CN108334822A - Kalman based on electric vehicle charging nonlinear-load feature and amendment wavelet transform filtering method - Google Patents
Kalman based on electric vehicle charging nonlinear-load feature and amendment wavelet transform filtering method Download PDFInfo
- Publication number
- CN108334822A CN108334822A CN201810052698.XA CN201810052698A CN108334822A CN 108334822 A CN108334822 A CN 108334822A CN 201810052698 A CN201810052698 A CN 201810052698A CN 108334822 A CN108334822 A CN 108334822A
- Authority
- CN
- China
- Prior art keywords
- wave
- electric vehicle
- nonlinear
- kalman
- vehicle charging
- 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.)
- Granted
Links
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)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
- Electric Propulsion And Braking For Vehicles (AREA)
Abstract
The invention discloses a kind of Kalman based on electric vehicle charging nonlinear-load feature and wavelet transform filtering method is corrected, is included the following steps：Threshold parameter, the hierarchy parameters of traditional wavelet algorithm are rationally corrected according to direct current fundamental wave, unstable state wave signal characteristic in electric vehicle charging nonlinear-load；Unstable state wave and higher hamonic wave are filtered out using wavelet transformation is corrected；Low-order harmonic is filtered out using Kalman filtering algorithm；Detect sign mutation point；It is smooth at Kalman filtering algorithm sign mutation point；After being detected at sign mutation point by previous step, time L behind decomposes to obtain the fundamental signal at sign mutation point using wavelet transformation decomposition replacement Kalman filtering algorithm is corrected.The present invention can improve the recognition accuracy of Kalman filtering algorithm, avoid " climbing delay " phenomenon of Kalman filtering algorithm at sign mutation.
Description
Technical field
It is especially a kind of to be filled based on electric vehicle the present invention relates to electric vehicle charging nonlinear-load technical field of filtering
The Kalman of electric nonlinear-load feature and amendment wavelet transform filtering method.
Background technology
For new-energy automobile based on electric vehicle, unconventional electric vehicle large-scale development needs a large amount of mating chargings
It stands support.The charger of electric automobile charging station belongs to novel high-power non-linear equipment, it can form 150 in charging process~
The high current of 600A, and excessively intensive concentration charging may cause charging station instantaneous load excessive.And electric vehicle electric power storage
Pond charging belongs to capacitive load, and load power factor is relatively low, and charging load embodies nonlinear characteristic.The complexity of charging process, leads
It causes that in charging process a large amount of harmonic wave, unstable state wave will be generated, the electrical energy measurement and operation of power networks of charging station is had an impact.
Currently, most commonly being filtered to nonlinear-load in electric vehicle charging process using Fourier algorithm
Cyclical signal is decomposed into the stacking pattern of different frequency component by wave in the way of fourier series, has response speed
Soon, the advantages that data-handling capacity is high, computational accuracy is high, real-time is good is suitable for detection harmonic wave, but Fourier algorithm is as one
Mathematics variation kind of overall importance the unstable state wave such as does not have localization analysis ability, cannot analyze shock wave, can not accurately identify
The DC charging signal of non-constant.The short time discrete Fourier transform (STFT) of the sliding window function of use can improve this limitation
Property, but its time frequency resolution immobilizes, and does not have adaptive ability.Wavelet Transformation Algorithm when window and the width of frequency window can be with
It adjusts, the frequency content of basis signal, automatically adjusts sampling density to handle jump signal, be suitble to the sudden change of reflection signal
It is tracked with time-varying, is particularly suitable for fluctuation harmonic wave, quickly changes the analysis of harmonic wave, jump signal and non-stationary signal, but can not divide
Analysis calculates the signal characteristic of fundamental wave, multiple harmonic.Electric vehicle was also once filled using individual Kalman filtering algorithm in this field
Nonlinear-load carries out in electric process, can identify decomposition each harmonic very well, but predicted by data with existing, work as signal
When there is quickly variation, certain buffer time is needed, still can not accurately identify unstable state wave, and to letter at sign mutation
There are serious " climbing delays " for the identification of number ingredient.Three kinds of methods cannot be satisfied requirement of the practical application to filtering.
