KR101752868B1

KR101752868B1 - A noise filtering system of periodically oscillating signals for optimal control

Info

Publication number: KR101752868B1
Application number: KR1020150093086A
Authority: KR
Inventors: 이용해; 이정훈
Original assignee: 한국디지탈콘트롤 주식회사
Priority date: 2015-06-30
Filing date: 2015-06-30
Publication date: 2017-07-04
Also published as: KR20170003782A

Abstract

The present invention relates to a periodic signal noise filter system for optimum control for dividing a periodic signal such as an output of an inverter into a periodic unit and reconstructing an output signal at an intermediate value of signals of a plurality of divided periods to filter the periodic signal. A sampling unit for sampling the signal on a cycle-by-cycle basis to generate sampling data; And a periodic filter unit filtering the sampling data of each period and filtering the sampled data of the same period in the front and rear periods of the period with the characteristic values of the sampled data.
According to the noise filter system as described above, noise (periodic signal) such as an inverter output signal is removed to accurately measure? (Phase difference) and a filtered output waveform is obtained. .

Description

[0001] The present invention relates to a noise filtering system for periodically oscillating signals for optimal control,

A periodic signal for optimal control, which is used for filtering a periodic signal such as an output signal of an inverter by dividing a periodic signal such as an output signal of an inverter into periodic signals, reconstructing an output signal with an intermediate value of signals of a plurality of periods, To a noise filter system.

Generally, an inverter is a device for converting direct current power into alternating current power. It is widely used to convert an alternating current power into a direct current power, control it by frequency, and then convert it into alternating current to output accurate speed control. In recent years, development of electric vehicles has become common, and DC electric power is widely used as a device for outputting three-phase current for driving a motor of an electric vehicle.

However, the three-phase current output from the inverter is measured and used for various control and detection. For example, a measured current signal is used to accurately control the switching frequency of the inverter [Patent Document 1]. In addition, the measured output signal of the inverter is used for power calculation, detection of contact failure of the cable connected to the inverter [Patent Document 2], and device defect [Patent Document 3].

However, there is a problem that noise is generated in the output signal of the inverter and the phase difference, which is an overall factor necessary for power calculation and the like, can not be properly measured. In other words, the inverter converts the alternating current to direct current in the converter, and converts it back to the desired alternating current through the PWM control in the inverter by the direct current power source. At this time, PWM control by switching of six transistors of inverter converts to three phase variable voltage and variable frequency AC to control variable speed of motor. Incidentally, switching noise occurs due to high-speed on / off of the six transistors.

Accurate current measurement is difficult due to the noise generated in this process. In addition, due to the noise included in the input voltage or the current of the inverter, it is possible to cause a malfunction or a trouble to a device common to the power source or a nearby peripheral device.

Therefore, it is necessary to remove the noise of the periodic signal such as the three-phase output signal of the inverter.

[Patent Document 1] Korean Published Patent Application No. 10-2013-0037054 (published on April 15, 2013) [Patent Document 2] Korean Patent Laid-Open Publication No. 10-2015-0026260 (published on April 11, 2015) [Patent Document 3] Korean Published Patent Application No. 10-2011-0086263 (Published on July 28, 2011)

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems, and it is an object of the present invention to provide a method and apparatus for classifying a periodic signal such as an output of an inverter in a cycle unit, reconstructing an output signal at an intermediate value of signals, And to provide a periodic signal noise filter system for optimal control that filters the signal.

In order to achieve the above object, the present invention provides a periodic signal noise filter system for optimal control, comprising: a sampling unit for sampling the periodic signal periodically and generating sampling data; And a periodic filter unit for filtering the sampling data of each period and filtering the sampled data of the same period in the front and rear periods of the period by the characteristic values with the sampled data.

According to another aspect of the present invention, there is provided a noise filter system of a periodic signal for optimal control, wherein the periodic filter unit filters the sampled data of a corresponding period and a median of sampled data of the same position .

According to another aspect of the present invention, there is provided a noise filter system for a periodic signal for optimal control, wherein the periodic filter includes filtering up to sampling data of M periods before and after the corresponding period.

The present invention is characterized in that in the periodic signal noise filter system for optimal control, the periodic filter unit is filtered according to the following [Equation 1].

[Equation 1]

(U, n) is the nth sampling data of period u, M is the size of the filter, and Median ({tilde over (s) ) Is a function that represents the median for the set {}.

