CN108334822B - Kalman and modified wavelet transform filtering method based on electric vehicle charging nonlinear load characteristics - Google Patents

Kalman and modified wavelet transform filtering method based on electric vehicle charging nonlinear load characteristics Download PDF

Info

Publication number
CN108334822B
CN108334822B CN201810052698.XA CN201810052698A CN108334822B CN 108334822 B CN108334822 B CN 108334822B CN 201810052698 A CN201810052698 A CN 201810052698A CN 108334822 B CN108334822 B CN 108334822B
Authority
CN
China
Prior art keywords
signal
kalman
charging
wavelet transform
waves
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201810052698.XA
Other languages
Chinese (zh)
Other versions
CN108334822A (en
Inventor
王立辉
祁顺然
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southeast University
Original Assignee
Southeast University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southeast University filed Critical Southeast University
Priority to CN201810052698.XA priority Critical patent/CN108334822B/en
Publication of CN108334822A publication Critical patent/CN108334822A/en
Application granted granted Critical
Publication of CN108334822B publication Critical patent/CN108334822B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Electric Propulsion And Braking For Vehicles (AREA)
  • Charge And Discharge Circuits For Batteries Or The Like (AREA)

Abstract

The invention discloses a Kalman and correction wavelet transform filtering method based on charging nonlinear load characteristics of an electric automobile, which comprises the following steps of: reasonably correcting threshold parameters and layering parameters of a traditional wavelet transformation algorithm according to the signal characteristics of direct current fundamental waves and unsteady state waves in the charging nonlinear load of the electric automobile; filtering unsteady state waves and higher harmonics by using modified wavelet transform; filtering low-order harmonic waves by using a Kalman filtering algorithm; detecting a signal mutation point; smoothing at a signal mutation point of a Kalman filtering algorithm; after the signal break point is detected in the last step, the fundamental wave signal at the signal break point is obtained by adopting modified wavelet transform decomposition instead of Kalman filtering algorithm decomposition at the subsequent time L. The method can improve the identification accuracy of the Kalman filtering algorithm, and avoids the phenomenon of 'climbing delay' of the Kalman filtering algorithm at a signal mutation position.

