channel CH for explaining the operation of the measuring apparatus 1.

  Hereinafter, embodiments of a measuring apparatus will be described with reference to the accompanying drawings.

  First, the configuration of the measuring apparatus 1 will be described with reference to the drawings.

  1 includes a sampling unit 2, a processing unit 3, an operation unit 4, a storage unit 5, and an output unit 6 as an example, and a plurality of alternating current signals input to the sampling unit 2 ( In this example, as an example, the frequency spectrums FSa, Fsb, Fsc, FSd (hereinafter not particularly distinguished) of four AC signals S1a, S1b, S1c, and S1d (hereinafter also referred to as “AC signal S1” unless otherwise distinguished). The “frequency spectrum FS” is sometimes calculated by executing a Fourier transform (specifically, a fast Fourier transform (FFT)) and output from the output unit 6.

  As an example, the sampling unit 2 includes a plurality of input channels CH so that a plurality of AC signals S1 can be simultaneously input and sampled. In this example, four input channels CHa, CHb, CHc, and CHd (hereinafter also referred to as “input channels CH” unless otherwise specified) are provided so that the four AC signals S1a to S1d can be input simultaneously. .

  In this example, the number of input channels CH is four as an example. However, if there are a plurality of input channels CH, two, three, or even five or more can be used, and as in this example, The AC signal S1 may be input to all the input channels CH, or the AC signal S1 may be input only to some of the plurality of input channels CH.

  Hereinafter, the configuration of each input channel CH will be described. As shown in FIG. 1, the input channels CHa, CHb, CHc, and CHd are input terminals 11a, 11b, 11c, and 11d (hereinafter also referred to as “input terminals 11” unless otherwise specified), and an A / D converter 12a. , 12b, 12c, 12d (hereinafter also referred to as “A / D converter 12” unless otherwise specified), and frequency detectors 13a, 13b, 13c, 13d (hereinafter also referred to as “frequency detector 13” unless otherwise specified). ) And thinning portions 14a, 14b, 14c, 14d (hereinafter also referred to as “thinning portion 14” unless otherwise specified), and are configured identically. Therefore, in the following, the configuration will be described more specifically by taking the input channel CHa among the input channels CHa to CHd as an example.

  One corresponding AC signal S1a among the plurality of AC signals S1 is input to the input terminal 11a. The A / D conversion unit 12a samples the AC signal S1a input to the input terminal 11a in synchronization with the sampling clock CLK common to the other A / D conversion units 12b to 12d, so that the instantaneous value of the AC signal S1a is obtained. The sampling process for outputting the instantaneous value data D1a indicating is executed.

  The frequency detector 13a detects the number of waveforms (or zero cross points) of the AC signal S1a per unit time defined in advance based on the instantaneous value data D1a, and also detects the number of detected waveforms (or zero cross points) and the unit. Based on the time, a frequency detection process for detecting the frequency fa of the AC signal S1a is executed. Since detecting the period is equivalent to detecting the frequency, detecting the frequency in this example includes detecting the period. Further, the frequency detection unit 13a outputs frequency data D2a indicating the detected frequency fa to the thinning unit 14a at a cycle of the frequency fa.

  The thinning-out unit 14a first acquires the frequency data D2a output from the frequency detection unit 13a, and the frequency fa indicated by the frequency data D2a is in any one of a plurality of predefined frequency ranges. A range specifying process for specifying whether or not it is included is executed. Next, the thinning-out unit 14a specifies the sampling number Na (the number of instantaneous values acquired per cycle of the AC signal S1a. Hereinafter, also referred to as the number Na) corresponding to the specified frequency range. Subsequently, the thinning-out unit 14a obtains the instantaneous value data D1a for one cycle of the AC signal S1a by acquiring the instantaneous value data D1a output from the A / D conversion unit 12a at the cycle of the sampling clock CLK. Execute the thinning process to acquire each. Finally, the thinning-out unit 14a uses the acquired Na pieces of instantaneous value data D1a as the acquired data group D3a, together with the frequency data D2a of the AC signal S1a, in the period of the AC signal S1a (that is, Na pieces of instantaneous data for one period). Whenever the acquisition of the value data D1a is completed, it is output to the processing unit 3. The frequency data D2a may be configured to be output to the processing unit 3 by the frequency detection unit 13a instead of the thinning unit 14a.

  The thinning portion 14a will be described more specifically with an example.

