Harmonic self-compensation control method and system for direct-current charging pile

Abstract

The invention provides a DC charging pile harmonic wave self-compensation control method and system, and the method and system comprise the steps: signal collection and preprocessing, harmonic wave detection and feature extraction, harmonic wave compensation strategy making, harmonic wave compensation execution, and system optimization and maintenance. According to the DC charging pile harmonic self-compensation control method and system provided by the invention, a dynamic self-adaptive harmonic compensation strategy is adopted, the compensation signal can be dynamically generated according to the difference value between the harmonic current and the limit, the response time of the compensation device is shortened, and the response time is shortened compared with that of a traditional active filter; the problems of two core pain points, insufficient compensation real-time performance and poor dynamic working condition adaptability in harmonic suppression are solved, the harmonic current change can be rapidly tracked, it is ensured that the compensation current and the harmonic current are equal in size and opposite in phase all the time, the harmonic suppression rate is high, the common connection point is stable, and the power grid stability is high. And the large-scale power failure risk caused by resonance is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of circuit, in particular to a control method and system for harmonic self-compensation of direct current charging pile. BACKGROUND

[0002] With the rapid growth of the number of electric vehicles, direct current charging piles as the core charging infrastructure, its large-scale access to the power grid has a significant impact on power quality. Because the charging pile contains a large number of nonlinear power electronic devices such as IGBT, diode rectifier, etc., a large amount of harmonic current is generated during operation, which causes voltage distortion of the power grid, increases the heating loss of equipment, and even causes relay protection misoperation. Among them, the six-pulse rectifier type charging pile has a simple structure and low cost, and reduces about 30%~50% compared with the twelve-pulse or higher pulse number scheme, and currently accounts for more than 70% of the market share, but its harmonic characteristics have a particularly prominent impact on the stability of the power grid.

[0003] In the power grid, nonlinear loads are prone to generate harmonics, which will impact the stability of the power system. When the charging pile is working, there are many nonlinear device loads, and accurate detection of harmonics is beneficial to the design of power filters for harmonic suppression.

[0004] Taking the six-pulse rectifier type charging pile as an example, the six-pulse rectifier type charging pile has relatively low design and manufacturing costs, and is one of the most widely used charging piles. Its equivalent topology structure is relatively simple, only the conventional rectification and filtering circuit, and the DC-DC power conversion circuit. Since the pulse number of the six-pulse rectifier is 6, it mainly generates sub-harmonics, and the effective value of the harmonic current is inversely proportional to the harmonic number. The higher the harmonic number, the smaller the corresponding harmonic current amplitude. However, the harmonic current produces a voltage drop on the grid impedance, resulting in an excessive THD of the common connection point voltage, such as GB / T 14549-93 requiring a voltage THD≤5%, which causes misoperation of other sensitive loads.

[0005] The most common harmonic analysis method is Fourier transform, i.e. FFT, which has strong frequency domain analysis capability but weak time domain analysis capability, and large errors occur when encountering nonlinear transient interference. Wavelet transform has strong time domain analysis capability, not only can strip out transient interference noise, but also can locate when the interference occurs. However, the frequency domain analysis capability of wavelet transform is weak, and the existing control method for harmonic self-compensation of direct current charging piles lacks adaptability in the use of fixed compensation parameters, such as the LC parameters of passive filters, which cannot cope with the dynamic changes of harmonic spectrum. The compensation delay of active filters is high, resulting in a phase deviation between the compensation current and the actual harmonic, reducing the compensation effect and causing inconvenience to actual use.

[0006] Therefore, it is necessary to provide a new DC charging pile harmonic self-compensation control method and system to solve the above technical problems. SUMMARY

[0007] To solve the above technical problems, the application provides a DC charging pile harmonic self-compensation control method and system.