Invention content
Technical problem to be solved by the present invention lies in provided and a kind of charged nonlinear-load feature based on electric vehicle
Kalman and amendment wavelet transform filtering method, can improve the recognition accuracy of Kalman filtering algorithm, avoid Kalman
" climbing delay " phenomenon of filtering algorithm at sign mutation.
In order to solve the above technical problems, the present invention provides a kind of karr for the nonlinear-load feature that charges based on electric vehicle
Graceful and amendment wavelet transform filtering method, includes the following steps：
(1) Traditional Wavelet is become according to direct current fundamental wave, unstable state wave signal characteristic in electric vehicle charging nonlinear-load
Threshold parameter, the hierarchy parameters of scaling method are rationally corrected；
(2) unstable state wave and higher hamonic wave are filtered out using amendment wavelet transformation；
(3) low-order harmonic is filtered out using Kalman filtering algorithm；
(4) sign mutation point is detected；
(5) smooth at Kalman filtering algorithm sign mutation point；After being detected at sign mutation point by step (4),
Time L behind decomposes to obtain the fundamental wave at sign mutation point using wavelet transformation decomposition replacement Kalman filtering algorithm is corrected
Signal.
Preferably, in step (1), the amendment to hierarchy parameters is specially：Fundamental frequency is when electric vehicle DC charging
0Hz increases n-layer, the formula of Decomposition order is on the basis of the original when calculating Decomposition order：
Wherein fsFor sample frequency, f0For the greatest common divisor of ripple signal frequency, μ=2；
Amendment to threshold parameter is specially：Two kinds of threshold values are averaged；
Wherein, Ds is treated data, and D be the data of acquisition, and β is to correct flexible strategy, and ε is the threshold value of selection；Utilize line
Property interpolating function, formula (2) is carried out smooth.
Preferably, in step (2), using correcting, wavelet transformation filters out unstable state wave and higher hamonic wave is specially：Fundamental wave is believed
Number in electric vehicle charging nonlinear-load frequency bottom part, unstable state wave and high order are filtered out using wavelet transformation is corrected
Harmonic wave；
F (t) is the low frequency signal of electric vehicle charging nonlinear-load in formula；χ () is scale space function；For the projection on scale, realization filters out unstable state wave and higher hamonic wave；er,k、er,mFor r rulers
Approximating parameter on degree；J is fundamental wave serial number, takes 1：M is the power series of Scale Discreteness；S () is Scale Space Filtering letter
Number embodies low pass signal characteristic.
Preferably, it in step (3), filters out low-order harmonic using Kalman filtering algorithm and specifically comprises the following steps：
(31) observation state is selected
Wherein y1,z,y2,zIt is one group, is the Orthogonal Decomposition waveform of fundamental wave and each harmonic；wiFor the amplitude of ith harmonic wave；
θiThe phase angle of ith harmonic wave；
(32) system dynamical equation and measurement equation are selected
Wherein, W (K) is observing matrix, YKFor related observation vector, ω with W (K)kFor process noise random sequence, vk
For observation noise,For transfer matrix, xK+1System dynamic matrix, N are electric vehicle charging nonlinear-load
Middle harmonic wave (ripple) frequency；
(33) identify, track the amplitude and phase angle of fundamental wave；
Preferably, in step (4), detection sign mutation point specifically comprises the following steps：
(41) for electric vehicle DC charging, current time data D is acquired0, the data D before time interval T, 2T, 3T1、
D2、D3If
It sets up, then this moment is sign mutation point, wherein σ1For determining threshold value；
(42) the amplitude Δ w calculated for electric vehicle alternating-current charging, acquisition current time Kalman Algorithm0, between the time
Every T1、2T1、3T1Preceding amplitude Δ w1、Δw2、Δw3If
It sets up, then this moment is sign mutation point, wherein σ2For determining threshold value.