  In order to calculate the frequency spectrum FSa of the AC signal S1a using FFT, the number Na of instantaneous value data D1a for one cycle of the AC signal S1a is preferably a power of two. Also, the time required for FFT becomes longer as the number Na of instantaneous value data D1a increases. For this reason, the processing unit 3 calculates the frequency spectrum FSa within the next cycle of the one cycle based on the instantaneous value data D1a for one cycle of the AC signal S1a acquired from the thinning unit 14a, and the output unit. In order to complete the output to 6, the number Na needs to be reduced in the AC signal S1a having the frequency fa included in the high frequency range.

  On the other hand, when the AC signal S1a has the frequency fa included in the low frequency range, the number Na of the instantaneous value data D1a is the same as that in the high frequency range. Based on the instantaneous value data D1a for one period of the AC signal S1a acquired from 14a, when the calculation of the frequency spectrum FSa is started from the beginning of the period next to the one period, the calculation of the frequency spectrum FSa and There is a lot of time from the completion of the output to the end of the cycle, which becomes a dead time. For this reason, when the AC signal S1a is in the frequency fa included in the low frequency range, the number Na of instantaneous value data D1a for one cycle used in the FFT is increased to calculate the frequency spectrum FSa with higher accuracy. Is preferred.

  For the above reason, in the measuring apparatus 1 of this example, as an example, the frequency range from 0 Hz to 2 kHz is set as the measurement frequency range, and this measurement frequency range is set to the first frequency in the frequency range from 0 Hz to 100 Hz. Range (low frequency range), second frequency range (medium frequency range) over 100 Hz to 500 Hz, and third frequency range (high frequency range) over 500 Hz to 2 kHz The number Na is 8192 (2 to the 13th power) in the first frequency range, the number Na is 2048 (2 to the 11th power) in the second frequency range, and the number Na is in the third frequency range. Is defined as 512 (2 to the 9th power).

  Therefore, in this example, the thinning-out unit 14a first acquires the frequency data D2a output from the frequency detection unit 13a, and the frequency fa indicated by the frequency data D2a is the first frequency range, the second frequency range, and Which frequency range of the third frequency range is included is specified. For example, when the frequency fa is 200 Hz, the second frequency range is specified. Next, the thinning unit 14a specifies the number Na corresponding to the specified frequency range. For example, when the second frequency range is specified, the number Na is specified as 2048.

  Subsequently, the thinning-out unit 14a equally thins out and inputs the instantaneous value data D1a output at the cycle of the sampling clock CLK from the A / D conversion unit 12a, so that the instantaneous value data D1a for one cycle of the AC signal S1a. Acquire Na. For example, when the frequency fa of the AC signal S1a is 200 Hz, 2048 pieces of instantaneous value data D1a for one cycle of the AC signal S1a are evenly thinned out and acquired. Finally, the thinning unit 14a outputs the acquired Na pieces of instantaneous value data D1a to the processing unit 3 together with the frequency data D2a as an acquired data group D3a.

  At this time, the thinning-out unit 14a may acquire Na pieces of instantaneous value data D1a obtained by thinning out the instantaneous value data D1a output from the A / D conversion unit 12a as it is, or perform A / D conversion. Using the instantaneous value data D1a output from the unit 12a, a pseudo waveform between the instantaneous value data D1a is obtained by interpolation processing, for example, from the rising zero cross point of the AC signal S1a to the next rising zero cross point. The Na instantaneous value data obtained from this pseudo waveform can be used as the acquired data group D3a.

  In the other input channels CHb, CHc, and CHd, the corresponding A / D conversion unit 12, the frequency detection unit 13, and the decimation unit 14 operate in the same manner as the above-described input channel CHa, so that each decimation unit 14b, Also from 14c and 14d, the acquired data composed of the number Nb, Nc and Nd of instantaneous value data D1b, D1c and D1d corresponding to the frequency range including the frequencies fb, fc and fd of the AC signals S1b, S1c and S1d. The groups D3b, D3c, and D3d are output to the processing unit 3 in the cycle of the AC signals S1b, S1c, and S1d together with the frequency data D2b, D2c, and D2d indicating the frequencies fb, fc, and fd.

  Note that the frequencies fa, fb, fc, and fd are also referred to as “frequency f” unless otherwise distinguished, and the numbers Na, Nb, Nc, and Nd are also referred to as “number N” unless otherwise distinguished. The instantaneous value data D1a, D1b, D1c, and D1d are also referred to as “instantaneous value data D1” when they are not particularly distinguished. It is also referred to as a group D3, and the frequencies fa, fb, fc, and fd are also referred to as “frequency f” when they are not particularly distinguished, and the frequency data D2a, D2b, D2c, and D2d are also referred to as “ Also referred to as “frequency data D2”.