[0008] The DC charging pile harmonic self-compensation control method provided by the application comprises the following operation steps:

[0009] Step one, signal acquisition and preprocessing: install a current sensor at the output end of the charging pile, real-time acquire the current signal output by the charging pile, and perform low-pass filtering on the acquired current signal to remove high-frequency noise while retaining harmonic components;

[0010] Step two, harmonic detection and feature extraction: Fourier transform the preprocessed current signal to obtain the preliminary amplitude or phase information of each harmonic, use wavelet transform to perform multi-scale decomposition on the current signal, extract transient interference signals and fundamental signals, input the denoised current signal into an Adaline neural network, and adjust the weight vector through an LMS algorithm, update the weight vector according to the difference between the expected output and the neuron output, and finally obtain the amplitude and phase of each harmonic;

[0011] Step three, develop a harmonic compensation strategy: calculate the total harmonic current generated by the charging pile according to the harmonic amplitude and phase output by the Adaline network, set the maximum harmonic current limit value allowed according to the grid standard or user demand, and calculate the difference between the harmonic current generated by the charging pile and the allowed limit value as the compensation of the harmonic current;

[0012] Step four, harmonic compensation execution: select the harmonic compensation device according to the size and frequency of the compensation current, use the calculated compensation current as the control signal to adjust the output of the compensation device, so that the compensation device generates a compensation current equal in size and opposite in phase to the harmonic current generated by the charging pile, real-time monitor the grid current after compensation, evaluate the compensation effect, and adjust the parameters of the compensation device according to the evaluation result to complete the closed-loop feedback control;

[0013] Step five, system optimization and maintenance: periodically evaluate the performance of the harmonic self-compensation system, adjust the parameters of the harmonic detection algorithm, compensation strategy or compensation device according to the evaluation result, and establish a fault warning mechanism.

[0014] Preferably, the wavelet transform multi-scale decomposition comprises the following operation steps:

[0015] Step 1, select a wavelet base function: after removing high-frequency noise from the current signal, select the wavelet base function as db20;

[0016] Step2, determining the decomposition layer: after selecting the wavelet base function, the decomposition layer is determined according to the signal characteristics, the fundamental frequency is selected as 50Hz-60Hz, the sampling frequency is selected as 10KHz-15Hz, the Nyquist frequency is selected as 5KHz-7.5KHz, and the frequency of each layer decomposition is halved, so that the fundamental wave is located in the lowest frequency sub-band;

[0017] Step3, wavelet decomposition: after the decomposition layer is determined, the current signal is decomposed layer by layer using the Mallat algorithm to obtain the approximation coefficient and the detail coefficient, and the sub-band energy is analyzed to obtain that the transient interference is concentrated in the high-frequency detail sub-band, and the fundamental wave signal is concentrated in the lowest frequency approximation sub-band;

[0018] Step4, threshold processing denoising: after the wavelet decomposition is completed, the threshold The median absolute deviation of the highest frequency detail sub-band is estimated as σ, the hard threshold and the soft threshold within the threshold are selected, and the sub-band is differentiated for processing, and the fundamental wave and the harmonic wave are retained, wherein σ is the noise standard deviation, and N is the signal length;

[0019] Step5, signal reconstruction: the transient signal and the fundamental wave signal in the threshold processing are retained, and the transient signal and the fundamental wave signal in the processing are reconstructed into time domain signals to obtain the denoised fundamental wave and transient component;

[0020] Step6, verification and optimization: the denoised fundamental wave and transient component are analyzed, and the signal ratio and mean square error formula are used for evaluation.

[0021] Preferably, in the step Step6, the evaluation mode includes transient signal loss and fundamental wave distortion, if the transient signal loss, the high-frequency sub-band threshold is reduced or the hard threshold is used, if the fundamental wave distortion, the number of coefficients retained in the low-frequency sub-band is increased.

[0022] Preferably, in the step one, the fundamental wave frequency range of the low-pass filter is 50Hz-60Hz, and the harmonic frequency range of the low-pass filter is 550Hz-650Hz.

[0023] Preferably, in the step one, the current signal output by the charging pile is collected in real time, and the sampling frequency is in the range of 10KHz-15KHz.

[0024] Preferably, in the step one, the cut-off frequency of the low-pass filter is greater than 700Hz.

[0025] Preferably, in the step Step3, the approximation coefficient is a low-frequency coefficient, and the detail coefficient is a high-frequency coefficient.

[0026] Preferably, in the step four, the harmonic compensation device includes an active power filter and a parallel capacitor bank.

[0027] Preferably, in the step five, periodically evaluating the performance of the harmonic self-compensation system includes periodically evaluating compensation accuracy, response speed and stability.

