An intelligent charging management system for electric vehicles and a communication method thereof

Abstract

An intelligent charging management system for electric vehicles includes: a distribution network master station system; a transformer substation, which is equipped with an intelligent communication management agent module and a first dual-mode communication module; and an intelligent charging pile, which includes a main controller, a second dual-mode communication module, a power quality monitoring circuit, and a power electronic switch. The dual-mode communication module is a PLC and wireless communication dual-mode communication module, which is equipped with a PINSF (Power Line Impulse Noise Filter). The communication process between the dual-mode communication modules includes a pilot training phase and a data transmission phase. In the pilot training phase, the PINSF acts as a channel estimator for channel estimation. In the data transmission phase, the PINSF acts as a channel equalizer for channel equalization. This invention is specifically optimized for low-voltage power line impulse noise environments. By using the PINSF, it suppresses the interference of impulse noise on channel estimation and equalization, ensuring the demodulation performance and communication reliability of the communication system.

Description

Technical Field

[0001] This invention belongs to the field of power system communication and smart grid control technology, and particularly relates to an intelligent charging management system for electric vehicles based on dual-mode communication and anti-impulse noise intelligent sensing, and its communication method. Background Technology

[0002] With the advancement of the "dual-carbon" strategy, new sources and loads such as distributed photovoltaics and electric vehicles are being connected to low-voltage power distribution networks on a large scale. Traditional electric vehicle charging management systems face the following problems: 1) Insufficient communication reliability; existing electric vehicle charging piles mostly use a single communication method, such as wireless communication, power line carrier communication (PLC), or fiber optic communication. A single communication method is easily affected by obstruction, interference, or changes in line topology in complex urban environments, leading to communication interruptions or delays; 2) Severe impulse noise interference; electric vehicle chargers, photovoltaic inverters, and other power electronic devices introduce high-intensity, non-Gaussian impulse noise into the power lines during operation, severely affecting the channel estimation and equalization performance of PLC communication, leading to communication failures. To improve communication quality, some have proposed applying a dual-mode communication architecture to electric vehicle charging systems, where the concentrator and charging pile interact via PLC or wireless communication. However, the switching logic between the two communication methods is simple, and the deep impact of power line impulse noise on communication quality is not considered, leaving room for improvement in communication reliability. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent charging management system for electric vehicles and its communication method based on dual-mode collaboration of power line carrier communication and wireless communication, which has the ability to resist impulse noise.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] An intelligent charging management system for electric vehicles, characterized in that it includes:

[0006] A power distribution network master station system, which is deployed in a cloud computing data center;

[0007] A distribution area concentrator is installed on the low-voltage side of the distribution transformer. The distribution area concentrator is communicatively connected to the distribution network master station system. The distribution area concentrator is equipped with an intelligent communication management agent module and a first dual-mode communication module.

[0008] The smart charging pile is installed on the user side or the website side. The smart charging pile includes a main controller, a second dual-mode communication module, a power quality monitoring circuit and a power electronic switch. The distribution area concentrator is connected to the smart charging pile through the first dual-mode communication module, the second dual-mode communication module and the smart charging pile.

[0009] The first dual-mode communication module and the second dual-mode communication module are PLC and wireless communication dual-mode communication modules. The first dual-mode communication module and the second dual-mode communication module are equipped with power line impulse noise suppression filters. The power line impulse noise suppression filters have a pilot training stage and a data transmission stage during the communication process between the first dual-mode communication module and the second dual-mode communication module. In the pilot training stage, the power line impulse noise suppression filters perform channel estimation as a channel estimator. In the data transmission stage, the power line impulse noise suppression filters perform channel equalization as a channel equalizer.

[0010] The channel estimation during the pilot training phase includes the following steps:

[0011] S101, The transmitting end sends a training pilot signal to the receiving end;

[0012] S102, The receiving end decomposes the training pilot signal into multiple sub-band pilot signals;

[0013] S103. The receiving end performs downsampling processing on the multi-channel subband pilot signals to obtain downsampled subband pilot signals. All downsampled subband pilot signals are arranged in columns to form a subband pilot signal matrix.

[0014] S104. Perform convolution operation on each downsampled subband pilot signal and the subband adaptive filter tap weights, and use the result as the subband estimated signal; after updating the subband adaptive filter tap weights in subsequent steps, use the new subband adaptive filter tap weights for convolution operation.

[0015] S105. The training pilot signal sent by the transmitting end passes through the actual channel and is received by the receiving end. The signal received by the receiving end is used as the expected signal.

[0016] S106. The receiving end decomposes the received desired signal into multiple sub-band desired signals;

[0017] S107. Perform downsampling processing on the multi-channel subband desired signal to obtain the downsampled subband desired signal;

[0018] S108. Calculate the difference between the estimated signal of each sub-band and the expected signal of each downsampled sub-band to obtain the error signal of each sub-band. Arrange all the sub-band error signals in columns to form an error signal matrix.

[0019] S109. Calculate the anti-pulse factor matrix; the anti-pulse factor matrix is ​​a diagonal matrix, and its diagonal elements are... In the formula Q is the projection order, and ω is the generalized cost function. The performance factor, δ, is the generalized cost function. scaling factor, generalized cost function , This represents the error signal of the kl-th sub-band, where N is the total number of sub-bands.

[0020] S110. Update the tap weights of the subband adaptive filter using the following formula:

[0021] In the formula, g(k) is the current subband adaptive filter tap weight, g(k+1) is the updated subband adaptive filter tap weight, μ is the convergence factor, U(k) is the subband pilot signal matrix, E(k) is the error signal matrix, and H(k) is the anti-pulse factor matrix.