  The processing unit 3 is configured using a computer, for example, and executes a spectrum calculation process for calculating a frequency spectrum FS for each AC signal S1 based on the acquired data group D3 output from the sampling unit 2. The operation unit 4 is provided with operation keys (not shown) for associating (grouping) an arbitrary plurality of input channels CH. When the operation key is operated to associate (group) a plurality of input channels CH, the operation unit 4 outputs association data Dre indicating the associated input channels CH to the processing unit 3. To do.

  The storage unit 5 is configured by a semiconductor memory, a hard disk device, or the like, and an operation program for the processing unit 3 is stored in the storage unit 5 in advance. Further, the storage unit 5 stores the association data Dre, the frequency data D2, the acquired data group D3, and the frequency spectrum FS for each AC signal S1 by the processing unit 3. For example, the output unit 6 includes a display device and displays (outputs) the frequency spectrum FS output from the processing unit 3 on the screen. The output unit 6 can also be configured by an external interface circuit instead of the display device. When this configuration is adopted, the output unit 6 inputs the frequency spectrum FS output from the processing unit 3 and is connected to the external interface circuit. This frequency spectrum FS is output to an external device.

  Next, the operation of the measuring apparatus 1 will be described.

  First, an operation when the operation key of the operation unit 4 is not performed and none of the input channels CH is associated will be described.

  In the state where AC signals S1a, S1b, S1c, and S1d are input to the input terminals 11a, 11b, 11c, and 11d of the input channels CHa, CHb, CHc, and CHd, the sampling unit 2 configures each input channel CH. As described above, the A / D conversion unit 12, the frequency detection unit 13, and the thinning-out unit 14 output the instantaneous value data D1 (sampling process) and the instantaneous value of each AC signal S1 synchronized with the common sampling clock CLK. Detection of the frequency f of each AC signal S1 based on the data D1 (frequency detection processing), specification of a frequency range including the frequency f of each AC signal S1 (range specification processing), and each AC corresponding to the specified frequency range The acquisition (decimation processing) of the instantaneous value data D1 of the number N for one cycle for the signal S1 is executed, Thinning portion 14 forces the channel CH is, the acquired data group D3 and frequency data D2 for the corresponding alternating signal S1 and outputs to the processing unit 3 at a period of the AC signal S1.

  The processing unit 3 executes a spectrum calculation process 50 shown in FIG. In the spectrum calculation process 50, the processing unit 3 first shifts to a waveform data waiting state (ST1), and then waits for a waveform data waiting state (ST1), a spectrum calculation assignment determination state in accordance with a transition condition to be described later. (ST2), transition to one of the first spectrum calculation state (ST3), the second spectrum calculation state (ST4), and the third spectrum calculation state (ST5), and execute a predefined operation for each state Thus, the frequency spectrum FS for each AC signal S1 is calculated and output from the output unit 6.

  In this case, in the waveform data waiting state (ST1), the processing unit 3 determines whether or not the frequency data D2 and the acquired data group D3 (hereinafter collectively referred to as waveform data) are output from the thinning unit 14 of each input channel CH. When it is detected that the frequency data D2 and the acquired data group D3 are output from the thinning unit 14 of any input channel CH, these are input and stored in the storage unit 5, and then the spectrum calculation is performed. The process proceeds to the assignment determination state (ST2).

  In the spectrum calculation assignment determination state (ST2), the processing unit 3 stores one or two or more acquired data groups D3 stored in the storage unit 5 (the acquired data group D3 in which the spectrum calculation process is interrupted and the spectrum calculation process). The number N or the frequency f (number N as an example in this example) is detected for both of the acquired data groups D3 that have not been processed (these acquired data groups D3 will be described later).

  As a result of this detection, the processing unit 3 determines that the acquisition data group D3 having the smallest number N of instantaneous value data D1 among the acquisition data group D3 stored in the storage unit 5 has the smallest number N (512 in this example). When it is detected that the acquired data group D3 is composed of the instantaneous value data D1 (that is, the acquired data group D3 for the AC signal S1 included in the third frequency range having the highest frequency f), It is determined that the calculation resource is assigned to the acquired data group D3 having the number N of 512 (that is, FFT is performed on the acquired data group D3 having the number N of 512), and the third spectrum calculation state (ST5) is entered. Transition.