[0028] A control method for harmonic self-compensation of a direct-current charging pile, comprising a signal acquisition and preprocessing module for real-time acquisition of a charging pile output current signal and low-pass filtering of the signal to retain harmonic components;

[0029] A harmonic detection and feature extraction module for Fourier transform of the preprocessed current signal to obtain preliminary amplitude or phase information of each harmonic, multi-scale decomposition of the current signal using wavelet transform to extract transient interference signals and fundamental wave signals, input of the denoised current signal into an Adaline neural network, and adjustment of a weight vector through an LMS algorithm;

[0030] A harmonic compensation strategy formulation module for calculation of total harmonic current according to harmonic amplitude or phase, setting of a limit value in combination with a power grid standard or user demand, and generation of a compensation current instruction;

[0031] A harmonic compensation execution module for selection of a compensation device, adjustment of output according to the compensation current instruction, generation of a compensation current equal in amplitude and opposite in phase to the harmonic current amplitude, and real-time optimization of compensation effect through closed-loop feedback;

[0032] A system optimization and maintenance module for periodic evaluation of harmonic self-compensation system performance, dynamic adjustment of detection algorithm, compensation strategy or compensation device parameters, and establishment of a fault early warning mechanism.

[0033] Compared with related technologies, the control method for harmonic self-compensation of a direct-current charging pile and the system provided by the present application have the following beneficial effects:

[0034] 1. The present application adopts a multi-technology fusion approach, FFT extraction of steady-state harmonic amplitude or phase, wavelet transform stripping of transient interference noise and positioning of interference time, overcoming the defects of single FFT spectrum leakage or wavelet transform frequency domain ambiguity, real-time adjustment of a weight vector through an LMS algorithm, elimination of the influence of nonlinear load dynamic changes on detection results, reduction of harmonic amplitude or phase extraction error, and improvement over traditional single methods, providing a more accurate data basis for harmonic compensation strategies, avoiding under-compensation or over-compensation due to detection errors, and solving a series of problems caused by single Fourier transform or single wavelet transform.

[0035] 2、The application adopts a dynamic self-adaptive harmonic compensation strategy, can generate a compensation signal according to the difference between the harmonic current and the limit, reduces the response time of the compensation device, shortens the response time of the traditional active filter, solves two core pain points in harmonic treatment, the problems of insufficient real-time compensation and poor dynamic working condition adaptability, can quickly track the harmonic current change, ensures that the compensation current and the harmonic current are always equal in size and opposite in phase, has a high harmonic suppression rate, and the point of common coupling is stable, the power grid stability is high, and the risk of large-scale power outage caused by resonance is avoided. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 A flowchart of the control method for harmonic self-compensation of the direct current charging pile provided by the application is provided.

[0037] Figure 2 A flowchart of multi-scale decomposition of wavelet transform is provided.

[0038] Figure 3 A system flowchart of the control method for harmonic self-compensation of the direct current charging pile provided by the application is provided. DETAILED DESCRIPTION

[0039] The application will be further described below in combination with the drawings and embodiments.

[0040] Please refer to Figures 1 to 3 , wherein, Figure 1 A flowchart of the control method for harmonic self-compensation of the direct current charging pile provided by the application is provided. Figure 2 A flowchart of multi-scale decomposition of wavelet transform is provided. Figure 3 A system flowchart of the control method for harmonic self-compensation of the direct current charging pile provided by the application is provided.

[0041] In some embodiments, as shown in Figure 1 , a control method for harmonic self-compensation of a direct current charging pile includes the following operation steps:

[0042] Step one, signal acquisition and pretreatment: install a current sensor at the output end of the charging pile, use the current sensor to collect the current signal output by the charging pile in real time, set the sampling frequency range to 10KHz-15KHz, convert the collected current signal to a digital signal, and perform low-pass filtering on the collected digital current signal, set the fundamental frequency range of the low-pass filter to 50Hz-60Hz, the harmonic frequency range to 550Hz-650Hz, and the cutoff frequency of the low-pass filter to 1.5kHz-2kHz, to remove high-frequency noise while retaining harmonic components;

[0043] Step two, harmonic detection and feature extraction: Fourier transform is performed on the preprocessed current signal to convert the time-domain signal into a frequency-domain signal, thereby obtaining the preliminary amplitude or phase information of each harmonic. Wavelet transform is used to perform multi-scale decomposition on the current signal to extract transient interference signals and fundamental signals. The denoised current signal is input into the Adaline neural network, and the weight vector is adjusted through the LMS algorithm. The weight vector is updated according to the difference between the expected output and the neuron output, and finally the amplitude and phase of each harmonic are obtained.

[0044] It should be noted that the Fourier transform uses the Fast Fourier Transform (FFT) algorithm to improve computational efficiency.