[0022] S111. Merge and interpolate the error signals of each sub-band to reconstruct the broadband error signal;

[0023] Repeat steps S101 to S111 until the mean square error between the current subband adaptive filter tap weight and the previous subband adaptive filter tap weight is less than a set threshold, or the maximum number of iterations is reached, then end.

[0024] The channel equalization during the data transmission phase of the intelligent charging management system for electric vehicles described above includes the following steps:

[0025] S201, The receiving end receives the signal;

[0026] S202, The receiving end decomposes the received signal into multiple sub-band signals;

[0027] S203. Perform downsampling processing on each sub-band signal to obtain the downsampled sub-band signal;

[0028] S204. Perform convolution operation on the downsampled subband signals and the equalizer filter head weights to obtain the subband equalized signal; after the equalizer filter tap weights are updated in subsequent steps, the new equalizer filter tap weights are used for convolution operation.

[0029] S205. The sub-band equalized signal is upsampled to obtain the reconstructed full-band signal;

[0030] S206. The full-band signal is fed into the decision circuit to obtain an estimate of the original signal. The estimate of the original signal is decomposed and downsampled to obtain multiple sub-band reference signals. All sub-band reference signals are arranged in columns to form a reference signal matrix.

[0031] S207. Subtract the subband reference signal and the subband equalization signal to obtain the subband error signal of the original signal. Arrange all the subband error signals of the original signal in columns to form the error signal matrix of the original signal.

[0032] S208. Calculate the anti-pulse factor matrix. The process of calculating the anti-pulse factor matrix in this step is the same as in step S109, except that the subband error signal in the calculation process is the subband error signal of the original signal obtained in step S207.

[0033] S209. Update the tap weights of the equalization filter using the following formula:

[0034] In the formula, ω0(k) represents the current tap weights of the equalization filter, ω0(k+1) represents the updated tap weights of the equalization filter, and μ is the convergence factor. Here, H(k) is the reference signal matrix, H(k) is the anti-pulse factor matrix obtained in step S208, and E0(k) is the error signal matrix of the original signal.

[0035] Repeat steps S201 to S209 until the mean square error between the current equalization filter tap weight and the previous equalization filter tap weight is less than the set threshold, or the maximum number of iterations is reached, then end.

[0036] As described above, the intelligent charging management system for electric vehicles also includes a photovoltaic inverter, an energy storage system, and a smart meter on the user side or website side. The intelligent charging pile is connected to the photovoltaic inverter, the energy storage system, and the smart meter via a power line and a local data interface.

[0037] The present invention also provides a communication method for the aforementioned intelligent charging management system for electric vehicles, comprising the following steps:

[0038] S1. After the system is powered on, the first dual-mode communication module and the second dual-mode communication module are started, activating the two links of PLC and wireless communication;

[0039] S2. The power line impulse noise suppression filter completes initialization;

[0040] S3. The intelligent communication management agent module monitors the communication status of the entire system in real time, including monitoring the physical layer performance indicators and internal indicators of the PLC link. The internal indicators are the broadband error signals obtained during the pilot training phase.

[0041] S4. The intelligent communication management agent module integrates the physical layer performance indicators and internal indicators of the PLC link to judge the link quality. If any indicator exceeds the preset threshold, the PLC link quality is considered to be degraded.

[0042] S5. When it is determined that the PLC link quality has deteriorated, the intelligent communication management agent module switches the link according to a preset strategy.

[0043] S6. If the PLC link quality is determined to be good, continue to use the PLC link for communication.

[0044] S7. The intelligent communication management agent module records events and reports them to the main station.

[0045] As described above, the physical layer performance indicators of the PLC link in the communication method include signal-to-noise ratio and bit error rate.

[0046] As described in the communication method above, in step S5, the intelligent communication management agent module decides whether to switch links based on the priority of the service. For high-priority services, it immediately switches to the wireless link; for low-priority services, it adjusts the parameters of PLISF so that PLISF can be restored to stability and then continues to use the PLC link for communication.

[0047] As described in the communication method above, the high-priority service includes charging control commands, and the low-priority service includes data upload services.

[0048] As can be seen from the above technical solutions, this invention uses a dual-mode communication module to realize data interaction between the distribution concentrator and the charging pile. The dual-mode communication module adopts a dual-mode communication architecture of power line carrier communication and wireless communication, simultaneously establishing a PLC link and a wireless link between the charging pile and the low-voltage distribution network concentrator. Based on link status detection and the internal state of PLISF, adaptive switching or parallel transmission is achieved to ensure communication reliability. The PLISF of this invention adopts a sub-band adaptive weight update mechanism based on a generalized cost function, which is specifically optimized for low-voltage power line impulse noise environment. By suppressing the interference of impulse noise on channel estimation and equalization through PLISF, the dual-mode communication module can successfully realize functions such as channel estimation and channel equalization in impulse noise environment, thereby indirectly ensuring the demodulation performance and communication reliability of the communication system and improving the availability of PLC communication in harsh electromagnetic environments. Attached Figure Description

[0049] To more clearly illustrate the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 is a structural block diagram of the intelligent charging management system according to an embodiment of the present invention;

[0051] Figure 2 This is a flowchart of channel estimation during the pilot training phase according to an embodiment of the present invention;

[0052] Figure 3This is a structural diagram of the PLINSF according to an embodiment of the present invention;

[0053] Figure 4 This is a flowchart of channel equalization during the data transmission phase of an embodiment of the present invention;

[0054] Figure 5 This is a flowchart of the communication method of the intelligent charging management system according to an embodiment of the present invention;

[0055] Figure 6 This is a control flowchart of an embodiment of the intelligent charging pile of the present invention;

[0056] Figure 7 The diagram shows the convergence curves of PLINSF and traditional APSSAF and IMSAF in the embodiments of the present invention.