Description

Kalman and modified wavelet transform filtering method based on electric vehicle charging nonlinear load characteristics
Technical Field
The invention relates to the technical field of charging nonlinear load filtering of electric automobiles, in particular to a Kalman and correction wavelet transform filtering method based on charging nonlinear load characteristics of electric automobiles.
Background
The new energy automobile mainly comprises an electric automobile, and the large-scale development of the ultra-conventional electric automobile needs a large number of matched charging station supports. A charger of an electric vehicle charging station belongs to novel high-power nonlinear equipment, 150-600A of large current can be formed in the charging process, and excessive intensive centralized charging can cause excessive instantaneous load of the charging station. And the charging of the storage battery of the electric automobile belongs to capacitive load, the load power factor is lower, and the charging load shows nonlinear characteristics. Due to the complexity of the charging process, a large amount of harmonic waves and unsteady waves are generated in the charging process, and influence is generated on electric energy metering and power grid operation of the charging station.
At present, the most common method is to filter nonlinear loads in the charging process of the electric vehicle by adopting a fourier algorithm, decompose periodic signals into a superposition form of different frequency components according to a fourier series mode, have the advantages of high response speed, high data processing capacity, high calculation precision, good real-time performance and the like, and are suitable for detecting harmonic waves. The Short Time Fourier Transform (STFT) of the sliding window function used can improve this limitation, but its frequency resolution is fixed and it has no adaptive capability. The widths of a time window and a frequency window of the wavelet transformation algorithm can be adjusted, the sampling density is automatically adjusted to process the mutation signal according to the frequency components of the signal, the method is suitable for reflecting the sudden change and time-varying tracking of the signal, and is particularly suitable for analyzing the fluctuation harmonic wave, the fast change harmonic wave, the mutation signal and the non-stationary signal, but the signal characteristics of the fundamental wave and the multiple harmonic wave cannot be analyzed and calculated. In the field, a separate kalman filter algorithm is used for carrying out nonlinear load in the charging process of the electric vehicle, which can well identify and decompose each harmonic wave, but the existing data is used for prediction, when the signal changes rapidly, a certain buffer time is needed, unstable waves cannot be identified accurately, and the identification of signal components at the sudden change position of the signal has serious 'climbing delay'. The three methods cannot meet the requirements of real application on filtering.
Disclosure of Invention
The invention aims to solve the technical problem of providing a Kalman and modified wavelet transform filtering method based on the charging nonlinear load characteristics of an electric automobile, which can improve the identification accuracy of a Kalman filtering algorithm and avoid the phenomenon of 'climbing delay' of the Kalman filtering algorithm at a signal mutation position.
In order to solve the technical problem, the invention provides a Kalman and modified wavelet transform filtering method based on charging nonlinear load characteristics of an electric vehicle, which comprises the following steps:
(1) reasonably correcting threshold parameters and layering parameters of a traditional wavelet transformation algorithm according to the signal characteristics of direct current fundamental waves and unsteady state waves in the charging nonlinear load of the electric automobile;
(2) filtering unsteady state waves and higher harmonics by using modified wavelet transform;
(3) filtering low-order harmonic waves by using a Kalman filtering algorithm;
(4) detecting a signal mutation point;
(5) smoothing at a signal mutation point of a Kalman filtering algorithm; and (4) after the signal break point is detected in the step (4), decomposing by adopting modified wavelet transform decomposition to replace a Kalman filtering algorithm at the subsequent time L to obtain a fundamental wave signal at the signal break point.
Preferably, in the step (1), the modifying of the layering parameter specifically comprises: the frequency of a wave is 0Hz when the electric automobile is charged by direct current, n layers are added on the original basis when the number of decomposition layers is calculated, and the formula of the number of decomposition layers is as follows:
Figure BDA0001552811460000021
wherein f issTo sample frequency, f0The greatest common divisor of the ripple signal frequency, mu is 2;
the threshold parameter is modified specifically as follows: averaging the two thresholds;
Figure BDA0001552811460000022
wherein Ds is the processed data, D is the collected data, beta is the correction weight, and epsilon is the selected threshold; equation (2) is smoothed using a linear interpolation function.
Preferably, in the step (2), the filtering of the unsteady state waves and the higher harmonics by using the modified wavelet transform specifically comprises: the fundamental wave signal is positioned at the bottom layer part of the charging nonlinear load frequency of the electric automobile, and unsteady state waves and higher harmonics are filtered by utilizing modified wavelet transform;
Figure BDA0001552811460000023
wherein f (t) is a low-frequency signal of the charging nonlinear load of the electric automobile; χ () is a scale space function;
Figure BDA0001552811460000024
filtering unsteady state waves and higher harmonics for projection on a scale; e.g. of the typer,k、er,mIs an approximation parameter on the r scale; j is the sequence number of the fundamental wave, 1 is taken: m is a power series of scale discretization; and S () is a scale space filtering function and embodies the low-pass characteristic of the signal.
Preferably, in the step (3), filtering the low-order harmonic by using a kalman filter algorithm specifically includes the following steps:
(31) selecting an observation State
Figure BDA0001552811460000031
Wherein y is1,z,y2,zAs a group, orthogonal decomposition waveforms of fundamental wave and each harmonic; w is aiIs the amplitude of the ith harmonic; thetaiPhase angle of the ith harmonic;
(32) selecting system dynamic equations and measurement equations
Figure BDA0001552811460000032
Wherein W (K) is an observation matrix, YKIs an observation vector, ω, related to W (K)kFor process noise random sequence, vkIn order to observe the noise, it is,
Figure BDA0001552811460000033
to shift the matrix, xK+1A system dynamic matrix, wherein N is the harmonic (ripple) frequency in the charging nonlinear load of the electric automobile;
(33) identifying and tracking the amplitude and phase angle of the fundamental wave;
Figure BDA0001552811460000034
preferably, in the step (4), the detecting the signal mutation point specifically comprises the following steps:
(41) for the direct current charging of the electric automobile, data D at the current moment is collected0Data D before time intervals T, 2T, 3T1、D2、D3If, if
Figure BDA0001552811460000035
Are all formed intoImmediately, this time is the signal discontinuity, where σ1Is a determined threshold;
(42) for alternating current charging of the electric automobile, collecting amplitude delta w calculated by Kalman algorithm at the current moment0Time interval T1、2T1、3T1Amplitude of front Δ w1、Δw2、Δw3If, if
Figure BDA0001552811460000041
All hold true, then this moment is the signal discontinuity, where σ2Is a determined threshold.
The invention has the beneficial effects that: reasonably correcting threshold parameters and layering parameters of a traditional wavelet transformation algorithm according to the signal characteristics of direct current fundamental waves and unsteady state waves in the charging nonlinear load of the electric automobile, wherein the obtained corrected wavelet algorithm is more accurate in layering of the acquired alternating current and direct current charging signals of the electric automobile, and the filtered unsteady state waves are more accurate; after the influence of unsteady waves is eliminated by correcting wavelet transformation, fundamental wave signals are identified by using a Kalman filtering algorithm, so that the identification accuracy of the Kalman filtering algorithm is improved; and the signal mutation is detected, and the wavelet change algorithm is used for replacing the signal obtained by decomposition of the Kalman filtering algorithm, so that the phenomenon of 'climbing delay' of the Kalman filtering algorithm at the signal mutation position is avoided.
Drawings
FIG. 1 is a schematic flow chart of the method of the present invention.
Detailed Description
As shown in fig. 1, a kalman and modified wavelet transform filtering method based on nonlinear load characteristics of electric vehicle charging includes the following steps:
step 1: reasonably correcting threshold parameters and layering parameters of a traditional wavelet transformation algorithm according to the signal characteristics of direct current fundamental waves and unsteady state waves in the charging nonlinear load of the electric automobile;
(11) modifying the layering parameters;
