TWI597419B - Calibration apparauts for flickermeter and calibration mehtod thereof - Google Patents

Description

Scintillation analyzer calibration device and calibration method

The invention relates to a calibration device and a calibration method for a scintillation analyzer, in particular to a calibration device and a calibration method for a scintillation analyzer for a wind power generation device.

In recent years, energy prices have risen, and countries around the world are actively promoting the development of renewable energy. Taiwan is no exception. As the price of equipment decreases, the number and capacity of wind turbine installations continue to increase. In the future, power will be generated by the power of private and manufacturers. The ratio will increase. Due to the start, stop, switch or operation of the wind turbine, the output voltage and current will fluctuate, which may cause voltage flicker. In general, when the wind turbine is connected in parallel to the high voltage line, the problem of voltage flicker is less serious. If connected to a medium voltage system, the voltage fluctuation caused by the change of wind speed can not be ignored. In order to solve the problem of voltage flicker caused by the wind turbine, IEC 61400-21 provides the relevant specifications and restrictions for the wind turbine to be connected in parallel with the power system.

The extent and extent of the voltage flicker generated by the wind turbine will vary depending on the wind turbine control structure, the installed capacity, the location of the joints, and the structure of the transmission and distribution network. When the inductive generator drives the fan in a motor mode, the surge current will be drawn from the system, causing a voltage drop of several seconds without proper control. If the wind farm contains a large number of fans, the order is generally The method of starting to reduce the impact on the system voltage, but in the weaker distribution network, the sequential start may cause a large voltage flicker, and the number of fans that can be installed at the location is relatively reduced. In addition, in the continuous operation of the fan, the torque will not be fixed due to different fluid effects. The above factors will affect the torque efficiency in the three-blade fan system. These torque changes directly cause parallel connection. The phenomenon of voltage and power disturbance at the point, especially in locations where the distribution network is weak, requires special consideration of the problem of voltage flicker and is limited to a certain range.

The voltage flashing of wind power has the characteristics of transmission, so it is necessary to consider the combined effect of each variable load and power source. Since countries have different definitions of voltage flicker, there are different ways to measure the severity of voltage flicker. Countries such as Japan, South Korea, and China use the ΔV ₁₀ value to estimate the amount of voltage flicker. In Europe and the United States, the IEC specification is used to indicate the degree of voltage flicker based on the statistical value of the voltage flicker severity. The standard measurement procedure states that the statistical value of the severity of voltage flicker can be divided into short-term severity ( P _st ) and long-term severity ( P _lt ).

The flicker analyzer proposed by the IEC can simulate and measure the illuminance changes caused by the low amplitude and low frequency changes of the main power supply. A measurement method for the voltage fluctuation of the scintillation analyzer has been appeared, and corresponding parameters are calculated accordingly. The flicker analyzer voltage measurement method using full time domain calculation can reduce the calculation error caused by system frequency change and the spectral leakage caused by fast Fourier transform (FFT). However, in order to convert an analog filter to a digital implementation, bi-linear transformation must be used and a lot of time is wasted to perform the calculation. In contrast, the measurement of voltage flicker can achieve satisfactory results by completely adopting the frequency domain approach, and has been applied to the arc furnace in a standardized IEC scintillation analyzer. The voltage measurement was verified. This approach can be quickly and well estimated for voltage flicker. However, if the frequency of the system is gradually changing and the frequency resolution is reduced, direct demodulation using fast Fourier transform in this approach will be less accurate and will cause spectral leakage effects. Therefore, a new calibration method for the scintillation analyzer must be proposed to compensate for the shortcomings of the two conventional methods and to achieve an optimal balance between accuracy and development cost.

In view of the above problems of the prior art, the object of the present invention is to provide a calibration apparatus and a calibration method for a scintillation analyzer to solve the problems of the conventional time domain approach technique and the frequency domain approach technique in the calibration method of the scintillation analyzer. .