[0057] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings. Detailed Implementation

[0058] The present invention will now be described in detail with reference to the accompanying drawings. In describing the embodiments of the present invention, for ease of explanation, the drawings illustrating the device structure will be partially enlarged, not according to general proportions. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of the present invention. It should be noted that the drawings are simplified and use non-precise proportions, intended only to facilitate and clearly illustrate the embodiments of the present invention. Additionally, in the description of this application, terms such as "first" and "second" are used only to distinguish descriptions and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Terms such as "positive," "negative," "bottom," "upper," "lower," "front," "rear," "left," and "right" indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, not indicating or implying that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention.

[0059] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "connected" and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can also refer to the internal connection of two components; and they can refer to a wireless connection or a wired connection. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0060] With the large-scale integration of new sources and loads such as distributed photovoltaics and electric vehicles into low-voltage distribution networks, traditional electric vehicle charging management systems face challenges such as communication reliability and anti-interference capabilities. This invention aims to build an integrated solution of "sensing-communication-control" to overcome the above-mentioned problems when electric vehicle charging piles are connected to smart distribution networks.

[0061] The hardware architecture of the intelligent charging management system for electric vehicles of this invention adopts a four-level collaborative design of "cloud-pipe-edge-device", such as... Figure 1 As shown, the intelligent charging management system for electric vehicles in this embodiment includes: a distribution network master station system deployed on the power grid side, substation concentrators deployed on the network side, and intelligent charging piles deployed on the user side or the site side. The distribution network master station system is located at the top level and is deployed in a cloud computing data center. It is responsible for performing global data aggregation, in-depth analysis, generation of optimized scheduling strategies, and monitoring and management of the entire network. The distribution network master station system interacts with all its subordinate substation concentrators via high-speed uplinks such as fiber optic or 5G wireless communication.

[0062] The distribution zone concentrator is installed on the low-voltage side of each distribution transformer. It incorporates an Intelligent Communication Management Agent (ICMA) module and a first dual-mode communication module. Through the ICMA module, the concentrator manages all terminal devices within its jurisdiction, aggregating various data and distributing control commands. The concentrator also possesses some edge computing capabilities, enabling it to execute preset localized control strategies when communication with the main station is interrupted, ensuring the basic operational reliability of the system.

[0063] The smart charging pile is an edge intelligent unit (EIU), which includes a main controller, a second dual-mode communication module, a power quality monitoring circuit, and a power electronic switch. In this embodiment, the first and second dual-mode communication modules have identical structures, both employing PLC and wireless communication technologies. The distribution area concentrator communicates with the smart charging pile through these modules, enabling communication between the smart charging pile and the concentrator via PLC or wireless communication. This embodiment's smart charging pile not only possesses data acquisition and communication functions but also integrates multi-dimensional sensing (including power quality, harmonics, photovoltaic output, and energy storage status) and local system control algorithms, enabling millisecond-level autonomous decision-making, such as voltage over-limit regulation and photovoltaic-energy storage-charging coordination, extending beyond data collection and fire alarms. As an edge intelligent unit, the charging pile achieves localized coordinated control of power quality, photovoltaics, and energy storage, achieving millisecond-level local response and reducing the control pressure on the main station.

[0064] On the user side or website side, there are also terminal devices such as photovoltaic inverters, energy storage systems and smart meters. The smart charging piles are connected to the above terminal devices through power lines and local data interfaces, and the above terminal devices are coordinated and scheduled in a unified manner.

[0065] In this embodiment, a novel sub-band adaptive filter is designed as a power line impulse noise suppression filter (hereinafter referred to as PLINSF) in the first and second dual-mode communication modules. By introducing PLINSF into the PLC link, power line impulse noise is suppressed, solving the problem of severe interference of impulse noise on communication quality in low-voltage power line environments and improving communication reliability in harsh power line environments.

[0066] PLINSF operates in two phases during the dual-mode communication module's communication process: a pilot training phase and a data transmission phase. In the pilot training phase, PLINSF (subband adaptive filter) acts as a channel estimator to obtain a relatively accurate channel estimate in impulse noise environments. In the data transmission phase, PLINSF acts as a channel equalizer to achieve channel equalization and reduce the impact of inter-symbol interference (ISI) in impulse noise environments. Through channel estimation and channel equalization, communication is ensured.

[0067] Figure 2 This is a flowchart illustrating channel estimation in PLINSF during the pilot training phase. Figure 2 As shown, and in combination Figure 3 The channel estimation process during the pilot training phase includes the following steps:

[0068] S101. The transmitting end sends a training pilot signal to the receiving end. The first dual-mode communication module of the substation concentrator and the second dual-mode communication module of the smart charging pile are the transmitting end and the receiving end, respectively. The training pilot signal sent by the transmitting end is a correlated random pilot signal sequence. , where n represents the time index, and the training pilot signal is also known to the receiver;

[0069] S102. The receiving end decomposes the known training pilot signal into multiple sub-band pilot signals. That is, when the transmitting end sends the training pilot signal to the receiving end, the receiving end simultaneously analyzes the known pilot signal sequence through the analysis filter bank. Decomposed into multiple subband pilot signals , Let i represent the pilot signal of the i-th subband, i=1,…,N, where N is the total number of subbands; the function of the analysis filter bank is to decompose the full-band signal into multiple subbands, thereby reducing the signal correlation in subsequent processing; the analysis filter banks that appear in subsequent steps are also used to decompose the full-band signal into subbands and reduce the correlation between signals.