[0045] Wavelet transform can provide time-frequency characteristics of signals at different scales, which helps to more accurately identify and analyze harmonic components.

[0046] It should be further noted that the Adaline neural network is a linear adaptive filter that can update the weight vector according to the difference between the expected output and the neuron output. Through continuous iterative training, the amplitude and phase information of each harmonic are ultimately obtained.

[0047] Step three, develop a harmonic compensation strategy: based on the harmonic amplitude and phase output by the Adaline network, calculate the total harmonic current generated by the charging pile, set the maximum allowable harmonic current limit according to the grid standard or user demand, and calculate the difference between the harmonic current generated by the charging pile and the allowable limit as the compensation of the harmonic current. The compensation amount is used to guide the subsequent harmonic compensation execution.

[0048] Step four, harmonic compensation execution: select the harmonic compensation device according to the size and frequency of the compensation current. The harmonic compensation device includes an active power filter and a parallel capacitor bank. The calculated compensation current is used as a control signal to adjust the output of the compensation device, so that the compensation device generates a compensation current that is equal in size and opposite in phase to the harmonic current generated by the charging pile. Real-time monitoring of the compensated grid current is performed to evaluate the compensation effect. Based on the evaluation results, the parameters of the compensation device, such as inductance and capacitance, are adjusted to complete closed-loop feedback control, which ensures that the compensation effect remains at an optimal state at all times.

[0049] Step five, system optimization and maintenance: periodically evaluate the performance of the harmonic self-compensation system, including the accuracy of harmonic detection, the effectiveness of compensation effect, and the stability of the system. This is done by comparing the grid current waveforms before and after compensation, calculating the harmonic distortion rate, and other methods. Based on the evaluation results, adjust the harmonic detection algorithm, compensation strategy, or parameters of the compensation device, and establish a fault warning mechanism.

[0050] Specifically, a multi-technology fusion method is adopted, FFT is used to extract the steady-state harmonic amplitude or phase, wavelet transform is used to strip the transient interference noise and locate the interference time, the defects of single FFT spectrum leakage or wavelet transform frequency domain ambiguity are overcome, the weight vector is adjusted in real time through LMS algorithm, the influence of nonlinear load dynamic change on the detection result is eliminated, the harmonic amplitude or phase extraction error is reduced, and the traditional single method is improved, which provides a more accurate data basis for harmonic compensation strategy, avoids insufficient compensation or overcompensation caused by detection error, and solves a series of problems caused by single Fourier transform or single wavelet transform.

[0051] Further, a dynamic adaptive harmonic compensation strategy is adopted, which can dynamically generate a compensation signal according to the difference between the harmonic current and the limit, reduce the response time of the compensation device, shorten the response time of the traditional active filter, solve the two core pain points in harmonic control, and solve the problems of insufficient real-time compensation and poor dynamic working condition adaptability. The present application can quickly track the change of harmonic current, ensure that the compensation current and the harmonic current are always equal in size and opposite in phase, the harmonic suppression rate is high, the point of common coupling is stable, the power grid stability is high, and the risk of large-scale power outage caused by resonance is avoided.

[0052] Embodiment one

[0053] Firstly, the period of performance evaluation is determined, for example, a comprehensive evaluation is performed once a month, and during the evaluation period, an oscilloscope is used to collect power grid current waveform data before and after compensation in real time at the output end of the charging pile;

[0054] Suppose the collection time is 1 minute each time, and the sampling frequency is 10 kHz, 600000 data points are obtained;

[0055] The collected current waveform data is subjected to fast Fourier transform (FFT), the amplitudes and phases of each harmonic are calculated, and then the harmonic distortion rate (THD) is calculated

[0056] For example, the distortion rate of the 5th harmonic is calculated, and the formula is:

[0057] Wherein, I1 is the fundamental current amplitude, I2, I3, I4 and I5 are the 2nd, 3rd, 4th and 5th harmonic current amplitudes respectively;

[0058] 1. At this time, the harmonic amplitudes and phases calculated by FFT and output by the Adaline neural network are compared, and the error rate is calculated. For example, the 5th harmonic amplitude error rate calculation formula is:

[0059] Suppose the 5th harmonic amplitude calculated by FFT is: The Adaline neural network output is I Adaline= 1.98A, the error rate is 1%, if the error rate exceeds the set threshold: 5%, it is considered that the harmonic detection is inaccurate;

[0060] 2, at this time, compare the harmonic distortion rate before and after compensation, evaluate the compensation effect. As in the above example, THD5 is reduced from 8% to 3% after compensation, which shows that the compensation effect is good, if the THD5 reduction amplitude after compensation does not reach the set target, such as the reduction amplitude is less than 50%, it is considered that the compensation effect is not good.