  Further, the processing unit 3 has the smallest number N (512 in this example) of the acquisition data group D3 having the smallest number N of instantaneous value data D1 among the acquisition data group D3 stored in the storage unit 5. Next, an acquisition data group D3 composed of the next smallest number N (2048 in this example) of instantaneous value data D1 (that is, acquisition data group D3 for the AC signal S1 whose frequency f is included in the second frequency range in this example). ), It is determined that the FFT calculation resource is allocated to the acquisition data group D3 having the number N of 2048 (that is, the FFT is performed on the acquisition data group D3 having the number N of 2048). Then, the process proceeds to the second spectrum calculation state (ST4).

  Further, the processing unit 3 has the smallest number N (512 in this example) of the acquisition data group D3 having the smallest number N of instantaneous value data D1 among the acquisition data group D3 stored in the storage unit 5. Next, the acquired data group D3 composed of the next smallest number N (in this example, 8192 which has the largest number N) of instantaneous value data D1 (that is, in the first frequency range in which the frequency f is the lowest in this example). When it is detected that the acquired AC signal S1 is the acquired data group D3), the FFT calculation resource is allocated to the acquired data group D3 whose number N is 8192 (that is, the acquired data whose number N is 8192). It is determined that FFT for the group D3 is performed), and the process proceeds to the first spectrum calculation state (ST3).

  In the first spectrum calculation state (ST3), the second spectrum calculation state (ST4), and the third spectrum calculation state (ST5), the processing unit 3 performs FFT on the number N of acquired data groups D3 corresponding to each spectrum calculation state. By executing, the frequency spectrum FS is calculated for the AC signal S1 indicated by the instantaneous value data D1 constituting the acquired data group D3, and is stored in the storage unit 5 and output from the output unit 6. Further, when the calculation, storage, and output of the frequency spectrum FS are completed in this way, the processing unit 3 shifts from each spectrum calculation state (ST3, ST4, ST5) to the spectrum calculation assignment determination state (ST2).

  In each spectrum calculation state, the processing unit 3 includes both the acquired data group D3 for which the spectrum calculation process is interrupted and the acquired data group D3 that has the same number as the acquired data group D3 and has not executed the spectrum calculation process. When stored in the storage unit 5, priority is given to the spectrum calculation process for the acquired data group D3 in which the spectrum calculation process is interrupted. The processing unit 3 stores both the acquired data group D3 in which the spectrum calculation process is interrupted and the acquired data group D3 that has the same number as the acquired data group D3 and in which the spectrum calculation process is interrupted in the storage unit 5. When it is determined, priority is given to the spectrum calculation process for the acquired data group D3 for which the spectrum calculation process has been interrupted first. In addition, the processing unit 3 stores both the acquired data group D3 in which the spectrum calculation process is not executed and the same number of the acquired data group D3 and the acquired data group D3 in which the spectrum calculation process is not executed are stored in the storage unit 5 Is prioritized, the spectrum calculation process for the acquired data group D3 stored earlier is prioritized.

  In addition, the processing unit 3 performs the operation of each input channel CH in any of the first spectrum calculation state (ST3) and the second spectrum calculation state (ST4) (when the frequency spectrum FS is being calculated). The presence / absence of output of waveform data (frequency data D2 and acquired data group D3) from the thinning unit 14 is detected. When the output of the waveform data is detected, the waveform data is input and stored in the storage unit 5. . Further, the processing unit 3 determines whether or not the number N of acquisition data groups D3 constituting the waveform data is smaller than the number N of acquisition data groups D3 for which the frequency spectrum FS is being calculated. The calculation of the frequency spectrum FS is interrupted, and the state shifts to the spectrum calculation assignment determination state (ST2). When the frequency spectrum FS is not small (same or large), the current spectrum calculation state is continued ( continue).

  When the processing unit 3 interrupts the calculation of the frequency spectrum FS currently being executed and shifts to the spectrum calculation assignment determination state (ST2), the processing unit 3 operates as described above in the second spectrum calculation state (ST4). Then, it is determined that the FFT calculation resource is allocated to the acquisition data group D3 having the smallest number N of the instantaneous value data D1 in the acquisition data group D3 stored in the storage unit 5, and this number N is supported. The state shifts to the spectrum calculation state (the third spectrum calculation state ST5 when the number is 512, and the second spectrum calculation state ST4 when the number is 2048).

  On the other hand, in the third spectrum calculation state (ST5), the processing unit 3 executes the FFT on the acquired data group D3 composed of 512 pieces of instantaneous value data D1 which is the minimum number N in the first place, and calculates the frequency spectrum FS. It is in the state of calculating. Therefore, the processing unit 3 detects the presence / absence of output of waveform data (frequency data D2 and acquired data group D3) from the thinning unit 14 of each input channel CH, and when detecting the output of the waveform data, This waveform data is inputted and stored in the storage unit 5. Whether or not the number N of acquisition data groups D3 constituting the waveform data is smaller than the number N of acquisition data groups D3 for which the frequency spectrum FS is being calculated. The discrimination is not executed, and the FFT for the currently executed acquisition data group D3 is continued. Therefore, when the processing unit 3 is in the third spectrum calculation state (ST5), the calculation of the currently executed frequency spectrum FS is completed and the storage and output of the calculated frequency spectrum FS are completed. Only, it shifts to the spectrum calculation assignment determination state (ST2).