the direct current charging time base wave frequency of the electric automobile is 0Hz, the meterWhen the number of decomposition layers is counted, f0Cannot be calculated at the fundamental frequency. In addition, the direct current has a form of a primary function and a secondary function, which greatly interferes with the wavelet decomposition level, so that n layers need to be added on the original basis, and the formula of the decomposition layer number is as follows:
Figure BDA0001552811460000042
wherein f issTo sample frequency, f0μ is 2, which is the greatest common divisor of the ripple signal frequency.
(12) Determination of threshold values
The invention reasonably corrects the classic soft threshold value method and the hard threshold value method, and adopts the discontinuity of the smooth change of the linear interpolation function, and the specific method is as follows:
firstly, averaging two thresholds;
Figure BDA0001552811460000043
wherein Ds is the processed data, D is the collected data, beta is the correction weight, and epsilon is the selected threshold.
And secondly, smoothing the equation (2) by utilizing a linear interpolation function.
Step 2: filtering unsteady state waves and higher harmonics by using modified wavelet transform;
the fundamental wave signal is positioned at the bottommost part of the charging nonlinear load frequency of the electric automobile, and unsteady state waves and higher harmonics can be filtered out by utilizing the modified wavelet transform.
Figure BDA0001552811460000051
Wherein f (t) is a low-frequency signal of the charging nonlinear load of the electric automobile; χ () is a scale space function;
Figure BDA0001552811460000052
for projection on a scale, realizing non-steady stateFiltering waves and higher harmonics; e.g. of the typer,k、er,mIs an approximation parameter on the r scale; j is the sequence number of the fundamental wave, 1 is taken: m is a power series of scale discretization; and S () is a scale space filtering function and embodies the low-pass characteristic of the signal.
And step 3: filtering low order harmonic wave by Kalman filtering algorithm
Selecting observation state
Figure BDA0001552811460000053
Wherein y is1,z,y2,zAs a group, orthogonal decomposition waveforms of fundamental wave and each harmonic; w is aiIs the amplitude of the ith harmonic; thetaiPhase angle of the ith harmonic.
Selecting system dynamic equation and measuring equation
Figure BDA0001552811460000054
Wherein W (K) is an observation matrix, YKIs an observation vector, ω, related to W (K)kFor process noise random sequence, vkIn order to observe the noise, it is,
Figure BDA0001552811460000055
to shift the matrix, xK+1And (3) a system dynamic matrix, wherein N is the harmonic (ripple) frequency in the charging nonlinear load of the electric automobile.
Identifying and tracing fundamental wave amplitude and phase angle
Figure BDA0001552811460000061
And 4, step 4: detecting signal discontinuities
Firstly, for the direct current charging of the electric automobile, the data D at the current moment is collected0Data D before time intervals T, 2T, 3T1、D2、D3If, if
Figure BDA0001552811460000062
All hold true, then this moment is the signal discontinuity, where σ1Is a determined threshold.
Secondly, for alternating current charging of the electric automobile, collecting the amplitude delta w calculated by the Kalman algorithm at the current moment0Time interval T1、2T1、3T1Amplitude of front Δ w1、Δw2、Δw3If, if
Figure BDA0001552811460000063
All hold true, then this moment is the signal discontinuity, where σ2Is a determined threshold.
And 5: smoothing of Kalman filtering algorithm signal discontinuities
And (4) after the signal break point is detected in the step (4), decomposing by adopting modified wavelet change decomposition to replace a Kalman filtering algorithm at the subsequent time L to obtain a fundamental wave signal at the signal break point.
According to the invention, the threshold parameter and the layering parameter of the traditional wavelet transform algorithm are reasonably corrected according to the signal characteristics of direct current fundamental wave and unsteady state wave in the charging nonlinear load of the electric automobile, the obtained corrected wavelet algorithm more accurately layers the acquired alternating current and direct current charging signals of the electric automobile, and the filtered unsteady state wave is more accurate; after the influence of unsteady waves is eliminated by correcting wavelet transformation, fundamental wave signals are identified by using a Kalman filtering algorithm, so that the identification accuracy of the Kalman filtering algorithm is improved; and the signal mutation is detected, and the wavelet change algorithm is used for replacing the signal obtained by decomposition of the Kalman filtering algorithm, so that the phenomenon of 'climbing delay' of the Kalman filtering algorithm at the signal mutation position is avoided.