[0070] S103. The receiving end performs downsampling processing on the decomposed multi-channel sub-band pilot signals to obtain downsampled sub-band pilot signals. k = n / N, where k is the index after downsampling, and the index in all signal processing steps after downsampling is represented by k; all downsampled subband pilot signals The subband pilot signal matrix U(k) is formed by arranging the signals in columns; downsampling is a conventional signal processing operation in the field of signal processing and is not an innovation of this invention, so it will not be elaborated here.

[0071] S104. Obtain each downsampled subband pilot signal The result y is obtained by convolving the subband pilot signals of each downsampled subband with the subband adaptive filter tap weights g. i (k) is used as the estimated signal for this subband; after updating the tap weights of the subband adaptive filter in subsequent steps, the new tap weights of the subband adaptive filter are used for convolution operation; the number of taps of the subband adaptive filter is L;

[0072] S105. The training pilot signal transmitted by the transmitting end passes through the actual channel and is received by the receiving end. The signal received by the receiving end is used as the desired signal. Expected signal It is a received signal that is mixed with strong impulse noise and background interference;

[0073] S106, The receiving end receives the desired signal Decomposed into multiple sub-band desired signals; desired signals The signal is fed into a set of parallel analysis filters and decomposed into several narrow-band sub-band desired signals with non-overlapping or slightly overlapping spectra. ;

[0074] S107. Perform downsampling processing on the multi-channel subband desired signal obtained in step S106 to obtain the downsampled subband desired signal. The overall computational complexity is reduced through downsampling operations;

[0075] S108, Calculate the estimated sub-band signals obtained in step S104. and the desired signals of each downsampled subband obtained in step 107 The difference is about to and Subtraction is performed to obtain the error signals of each subband. All sub-band error signals (error vectors) are arranged in columns to form the error signal matrix E(k);

[0076] S109. Calculate the anti-pulse factor matrix. Anti-pulse factor matrix It is a diagonal matrix. Anti-pulse factor matrix diagonal elements In the formula , i and l are the subscripts of the elements in the anti-pulse factor matrix;

[0077] This embodiment constructs a generalized cost function. For the generalized cost function By performing differentiation calculations, the anti-pulse factor matrix is ​​obtained;

[0078] Generalized cost function In the formula, ω is the performance factor and δ is the scaling factor. The performance factor is used to adjust the performance of the cost function, and the scaling factor is used to adjust the numerical stability. The values ​​of ω and δ are empirical values ​​and can be selected according to different environments and requirements, thus providing a basis for the environmental adaptability of the dual-mode communication module. ω can usually be 0, 1, -2, and the value range of δ can be [0.05, 0.5]. Q is the projection order, which is the sum of the number of signals at the current time and the number of historical signals used. That is, each time input, not only the signal at the current time is input, but also the input signals at the previous Q-1 times are combined. This represents the error signal of the kl-th sub-band;

[0079] S110. Update the tap weights of the sub-band adaptive filter;

[0080] This embodiment uses the stochastic gradient descent method to update the tap weights of the subband adaptive filter. The update formula is as follows:

[0081] In the formula, g(k) is the current subband adaptive filter tap weight, g(k+1) is the updated subband adaptive filter tap weight, μ is the convergence factor, which is an empirical value. The larger the convergence factor value, the faster the convergence, but it may cause oscillation or even divergence. The smaller the value, the more stable the convergence, but the slower the convergence speed. The value can be selected according to the requirements. In this embodiment, the convergence factor is 0.001. U(k) is the subband pilot signal matrix, E(k) is the error signal matrix, and H(k) is the anti-pulse factor matrix.

[0082] S111, Reconstructing the error signal; In this embodiment, a conventional integrated filter bank is used to merge and interpolate the error signals of each sub-band to reconstruct the complete broadband error signal e(k); During system communication, the broadband error signal e(k) is used as the basis for switching the dual-mode communication link. When the broadband error signal e(k) is too large, it is considered that there is impulse noise in the system, which will have a significant impact on the PLC communication link, and the system can be switched to a wireless link.

[0083] During the pilot training phase, repeat steps S101 to S111 until the mean square error between the current subband adaptive filter tap weight and the previous subband adaptive filter tap weight is less than a set threshold (e.g., less than 0.0001), or until the maximum number of iterations (e.g., 1000) is reached.

[0084] During the pilot training phase, the actual channel g0 is estimated by continuously updating the tap weights of the subband adaptive filter, laying the foundation for stable data transmission in the future.

[0085] Figure 4 This is a flowchart illustrating channel equalization performed by the PLINSF during the data transmission phase. In this phase, the designed sub-band adaptive filter (PLINSF) acts as an equalizer for channel equalization. To avoid confusion with the sub-band adaptive filter tap weights used in the pilot training phase, the sub-band adaptive filter tap weights are defined as the equalization filter tap weights in the following description of the data transmission phase steps. Figure 4 As shown, and in combination Figure 3 The channel equalization process during data transmission includes the following steps:

[0086] S201. The receiving end receives the signal. In the actual communication process, the transmitting end sends a signal to the receiving end. The full-band signal d0(n) received by the receiving end is a signal with noise pollution, that is, the received signal is a signal mixed with strong impulse noise and background interference. The signal sent by the transmitting end is defined as the original signal.

[0087] S202, The receiving end decomposes the received full-band signal d0(n) into multiple sub-band signals d 0i At the receiving end, the full-band signal d0(n) is decomposed through an analysis filter bank to obtain the multi-subband signal d. 0i (n);

[0088] S203, for each sub-band signal d 0i (n) is downsampled to obtain the downsampled subband signal d. 0i (k), k=n / N;

[0089] S204, d of each downsampled subband signal 0i (k) After filtering by an equalization filter, the sub-band equalized signal is obtained, which is the downsampled sub-band signal d. 0i The subband equalization signal y is obtained by convolving (k) with the equalization filter head weight ω0. 0i (k); After the subsequent steps update the equalization filter tap weights, the new equalization filter tap weights are used for filtering.