Beneficial effects of the present invention are：According to direct current fundamental wave, unstable state wave signal in electric vehicle charging nonlinear-load
Feature rationally corrects threshold parameter, the hierarchy parameters of traditional wavelet algorithm, and obtained amendment wavelet algorithm is to adopting
The electric vehicle alternating current-direct current charging signals layering of collection is more accurate, and the unstable state wave filtered out is more accurate；Become by correcting small echo
After changing the influence for eliminating unstable state wave, Kalman filtering algorithm identification fundamental signal is reused, Kalman filtering algorithm is improved
Recognition accuracy；Sign mutation is detected, and the signal that Kalman filtering algorithm decomposes is replaced using Wavelet transformation algorithm,
Avoid " climbing delay " phenomenon of Kalman filtering algorithm at sign mutation.
Description of the drawings
Fig. 1 is the method flow schematic diagram of the present invention.
Specific implementation mode
As shown in Figure 1, a kind of Kalman and amendment wavelet transformation filter based on electric vehicle charging nonlinear-load feature
Wave method includes the following steps：
Step 1：According to direct current fundamental wave, unstable state wave signal characteristic in electric vehicle charging nonlinear-load to Traditional Wavelet
Threshold parameter, the hierarchy parameters for becoming scaling method are rationally corrected；
(11) to the amendment of hierarchy parameters；
Fundamental frequency is 0Hz when electric vehicle DC charging, when calculating Decomposition order, f0It cannot be calculated with fundamental frequency.
And there are the forms of linear function and quadratic function in direct current, have very big interference to wavelet decomposition level, so in original base
It needs to increase n-layer on plinth, the formula of Decomposition order is：
Wherein fsFor sample frequency, f0For the greatest common divisor of ripple signal frequency, μ=2.
(12) determination of threshold value
The present invention reasonably corrects classical Soft thresholding and hard threshold method, and smooth using linear interpolation function
The discontinuity of variation, the specific method is as follows：
1. being averaged first to two kinds of threshold values；
Wherein, Ds is treated data, and D be the data of acquisition, and β is to correct flexible strategy, and ε is the threshold value of selection.
2. utilizing linear interpolation function, (2) formula is carried out smooth.
Step 2：Unstable state wave and higher hamonic wave are filtered out using wavelet transformation is corrected；
Fundamental signal is in electric vehicle charging nonlinear-load frequency bottom part, can be with using wavelet transformation is corrected
Filter out unstable state wave and higher hamonic wave.
F (t) is the low frequency signal of electric vehicle charging nonlinear-load in formula；χ () is scale space function；For the projection on scale, realization filters out unstable state wave and higher hamonic wave；er,k、er,mFor r rulers
Approximating parameter on degree；J is fundamental wave serial number, takes 1：M is the power series of Scale Discreteness；S () is Scale Space Filtering letter
Number embodies low pass signal characteristic.
Step 3：Low-order harmonic is filtered out using Kalman filtering algorithm
1. selecting observation state
Wherein y1,z,y2,zIt is one group, is the Orthogonal Decomposition waveform of fundamental wave and each harmonic；wiFor the amplitude of ith harmonic wave；
θiThe phase angle of ith harmonic wave.
2. selecting system dynamical equation and measurement equation
Wherein, W (K) is observing matrix, YKFor related observation vector, ω with W (K)kFor process noise random sequence, vk
For observation noise,For transfer matrix, xK+1System dynamic matrix, N are electric vehicle charging nonlinear-load
Middle harmonic wave (ripple) frequency.
3. the amplitude and phase angle of identification, tracking fundamental wave
Step 4：Detect sign mutation point
1. for electric vehicle DC charging, current time data D is acquired0, the data D before time interval T, 2T, 3T1、
D2、D3If
It sets up, then this moment is sign mutation point, wherein σ1For determining threshold value.
2. for electric vehicle alternating-current charging, the amplitude Δ w of acquisition current time Kalman Algorithm calculating0, time interval
T1、2T1、3T1Preceding amplitude Δ w1、Δw2、Δw3If
It sets up, then this moment is sign mutation point, wherein σ2For determining threshold value.
Step 5：It is smooth at Kalman filtering algorithm sign mutation point
After at the sign mutation point detected by step 4, time L behind decomposes replacement using Wavelet transformation is corrected
Kalman filtering algorithm decomposes to obtain the fundamental signal at sign mutation point.