Based on the above object, the present invention provides a calibration apparatus suitable for calibrating a scintillation analyzer of a wind power generation apparatus, comprising a calibration signal generation module, a digital-to-analog converter (DAC), a power amplifier, An analog-to-digital converter (ADC), a bandpass filter (BPF), and a signal processing module. The calibration signal generation module is configured to generate and output a digital calibration signal. The digital to analog converter is electrically coupled to the calibration signal generation module for converting the digital calibration signal into an analog calibration signal. The power amplifier is electrically connected to the digital to analog converter to adjust the analog calibration signal to a specific amplitude and output to the scintillation analyzer to be tested. The analog to digital converter is electrically coupled to the power amplifier for sampling the analog calibration signal and converting it to a sampled signal comprising a plurality of digital values. The band pass filter is electrically connected to the analog to digital converter for receiving the sampled signal and filtering out a plurality of harmonic signals generated by the sampling in the sampled signal and noise outside the band pass filter band. The signal processing module is electrically connected to the band pass filter to receive the filtered sample signal and solve the first The frequency and the first amplitude; in addition, the signal processing module is further electrically connected to the calibration signal generating module to adjust the digital calibration signal by the first frequency and the first amplitude. The signal processing module is further electrically connected to the digital to analog converter and the analog to digital converter for synchronizing by the first frequency to digital analog to analog converter and the analog to digital converter. The calibration signal generation module includes a scintillation measurement model of the wind power generation device, and the scintillation measurement model of the wind power generation device is composed of the IEC 61000-4-15 standard.

Preferably, the lower limit of the passband bandwidth of the bandpass filter is in the range of 0 to 1 Hz and the upper limit is in the range of 40 to 70 Hz.

Preferably, the signal processing module estimates the first frequency according to the following formula: ,among them The estimated first frequency, Δt is the sampling period, and y(n) is the filtered sampling signal at the nth sampling period; the signal processing module estimates the digit to analog converter and analog to digital according to the following formula Converter sync signal frequency Where f _s is the sync signal frequency, M is the analog to the total number of sample cycles of the digital converter, N is the total sampled data points obtained by the analog to digital converter within the total number of sample cycles; and the signal The processing module brings the estimated first frequency into an adaptive linear neural network (ADALINE) to estimate the first amplitude.

According to still another object of the present invention, a calibration method is provided, which is suitable for calibrating a scintillation analyzer of a wind power generator, and the calibration method comprises the following steps: A calibration signal is used to generate a module to generate and output a digital calibration signal. The digital calibration signal is converted to an analog calibration signal using a digital to analog converter. The analog amplifier is adjusted to a specific amplitude by a power amplifier and output to the scintillation analyzer to be tested. The analog calibration signal is sampled using an analog to digital converter and converted to a sampled signal containing a plurality of digital values. A band pass filter is used to receive the sampled signal, and a plurality of harmonic signals generated by sampling in the sampled signal and noise outside the band pass filter band are filtered out. The signal processing module is configured to receive the filtered sampling signal and solve the first frequency and the first amplitude, and then adjust the digital calibration signal by the first frequency and the first amplitude to meet the testing requirements of the scintillation analyzer, and further Synchronization is performed by a first frequency log to analog converter and an analog to digital converter.

Preferably, the step of generating and outputting the digital calibration signal further comprises outputting the analog calibration signal to the scintillation analyzer of the wind power generator.

Preferably, the step of generating and outputting the digital calibration signal further comprises bringing a plurality of different input phases into a scintillation measurement model consisting of the IEC 61000-4-15 standard to generate a digital calibration signal suitable for calibrating the scintillation analyzer. .

Preferably, the step of decoding the frequency of the sampled signal and the amplitude further comprises estimating the first frequency by the equation (1) and estimating by introducing the estimated first frequency into the adaptive linear neural network. First amplitude.

Preferably, the calibration method further comprises: a second frequency and a second output of the scintillation analyzer output corresponding to the input phase and the first frequency estimated by the sampling signal and the first amplitude and the input analog calibration signal by the known flicker measurement model The amplitudes are aligned for calibration of the scintillation analyzer.

As described above, the calibration apparatus and calibration method of the scintillation analyzer according to the present invention may have one or more of the following advantages:

(1) The calibration device and the calibration method of the scintillation analyzer of the present invention can obtain the frequency and amplitude of the scintillation analyzer to be corrected without performing bilinear conversion, and the time consumed for the calculation can be greatly reduced.