Claims (3)

1. The Kalman and correction wavelet transform filtering method based on the charging nonlinear load characteristics of the electric automobile is characterized by comprising the following steps of:
(1) reasonably correcting threshold parameters and layering parameters of a traditional wavelet transformation algorithm according to the signal characteristics of direct current fundamental waves and unsteady state waves in the charging nonlinear load of the electric automobile;
(2) filtering unsteady state waves and higher harmonics by using modified wavelet transform; the fundamental wave signal is positioned at the bottom layer part of the charging nonlinear load frequency of the electric automobile, and unsteady state waves and higher harmonics are filtered by utilizing modified wavelet transform;
Figure FDA0003041395290000011
wherein f (t) is a low-frequency signal of the charging nonlinear load of the electric automobile; χ () is a scale space function;
Figure FDA0003041395290000012
filtering unsteady state waves and higher harmonics for projection on a scale; e.g. of the typer,k、er,mIs an approximation parameter on the r scale; j is the sequence number of the fundamental wave, and 1 is taken; m is a power series of scale discretization; s () is a scale space filtering function and embodies the low-pass characteristic of a signal;
(3) filtering low-order harmonic waves by using a Kalman filtering algorithm; the method specifically comprises the following steps:
(31) selecting an observation State
Figure FDA0003041395290000013
Wherein y is1,z,y2,zAs a group, orthogonal decomposition waveforms of fundamental wave and each harmonic; w is aiIs the amplitude of the ith harmonic; thetaiPhase angle of the ith harmonic;
(32) selecting system dynamic equations and measurement equations
Figure FDA0003041395290000014
Wherein W (K) is an observation matrix, YKIs an observation vector, ω, related to W (K)kFor process noise random sequence, vkIn order to observe the noise, it is,
Figure FDA0003041395290000015
to shift the matrix, xK+1System dynamic matrix, fNCharging the electric vehicle with harmonic frequency in the nonlinear load;
(33) identifying and tracking the amplitude and phase angle of the fundamental wave;
Figure FDA0003041395290000021
(4) detecting a signal mutation point;
(5) smoothing at a signal mutation point of a Kalman filtering algorithm; and (4) after the signal break point is detected in the step (4), decomposing by adopting modified wavelet transform decomposition to replace a Kalman filtering algorithm at the subsequent time L to obtain a fundamental wave signal at the signal break point.
2. The Kalman and modified wavelet transform filtering method based on nonlinear load characteristics for electric vehicle charging according to claim 1, characterized in that in step (1), the modification of the hierarchical parameters specifically comprises: the frequency of a wave is 0Hz when the electric automobile is charged by direct current, n layers are added on the original basis when the number of decomposition layers is calculated, and the formula of the number of decomposition layers is as follows:
Figure FDA0003041395290000022
wherein f issTo sample frequency, f0The greatest common divisor of the ripple signal frequency, mu is 2;
the threshold parameter is modified specifically as follows: averaging the two thresholds;
Figure FDA0003041395290000023
wherein Ds is the processed data, D is the collected data, beta is the correction weight, and epsilon is the selected threshold; equation (2) is smoothed using a linear interpolation function.
3. The Kalman and modified wavelet transform filtering method based on nonlinear load characteristics for electric vehicle charging according to claim 1, wherein in the step (4), the step of detecting the signal discontinuity specifically comprises the following steps:
(41) for the direct current charging of the electric automobile, data D at the current moment is collected0Data D before time intervals T, 2T, 3T1、D2、D3If, if
Figure FDA0003041395290000024
All hold true, then this moment is the signal discontinuity, where σ1Is a determined threshold;
(42) for alternating current charging of the electric automobile, collecting amplitude delta w calculated by Kalman algorithm at the current moment0Time interval T1、2T1、3T1Amplitude of front Δ w1、Δw2、Δw3If, if
Figure FDA0003041395290000031
All hold true, then this moment is the signal discontinuity, where σ2Is a determined threshold.
CN201810052698.XA 2018-01-19 2018-01-19 Kalman and modified wavelet transform filtering method based on electric vehicle charging nonlinear load characteristics Active CN108334822B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810052698.XA CN108334822B (en) 2018-01-19 2018-01-19 Kalman and modified wavelet transform filtering method based on electric vehicle charging nonlinear load characteristics

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810052698.XA CN108334822B (en) 2018-01-19 2018-01-19 Kalman and modified wavelet transform filtering method based on electric vehicle charging nonlinear load characteristics

Publications (2)

Publication Number Publication Date
CN108334822A CN108334822A (en) 2018-07-27
CN108334822B true CN108334822B (en) 2021-07-27

Family

ID=62925209

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810052698.XA Active CN108334822B (en) 2018-01-19 2018-01-19 Kalman and modified wavelet transform filtering method based on electric vehicle charging nonlinear load characteristics

Country Status (1)

Country Link
CN (1) CN108334822B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110163148B (en) * 2019-05-21 2021-02-09 东南大学 Self-adaptive identification method for direct-current charging distortion signal of electric vehicle
CN111695618B (en) * 2020-06-01 2023-04-07 清华大学深圳国际研究生院 Electric vehicle motor fault detection method based on OBD data