[0061] 3, after comparison, observe the running record of the system in the evaluation period, count whether abnormal fluctuations, protection mechanism triggering times, etc. appear, for example, if the system appears more than 3 times of abnormal fluctuations or protection mechanism triggering in a month, it is considered that the system stability is poor; and the evaluation data, index calculation results and evaluation conclusion are arranged into a report.

[0062] Further, for the fault early warning mechanism;

[0063] Operation I, determine the monitoring parameters: according to the system characteristics and the possible fault types, determine the parameters that need to be monitored, such as current anomaly, voltage fluctuation, temperature overload, etc.

[0064] Operation II, set the monitoring threshold:

[0065] S1, current anomaly detection: set the threshold of current anomaly, assuming that the rated current of the charging pile is 50A, the current anomaly threshold is set to 60A, when the current exceeds 60A and the duration exceeds 5 seconds, the early warning mechanism is triggered;

[0066] S2, voltage fluctuation monitoring: set the threshold of voltage fluctuation, when the voltage changes more than 10% of the rated voltage within 1 second, or the deviation of the voltage from the rated voltage exceeds ±5%, the early warning mechanism is triggered;

[0067] S3, temperature monitoring: set the temperature threshold of the sensor to 80℃, when the temperature sensor detects that the component temperature exceeds 80℃, the early warning mechanism is triggered.

[0068] Operation III, establish the early warning system: install corresponding sensors and monitoring equipment, including current sensor, voltage sensor, temperature sensor, etc., real-time collect the data of monitoring parameters.

[0069] Operation IV, early warning signal processing:

[0070] Sound alarm: install a sound alarm in the control room, when the early warning system sends an early warning signal, the sound alarm sends a loud alarm sound to remind the operator;

[0071] Light indication: Set corresponding indicator lights on the control panel. When a warning signal is issued, the indicator light turns on and flashes to indicate the fault type or location. When the current is abnormal, the red indicator light turns on. When the voltage fluctuates, the yellow indicator light turns on. When the temperature is too high, the blue indicator light turns on.

[0072] Remote notification: Send warning information to relevant personnel through SMS, email or APP, etc. When the current anomaly warning occurs, the system automatically sends an SMS to the maintenance personnel's mobile phone: "Charging pile current anomaly, current [X] A, exceeds threshold [Y] A, please handle in time."

[0073] Operation five, regular maintenance and inspection: Regularly check and maintain the warning system, conduct a comprehensive check once a month, check the accuracy of the sensor, the integrity of the monitoring equipment, etc., and use standard current source, voltage source, etc. to calibrate the sensor to ensure its measurement accuracy. After the inspection is completed, record the detailed information of each warning, including warning time, warning type, handling result, etc.

[0074] Example two

[0075] Step one: signal acquisition and preprocessing: Install a high-precision current sensor at the output end of the charging pile, with a range of 0-100A and an accuracy of ±0.5%. At this time, the current signal output by the charging pile is collected in real time, with a sampling frequency of 10kHz. The collected current signal is low-pass filtered with a cutoff frequency of 5kHz to remove high-frequency noise while retaining harmonic components;

[0076] Step two, harmonic detection and feature extraction;

[0077] a1, Fourier transform: Perform fast Fourier transform (FFT) on the preprocessed current signal to obtain the preliminary amplitude or phase information of each harmonic. Assuming that the amplitude of the 5th harmonic in the FFT result is I5=2A and the phase is θ5=30°;

[0078] a2, wavelet transform: Use wavelet transform to perform multi-scale decomposition on the current signal to extract transient interference signals and fundamental signals;

[0079] a3, Adaline neural network and LMS algorithm: Input the denoised current signal into the Adaline neural network and adjust the weight vector through the LMS algorithm. After multiple iterations, the 5th harmonic amplitude output by the Adaline network is I5=1.98A and the phase is θ5=30.2°;

[0080] Step three, develop a harmonic compensation strategy:

[0081] 1、According to the amplitude and phase of each harmonic output by the Adaline network, the total harmonic current generated by the charging pile is calculated. Assuming that only the 5th harmonic is considered, the effective value of the total harmonic current is:

[0082]

[0083] where I h is the amplitude of the hth harmonic, in this embodiment,

[0084] 2、Set the limit value: according to the grid standard or user demand, set the maximum harmonic current limit value allowed: 限 = 1.5 A.