  Thereby, in this measuring apparatus 1, the processing unit 3 is calculating the frequency spectrum FS for the AC signal S1 indicated by the instantaneous value data D1 of the one acquired data group D3 as a basic operation. When another acquisition data group D3 having a number N smaller than the number N of the one acquisition data group D3 is output from the sampling unit 2 (in this example, when the processing unit 3 stores this in the storage unit 5) ) Interrupts the frequency spectrum FS for the AC signal S1 indicated by the instantaneous value data D1 of this one acquired data group D3, and the AC signal S1 indicated by the instantaneous value data D1 of this other acquired data group D3. The frequency spectrum FS is preferentially executed. When the calculation of the frequency spectrum FS for the AC signal S1 executed with priority is completed, the calculation of the frequency spectrum FS for the interrupted AC signal S1 is resumed.

  With reference to FIG. 3, the basic operation of the processing unit 3 will be described with a specific example. In this example, as an example, the frequency fa of the AC signal S1a input to the input channel CHa is included in the second frequency range (medium frequency range), and the frequency fb of the AC signal S1b input to the input channel CHb is It is assumed that the frequency fc of the AC signal S1c included in the first frequency range (low frequency range) and input to the input channel CHc is included in the third frequency range (high frequency range). Further, in the waveform data waiting state (ST1), the processing unit 3 first completes the storage of the acquired data group D3a for the AC signal S1a in the storage unit 5, and at this time, the storage unit 5 stores other data It is assumed that the acquired data group D3 is not stored.

  In this case, after the storage of the acquired data group D3a in the storage unit 5 is completed, the processing unit 3 shifts to a spectrum calculation assignment determination state (ST2), and the number Na (2048) of the acquired data group D3a for the AC signal S1a. Is determined to be the smallest, and it is determined that FFT operation resources are allocated to the acquired data group D3a. Thereby, the process part 3 transfers to a 2nd spectrum calculation state (ST4), and performs FFT with respect to the acquisition data group D3a. During the execution of the FFT for the acquired data group D3a, as shown in FIG. 3, the processing unit 3 completes the storage of the acquired data group D3b with the number Nb of 8192 for the AC signal S1b in the storage unit 5. However, if the number Nb of the acquired data group D3b is larger than the number Na of the acquired data group D3a that is currently executing FFT, it is determined in the second spectrum calculation state (ST4). For this reason, the processing unit 3 continues to perform the FFT on the acquired data group D3a, calculates the frequency spectrum FSa of the AC signal S1a, and completes the FFT.

  Thereafter, the processing unit 3 shifts from the second spectrum calculation state (ST4) to the spectrum calculation assignment determination state (ST2). In this case, since the storage unit 5 stores only the acquisition data group D3b having the number Nb of 8192, the processing unit 3 has the minimum number Nb (8192) of the acquisition data group D3b for the AC signal S1b. It is determined that the FFT calculation resource is allocated to the acquired data group D3b. Thereby, the process part 3 transfers to a 1st spectrum calculation state (ST3), and performs FFT with respect to the acquisition data group D3b.

  In the example illustrated in FIG. 3, the processing unit 3 completes the storage of the acquired data group D3c having 512 Nc pieces of the AC signal S1c acquired during the execution of the FFT for the acquired data group D3b in the storage unit 5. . In the first spectrum calculation state (ST3), the processing unit 3 determines that the number Nc of the acquired data group D3c is smaller than the number Nb of the acquired data group D3b performing the FFT. For this reason, the processing unit 3 interrupts the execution of the FFT on the acquired data group D3b, and shifts from the first spectrum calculation state (ST3) to the spectrum calculation assignment determination state (ST2). In this case, the processing unit 3 determines that the number Nc (512) is the minimum, and determines that an FFT calculation resource is allocated to the number Nc of acquired data groups D3c. Thereby, the processing unit 3 shifts to the third spectrum calculation state (ST5), calculates the frequency spectrum FSc of the AC signal S1c by executing FFT on the acquired data group D3c, and completes this FFT. .