According to another aspect of the present invention, there is provided a periodic signal noise filter system for optimal control, the system comprising: a period filter for filtering a signal filtered by a period with a characteristic value, And further comprising:

Further, the present invention is characterized in that in the periodic signal noise filter system for optimal control, the section filter unit performs filtering according to the following [Expression 2].

[Equation 1]

(T, n) is a value filtered by the period of the n-th sampling data of the period t, and s "(t, n) And L is the size of the filter.

As described above, according to the noise filter system of the periodic signal for optimal control according to the present invention, the noise (noise) of the periodic signal such as the inverter output signal is removed to accurately measure? (Phase difference) This makes it possible to compensate for the deficiency of the periodic signal analysis and enable precise measurement. In particular, by eliminating the noise of the inverter output signal in the inverter system and accurately calculating the θ (phase difference) and deriving the filtered output waveform, it compensates for the lack of data analysis for the existing power management of the inverter applied equipment Precise measurement can be done.

In addition, according to the noise filter system of the periodic signal for optimal control according to the present invention, the phase difference and the three-phase output waveform can be accurately derived, thereby solving the phase incomparability, difficulty in improving the power factor and signal analysis problem, Various advantages such as speed, operation efficiency, and torque measurement can be secured.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of an entire system for implementing the present invention. FIG.
2 is a block diagram of a configuration of a periodic signal noise filter system for optimal control according to an embodiment of the present invention;
FIG. 3 is a graph illustrating the division of a 3-phase signal according to an embodiment of the present invention into periods.
Figure 4 illustrates sampling for a three-phase signal of one period in accordance with one embodiment of the present invention.
FIG. 5 and FIG. 6 are diagrams illustrating the division of three-phase signals by a period when the filter size is 2 according to an exemplary embodiment of the present invention, and sampled data at the same position in each cycle. FIG.
FIG. 7 is a diagram showing an example of filtering by temporally adjacent data according to an embodiment of the present invention, where the filter size is 3; FIG.
8 is a graph illustrating an example of applying a one-dimensional delay filter according to an embodiment of the present invention.
FIG. 9 is a graph showing the result of measuring three-phase currents according to the prior art. FIG.
10 is a graph showing the results of measurement of three-phase current according to the present invention.
11 is a graph of the results of applying confidence interval averages according to the present invention.
12 is a graph of a result of applying a one-dimensional delay filter according to the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the drawings.

In the description of the present invention, the same parts are denoted by the same reference numerals, and repetitive description thereof will be omitted.

First, examples of the configuration of the entire system for carrying out the present invention will be described with reference to Fig.

1, the overall system for implementing the present invention includes an inverter 10, a three-phase power cable 20 for supplying power output through the inverter 10 to the load 60, a three-phase power cable A current sensor 30 provided on the main body 20, and a noise filter system 40 for filtering the measured three-phase current for noise removal. In addition, it further comprises a filtered current and a power calculation system 50 for calculating power using the phase difference.

The inverter 10 converts the DC power generated from the battery into three-phase AC power. The inverter 10 shown in Fig. 1 is an example, and as another example, the inverter 10 may be an inverter that rectifies the AC power and converts the rectified power to AC power by generation of a pulse width modulation (PWM) signal.

The inverter 10 converts the DC power supplied to the battery to AC power and supplies the AC power to the load 60. The AC power converted by the inverter 10 is preferably three-phase AC power. The inverter 10 is composed of a capacitor and a plurality of insulated gate bipolar transistors (IGBTs). The IGBT performs PWM (Pulse Width Modulation) switching in accordance with a control signal to phase-convert the direct current power supply to the load 60 .

The load 60 is an electric device that operates by receiving a three-phase power source such as a motor.

The three-phase power cable 20 is a power cable that supplies power output through the inverter 10 to the load 60.

The current sensor 30 is disposed on each output line of the three-phase power cable 20 formed between the inverter 10 and the load 60 to measure the current for the three phases. That is, the current sensor 30 detects the three-phase current signals (u-phase current signal, v-phase current signal, w-phase current signal) supplied to the load 60 and transmits them to the noise filter system 40.

The noise filter system 40 receives the three-phase current signals measured at the current sensor 30, performs a filtering operation to remove noise, and transmits the filtered three-phase current value to the power calculation system 50.

In particular, the noise filter system 40 divides a periodic signal such as an output signal of the inverter (or a three-phase current signal) into periods, reconstructs the output signal with an intermediate value of signals of a plurality of periods, .