(2) The calibration device and the calibration method of the scintillation analyzer of the present invention have robust characteristics for the frequency variation and resolution of the calibration system, and can reduce the error of the measurement frequency or amplitude caused by the spectrum leakage.

For a better understanding and understanding of the technical features and the efficacies of the present invention, the preferred embodiments and the detailed description are as follows.

1‧‧‧ calibration device

10‧‧‧Calibration Signal Generation Module

20‧‧‧Digital to analog converter

30‧‧‧Power Amplifier

40‧‧‧ Analog to Digital Converter

50‧‧‧ bandpass filter

60‧‧‧Signal Processing Module

70‧‧‧Flicker Analyzer

S10~S70‧‧‧Steps

The above and other features and advantages of the present invention will become more apparent from the detailed description of the exemplary embodiments thereof. Figure 2 is a flow chart of an embodiment of a calibration method for a scintillation analyzer according to the present invention; and Figure 3 is a diagram showing a power system frequency versus relative error for a scintillation analyzer applied to a wind power generation device according to the present invention. Fig. 4 is a graph showing the scintillation severity versus relative error of a scintillation analyzer applied to a wind power generator according to the present invention.

As used herein, the term "and/or" includes one or more related items. Any or all combinations of these. When the phrase "at least one of" is preceded by a list of elements, the entire list of elements is modified instead of the individual elements in the list.

Please refer to Fig. 1, which is a functional block diagram of a calibration device for a scintillation analyzer according to the present invention. In the first diagram, the present invention provides a calibration apparatus 1 for a scintillation analyzer, comprising a calibration signal generation module 10, a digital to analog converter 20, a power amplifier 30, an analog to digital converter 40, and a bandpass filter 50. And signal processing module 60. The calibration signal generation module 10 is configured to generate and output a digital calibration signal. The digital to analog converter 20 is electrically coupled to the calibration signal generation module 10 for converting the digital calibration signal into an analog calibration signal. The power amplifier 30 is electrically connected to the digital to analog converter 20 for adjusting the analog calibration signal to a specific amplitude and outputting to the scintillation analyzer 70 to be tested. The analog to digital converter 40 is electrically coupled to the power amplifier 30 for sampling the analog calibration signal and converting it to a sampled signal comprising a plurality of digital values. The bandpass filter 50 is electrically coupled to the analog to digital converter 40 for receiving the sampled signal and filtering out a plurality of harmonic signals generated by the sampling in the sampled signal and noise outside the bandpass filter band. The signal processing module 60 is electrically connected to the band pass filter 50 to receive the filtered sample signal and solve the first frequency and the first amplitude. In addition, the signal processing module 60 is further electrically connected to the calibration signal generating mode. Group 10, for adjusting the digital calibration signal by the first frequency and the first amplitude. The signal processing module 60 is further electrically coupled to the digital to analog converter 20 and the analog to digital converter 40 for synchronization by the first frequency to digital to analog converter 20 and analog to digital converter 40.

The following is a scintillation analyzer for measuring a wind power generator using the calibration device and calibration method of the scintillation analyzer of the present invention to verify the feasibility of the present invention, but not limited thereto.

In order to calibrate the scintillation analyzer of the wind power generation device, the lower limit of the passband bandwidth of the band pass filter is 0 to 1 Hz and the upper limit range is 40 to 70 Hz. And the calibration signal generation module is constructed according to the IEC 61000-4-15* standard. According to the IEC 61000-4-15 standard, Table 1 below shows the state of instability test for A-level performance.

In Table 1, f _nom represents the nominal frequency, U _din represents the specified input voltage obtained by the specified supply voltage by the converter ratio, and P _st represents the flicker. Severity (flicker severity). *Reference 1: Wind Turbines-Part 21: Measurement and Assessment of Power Quality Characteristics of Grid Connected Wind Turbines , IEC Std. 61400-21, 2008.