  Thereafter, the processing unit 3 shifts from the third spectrum calculation state (ST5) to the spectrum calculation assignment determination state (ST2). In this case, the storage unit 5 stores only the acquired data group D3b for the AC signal S1b in which the FFT is interrupted. For this reason, the processing unit 3 determines that the number Nb (8192) of the acquired data group D3b is the smallest, and determines that an FFT computation resource is allocated to the acquired data group D3b of the number Nb. Thereby, the processing unit 3 shifts to the first spectrum calculation state (ST3), restarts the FFT for the acquired data group D3b, calculates the frequency spectrum FSb of the AC signal S1b, and completes this FFT.

  Next, an operation on the operation key of the operation unit 4 is performed, and as an example, the input channel CHa, CHb, CHd of each input channel CH is changed to one of the input channels CHa, CHb, CHd. (In this example, as an example, the operation when the synchronization with the input channel CHb) is established will be described. In the above example, none of the input channels CH is associated, and the frequency spectrum FS calculated for the AC signal S1 input to each input channel CH has been configured to be output alone. In the example, the frequency spectrum FS of each AC signal S1 input to the associated input channels CHa, CHb, and CHd is output as it is, and the result calculated between each other (result of the calculation between channels). Is also output. Since the basic operation is the same as the configuration in which the above association is not performed, description of the same operation is omitted, and different operations are mainly described. Further, as an example of the calculation between each other, a calculation for calculating the total value of the frequency spectrum FS of each AC signal S1 input to the associated input channels CHa, CHb, and CHd will be described as an example.

  In this case, the operation unit 4 outputs association data Dre indicating that the input channels CHa, CHb, and CHd among the input channels CH are synchronized (associated) with the input channel CHb to the processing unit 3. The processing unit 3 stores the association data Dre in the storage unit 5.

  For example, the input terminals 11a, 11b, and 11d of the input channels CHa, CHb, and CHd receive, for example, three phase voltage signals for one three-phase circuit as AC signals S1a, S1b, and S1d. In the state where an AC voltage (voltage having a higher frequency than that of the signal for the three-phase circuit) is input as the AC signal S1c to the input terminal 11c, the AC voltage for the other circuit unrelated to the signal for the three-phase circuit is When performing an operation for calculating the total value of the frequency spectra FSa, FSb, FSd of the AC signals S1a, S1b, S1d, the operator of the measuring apparatus 1 operates the operation unit 4 as described above. The input channels CHa, CHb, and CHd among the input channels CH are synchronized with the input channel CHb (functions). With make) it can be output to the processing section 3 the association data Dre indicating a.

  In this case, since the AC signals S1a, S1b, and S1d input to the input channels CHa, CHb, and CHd have the same frequency, the thinning-out portions 14a, 14b, and 14d of these three input channels CHa, CHb, and CHd, 14d is an acquisition data group D3a, D3b, D3d composed of instantaneous value data D1a, D1b, D1d of the same number Na, Nb, Nd, and the same frequency fa, fb, included in the same frequency range (low frequency range) The frequency data D2a, D2b, D2d indicating fd is output to the processing unit 3. On the other hand, the thinning-out part 14c of the remaining input channels CHc processes the acquired data group D3c composed of the number Nc of instantaneous value data D1c and the frequency data D2c indicating the frequency fc included in the frequency range (for example, high frequency range). Output to part 3.

  The processing unit 3 in this example performs basically the same operation as the above example in which the association (grouping) is not performed for each input channel CH, and the frequency of the AC signal S1 input to each input channel CH. Each spectrum FS is calculated, but when the frequency spectrum FS for all the associated (grouped) input channels CH is completed, a period for executing an operation combining two or more frequency spectra FS is secured. Is different.

  The operation of the processing unit 3 will be described by taking the example shown in FIG. 4 as an example. In the example of the figure, first, each thinning unit 14 of the input channels CHa, CHb, and CHd outputs the acquired data group D3 and the frequency data D2 to the processing unit 3 in this order. As a result, the processing unit 3 The storage of the acquired data group D3 and the frequency data D2 in the storage unit 5 is completed in the order of the channels CHa, CHb, and CHd, and the FFT on the acquired data group D3 is executed in the order of the input channels CHa, CHb, and CHd. On the other hand, the decimation unit 14c of the input channel CHc processes the acquisition data group D3c and the frequency data D2c during the execution of the FFT on the acquisition data group D3b of the input channel CHb after completion of the FFT on the acquisition data group D3a of the input channel CHa. Output to part 3.