The power calculation system 50 receives the filtered signal from the noise filter system 40 and calculates the power.

In the example of FIG. 1, the noise filter system 40 according to the present invention is applied not only to the inverter but also to all the signals having periodicity. Further, not only a three-phase periodic signal but also a single (single-phase) signal can be applied.

Hereinafter, for convenience of explanation, the description will be limited to the three phases of the inverter.

Next, a periodic signal noise filter system for optimal control according to an embodiment of the present invention will be described in more detail with reference to FIG.

As shown in FIG. 2, the noise filter system 40 according to the present invention includes a sampling unit 41 for sampling a periodic signal such as a three-phase signal on a periodic basis, and a filter for periodically filtering the sampled signal And a periodic filter unit 42. In addition, a delay filter section 44 for applying a first order lag filter and a section filter section 43 for filtering the filtered signal with characteristic values such as average or sum of adjacent section data, As shown in FIG.

The sampling unit 41 samples each of the periodic signals such as the three-phase signal measured by the sensor 30 on a cycle-by-cycle basis. At this time, the size of the sampling per one cycle is denoted by N. Also, the n-th sampling data at period t is denoted by s (t, n).

Assuming a frequency of 60 Hz, one cycle per 0.01667 sec is generated. Therefore, the signals are classified by 0.01667 sec. An example in which the signals are divided into cycles is shown in Fig.

And samples the signal with a sampling size N for each period. Therefore, the sampling data is displayed as follows.

S (1, 1), s (1,2), ...., s (1, N) at period T =

S (2,1), s (2,2), ...., s (2, N) at period T = 2,

...

(T, 1), s (t, 2), ...., s (t, N) at period T = t,

Data sampled at sampling rate N at period T = t is shown in FIG.

Next, the period filter unit 42 filters the sampling data of each cycle, and filters the data at a median between the sampling data at the same position in the cycle before and after the cycle. On the other hand, other characteristic values such as an average value can be applied in addition to the intermediate value. However, it is preferable to use an intermediate value.

That is, the intermediate value is calculated in the group of the sampling data of the cycle t and the sampling data (sampling data of the same order in the cycle) located at the same position in the cycles before and after the cycle t. Then, the calculated intermediate value is set as a filtered value of the sampling data of the period t.

This can be expressed by the following equation.

[Equation 1]

Here, s' (t, n) is the filtered value (filtered sampling data) of the nth sampling data of period t. Also, s (u, n) is nth sampling data of period u. In addition, M is the size of the filter and is a predetermined value. Median ({}) is a function that represents the median for the set {}.

Preferably, the filter size M is set to 5. The size of the filter indicates whether to include up to sampling data of the period before and after the cycle t as a filter. That is, a filtered value is obtained from the sampling data of the cycle t and the intermediate value in the sampling data on the same position of the Mth cycle before and after the cycle t.

Here, the reason for applying the median rather than the average value is that the peak value may go up or down unexpectedly in a certain period. Since these peaks can affect the size, the median, Median, is applied.

For example, the T ₁ of s (1,1) value is 1.01, the s (2,1) value of T ₂ 1.05, s (3,1) value of T ₃ is 1.05, T ₄ s (4,1 ) Value is 1.03, s (m, 1) of the period T _m1 to which the median is applied becomes 1.04.

Next, with reference to FIG. 5 and FIG. 6, an example in which the filter size is 2 will be described. Specifically, a method of filtering the sampling data of the cycle t will be described.

As shown in FIG. 5, the periodic signals such as three-phase signals before and after the period t are divided and sampled according to periods. At this time, the sampling data corresponding to each cycle is as shown in FIG. Apply a median in the set of sampling data corresponding to the same position in each period. The value to which the median applies is the filtered value. That is, the sampling data connected by the dotted line in FIG. 6 are data at the same position. An intermediate value of the sampling data connected by the dotted line is obtained and set as the sampling data value filtered by the corresponding intermediate value.

On the other hand, in the case of the signal of the first cycle, since there is no previous cycle of the cycle, only the subsequent cycle is filtered without the previous cycle. For example, if the median is applied to the periodic signal before the filter is applied and the tolerance of the applied error is set to 1% or less, T _m1 is data from T ₁ to T ₅ , and T _m2 T ₁ to T ₇ and T _m3 to T ₁ to T ₈ .... T _m10 applies sampling data from T ₅ to T ₁₅ . And apply five sampling data back and forth to the reference period.