The present invention is according to direct current fundamental wave, unstable state wave signal characteristic in electric vehicle charging nonlinear-load to Traditional Wavelet
Threshold parameter, the hierarchy parameters for becoming scaling method are rationally corrected, and obtained wavelet algorithm of correcting hands over the electric vehicle of acquisition
The layering of DC charging signal is more accurate, and the unstable state wave filtered out is more accurate；Unstable state wave is eliminated by correcting wavelet transformation
Influence after, reuse Kalman filtering algorithm identification fundamental signal, improve the recognition accuracy of Kalman filtering algorithm；Inspection
Sign mutation is surveyed, and the signal that Kalman filtering algorithm decomposes is replaced using Wavelet transformation algorithm, avoids Kalman's filter
" climbing delay " phenomenon of wave algorithm at sign mutation.
Claims (5)
1. Kalman and amendment wavelet transform filtering method, feature based on electric vehicle charging nonlinear-load feature exist
In including the following steps：
(1) traditional wavelet is calculated according to direct current fundamental wave, unstable state wave signal characteristic in electric vehicle charging nonlinear-load
Threshold parameter, the hierarchy parameters of method are rationally corrected；
(2) unstable state wave and higher hamonic wave are filtered out using amendment wavelet transformation；
(3) low-order harmonic is filtered out using Kalman filtering algorithm；
(4) sign mutation point is detected；
(5) smooth at Kalman filtering algorithm sign mutation point；After being detected at sign mutation point by step (4), at it
Time L afterwards decomposes to obtain the letter of the fundamental wave at sign mutation point using wavelet transformation decomposition replacement Kalman filtering algorithm is corrected
Number.
2. the Kalman as described in claim 1 based on electric vehicle charging nonlinear-load feature and amendment wavelet transformation filter
Wave method, which is characterized in that in step (1), the amendment to hierarchy parameters is specially：Fundamental frequency when electric vehicle DC charging
Increase n-layer on the basis of the original, the formula of Decomposition order is when calculating Decomposition order for 0Hz：
Wherein fsFor sample frequency, f0For the greatest common divisor of ripple signal frequency, μ=2；
Amendment to threshold parameter is specially：Two kinds of threshold values are averaged；
Wherein, Ds is treated data, and D be the data of acquisition, and β is to correct flexible strategy, and ε is the threshold value of selection；It is inserted using linear
Value function carries out formula (2) smooth.
3. the Kalman as described in claim 1 based on electric vehicle charging nonlinear-load feature and amendment wavelet transformation filter
Wave method, which is characterized in that in step (2), wavelet transformation filters out unstable state wave and higher hamonic wave is specially using correcting：Fundamental wave
Signal is in electric vehicle charging nonlinear-load frequency bottom part, and unstable state wave and height are filtered out using wavelet transformation is corrected
Subharmonic；
F (t) is the low frequency signal of electric vehicle charging nonlinear-load in formula；χ () is scale space function；For the projection on scale, realization filters out unstable state wave and higher hamonic wave；er,k、er,mFor r rulers
Approximating parameter on degree；J is fundamental wave serial number, takes 1：M is the power series of Scale Discreteness；S () is Scale Space Filtering letter
Number embodies low pass signal characteristic.
4. the Kalman as described in claim 1 based on electric vehicle charging nonlinear-load feature and amendment wavelet transformation filter
Wave method, which is characterized in that in step (3), filter out low-order harmonic using Kalman filtering algorithm and specifically comprise the following steps：
(31) observation state is selected
Wherein y1,z,y2,zIt is one group, is the Orthogonal Decomposition waveform of fundamental wave and each harmonic；wiFor the amplitude of ith harmonic wave；θiI-th
The phase angle of subharmonic；
(32) system dynamical equation and measurement equation are selected
Wherein, W (K) is observing matrix, YKFor related observation vector, ω with W (K)kFor process noise random sequence, vkTo see
Noise is surveyed,For transfer matrix, xK+1System dynamic matrix, N are humorous in electric vehicle charging nonlinear-load
Wave (ripple) frequency；
(33) identify, track the amplitude and phase angle of fundamental wave；
5. the Kalman as described in claim 1 based on electric vehicle charging nonlinear-load feature and amendment wavelet transformation filter
Wave method, which is characterized in that in step (4), detection sign mutation point specifically comprises the following steps：
(41) for electric vehicle DC charging, current time data D is acquired0, the data D before time interval T, 2T, 3T1、D2、
D3If
It sets up, then this moment is sign mutation point, wherein σ1For determining threshold value；
(42) the amplitude Δ w calculated for electric vehicle alternating-current charging, acquisition current time Kalman Algorithm0, time interval T1、
2T1、3T1Preceding amplitude Δ w1、Δw2、Δw3If
It sets up, then this moment is sign mutation point, wherein σ2For determining threshold value.