  In this case, the processing unit 3 performs FFT on the acquired data group D3a of the input channel CHa that has been initially stored in the storage unit 5, and completes this FFT. Next, execution of FFT is started on the acquired data group D3b of the input channel CHb that has been stored in the storage unit 5 next. Subsequently, during the execution of the FFT for the acquired data group D3b, the processing unit 3 stores the acquired data group D3c of the input channel CHc (the acquired data group D3c having a number Nc smaller than the number Nb of the acquired data group D3b). The storage in the unit 5 is completed. In this case, the processing unit 3 preferentially executes the FFT for the acquisition data group D3c having a small number Nc. Therefore, the processing unit 3 interrupts the FFT for the acquisition data group D3b and executes the FFT for the acquisition data group D3c. Complete the FFT.

  Subsequently, the processing unit 3 restarts the FFT for the interrupted acquisition data group D3b to complete the FFT, and then the last of the acquisition data groups D3a, D3b, and D3d of the input channels CHa, CHb, and CHd. Then, the FFT is performed on the acquired data group D3d of the input channel CHd that has been stored in the storage unit 5 to complete the FFT. Each time the FFT on the acquired data groups D3a, D3b, and D3d of the associated (grouped) input channels CHa, CHb, and CHd is completed, the processing unit 3 sets each of the associated input channels CHa, CHb, and CHd. It is determined whether or not the FFT for the acquired data groups D3a, D3b, and D3d is completed, and it is determined that the FFT for the acquired data groups D3a, D3b, and D3d for all the associated input channels CHa, CHb, and CHd is completed. Sometimes, a calculation (computation between input channels CH) is performed by combining two or more frequency spectra FS for the associated input channels CH.

  In the example shown in FIG. 4, when the FFT for the acquisition data group D3d of the input channel CHd is completed, the FFT for the acquisition data groups D3a, D3b, D3d of all the associated input channels CHa, CHb, CHd is completed. To do. For this reason, the processing unit 3 performs an inter-channel operation (combining two or more frequency spectra FSa, FSb, and FSd for the associated input channels CHa, CHb, and CHd after completion of the FFT for the acquired data group D3d of the input channel CHd ( In this example, the inter-channel calculation for calculating the total value of the frequency spectra FSa, FSb, FSd of the AC signals S1a, S1b, S1d input to the input channels CHa, CHb, CHd is executed, and the calculation result is obtained. The data is stored in the storage unit 5 and output from the output unit 6 together with the frequency spectra FSa, FSb, and FSd.

  Thus, in this measuring apparatus 1, the sampling unit 2 calculates the number N of instantaneous value data D1 for one period as the frequency range including the frequency f of the input AC signal S1 becomes a lower frequency range. Execute the decimation process to increase.

  Therefore, according to this measuring apparatus 1, since the number N of instantaneous value data D1 for one cycle is small for the AC signal S1 (a signal having a short cycle) whose frequency is included in a high frequency range, the number N FFT for the instantaneous value data D1 can be reliably completed within one period, and for the AC signal S1 (a signal having a long period) including a frequency in a low frequency range, the instantaneous value data D1 for one period is included. Since the number N of N is increased, it is possible to reliably complete FFT with higher accuracy within one cycle while avoiding generation of dead time. That is, according to this measuring apparatus 1, calculation and output of the frequency spectrum FS can be surely performed regardless of the frequency f of the input AC signal S1, and the generation of dead time within one cycle is possible. It can be avoided reliably. Here, the high-accuracy FFT means an FFT that can include higher-order harmonic components because the number N of instantaneous value data D1 is large. When the number N is large, the quantization error in the A / D converter 12 is averaged, and this also reduces the error (that is, high accuracy is achieved).

  When the frequency f of the AC signal S1 input to the sampling unit 2 is known, the operator of the measuring apparatus 1 manually sets the number N of instantaneous value data D1 for one cycle for each AC signal S1. You can also However, like the sampling unit 2 in the measurement apparatus 1 of the above example, a sampling process for sampling a plurality of AC signals S1 with a common sampling clock CLK to obtain each instantaneous value data D1, and the obtained instantaneous value data D1. The frequency detection processing for detecting the frequency f of each AC signal S1 based on the frequency, the range specification processing for specifying the frequency range including the frequency f of each AC signal S1 based on the detected frequency f, and the sampling processing Further, by adopting a configuration in which the instantaneous value data D1 for each AC signal S1 is thinned out to the number N for one cycle corresponding to the frequency range specified by the range specifying process, the AC signal S1 Individual data corresponding to the frequency range including the frequency f of the AC signal S1 for the instantaneous value data D1 It can be automatically thinned out in N.