The instability test procedure for the scintillation analyzer for each measurement is as follows:

1. Select a parameter to be measured, such as the root-mean-square (RMS) voltage.

2. Set all the parameter values of test state one, within the range of influence parameters, such as U _din : 0% to 200%, and take an average of five points to test to verify the stability of the parameters to be measured.

3. Set all parameter values for test state two and repeat step 2.

4. Set all parameter values for test state three and repeat step 2.

To measure the frequency of the power system, the present invention employs an autoregressive (AR) model based on techniques with high frequency resolution**. It is assumed that the signal filtered by the band pass filter 50 in Fig. 1 can be expressed as Where A ₀ is the signal amplitude, ψ ₀ is the starting phase angle, and f ₀ is the frequency of the power system component. **Reference 2: CI Chen and GW Chang, "Virtual Instrumentation and Educational Platform for Time-Varying Harmonics and Interharmonics Detection," IEEE Trans. on Industtial Electronics, Vol. 57, No 10, Oct. 2010, pp. 3334- 3342.

According to the autoregressive model, when the signal y(n) in the equation (3) enters the signal processing module 60, it can be expressed as a linear prediction theory. Where the parameter a _p is a real number and P is the estimated progression of the autoregressive model. Relatively estimated signal (n) can be expressed as In the formula (5), ε(n) represents an error. If the error can be minimized for any n value, then the true signal y(n) and the estimated signal (n) will be the same. It can be known from the above reference 2 that the relationship between the parameter a _p and the transfer function of the autoregressive model can be expressed as a ₀ z ^P + a ₁ z ^{P -1} +... + a _{P -1} z + a _P = 0, (6) where .

Therefore, if the above parameter a _p can be obtained, the estimation of the frequency of the power system can be easily performed. Further, since the power signal y(n) is filtered via a band pass filter, the value of P can be set to 2. Under these conditions, z ₁ and z ₂ will be conjugate complex numbers. In practical applications, a ₀ = a ₂ and a ₁ =1 can be conveniently set to simplify the calculation. The power signal y(n) is below the N sample data points, and the total estimation error will be equal to Under the condition that the measurement error is minimized, the following equation must be established. . Using equation (8), we can solve for a ₀ as follows . Then, taking a ₀ in the equation (9) into the equation (6), the estimated frequency of the power signal can be obtained. for ,among them The filtered sampling signal is the estimated power signal frequency, Δt is the sampling period, and y(n) is the nth sampling period. Finally, the synchronizing signal frequency f _s of the digital to analog converter and the analog to digital converter can be obtained by the following equation for synchronization. Where M is the total number of sampling cycles of the analog to digital converter, and N is the total number of sampled data points obtained by the analog to digital converter within the total number of sampling cycles.

A commonly used method for obtaining a voltage amplitude envelope is a combination of a square multiplier and a specific band pass filter indirect demodulation method. However, multiplication will produce more flickering noise. It can be seen from the above reference 2 that the adaptive linear neural network can be adapted to acquire the voltage amplitude envelope. However, estimating the error of the power system frequency can impair the accuracy of the adaptive linear neural network and slow down the rate of convergence. This part can be obtained by observing the real signal y(n) at the kth timing and the estimated signal obtained by the adaptive linear neural network. (n) the estimated error e(k) (12) wherein Δω ₀ , Δω _2m-1 and Δω _2m are the true power signal angular frequency ω ₀ and the true weight ω _2m-1 and ω _2m and the estimated power signal angular frequency ω ₀ and the estimated weight, respectively The difference between ω _2m-1 and ω _2m . From equation (12), it can be concluded that the estimated error e(k) is affected by Δω ₀ in the first sum term and Δω _2m-1 and Δω _2m in the second sum term.

To solve the problem of acquisition of power signal frequencies in squared multiplication demodulation methods and traditional adaptive linear neural networks, conventional techniques use frequency domain interpolation techniques. However, due to the limitation of estimation accuracy caused by the use of fast Fourier transform, a large number of sampled data points are required to achieve better frequency resolution. Therefore, the above-mentioned autoregressive model-based estimation method with high frequency resolution is adopted to solve this problem, and its procedure and accuracy have been verified, for example, as disclosed in reference GWChang and CIChen," An Accurate Time-Domain Procedure for Harmonics and Interharmonics Detection, "IEEE Trans. on Power Delivery, Vol. 25, No. 3, July 2011, pages 1787 to 1795. When the estimated frequency of the power signal in the equation (10) is obtained, it can be brought into the analysis model of the adaptive linear neural network, and the estimation error caused by the frequency deviation of the power signal in the equation (12) can be obtained. minimize.