[0085] 3、Calculate the compensation current: calculate the difference between the harmonic current generated by the charging pile and the allowed limit value as the compensation amount of the harmonic current:

[0086] I 补 = I 有效 -I 限 = 1.98 A - 1.5 A = 0.48 A

[0087] Step four, harmonic compensation execution: according to the 5th harmonic, the frequency is 250 Hz, select the active filter, take the calculated compensation current 0.48 A as the control signal, adjust the output of the active filter, make it generate a compensation current equal in size and opposite in phase to the harmonic current generated by the charging pile, at this time, the real-time monitored grid current after compensation is evaluated, and the compensation effect is evaluated. Assuming that after compensation, the total harmonic current is reduced to 1.45 A, which is lower than the allowed limit, according to the evaluation result, the parameters of the active filter can be further adjusted to optimize the compensation effect.

[0088] Step five, system optimization and maintenance: regularly evaluate the performance of the harmonic self-compensation system, including the accuracy of harmonic detection, the effectiveness of compensation effect and the stability of the system, according to the evaluation result, adjust the harmonic detection algorithm, such as the window function selection of FFT, the base function selection of wavelet transform, the compensation strategy, such as the calculation method of compensation current, the start-up condition of compensation device or the parameters of compensation device, such as the inductance value and capacitance value of active filter, and establish a warning mechanism for faults.

[0089] Implementation results: by implementing the above harmonic self-compensation control method, the harmonic current generated by the charging pile is effectively suppressed, the stability and power quality of the grid are significantly improved, at the same time, the adaptive ability and closed-loop feedback control mechanism of the system ensure the persistence and stability of the compensation effect.

[0090] In some embodiments, reference Figure 1 and Figure 2As shown, a control method for harmonic self-compensation of a direct current charging pile, wavelet transform multi-scale decomposition includes the following operation steps:

[0091] Step 1, selecting a wavelet base function: after removing high-frequency noise from the current signal, select the wavelet base function as db20;

[0092] Step 2, determine the decomposition level: after selecting the wavelet base function, determine the decomposition level according to the signal characteristics, select the fundamental frequency as 50Hz-60Hz, the sampling frequency as 10KHz-15Hz, and the Nyquist frequency as 5KHz-7.5KHz, halve the frequency of each layer decomposition, so that the fundamental wave is located in the lowest frequency subband;

[0093] Step 3, wavelet decomposition: after determining the decomposition level, use the Mallat algorithm to decompose the current signal layer by layer to obtain the approximation coefficient and the detail coefficient, and analyze the subband energy to obtain that the transient disturbance is concentrated in the high-frequency detail subband and the fundamental wave signal is concentrated in the lowest frequency approximation subband;

[0094] Step 4, threshold processing denoising: after wavelet decomposition, select the threshold Estimate the median absolute deviation of the highest frequency detail subband as σ, select the hard threshold and soft threshold within the threshold, and process the subbands differently to retain the fundamental wave and harmonic wave, wherein σ is the noise standard deviation and N is the signal length;

[0095] Step 5, signal reconstruction: retain the transient signal and fundamental wave signal within the threshold processing, reconstruct the transient signal and fundamental wave signal within the processing into time domain signals to obtain the denoised fundamental wave and transient component;

[0096] Step 6, verification and optimization: analyze the denoised fundamental wave and transient component, and evaluate using the signal ratio and mean square error formula.

[0097] Among them, the signal-to-noise ratio (SNR) is:

[0098] Mean square error (MSE):

[0099] Specifically, by selecting the wavelet base function db20 and determining the decomposition level, the current signal can be accurately decomposed into different subbands, so that the transient disturbance is concentrated in the high-frequency detail subband and the fundamental wave signal is concentrated in the lowest frequency approximation subband. On this basis, the threshold processing denoising method is used to process different subbands differently, which can effectively remove noise while retaining the fundamental wave and harmonic wave, significantly improving the signal quality.

[0100] Further, the over-verification and optimization step uses the signal ratio and mean square error formula to evaluate the denoised signal, which can intuitively understand the denoising effect. According to the evaluation result, the parameters in the processing flow, such as threshold selection, decomposition layer number, etc. can be adjusted and optimized, further improving the denoising effect and the accuracy of signal processing, making the whole processing process more scientific and reasonable.