[0090] S205, Pair band equalization signal y 0i (k) Upsample the signal through the synthesis filter bank to obtain the reconstructed full-band signal y(n), where the k index represents the index of the sampled signal and the n index represents the index of the reconstructed full-band signal; the function of the synthesis filter is to restore the multi-subband signal to the full-band signal.

[0091] S206. The reconstructed full-band signal y(n) is fed into the decision circuit to obtain an estimate of the original signal. The estimation of the original signal By analyzing the filter bank and decomposing it, and then downsampling it, the multi-subband reference signal is obtained. All sub-band reference signals are arranged in columns to form a reference signal matrix. The decision unit's function is to determine symbols and provide reference signals. The reconstructed full-band signal (continuous signal) may still contain residual distortion. The full-band signal is fed into the decision unit, which maps the reconstructed full-band signal to the most likely discrete transmitted symbols as an estimate of the original transmitted data (symbol decision). The estimated signal after decision is used as a known expected signal. After decomposition and downsampling, it is used to calculate the error signal in each sub-band, thereby driving the equalization filter weight update, enabling the equalization filter to continuously track channel changes and resist impulse noise (providing reference signals). The analysis filter bank in this step is the same as the analysis filter bank in step S102, both of which are used to decompose the full-band signal into multiple sub-bands. The decision unit is a conventional technology in the field of signal processing and is not an innovation of this invention, so it will not be elaborated here.

[0092] S207. Calculate the error signal of the original signal using the sub-band reference signal. and sub-band equalized signal y 0i (k) Perform a subtraction operation to obtain the subband error signal e of the original signal. 0i (k), the sub-band error signal e of all original signals 0i (k) The error signal matrix E0(k) is formed by arranging the original signal in columns;

[0093] S208. Calculate the anti-pulse factor matrix. The process of calculating the anti-pulse factor matrix in this step is the same as in step S109, except that the subband error signal in the calculation process is the subband error signal of the original signal obtained in step S207.

[0094] S209. Update the tap weights of the equalization filter using the following formula:

[0095] In the formula, ω0(k) represents the current tap weights of the equalization filter, ω0(k+1) represents the updated tap weights of the equalization filter, and μ is the convergence factor. Let H(k) be the reference signal matrix, H(k) be the anti-pulse factor matrix, and E0(k) be the error signal matrix of the original signal.

[0096] During the signal transmission phase, steps S201 to S209 are repeated until the mean square error between the current equalizer tap weight and the previous equalizer tap weight is less than a set threshold (e.g., less than 0.0001), or until the maximum number of iterations (e.g., 1000) is reached. At the end, the estimate of the original signal obtained in step S206 is used as the recovered original transmitted data.

[0097] This invention introduces PLINSF into a dual-mode communication module, innovatively applying it to power line communication systems to address impulse noise, including channel estimation and equalization. During data transmission, PLINSF utilizes channel knowledge obtained during pilot training (i.e., an estimate of the actual channel) to eliminate channel distortion and resist impulse noise in unknown data transmission, achieving impulse noise immunity. Existing dual-mode communication modules lack impulse noise immunity; impulse noise causes abrupt changes in the error signal of traditional adaptive filters, leading to filter weight divergence, channel estimation / equalization failure, and ultimately communication interruption. This invention, by introducing a generalized cost function and an anti-impulse factor matrix into PLINSF, enables the filter / equalizer to maintain convergence stability under impulse noise conditions, thereby improving channel estimation accuracy and equalization performance, ultimately achieving reliable operation of the communication system under impulse noise environments. This invention does not directly "eliminate" impulse noise, but rather suppresses the interference of impulse noise on the channel estimation and equalization process by fusing PLISF. Regardless of the stage, due to the nonlinearity of the anti-impulse factor matrix, the PLISF structure in the dual-mode communication module can suppress the impact of impulse noise to a certain extent, whether it is used as a channel estimator or a channel equalizer, thereby indirectly ensuring the demodulation performance and communication reliability of the communication system.

[0098] Furthermore, during the communication process, the internal state of the filter (wideband error signal e(k)) is used as the intelligent sensing basis for link switching, that is, the switching between wireless communication and PLC communication links, forming a closed-loop communication guarantee system, constructing a highly integrated system-level solution, solving the coordination problem between existing filter modules and communication systems. Existing dual-mode communication modules do not have this intelligent sensing switching mode.

[0099] Figure 5 This is a communication flowchart of the intelligent charging management system in this embodiment, as follows: Figure 5As shown, the communication method of the intelligent charging management system in this embodiment includes the following steps:

[0100] S1. After the system is powered on, the dual-mode communication module starts (including the first dual-mode communication module of the distribution area concentrator and the second dual-mode communication module of the smart charging pile), activating the two links of power line carrier communication (PLC) and wireless communication, preparing for the operation of the intelligent communication management agent module (ICMA) and the power line impulse noise suppression filter (PLINSF).

[0101] S2, PLINSF completes initialization; through autonomous training, PLINSF has completed the initial convergence of tap weights, and while suppressing the impact of impulse noise in the power line, it completes channel estimation or channel equalization tasks to ensure normal communication.