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 true CN108334822A (en) | 2018-07-27 |
| CN108334822B 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) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110163148A (en) * | 2019-05-21 | 2019-08-23 | 东南大学 | A kind of electric car DC charging distorted signal self-adaptive identification method |
| CN111695618A (en) * | 2020-06-01 | 2020-09-22 | 清华大学深圳国际研究生院 | Electric vehicle motor fault detection method based on OBD data |
Citations (4)
| 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 |
-
2018
- 2018-01-19 CN CN201810052698.XA patent/CN108334822B/en active Active
Patent Citations (4)
| 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)
| Title |
|---|
| 崔冰波等: "EMD阈值滤波在光纤陀螺漂移信号去噪中的应用", 《光学学报》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110163148A (en) * | 2019-05-21 | 2019-08-23 | 东南大学 | A kind of electric car DC charging distorted signal self-adaptive identification method |
| 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 |
| CN111695618A (en) * | 2020-06-01 | 2020-09-22 | 清华大学深圳国际研究生院 | Electric vehicle motor fault detection method based on OBD data |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108334822B (en) | 2021-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103177188A (en) | Electric system load dynamic characteristic classifying method based on characteristic mapping | |
| CN101984360B (en) | Normalized leakage LMS self-adaptive mobile target detector based on FRFT | |
| CN1200177A (en) | Method of locating a single-phase ground fault in a power distribution network | |
| CN101806832A (en) | Measuring method for frequencies of low-frequency signals | |
| CN104320157B (en) | A kind of power line bi-directional Glenn shunt upward signal detection method | |
| Fan et al. | Transmission line fault location using deep learning techniques | |
| CN108334822A (en) | Kalman based on electric vehicle charging nonlinear-load feature and amendment wavelet transform filtering method | |
| CN110163148B (en) | Self-adaptive identification method for direct-current charging distortion signal of electric vehicle | |
| CN102095929A (en) | Method for rapidly measuring frequency of alternating-current signals | |
| CN104833852A (en) | Power system harmonic signal estimation and measurement method based on artificial neural network | |
| CN106137184B (en) | Electrocardiosignal QRS complex detection method based on wavelet transformation | |
| CN103904652B (en) | Power supply network harmonic suppression method and system capable of resisting impulse noise interference | |
| CN108197073A (en) | A kind of improved electric vehicle rechargeable electrical energy signature analysis method | |
| CN113514743A (en) | Construction method of GIS partial discharge pattern recognition system based on multi-dimensional features | |
| CN110287853B (en) | Transient signal denoising method based on wavelet decomposition | |
| Aslan et al. | ANN based fault location for medium voltage distribution lines with remote-end source | |
| CN104808060A (en) | Method for digitally measuring the phase difference of electrical signals | |
| CN115128400A (en) | Distribution network fault type identification and fault route selection comprehensive research and judgment method and system | |
| CN104808055A (en) | Electrical signal frequency digitized measurement method | |
| CN112269058A (en) | Electric automobile direct current charging signal feature extraction method | |
| CN114184838A (en) | Power system harmonic detection method, system and medium based on SN mutual convolution window | |
| CN104382586A (en) | Electric shock signal detecting method and device | |
| CN117872039B (en) | Line fault location method and system based on improved RBF network | |
| CN110441663B (en) | Method for judging direct current series arc fault based on frequency domain stage ratio value | |
| Siavashi et al. | Detection of voltage sag using unscented Kalman smoother |
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 |