[0101] In some embodiments, as shown in the reference Figure 3 A control method for harmonic self-compensation of a direct current charging pile, a system thereof includes a signal acquisition and preprocessing module for real-time acquisition of a charging pile output current signal and low-pass filtering of the signal to retain harmonic components;

[0102] A harmonic detection and feature extraction module for Fourier transform of the preprocessed current signal to obtain preliminary amplitude or phase information of each harmonic, multi-scale decomposition of the current signal using wavelet transform to extract transient interference signals and fundamental wave signals, input of the denoised current signal into an Adaline neural network, and adjustment of a weight vector through an LMS algorithm;

[0103] A harmonic compensation strategy formulation module for calculation of total harmonic current according to harmonic amplitude or phase, setting of a limit value in combination with a power grid standard or user demand, and generation of a compensation current instruction;

[0104] A harmonic compensation execution module for selection of a compensation device, adjustment of output according to the compensation current instruction, generation of a compensation current equal in amplitude and opposite in phase to the harmonic current amplitude, and real-time optimization of compensation effect through closed-loop feedback;

[0105] A system optimization and maintenance module for periodic evaluation of harmonic self-compensation system performance, dynamic adjustment of detection algorithm, compensation strategy or compensation device parameters, and establishment of a fault early warning mechanism.

[0106] Specifically, the harmonic detection and feature extraction module comprehensively uses two methods of Fourier transform and wavelet transform, Fourier transform can quickly obtain preliminary amplitude or phase information of each harmonic, providing basic data for harmonic analysis; multi-scale decomposition of wavelet transform can effectively extract transient interference signals and fundamental wave signals, decompose the signal to different scales, facilitate detailed analysis of different frequency components, and input the denoised current signal into an Adaline neural network and adjust the weight vector through an LMS algorithm, which can further improve the accuracy and adaptability of harmonic detection. The neural network can automatically learn and adapt to signal changes, has better detection capability for complex harmonic signals, and reduces manual intervention and errors.

[0107] Further, the wave compensation strategy module calculates total harmonic current according to harmonic amplitude or phase, and generates compensation current instruction in combination with grid standard or user demand setting limit, can meet the requirements of different users and different grid environment, realizes personalized harmonic compensation, and dynamically adjusts according to the real-time detected harmonic condition, ensures that the harmonic can be effectively inhibited under different working conditions, and improves the power quality of the grid.

[0108] The circuit and control involved in the application are prior art, and will not be described in detail.

[0109] The above is only an embodiment of the application, and does not limit the patent range of the application, and any equivalent structure or equivalent flow transformation using the content of the specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection range of the application.

Claims

1. A control method for harmonic self-compensation of a direct current charging pile, characterized in that, The method comprises the following steps: Step 1, signal acquisition and preprocessing: install a current sensor at the output end of the charging pile, real-time collect the current signal output by the charging pile, and perform low-pass filtering on the collected current signal to remove high-frequency noise while retaining harmonic components; Step 2, harmonic detection and feature extraction: Fourier transform the preprocessed current signal to obtain the preliminary amplitude or phase information of each harmonic, use wavelet transform to perform multi-scale decomposition on the current signal, extract transient interference signals and fundamental signals, input the denoised current signal into an Adaline neural network, and adjust the weight vector through an LMS algorithm, update the weight vector according to the difference between the expected output and the neuron output, and finally obtain the amplitude and phase of each harmonic; Step 3, develop a harmonic compensation strategy: calculate the total harmonic current generated by the charging pile according to the harmonic amplitude and phase output by the Adaline network, set the maximum allowable harmonic current limit according to the grid standard or user demand, and calculate the difference between the harmonic current generated by the charging pile and the allowable limit as the compensation of the harmonic current; Step 4, harmonic compensation execution: select the harmonic compensation device according to the size and frequency of the compensation current, take the calculated compensation current as the control signal, adjust the output of the compensation device, make the compensation device generate a compensation current equal in size and opposite in phase to the harmonic current generated by the charging pile, real-time monitor the grid current after compensation, evaluate the compensation effect, and adjust the parameters of the compensation device according to the evaluation result to complete the closed-loop feedback control; Step 5, system optimization and maintenance: periodically evaluate the performance of the harmonic self-compensation system, adjust the harmonic detection algorithm, compensation strategy or compensation device parameters according to the evaluation result, and establish a fault warning mechanism.