[0102] S3, the intelligent communication management agent module monitors the communication status of the entire system in real time, including monitoring the physical layer performance indicators of the PLC link, such as signal-to-noise ratio (SNR) and bit error rate (BER), for routine link quality assessment. SNR and BER are external indicators for evaluating channel quality. In addition to monitoring the physical layer performance indicators of the PLC link, it also monitors the internal state of PLINSF (i.e., the broadband error signal e(k)). This invention uses the broadband error signal e(k) obtained during the pilot training phase as an internal indicator for evaluating channel quality. The broadband error signal e(k) obtained during the pilot training phase directly reflects the filter's suppression effect on the current noise and its convergence state. When the impulse noise is abnormally severe, the broadband error signal e(k) will remain high, which provides a more essential and a priori internal criterion for link quality assessment.

[0103] S4. The intelligent communication management agent module combines external indicators (signal-to-noise ratio, bit error rate) and internal status (continuous exceedance of broadband error signal) to judge the link quality. If any indicator exceeds the preset threshold (such as too low signal-to-noise ratio, too high bit error rate, or continuous exceedance of error norm), the PLC link quality is considered to be degraded. The preset thresholds for each indicator can be set based on experience and are not limited here.

[0104] S5. When it is determined that the PLC link quality has deteriorated, that is, when it is confirmed that the current PLC link cannot meet the requirements of reliable communication, the intelligent communication management agent module switches the link according to the preset strategy.

[0105] Specifically, when deciding whether to switch communication links, this invention does not switch communication links every time the PLC link quality is determined to be degraded. Instead, it makes the optimal decision based on the priority of the service and real-time requirements: for high-priority services such as charging control commands, to ensure absolute reliability and real-time performance, it immediately switches to the wireless link; for low-priority services, such as data uploads, which can tolerate short delays, by adjusting the performance factor ω of the generalized cost function in PLISF, it attempts to make PLISF adapt to the current noise environment and restore stability before continuing to use the PLC link for communication. An adaptive optimization closed loop is formed by dynamically adjusting the filter parameters.

[0106] S6. When the PLC link quality is determined to be good, that is, when it is confirmed that the current PLC link can meet the requirements of reliable communication, then continue to use the PLC link with lower cost and no additional cost for communication.

[0107] S7. The intelligent communication management agent module records events and reports them to the master station. Regardless of the operation performed (maintenance, switching, or parameter adjustment), the intelligent communication management agent module will fully record the event (including link quality data, filter status, decision actions, and results) and report it to the distribution network master station system for big data analysis, network optimization, and operation and maintenance management.

[0108] As can be seen from the above communication steps, the communication method of the present invention has the following characteristics compared with the closed-loop control mechanism of the traditional communication method: (1) Deep perception: The internal state of the filter is incorporated into the link quality assessment system, realizing a deeper perception of communication obstacles. The internal state of the filter serves as the internal criterion for the quality assessment of the communication link, realizing the closed-loop coordination of "signal processing-communication switching"; (2) Intelligent decision-making: The "switching" or "optimization" strategy is dynamically selected according to the service priority, realizing the best balance between cost and reliability; (3) Active optimization: By dynamically adjusting the filter parameters, the system is given the ability to self-optimize and adapt to the time-varying environment.

[0109] The intelligent charging management system of this invention continuously monitors the physical layer performance indicators of the PLC and wireless links and the internal state of the PLINSF filter in real time through the intelligent communication management agent module. Based on the preset decision logic, when the quality of a link deteriorates, it automatically and seamlessly switches the high-priority service flow to another high-quality link. This dual-mode intelligent switching strategy provides redundancy backup and intelligent optimization mechanism for the transmission of key control commands, fundamentally ensuring that the communication link is "never interrupted".

[0110] Based on the dual-mode communication ensuring communication quality, the smart charging pile of this invention acts as an edge intelligent unit to manage and control the entire charging pile system, such as... Figure 7 As shown, the charging management process of smart charging piles is as follows:

[0111] S301. Real-time local data acquisition: The main controller of the smart charging pile continuously acquires the voltage and current waveforms of the power grid through its built-in high-precision ADC power quality monitoring module. It outputs residual signals through sub-band adaptive filters, extracts harmonic content and voltage fluctuation indicators, and calculates key parameters such as voltage RMS value, frequency, and harmonic distortion rate in real time. At the same time, it obtains the real-time power generation of the photovoltaic inverter, the current state of charge of the energy storage system, and the operating status of the charging pile itself through the local communication interface, thereby constructing a comprehensive local operating situation awareness picture.

[0112] S302. Entering the state assessment and decision triggering stage, the main controller compares the monitored effective voltage value with the safety range specified by the national standard in real time. Once a voltage over-limit event is detected (such as the voltage exceeding the upper limit due to photovoltaic backfeeding), the local collaborative control strategy is immediately triggered autonomously. This process bypasses the cloud backhaul path, achieving a millisecond-level rapid response and becoming the first solid line of defense to ensure grid security.

[0113] S303. Execute a multi-priority collaborative control strategy. The principle is to restore grid stability most effectively while minimizing the impact on the user's charging experience. The first priority is to directly adjust the charging power for peak shaving. The EIU sends instructions to the power module of the charging pile to reduce the output power in predetermined steps. If the voltage problem persists, the second priority strategy is activated to schedule the energy storage system for local consumption. When there is excess photovoltaic power generation, the energy storage charging is activated to absorb the excess power. If the voltage is too low due to peak load, the third priority is entered, and the energy storage is called to discharge to support the local voltage. After each control action is executed, the system will enter a closed-loop verification stage to re-monitor the grid voltage to evaluate the control effect and decide whether to proceed with the next operation.

[0114] The S304 main controller fully records all operation logs, event details, and processing results, and packages and reports them to the power grid master station for global optimization and post-event analysis, thus forming a complete local intelligent closed loop of perception-decision-execution-verification-reporting.