  Further, in the measurement apparatus 1, the processing unit 3 preferentially executes the calculation of the frequency spectrum FS for the AC signal S1 in which the frequency f is included in the higher frequency range in the spectrum calculation process. Therefore, according to this measuring apparatus 1, since the number N of the instantaneous value data D1 is large, the time required to complete the FFT becomes long, and the number of the instantaneous value data D1 is calculated by the FFT calculation of the AC signal S1 having a low frequency f. It is possible to prevent the occurrence of a situation in which the FFT calculation of the AC signal S1 (signal having a high frequency f) with a small N is awaited, and thus it is preferable to calculate the frequency spectrum FS at shorter time intervals. It is possible to perform an FFT operation on the AC signal S1 having a high frequency f without delay.

  Further, in the measurement apparatus 1, the processing unit 3 combines two or more of these frequency spectra FS when calculation of the frequency spectra FS for all the associated AC signals S1 (included in the group) is completed. Perform the operation. Therefore, according to this measurement apparatus 1, while the plurality of AC signals S1 input to the sampling unit 2 include, for example, three phase voltage signals for one three-phase circuit as described above, Even in the case where an AC signal unrelated to these phase voltage signals is included, the frequency for each phase voltage signal is obtained by associating (grouping) the three phase voltage signals as the AC signal S1. In a state where the spectra FS are aligned, it is possible to reliably execute the calculation combining the frequency spectra FS.

  In addition, in said measuring apparatus 1, although the frequency f of alternating current signal S1 is employ | adopted as the structure included in either of the 1st-3rd three frequency ranges, as long as the number of frequency ranges is more than one, There may be two, or four or more. The number N of instantaneous value data D1 corresponding to each frequency range may be a power of 2, and is limited to the above example (2 13, 2 11, 2 9). Instead, it may be an arbitrary power of 2.

  In the measurement apparatus 1 described above, the configuration in which only one set of the associated (grouped) AC signals S1 is included in the plurality of AC signals S1 has been described. Of course, it may be configured to include two or more sets of signals S1.

1 Measuring device
2 Sampling part
3 Processing Unit D1 Instantaneous Value Data FS Frequency Spectrum
N number S1 AC signal

Claims (3)

By sampling by inputting a plurality of AC signals, the instantaneous values of the plurality of AC signals are acquired so that the number of one cycle is a power of 2 and corresponding to each frequency. A sampling section to output,
A processing unit for performing a spectrum calculation process for calculating a frequency spectrum for the plurality of AC signals by performing a Fourier transform on the instantaneous values for the plurality of AC signals output from the sampling unit;
The sampling unit is a measuring device that increases the number of the one period as the frequency range including the frequency of the AC signal becomes a lower frequency range ,
The sampling unit samples the plurality of AC signals with a common sampling clock to obtain each instantaneous value, and a frequency for detecting the frequencies of the plurality of AC signals based on the obtained instantaneous values. Detection processing, range specification processing for specifying the frequency range including the frequencies of the plurality of AC signals based on the detected frequencies, and instantaneous values for the plurality of AC signals obtained by the sampling processing. measurement apparatus for performing the thinning process for thinning the number of the one cycle corresponding to the frequency range specified by the range specifying process. By sampling by inputting a plurality of AC signals, the instantaneous values of the plurality of AC signals are acquired so that the number of one cycle is a power of 2 and corresponding to each frequency. A sampling section to output,
A processing unit for performing a spectrum calculation process for calculating a frequency spectrum for the plurality of AC signals by performing a Fourier transform on the instantaneous values for the plurality of AC signals output from the sampling unit;
The sampling unit is a measuring device that increases the number of the one period as the frequency range including the frequency of the AC signal becomes a lower frequency range ,
Wherein the processing unit is configured in the spectrum calculation process, measurement devices run with priority calculation of the frequency spectrum for said AC signal including said frequency to a higher the frequency range. By sampling by inputting a plurality of AC signals, the instantaneous values of the plurality of AC signals are acquired so that the number of one cycle is a power of 2 and corresponding to each frequency. A sampling section to output,
A processing unit for performing a spectrum calculation process for calculating a frequency spectrum for the plurality of AC signals by performing a Fourier transform on the instantaneous values for the plurality of AC signals output from the sampling unit;
The sampling unit is a measuring device that increases the number of the one period as the frequency range including the frequency of the AC signal becomes a lower frequency range ,
The plurality of AC signals are grouped into one or more groups of the plurality of AC signals,
Wherein the processing unit, when calculating the frequency spectrum for all of the AC signal included in the group has been completed, measurement device to perform the operations that combine the frequency spectrum more.

Images