According to still another object of the present invention, a calibration method for a scintillation analyzer is provided, which is suitable for calibrating a scintillation analyzer of a wind power generator. The calibration method comprises the following steps: Step S10 is a scintillation measurement using the IEC 61000-4-15 standard. Model to generate and output digital calibration signals. Step S20 is to convert the digital calibration signal into an analog calibration signal using a digital to analog converter. Step S30 is to use the power amplifier to adjust the analog calibration signal to a specific amplitude and output it to the scintillation analyzer to be tested. Step S40 is to sample the analog calibration signal by using a analog-to-digital converter and convert it into a sampling signal including a plurality of digital values. Step S50 uses a band pass filter to receive the sampled signal, and filters out a plurality of harmonic signals generated by the sampling in the sampled signal and noise outside the band pass filter band. Step S60 is to use a signal processing module to receive the filtered sample signal and solve the first frequency and the first amplitude, and then adjust the digital calibration signal by the first frequency and the first amplitude to conform to the test of the scintillation analyzer. The demand is synchronized by the first frequency-to-digital to analog converter and the analog to digital converter.

Preferably, the step of decoding the frequency of the sampled signal and the amplitude further comprises the step S51 of: estimating the first frequency by the equation (1) and bringing the estimated first frequency into the adaptive linear neural network. The road is used to estimate the first amplitude.

Preferably, the calibration method further comprises the step S70: the second frequency of the scintillation analyzer output corresponding to the input phase and the first frequency estimated by the sampling signal and the first amplitude and the input analog calibration signal by the known flicker measurement model. The second amplitude is compared for calibration of the scintillation analyzer.

Please refer to FIG. 3 and FIG. 4 together. FIG. 3 is a diagram showing a power system frequency versus relative error diagram of a scintillation analyzer applied to a wind power generator according to the calibration apparatus and calibration method of the scintillation analyzer according to the present invention. 4 is a scintillation severity versus relative error map of a scintillation analyzer applied to a wind power generator according to the calibration apparatus and calibration method of the scintillation analyzer of the present invention. In Figures 3 and 4, the relative error is defined as Where Relative Error represents the relative error, Actual Value represents the true value, and Estimated Value represents the estimated value. It can be seen from Figures 3 and 4 that a large offset from the main fundamental frequency causes an increase in the relative error. Due to the high resolution of the autoregressive model for power signal frequency estimation, the relative error of the three test states in Figure 3 is lower than the maximum allowable value of 0.03% required by the IEC 61000-4-7 standard. Similarly, for the estimation of the scintillation severity P _st , the relative errors of the three test states are also lower than the maximum allowable value of 5% required by the above criteria. It can thus be confirmed that the calibration device and the calibration method of the scintillation analyzer of the present invention are indeed applicable to the determination of the characteristics of the disturbance signal and are suitable as a standard for calibrating a commercial scintillation analyzer.

The present invention has been particularly shown and described with reference to the exemplary embodiments thereof, and it is understood by those of ordinary skill in the art Various changes in form and detail can be made in the context of the category.

1‧‧‧ calibration device

10‧‧‧Calibration Signal Generation Module

20‧‧‧Digital to analog converter

30‧‧‧Power Amplifier

40‧‧‧ Analog to Digital Converter

50‧‧‧ bandpass filter

60‧‧‧Signal Processing Module

70‧‧‧Flicker Analyzer

Claims (8)