[0115] After dual-mode communication ensures communication quality, the edge intelligent unit (EIU) continues to manage and control the entire charging pile system, and can autonomously maintain the stable operation of the distribution network end without relying on cloud commands.

[0116] This invention introduces a dual-mode communication architecture of PLC and wireless in intelligent charging piles for electric vehicles, and adopts a PLINSF filter structure at both the PLC transmitting and receiving ends. This effectively solves problems such as severe pulse noise interference in low-voltage power distribution networks, instability of single communication modes, and lack of intelligent sensing extension in charging piles, and has the following beneficial effects:

[0117] Firstly, regarding impact noise resistance, the PLINSF filter used in this invention enables channel estimation and channel equalization to be achieved under impact noise conditions, thereby ensuring communication quality. The inventors used a segment of actual speech signal as the pilot training signal for channel estimation. At the receiving end, with a sub-band number N of 4, the PLINSF filter, with performance factors ω=1, ω=0, and ω=-2 in the generalized cost function, [achieves good performance]. Figure 7 As can be seen, under the influence of impulse noise, the weight convergence speed and convergence stability of this filter are significantly better than those of the traditional APSSAF (Radial Projection Subband Adaptive Filter) and IMSAF (Improved Subband Adaptive Filter).

[0118] Secondly, regarding communication link reliability, the dual-mode communication mechanism proposed in this invention enables the system to dynamically switch or transmit in parallel between the PLC and wireless based on real-time link quality. When the PLC channel environment is good, the PLC is used as the main link to reduce communication costs. When the PLC is affected by strong impulse noise interference or impedance changes, the system automatically switches to the wireless link as the main communication channel, thereby avoiding communication interruption and data loss. Since PLINSF significantly improves the availability of the PLC under most noise conditions, the system's reliance on the wireless link is reduced, thereby reducing wireless traffic consumption and additional communication costs. This dual-mode architecture ensures communication continuity while also optimizing operating costs. Moreover, this invention monitors the signal-to-noise ratio, bit error rate, and PLINSF internal parameters of the PLC and wireless in real time through an intelligent communication management agent module, and achieves seamless switching based on these parameters. In contrast, existing technologies only perform simple switching based on "link cost" or "signal strength," lacking adaptability to power line impulse noise.

[0119] Furthermore, in terms of intelligent sensing expansion, this invention not only ensures communication for electric vehicle charging but also enables charging piles to function as edge sensing nodes in low-voltage distribution networks. Specifically, while uploading charging status and electricity metering data, the charging pile can also collect electricity meter data and power quality indicators (such as voltage fluctuations, harmonic components, and transient interference), and reliably upload them to the main station system via dual-mode communication. In addition, when the photovoltaic inverter in the area where the charging pile is located is connected to the grid, this invention can also synchronously sense photovoltaic output information and form source-load coordination with the charging strategy, providing more accurate data support for grid dispatch. Thus, the charging pile is not only an energy supply device but also an edge sensing node for distribution network operation data, serving various application scenarios such as smart meter reading, distribution area load forecasting, distributed photovoltaic control, and energy efficiency analysis. This invention deeply integrates communication with functions such as power quality monitoring, photovoltaic output sensing, and energy storage dispatch, realizing the integration of "sensing-communication-control," while existing technologies only support basic data transmission and charging management.

[0120] Furthermore, at the engineering application level, this invention ensures robustness while also considering the complexity of actual hardware implementation. PLINSF is an adaptive filter of the affine projection class, but with the support of sub-band structure, its computational complexity is significantly lower than that of the full-band affine projection series structure. It can run in real time on DSP or embedded MCU platforms commonly used in charging piles, and can be further accelerated by FPGA or ASIC if necessary. This makes the invention highly deployable and scalable without significantly increasing hardware costs. At the same time, PLINSF parameters (such as the number of sub-bands, projection order, and robustness factor) are adjustable and can be flexibly configured according to different transformer area environments to adapt to various low-voltage distribution network application scenarios.

[0121] In summary, this invention has the advantages of strong resistance to impact noise, high communication reliability, comprehensive intelligent sensing functions, and controllable implementation cost. Compared with existing solutions that rely solely on single-mode communication or traditional filtering algorithms, it can not only significantly reduce the bit error rate and interruption rate, but also provide multi-dimensional power grid operation data, improving the observability and controllability of low-voltage distribution networks, thus having significant engineering application value and promotion significance.