2. The control method of harmonic self-compensation of direct current charging pile according to claim 1, characterized in that, The wavelet transform multi-scale decomposition comprises the following steps: Step 1, select a wavelet base function: after removing high-frequency noise from the current signal, select db20 as the wavelet base function; Step 2, determine the decomposition level: after selecting the wavelet base function, determine the decomposition level according to the signal characteristics, select the fundamental frequency as 50Hz-60Hz, the sampling frequency as 10KHz-15Hz, and the Nyquist frequency as 5KHz-7.5KHz, and halve the frequency of each layer of decomposition so that the fundamental wave is located in the lowest frequency subband; Step 3, wavelet decomposition: after determining the decomposition level, use the Mallat algorithm to decompose the current signal layer by layer to obtain the approximation coefficients and detail coefficients, and analyze the subband energy to obtain that the transient interference is concentrated in the high-frequency detail subband and the fundamental signal is concentrated in the lowest frequency approximation subband; Step 4, threshold processing denoising: after wavelet decomposition, select threshold The median absolute deviation of the highest frequency detail subband is estimated as σ, the hard threshold and the soft threshold in the threshold are selected, and the subband is processed differently to retain the fundamental wave and the harmonic wave, wherein σ is the noise standard deviation, and N is the signal length. Step 5, signal reconstruction: retain the transient signals and fundamental signals within the threshold processing, reconstruct the transient signals and fundamental signals within the processing into time domain signals to obtain the denoised fundamental wave and transient components; Step 6, verification and optimization: analyze the denoised fundamental wave and transient components, and evaluate them using the signal ratio and mean square error formulas.

3. The control method of harmonic self-compensation of direct current charging pile according to claim 2, characterized in that, In the step Step6, the evaluation mode includes transient signal loss and fundamental distortion, if transient signal loss, reduce high frequency sub-band threshold or use hard threshold, if fundamental distortion, increase the number of low frequency sub-band reserved coefficients.

4. The control method of harmonic self-compensation of direct current charging pile according to claim 3, characterized in that, In the step one, the fundamental frequency range of the low-pass filter is 50Hz-60 Hz, and the harmonic frequency range of the low-pass filter is 550Hz-650Hz.

5. The control method of harmonic self-compensation of direct current charging pile according to claim 4, characterized in that, In the step one, the current signal output by the charging pile is collected in real time, and the sampling frequency ranges from 10KHz to 15KHz.

6. The control method of harmonic self-compensation of direct current charging pile according to claim 5, characterized in that, In the step one, the cutoff frequency of the low-pass filter is greater than 700Hz.

7. The control method of harmonic self-compensation of direct current charging pile according to claim 6, characterized in that, In the step Step3, the approximate coefficient is a low frequency coefficient, and the detail coefficient is a high frequency coefficient.

8. The control method of harmonic self-compensation of direct current charging pile according to claim 7, characterized in that, In the step four, the harmonic compensation device includes an active power filter and a parallel capacitor bank.

9. The control method of harmonic self-compensation of direct current charging pile according to claim 8, characterized in that, In the step five, periodically evaluating the performance of the harmonic self-compensation system includes periodically evaluating the compensation accuracy, response speed and stability.

10. System suitable for the control method of harmonic self-compensation of a direct current charging column according to any one of claims 1-9, characterized by, The signal acquisition and preprocessing module is used to collect the current signal output by the charging pile in real time, and to perform low-pass filtering on the signal to retain the harmonic component. The harmonic detection and feature extraction module is used to perform Fourier transform on the preprocessed current signal to obtain the preliminary amplitude or phase information of each harmonic, use wavelet transform to perform multi-scale decomposition on the current signal, extract transient interference signal and fundamental signal, input the denoised current signal into Adaline neural network, and adjust the weight vector through LMS algorithm. The harmonic compensation strategy formulation module is used to calculate the total harmonic current according to the harmonic amplitude or phase, set the limit value according to the power grid standard or user demand, and generate the compensation current instruction. The harmonic compensation execution module is used to select the compensation device, adjust the output according to the compensation current instruction, generate the compensation current with equal amplitude and opposite phase of the harmonic current, and optimize the compensation effect in real time through closed-loop feedback. The system optimization and maintenance module is used to periodically evaluate the performance of the harmonic self-compensation system, dynamically adjust the detection algorithm, compensation strategy or compensation device parameters, and establish a fault warning mechanism.

Images