A calibration device suitable for calibrating a scintillation analyzer of a wind power generation device, the calibration device comprising: a calibration signal generation module for generating and outputting a digital calibration signal; a digital to analog converter electrically connected to the Calibrating a signal generating module to convert the digital calibration signal into an analog calibration signal; a power amplifier electrically connected to the digital to analog converter to adjust the analog calibration signal to a specific amplitude and output to Calibrating the scintillation analyzer; a analog to digital converter electrically coupled to the power amplifier to sample the analog calibration signal and convert it to a sample signal comprising a plurality of digital values; a band pass filter Electrically connecting to the analog to digital converter to receive the sampling signal and filtering out a plurality of harmonic signals generated by sampling in the sampling signal and noise outside the band pass filter band; and a signal processing mode The group is electrically connected to the band pass filter to receive the filtered sample signal and solve a first frequency and a first amplitude The signal processing module is further electrically connected to the calibration signal generating module to adjust the digital calibration signal by the first frequency and the first amplitude; the signal processing module is electrically connected to the digital to analog conversion And the analog to digital converter for synchronizing the digital to analog converter and the analog to digital converter by the first frequency; wherein the calibration signal generating module comprises a scintillation measurement of a wind power generation device Model, and the scintillation measurement model of the wind power generation device is IEC The 61000-4-15 standard is constructed in which the lower limit of the passband bandwidth of the bandpass filter is 0 to 1 Hz and the upper limit is 40 to 70 Hz. The calibration device of claim 1, wherein the signal processing module estimates the first frequency according to the following formula ,among them For the first frequency, Δt is a sampling period, y(n) is the filtered sampling signal at the nth sampling period; and the signal processing module estimates the digital to analog converter according to the following formula And analog to one of the digital converters Where f _{s is} the frequency of the synchronization signal, M is the analog to one of the total number of sampling cycles of the digital converter, and N is one of the total sampled data points obtained by the analog to digital converter within the total number of sampling cycles . The calibration device of claim 2, wherein the signal processing module brings the first frequency into an adaptive linear neural network to estimate the first amplitude. A calibration method for calibrating a scintillation analyzer of a wind power generator, the calibration method comprising the steps of: generating and outputting a digital calibration signal using a calibration signal generation module; calibrating the digit using a digital to analog converter Signal conversion into a class Ratio calibration signal; the analog calibration signal is adjusted to a specific amplitude by a power amplifier; the analog calibration signal is sampled by an analog to digital converter, and converted into a sample signal including one of a plurality of digital values; Filtering to receive the sampling signal, and filtering out a plurality of harmonic signals generated by sampling in the sampling signal and noise outside the band pass filter band, wherein a limit of a passband bandwidth of the band pass filter The range is 0 to 1 Hz and the upper limit is 40 to 70 Hz; and a signal processing module is used to receive the filtered sample signal and solve a first frequency and a first amplitude, and then the first frequency And the first amplitude is used to adjust the digital calibration signal to meet the test requirements of the scintillation analyzer, and the digital to analog converter and the analog to digital converter are synchronized by the first frequency. The calibration method of claim 4, wherein the step of adjusting the amplitude of the analog calibration signal further comprises the step of outputting the analog calibration signal having the specific amplitude to a scintillation analyzer of a wind power generation device. The calibration method of claim 5, wherein the step of generating and outputting the digital calibration signal comprises the steps of: respectively bringing a plurality of different input phases into a flashing amount formed by an IEC 61000-4-15 standard. The model is measured to generate the digital calibration signal suitable for calibrating the scintillation analyzer. The calibration method of claim 6, wherein the step of solving the first frequency and the first amplitude of the sampling signal comprises the following steps: estimating the first frequency by using the following formula ,among them To estimate the first frequency, Δt is the sampling period, and y(n) is the filtered sampling signal at the nth sampling period; the following formula is used to estimate the digit to the analog converter and the analogy to Frequency of the sync signal of the digital converter Where f _{s is} the frequency of the synchronization signal, M is the analog to one of the total number of sampling cycles of the digital converter, and N is the total number of sampled data points obtained by the analog to digital converter within the total number of sampling cycles; And estimating the first amplitude by bringing the estimated first frequency into an adaptive linear neural network. The calibration method of claim 7 further includes the following steps: calibrating the input phase by the known scintillation measurement model and the first frequency estimated by the sampling signal and the first amplitude and input analogy A second frequency of the scintillation analyzer output of the signal is compared to a second amplitude for calibration of the scintillation analyzer.