Next, the section filter unit 43 filters the filtered signal s' (t, n) by a characteristic value such as an average or a sum of data per section (per hour). That is, after applying the median filter, the average or sum of the data per hour of the wave (waveform) is obtained. This is filtered by the following equation.

&Quot; (2) "

&Quot; (3) "

S '(t, n) is a filtered signal of the periodicity by the median, and s' (t, n) represents a filtered signal by the data before and after the interval (T, n), and Equation (3) is a signal obtained by filtering the average of neighbor signals including s' (t, n) That is, the maximum distance (time) of the adjacent signal in time (the adjacent signal in the segment) in time of the corresponding signal.

Specifically, according to Equation (2), the sum of the sizes of data of the Lth and previous data at an arbitrary point of the waveform passed through the median filter in the original signal is divided by 2L, and a result is obtained. The value becomes the data value of an arbitrary point after applying the average value filtering.

The purpose of filtering by temporal (interval) characteristic values is to add and average data before and after a certain point to remove the noise of the waveform because there is some noise in the waveform after applying the periodic median filter. That is, the median filtering of the periodicity is divided into periods, but the interval filtering is a filtering that is not related to the period.

As shown in FIG. 7, for the n-th signal, the three signals before and after are averaged to filter the n-th signal. (T, n + 1), s (t, n + 1) and s (t, n + 2) ), s (t, n + 3). And their values are filtered to 33.25 which is an average of 38, 36, 35, 32, 30.5, and 28. Is an average according to the above equation (2).

Next, the delay filter unit 44 performs filtering by applying a 1st order lag filter. At this time, the output data is calculated as follows. The delay filter is also a filtering method to reduce the noise of the waveform.

&Quot; (4) "

Here, smooth and divider are values determined in advance as coefficients of the delay filter. The coefficients smooth and divider vary depending on how the value is given. in Old Data refers to the filtered data of the sampling data just before the time, and in Read Data is the current data. That is, the former is the (n-1) -th filtered data and the latter is the n-th sampled data.

The results of applying the first-order delay filter are as shown in FIG.

As shown in FIG. 8, the frequency characteristic of the periodic signal is not lost.

Next, the effects of the present invention will be described with reference to Figs. 9 to 12 through experiments of the present invention.

The noise filter system according to the present invention derives a filtered output waveform by measuring θ (phase difference) by removing noise of a periodic signal such as an inverter output. This makes it possible to compensate for the lack of data analysis (phase comparisons, difficulty in improving power factor, signal analysis problems, etc.) for existing power management of inverters and precision measurement. In addition, various applications such as rotation speed, operation efficiency, and torque measurement can be secured.

FIG. 9 is a result of measuring a three-phase current signal according to the related art, and FIG. 10 is a result of filtering a three-phase current signal according to the present invention.

11 is a result of applying a confidence interval average, and FIG. 12 is a result of applying a first-order delay filter.

Although the present invention has been described in detail with reference to the above embodiments, it is needless to say that the present invention is not limited to the above-described embodiments, and various modifications may be made without departing from the spirit of the present invention.

10: Inverter 20: Three phase power cable
30: Current sensor 40: Noise filter system
41: Sampling unit 42: Periodic filter unit
43: section filter section 44: delay filter section
50: Power calculation system 60: Load

Claims

1. A noise filter system for periodic signals,
A sampling unit for sampling the periodic signal and sampling the periodic signal in a period unit of the periodic signal to generate sampling data having a sampling size of N for each period; And
And a periodic filter unit filtering the sampling data of each cycle and filtering the nth sampling data of the cycle (1? N? N) by the characteristic values of the nth sampling data of the cycle before and after the corresponding cycle / RTI > for a periodic signal.

The method according to claim 1,
Wherein the periodic filter unit filters the n-th sampling data of the period and the median of the n-th sampling data of the period before and after the corresponding period.

The method according to claim 1,
Wherein the periodic filter unit includes filtering up to the sampling data of the Mth period before and after the corresponding period to filter the periodic signal.

The method according to claim 1,
Wherein the periodic filter unit is filtered according to the following equation (1): " (1) "
[Equation 1]

The system of claim 1,
Further comprising an interval filter unit for filtering the signal filtered by the period by a characteristic value including an average or a sum of data of the signal about the time before and after the signal.

6. The method of claim 5,
Wherein the section filter unit performs filtering according to the following equation (2): " (2) "
[Equation 1]