[0122] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent charging management system for electric vehicles, characterized in that, include: A power distribution network master station system, which is deployed in a cloud computing data center; A distribution area concentrator is installed on the low-voltage side of the distribution transformer. The distribution area concentrator is communicatively connected to the distribution network master station system. The distribution area concentrator is equipped with an intelligent communication management agent module and a first dual-mode communication module. The smart charging pile is installed on the user side or the website side. The smart charging pile includes a main controller, a second dual-mode communication module, a power quality monitoring circuit and a power electronic switch. The distribution area concentrator is connected to the smart charging pile through the first dual-mode communication module and the second dual-mode communication module. The first dual-mode communication module and the second dual-mode communication module are PLC and wireless communication dual-mode communication modules. The first dual-mode communication module and the second dual-mode communication module are equipped with power line impulse noise suppression filters. The power line impulse noise suppression filters have a pilot training stage and a data transmission stage during the communication process between the first dual-mode communication module and the second dual-mode communication module. In the pilot training stage, the power line impulse noise suppression filters perform channel estimation as a channel estimator. In the data transmission stage, the power line impulse noise suppression filters perform channel equalization as a channel equalizer. The channel estimation during the pilot training phase includes the following steps: S101, The transmitting end sends a training pilot signal to the receiving end; S102, The receiving end decomposes the training pilot signal into multiple sub-band pilot signals; S103. The receiving end performs downsampling processing on the multi-channel subband pilot signals to obtain downsampled subband pilot signals. All downsampled subband pilot signals are arranged in columns to form a subband pilot signal matrix. S104. Perform convolution operation on each downsampled subband pilot signal and the subband adaptive filter tap weights, and use the result as the subband estimated signal; after updating the subband adaptive filter tap weights in subsequent steps, use the new subband adaptive filter tap weights for convolution operation. S105. The training pilot signal sent by the transmitting end passes through the actual channel and is received by the receiving end. The signal received by the receiving end is used as the expected signal. S106. The receiving end decomposes the received desired signal into multiple sub-band desired signals; S107. Perform downsampling processing on the multi-channel subband desired signal to obtain the downsampled subband desired signal; S108. Calculate the difference between the estimated signal of each sub-band and the expected signal of each downsampled sub-band to obtain the error signal of each sub-band. Arrange all the sub-band error signals in columns to form an error signal matrix. S109. Calculate the anti-pulse factor matrix; the anti-pulse factor matrix is ​​a diagonal matrix, and its diagonal elements are... In the formula Q is the projection order, and ω is the generalized cost function. The performance factor, δ, is the generalized cost function. scaling factor, generalized cost function , This represents the error signal of the kl-th sub-band, where N is the total number of sub-bands. S110. Update the tap weights of the subband adaptive filter using the following formula: In the formula, g(k) is the current subband adaptive filter tap weight, g(k+1) is the updated subband adaptive filter tap weight, μ is the convergence factor, U(k) is the subband pilot signal matrix, E(k) is the error signal matrix, and H(k) is the anti-pulse factor matrix. S111. Merge and interpolate the error signals of each sub-band to reconstruct the broadband error signal; Repeat steps S101 to S111 until the mean square error between the current subband adaptive filter tap weight and the previous subband adaptive filter tap weight is less than a set threshold, or the maximum number of iterations is reached, then end.

2. The intelligent charging management system for electric vehicles as described in claim 1, characterized in that: Channel equalization during the data transmission phase includes the following steps: S201, The receiving end receives the signal; S202, The receiving end decomposes the received signal into multiple sub-band signals; S203. Perform downsampling processing on each sub-band signal to obtain the downsampled sub-band signal; S204. Perform convolution operation on the downsampled subband signals and the equalizer filter head weights to obtain the subband equalized signal; after the equalizer filter tap weights are updated in subsequent steps, the new equalizer filter tap weights are used for convolution operation. S205. The sub-band equalized signal is upsampled to obtain the reconstructed full-band signal; S206. The full-band signal is fed into the decision circuit to obtain an estimate of the original signal. The estimate of the original signal is decomposed and downsampled to obtain multiple sub-band reference signals. All sub-band reference signals are arranged in columns to form a reference signal matrix. S207. Subtract the subband reference signal and the subband equalization signal to obtain the subband error signal of the original signal. Arrange all the subband error signals of the original signal in columns to form the error signal matrix of the original signal. S208. Calculate the anti-pulse factor matrix. The process of calculating the anti-pulse factor matrix in this step is the same as in step S109, except that the subband error signal in the calculation process is the subband error signal of the original signal obtained in step S207. S209. Update the tap weights of the equalization filter using the following formula: In the formula, ω0(k) represents the current tap weights of the equalization filter, ω0(k+1) represents the updated tap weights of the equalization filter, and μ is the convergence factor. Here, H(k) is the reference signal matrix, H(k) is the anti-pulse factor matrix obtained in step S208, and E0(k) is the error signal matrix of the original signal. Repeat steps S201 to S209 until the mean square error between the current equalization filter tap weight and the previous equalization filter tap weight is less than the set threshold, or the maximum number of iterations is reached, then end.

3. The intelligent charging management system for electric vehicles as described in claim 1, characterized in that: A photovoltaic inverter, an energy storage system, and a smart meter are also installed on the user side or the website side. The smart charging pile is connected to the photovoltaic inverter, the energy storage system, and the smart meter via a power line and a local data interface.

4. The communication method of the intelligent charging management system for electric vehicles as described in any one of claims 1 to 3, characterized in that, Includes the following steps: S1. After the system is powered on, the first dual-mode communication module and the second dual-mode communication module are started, activating the two links of PLC and wireless communication; S2. The power line impulse noise suppression filter completes initialization; S3. The intelligent communication management agent module monitors the communication status of the entire system in real time, including monitoring the physical layer performance indicators and internal indicators of the PLC link. The internal indicators are the broadband error signals obtained during the pilot training phase. S4. The intelligent communication management agent module integrates the physical layer performance indicators and internal indicators of the PLC link to judge the link quality. If any indicator exceeds the preset threshold, the PLC link quality is considered to be degraded. S5. When it is determined that the PLC link quality has deteriorated, the intelligent communication management agent module switches the link according to a preset strategy. S6. If the PLC link quality is determined to be good, continue to use the PLC link for communication. S7. The intelligent communication management agent module records events and reports them to the main station.

5. The communication method as described in claim 4, characterized in that: The physical layer performance metrics of the PLC link include signal-to-noise ratio and bit error rate.

6. The communication method as described in claim 4, characterized in that: In step S5, the intelligent communication management agent module decides whether to switch links based on the priority of the service. For high-priority services, it immediately switches to the wireless link; for low-priority services, it adjusts the parameters of the power line impulse noise suppression filter so that the power line impulse noise suppression filter can be restored to stability before continuing to use the PLC link for communication.

7. The communication method as described in claim 6, characterized in that: The high-priority services include charging control commands, and the low-priority services include data upload services.

Images