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105425039A (en) * 2015-12-29 2016-03-23 南京因泰莱电器股份有限公司 Harmonic detection method based on adaptive kalman filtering
CN106645929A (en) * 2016-09-30 2017-05-10 国网江苏省电力公司电力科学研究院 Improved electric vehicle charging non-linear load harmonic wave electric energy metering method
CN106908663A (en) * 2017-03-07 2017-06-30 国网江苏省电力公司电力科学研究院 A kind of charging electric vehicle harmonic identification method based on wavelet transformation
CN106980044A (en) * 2017-03-22 2017-07-25 西南交通大学 A kind of Harmonious Waves in Power Systems current estimation method for adapting to wind power integration

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105425039A (en) * 2015-12-29 2016-03-23 南京因泰莱电器股份有限公司 Harmonic detection method based on adaptive kalman filtering
CN106645929A (en) * 2016-09-30 2017-05-10 国网江苏省电力公司电力科学研究院 Improved electric vehicle charging non-linear load harmonic wave electric energy metering method
CN106908663A (en) * 2017-03-07 2017-06-30 国网江苏省电力公司电力科学研究院 A kind of charging electric vehicle harmonic identification method based on wavelet transformation
CN106980044A (en) * 2017-03-22 2017-07-25 西南交通大学 A kind of Harmonious Waves in Power Systems current estimation method for adapting to wind power integration

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
EMD阈值滤波在光纤陀螺漂移信号去噪中的应用;崔冰波等;《光学学报》;20150210;全文 *

Also Published As

Publication number Publication date
CN108334822A (en) 2018-07-27

Similar Documents

Publication Publication Date Title
CN107590317B (en) Generator dynamic estimation method considering model parameter uncertainty
CN106845010B (en) Low-frequency oscillation dominant mode identification method based on improved SVD noise reduction and Prony
CN103944174B (en) The low-frequency oscillation on-line identification method of algorithm of making an uproar is filtered based on cross-correlation function
CN109521275B (en) Synchronous phasor determination method, system, device and readable storage medium
CN111046327B (en) Prony analysis method suitable for low-frequency oscillation and subsynchronous oscillation identification
CN110967599A (en) Electric energy quality disturbance detection and positioning algorithm
CN107543962B (en) Calculation method of dominant inter-harmonic frequency spectrum distribution
CN109633266B (en) Frequency measurement method, system, device and computer readable storage medium
CN108334822B (en) Kalman and modified wavelet transform filtering method based on electric vehicle charging nonlinear load characteristics
CN111077386A (en) Early fault signal noise reduction method for electrical equipment
CN113032716A (en) Harmonic and inter-harmonic analysis method based on windowed interpolation and Prony algorithm
CN108197073B (en) Improved electric vehicle charging electric energy signal characteristic analysis method
CN110163148B (en) Self-adaptive identification method for direct-current charging distortion signal of electric vehicle
CN105486921A (en) Kaiser third-order mutual convolution window triple-spectrum-line interpolation harmonic wave and inter-harmonic wave detection method
CN115618213A (en) Charger voltage disturbance analysis method, system, equipment and storage medium
CN114065114A (en) Method and system for predicting metering error of capacitive voltage transformer
CN110659620A (en) Filtering noise reduction method based on fuzzy control
CN110287853B (en) Transient signal denoising method based on wavelet decomposition
CN110098610B (en) Real-time identification method and system for oscillation leading mode of power system under fault disturbance
CN110045250B (en) Method and system for judging insulation state of power cable
US20170363670A1 (en) Noise spectrum analysis for electronic device
CN109254202B (en) Synchronous phasor measurement device applied to power distribution network
CN113358922B (en) Fundamental wave, harmonic wave and inter-harmonic wave electric energy power measurement method of unsteady state power grid signal
CN114184838A (en) Power system harmonic detection method, system and medium based on SN mutual convolution window
CN112269058A (en) Electric automobile direct current charging signal feature extraction method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant