Ammeter carrier communication module adaptive frequency adjustment method, medium and system

Abstract

The invention provides a self-adaptive frequency adjustment method, medium and system for an ammeter carrier communication module, and belongs to the technical field of intelligent ammeters, and the method comprises the steps: firstly obtaining and sampling a power line carrier communication signal, obtaining communication time distribution characteristics within 24 hours through time domain analysis, and carrying out frequency domain analysis to obtain noise spectrum characteristics; and establishing a time-frequency characteristic incidence matrix. 24 hours are divided into a plurality of transmission time windows based on the matrix, and the optimal frequency range of each window is determined and sub-band division is carried out. And performing channel quality evaluation on the sub-bands, substituting evaluation data into the noise characteristic equation set, solving to obtain a channel transmission characteristic curve, and selecting an optimal transmission sub-band for communication according to the channel transmission characteristic curve. The system monitors communication quality in real time, and performs time-frequency characteristic analysis and optimal sub-band selection again when a quality parameter is lower than a threshold value, so that adaptive frequency adjustment is realized, and the problem that current electric meter carrier communication is difficult to adapt to a complex power line channel environment is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of smart meters, and specifically relates to a method, medium and system for adaptive frequency adjustment of a meter carrier communication module. Background Art

[0002] Power line carrier communication uses the power line as the transmission medium and realizes data transmission by superimposing a high-frequency carrier signal on the power frequency (50 / 60 Hz), and is commonly used for smart meter data transmission. However, the power line carrier communication system faces severe noise interference problems. There are various electrical noise sources on the power line, such as switching power supplies, motors, etc., which will generate interference signals such as broadband white noise, narrowband harmonic noise, and short-time impulse noise, seriously affecting the communication quality, resulting in a time-varying channel environment and non-stationary noise characteristics, bringing great challenges to the performance of the communication system. How to adapt to the complex power line channel environment and improve the reliability and stability of communication is a key technical problem that urgently needs to be solved in the field of power line carrier communication. Summary of the Invention

[0003] In view of this, the present invention provides a method, medium and system for adaptive frequency adjustment of a meter carrier communication module, which can solve the technical problem that the carrier communication of the meter in the prior art is difficult to adapt to the complex power line channel environment.

[0004] The present invention is implemented as follows:

[0005] In the first aspect of the present invention, a method for adaptive frequency adjustment of a meter carrier communication module is provided, which acquires a power line carrier communication signal and samples it to obtain a sampled data sequence; performs time-domain analysis on the sampled data sequence to obtain a time distribution feature; performs frequency-domain analysis on the sampled data sequence to obtain a power line noise spectrum feature; establishes a time-frequency feature correlation matrix; divides 24 hours of each day into multiple transmission time windows according to the time-frequency feature correlation matrix; divides the optimal frequency range within each transmission time window into sub-bands and performs channel quality evaluation; substitutes the channel quality evaluation data into a noise feature equation set for solution to obtain a channel transmission characteristic curve; selects an optimal transmission sub-band for carrier communication according to the channel transmission characteristic curve, and re-performs analysis and selection when the communication quality parameter is lower than a preset threshold.

[0006] Among them, the time distribution feature includes a communication intensive period and a communication sparse period; the power line noise spectrum feature includes the frequency interference distribution in different periods; the time-frequency feature correlation matrix includes the corresponding relationship between the time distribution feature and the power line noise spectrum feature.

[0007] Among them, each transmission time window has a corresponding optimal frequency range; the channel quality evaluation data includes a signal-to-noise ratio, a bit error rate, and a channel attenuation.

[0008] Among them, the noise feature equation set includes a frequency response equation, a channel attenuation equation, and a signal integrity equation.

[0009] Among them, the frequency response equation characterizes the amplitude change characteristics of signals in different frequency bands.

[0010] Among them, the channel attenuation equation describes the energy loss of signals during transmission.

[0011] Among them, the signal integrity equation represents the waveform distortion degree during signal transmission.

[0012] Among them, when the communication quality parameter is lower than the preset threshold, return to execute the steps of time-domain analysis, frequency-domain analysis, establishing a time-frequency feature correlation matrix, dividing the transmission time window, performing sub-band division and channel quality evaluation, solving the feature equation set, and selecting the optimal transmission sub-band, and re-perform time-frequency feature analysis and optimal sub-band selection. Specifically, it includes the following steps:

[0013] S10. Obtain a power line carrier communication signal, sample the power line carrier communication signal to obtain a sampled data sequence; specifically: use a sampling device (such as a high-speed ADC) to sample the power line carrier communication signal; the sampling rate needs to be more than 2 times the communication bandwidth to meet the Nyquist sampling theorem; during the sampling process, windowing techniques such as Hanning window or Hamming window can be used to suppress spectral leakage; after obtaining the sampled data sequence x(n), subsequent time-domain analysis can be carried out; the main purpose of this step is to obtain the original power line carrier communication signal to be analyzed.

[0014] S20. Perform time-domain analysis on the sampled data sequence to obtain the time distribution characteristics of power line carrier communication within 24 hours a day, and the time distribution characteristics include communication-intensive periods and communication-sparse periods; specifically: calculate the number of communication requests N(t) within each time window (such as 300 seconds); record the actual communication duration L i (t) of each type of service; according to the communication density calculation formula calculate the communication density of each time window, where T is the statistical time window, α and β are weight coefficients, n is the total number of service types, and M i (t) is the maximum allowable communication duration of the i-th type of service; according to the change trend of the communication density D(t), divide the 24 hours of each day into communication-intensive periods and communication-sparse periods; the purpose of this step is to obtain the time distribution characteristics of power line carrier communication and provide a basis for subsequent frequency-domain analysis and time-frequency correlation.

[0015] S30. Perform frequency-domain analysis on the sampled data sequence to obtain the power line noise spectrum characteristics, where the power line noise spectrum characteristics include the frequency interference distribution in different time periods. Specifically: perform FFT transformation on the sampled data sequence x(n) and calculate the power spectral density. where Ξ is the number of FFT points (default 1024); collect 100 groups of noise data during the communication idle period, perform FFT analysis on each group of data and take the average to obtain the background noise power spectrum P bg (f); measure the power spectrum of the known signal source and compare it with the theoretical value, and calculate the correction coefficient γ = P measured / P theoretical ; according to the corrected power spectral density formula calculate the final noise power spectrum characteristics. The purpose of this step is to obtain the spectrum characteristics of the power line noise and provide a basis for subsequent time-frequency correlation analysis.

[0016] S40. Establish a time-frequency feature correlation matrix, where the time-frequency feature correlation matrix includes the correspondence between the time distribution feature and the power line noise spectrum feature. Specifically: construct a two-dimensional time-frequency feature correlation matrix M, with rows representing time features (time windows) and columns representing frequency features (noise power spectra); map the time distribution feature D(t) obtained in step S20 to the matrix rows to represent the communication density of each time window; map the noise power spectrum P(f) obtained in step S30 to the matrix columns to represent the noise characteristics of different frequency bands; the matrix element M i,j represents the correlation degree between time window i and frequency band j, and can be quantified using indicators such as correlation coefficient and mutual information. The purpose of this step is to establish a correlation mapping relationship between the two dimensions of time and frequency and provide a basis for subsequent adaptive frequency modulation.

[0017] S50. According to the time-frequency feature correlation matrix, divide the 24 hours of each day into multiple transmission time windows, and each of the transmission time windows has a corresponding optimal frequency range. Specifically: perform clustering analysis on the time-frequency feature correlation matrix M to divide the 24 hours into K transmission time windows; for each time window k, find the frequency range [f k,min , f k,max with the highest correlation degree as the optimal transmission frequency of this window; merge and optimize the optimal frequency ranges of each time window to minimize the frequency overlap of adjacent windows. The purpose of this step is to reasonably divide the transmission time windows and determine the optimal frequency range for each window according to the time-frequency feature correlation relationship, and provide a basis for subsequent adaptive frequency modulation.

[0018] S60. Subdivide the optimal frequency range within each of the said transmission time windows into sub - bands, and perform channel quality evaluation data, where the channel quality evaluation data includes signal - to - noise ratio, bit error rate, and channel attenuation. Specifically: Divide the optimal frequency range [f k,min , f k,max of each transmission time window into M sub - bands at equal intervals; for each sub - band, collect channel quality evaluation metrics, including signal - to - noise ratio SNR, bit error rate BER, and channel attenuation PER; record the channel quality evaluation data of each sub - band in a matrix, where the rows represent time windows and the columns represent sub - bands. The purpose of this step is to evaluate the channel quality status of different sub - bands within each transmission time window, providing a basis for subsequent adaptive modulation and frequency selection.

[0019] S70. Substitute the results of the channel quality evaluation data into a pre - fitted noise characteristic equation set for solution to obtain the channel transmission characteristic curve of each of the said transmission time windows. The purpose of this step is to establish a mathematical model of the power line carrier communication channel, describing various characteristics such as frequency response, attenuation characteristics, and signal integrity, providing a theoretical basis for subsequent adaptive modulation and frequency selection.

[0020] S80. Select the optimal transmission sub - band according to the channel transmission characteristic curve, perform carrier communication within the said transmission time window, and monitor the communication quality parameters in real - time. When the communication quality parameters are lower than the preset threshold, return to execute steps S20 to S80 to re - perform time - frequency characteristic analysis and optimal sub - band selection. Specifically: Select the sub - band with the optimal channel transmission characteristics for carrier communication; monitor the communication quality parameters in real - time, including signal - to - noise ratio SNR, bit error rate BER, and packet loss rate PER, and calculate the comprehensive communication quality score Q(t)=w1SNR(t)+w2BER(t)+w3PER(t); perform statistical analysis on the historical communication quality scores Q hist (i) to determine the preset threshold θ th = μ - kσ; when the real - time monitored communication quality score Q(t)<θ th , it indicates that the current channel no longer meets the requirements, and it is necessary to re - analyze the time - frequency characteristics and select the optimal sub - band. At this time, return to step S20 to continue execution. The purpose of this step is to achieve adaptive frequency adjustment of power line carrier communication, dynamically select the optimal transmission sub - band according to the real - time channel conditions, and ensure that the communication quality meets the expected requirements.

[0021] Based on the above technical solutions, an adaptive frequency adjustment method for an electric meter carrier communication module of the present invention can also be improved as follows:

[0022] Among them, the noise characteristic equation set includes a frequency response equation, a channel attenuation equation, and a signal integrity equation, where:

[0023] The frequency response equation characterizes the amplitude change characteristics of signals in different frequency bands,

[0024] The channel attenuation equation describes the energy loss of signals during transmission,

[0025] The signal integrity equation represents the degree of waveform distortion during signal transmission.

[0026] Furthermore, the communication-intensive period and the communication-sparse period are distinguished according to the communication intensity, and the calculation formula for the communication intensity is as follows:

[0027] The calculation formula for the communication intensity is as follows:

[0028]

[0029] In the formula, D(t) is the communication intensity at time t (range 0 - 1); N(t) is the number of communication times per unit time; T is the statistical time window (default value 300s); L i (t) is the actual communication duration of the i-th type of service; M i (t) is the maximum allowable communication duration of the i-th type of service; α, β are weight coefficients (default values are 0.6 and 0.4 respectively); n is the total number of service types. A service refers to various data streams transmitted and exchanged in data communication.

[0030] The general usage of the above formula is: count the total number of communication requests N(t) every 300s; record the communication duration L i (t) of each type of service; calculate the value of D(t) according to the formula.

[0031] Each equation in the noise characteristic equation set is described in detail as follows:

[0032] 1. Frequency response equation:

[0033]

[0034] In the formula, H(f) is the frequency response function; A0 is the reference gain (default value 1); α is the frequency attenuation coefficient (range 0.001 - 0.1); f k is the k-th resonance frequency point; Q k is the quality factor; δ i (f) is the i-th interference component; K is the number of resonance points; M is the number of interference sources.

[0035] 2. Channel attenuation equation:

[0036]

[0037] Where, L(d,f) is the path loss (dB) at distance d and frequency f; L0 is the path loss at the reference distance d0; η is the path loss exponent (range 2 - 4); κ is the frequency-dependent attenuation coefficient; W i is the attenuation weight of the i-th material; μ i is the attenuation coefficient of the i-th material; N is the number of material types on the transmission path.

[0038] 3. Signal integrity equation:

[0039]

[0040] Where, S(t,f) is the signal integrity index; V(t) is the time-domain voltage amplitude; V0 is the nominal voltage value; P(f) is the frequency-domain power; P0 is the nominal power value; x n is the actual sampling value; is the ideal sampling value; λ is the frequency-domain weight coefficient (range 0.1 - 1); ξ is the distortion weight coefficient (range 0.1 - 1); Ψ is the number of sampling points.

[0041] Parameter acquisition method:

[0042] 1. Acquisition of f k and Q k :

[0043] Step 1: Measure the system frequency response by using a network analyzer to sweep the frequency (range 1 - 500 kHz);

[0044] Step 2: Locate the resonance point through the peak detection algorithm:

[0045] Step 3: Calculate the 3dB bandwidth to determine the quality factor:

[0046] 2. Acquisition of W i and μ i :

[0047] Step 1: Conduct attenuation tests on a line with known material types;

[0048] Step 2: Use the least squares method to fit the test data to obtain the attenuation coefficient;

[0049] Step 3: Solve through matrices:

[0050]

[0051] The steps for obtaining the preset threshold specifically include:

[0052] 1. Collection of historical communication quality parameters:

[0053] Qhist (i) = w1SNR(i) + w2BER(i) + w3PER(i);

[0054] In the formula, Q hist (i) is the comprehensive communication quality score of the i-th historical sample point; SNR(i) is the signal-to-noise ratio (dB); BER(i) is the bit error rate; PER(i) is the packet loss rate; w1, w2, and w3 are weight coefficients (default values are 0.4, 0.3, and 0.3 respectively).

[0055] 2. Sample statistical analysis:

[0056]

[0057] In the formula, θ th is the finally determined preset threshold; μ is the historical sample mean; σ is the standard deviation; is the confidence coefficient (default value 3); Γ is the number of historical samples.

[0058] Among them, N(t) is the number of communications, and the acquisition method is: set a counter in the communication module; increment the counter when a communication request is detected; reset the counter after each statistical period ends.

[0059] Among them, T is the statistical time window, a system preset parameter: used to analyze the business response time requirement; a reasonable statistical period can be set by weighing, default 300s.

[0060] Among them, L i (t) is the actual communication duration, and the acquisition method is: record the start time stamp t of each communication start ; record the end time stamp t of the communication end ; calculate L i (t) = t end -t start .

[0061] Among them, P bg (f) is the background noise power spectrum, and the specific acquisition steps are: collect 100 pieces of noise data during the communication idle period; perform FFT analysis on each group of data; take the average value as the background noise spectrum.

[0062] Among them, γ is the correction coefficient, and the acquisition method is: measure the power spectrum of a known signal source and compare it with the theoretical value; calculate the correction coefficient γ = P measured / P theoretical .

[0063] Among them, A0 is the reference gain, and the acquisition method is: select a reference frequency point (usually 10kHz); measure the signal gain at this frequency point; perform normalization processing to obtain A0.

[0064] Among them, α is the frequency attenuation coefficient, and the acquisition method is as follows: measure the signal attenuation at different frequency points; use the least squares method to fit the attenuation curve; extract the exponential attenuation coefficient α.

[0065] Among them, the acquisition method of L0 is as follows: measure the signal intensity at the standard distance d0 (usually 1m); calculate the loss by comparing with the transmission power; repeat the measurement and take the average value.

[0066] Among them, the acquisition method of η is as follows: measure the signal intensity at different distance points; plot the logarithmic relationship diagram of distance-loss; obtain the slope η through linear regression.

[0067] Among them, the acquisition method of κ is as follows: fix the distance, measure the attenuation at different frequencies; establish a quadratic relationship model of frequency-attenuation; fit to obtain the coefficient κ.

[0068] Furthermore, the default value of the preset threshold is 0.6.

[0069] The second aspect of the present invention provides a computer-readable storage medium, wherein program instructions are stored in the computer-readable storage medium, and when the program instructions run on a computer, they are used to execute the above-mentioned adaptive frequency adjustment method for an electric meter carrier communication module.

[0070] The third aspect of the present invention provides an adaptive frequency adjustment system for an electric meter carrier communication module, which includes the above-mentioned computer-readable storage medium.

[0071] Compared with the prior art, the beneficial effects of an adaptive frequency adjustment method, medium and system for an electric meter carrier communication module provided by the present invention are as follows:

[0072] 1. Comprehensively explore the time-frequency characteristics of the power line channel: The present invention conducts dual analysis of the power line carrier communication signal in the time domain and frequency domain, extracts the communication time distribution characteristics and noise spectrum characteristics, and constructs an accurate time-frequency correlation model. This time-frequency characteristic analysis method can reflect the dynamic change law of the power line channel more than the existing single-dimensional adaptive scheme.

[0073] 2. Implement the optimization strategy of adaptive frequency modulation: Based on the constructed time-frequency correlation model, the present invention proposes an optimization strategy for adaptive frequency modulation. In each transmission time window, the optimal transmission frequency band is dynamically selected to avoid noise interference to the greatest extent, and the communication reliability is greatly improved. This adaptive frequency modulation mechanism can adapt to the complex and changeable power line channel environment more than the existing single fixed frequency band scheme.

[0074] 3. Low-power design for small embedded devices: In the algorithm implementation process of the present invention, mathematical modeling means such as noise characteristic equations and matrix calculations are adopted, which can operate efficiently on small chips (such as single-chip microcontrollers) and have the characteristics of low power consumption and low cost. This makes this method particularly suitable for small embedded devices such as smart meters, meeting the stringent requirements of power grid automation for communication systems.

[0075] In summary, the adaptive frequency regulation method for the electric meter carrier communication module proposed by the present invention realizes a significant improvement in the performance of the power line carrier communication system through key technological innovations such as time-frequency characteristic analysis and adaptive frequency modulation. This method makes full use of the time-frequency characteristics of the power line channel, dynamically adjusts the transmission frequency band to adapt to the complex noise environment, and greatly improves the communication reliability and stability. At the same time, this method is processed through mathematical models and matrix operations, without using complex neural networks or deep learning algorithms, and adopts a low-power and low-cost design in implementation, which is very suitable for small embedded application scenarios such as smart meters, and solves the technical problem that the carrier communication of electric meters in the prior art is difficult to adapt to the complex power line channel environment. Brief Description of the Drawings

[0076] Figure 1 is the flowchart of the method provided by the present invention;

[0077] Figure 2 is the change curve of the 24-hour communication density in the embodiment;

[0078] Figure 3 is the power line noise power spectrum characteristic in the embodiment;

[0079] Figure 4 is the curve graph of the channel transmission characteristics at different time periods. Detailed Embodiments

[0080] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0081] As Figure 1 shown, it is the flowchart of an adaptive frequency regulation method for an electric meter carrier communication module provided by the present invention. This method includes the following steps:

[0082] S10. Obtain the power line carrier communication signal, sample the power line carrier communication signal, and obtain a sampled data sequence;

[0083] S20. Perform time-domain analysis on the sampled data sequence to obtain the time distribution characteristics of the power line carrier communication within 24 hours a day. The time distribution characteristics include communication-intensive periods and communication-sparse periods;

[0084] S30. Perform frequency-domain analysis on the sampled data sequence to obtain the power line noise spectrum characteristics, which include the frequency interference distribution in different time periods.

[0085] S40. Establish a time-frequency feature correlation matrix, which includes the corresponding relationship between the time distribution feature and the power line noise spectrum characteristics.

[0086] S50. According to the time-frequency feature correlation matrix, divide the 24 hours of each day into multiple transmission time windows, and each transmission time window has a corresponding optimal frequency range.

[0087] S60. Perform sub-band division on the optimal frequency range within each transmission time window and conduct channel quality evaluation data, which includes signal-to-noise ratio, bit error rate, and channel attenuation.

[0088] S70. Substitute the results of the channel quality evaluation data into the pre-fitted noise feature equation set for solution to obtain the channel transmission characteristic curve of each transmission time window.

[0089] S80. Select the optimal transmission sub-band according to the channel transmission characteristic curve, perform carrier communication within the transmission time window, and monitor the communication quality parameters in real time. When the communication quality parameters are lower than the preset threshold, return to execute steps S20 to S80 to re-perform time-frequency feature analysis and optimal sub-band selection.

[0090] The following describes the specific implementation manners of the above steps in detail:

[0091] The specific implementation manner of step S10 is as follows: First, sample the power line carrier communication signal using a sampling device (such as a high-speed ADC). The sampling rate f s needs to be more than twice the communication bandwidth B to meet the Nyquist sampling theorem, that is, f s ≥ 2B. During the sampling process, windowing techniques such as the Hanning window w(n) or the Hamming window h(n) can be used to suppress spectral leakage. The mathematical expression of the Hanning window is: where NL is the window length. The mathematical expression of the Hamming window is: After windowing the sampled data sequence x(n), the original power line carrier communication signal to be analyzed can be obtained.

[0092] The specific implementation manner of step S20 is as follows: Perform time-domain statistical analysis on the sampled data sequence x(n) with the goal of extracting the time distribution characteristics of the power line carrier communication within 24 hours of each day. The specific method is as follows:

[0093] 1) Calculate the number of communication requests N(t) within each time window (e.g., 300 seconds). Set a counter in the communication module. When a communication request is detected, increment the counter by 1, and reset the counter after each statistical period ends.

[0094] 2) Record the actual communication duration L i (t) of each type of service. By recording the start timestamp t start and end timestamp t end of each communication, calculate L i (t) = t end - t start .

[0095] 3) According to the communication density calculation formula:

[0096]

[0097] Calculate the communication density of each time window. Among them, T is the statistical time window (default 300 seconds), α and β are weight coefficients (default values are 0.6 and 0.4 respectively), n is the total number of service types, M i (t) is the maximum allowable communication duration of the i-th type of service.

[0098] 4) According to the change trend of the communication density D(t), divide the 24 hours of each day into communication-intensive periods and communication-sparse periods. The communication-intensive periods correspond to the high communication density intervals, and the communication-sparse periods correspond to the low communication density intervals.

[0099] The purpose of this step is to obtain the time distribution characteristics of power line carrier communication, providing a basis for subsequent frequency domain analysis and time-frequency correlation.

[0100] The specific implementation of step S30 is as follows: Perform frequency domain analysis on the sampled data sequence x(n), and the goal is to extract the spectral characteristics of power line noise. The specific method is as follows:

[0101] 1) Perform N-point FFT transformation on the sampled data sequence and calculate the power spectral density:

[0102]

[0103] Among them, Ξ is the number of FFT points (default 1024).

[0104] 2) Collect M groups of noise data during the communication idle period, perform FFT analysis on each group of data and take the average value to obtain the background noise power spectrum P bg (f).

[0105] 3) Calculate the correction coefficient γ. Measure the power spectrum P measured of the known signal source and compare it with the theoretical value P theoretical(f) is compared to calculate γ = P measured (f) / P theoretical (f).

[0106] 4) According to the corrected power spectral density formula:

[0107]

[0108] Calculate the final noise power spectral characteristic P(f).

[0109] The purpose of this step is to obtain the spectral characteristics of power line noise and provide a basis for subsequent time-frequency correlation analysis.

[0110] The specific implementation of step S40 is as follows: Establish a time-frequency feature correlation matrix, and perform correlation analysis on the time distribution feature obtained in step S20 and the noise spectral feature obtained in step S30. The specific method is as follows:

[0111] 1) Construct a two-dimensional time-frequency feature correlation matrix M, where the rows represent time features (time windows) and the columns represent frequency features (noise power spectra).

[0112] 2) Map the time distribution feature D(t) obtained in step S20 to the rows of the matrix, representing the communication density of each time window.

[0113] 3) Map the noise power spectrum P(f) obtained in step S30 to the columns of the matrix, representing the noise characteristics of different frequency bands.

[0114] 4) The matrix element M i,j represents the correlation degree between time window i and frequency band j. The correlation degree can be quantified using indicators such as the correlation coefficient r i,j or mutual information I i,j etc. The calculation of the correlation coefficient or mutual information is all common existing technologies.

[0115] The purpose of this step is to establish a correlation mapping relationship between the two dimensions of time and frequency and provide a basis for subsequent adaptive frequency modulation.

[0116] The specific implementation of step S50 is as follows: According to the time-frequency feature correlation matrix M, divide 24 hours of each day into multiple transmission time windows and determine the optimal frequency range for each window. The specific steps are as follows:

[0117] 1) Perform clustering analysis on the time-frequency feature correlation matrix M to divide 24 hours into K transmission time windows. The clustering method can use the k-means algorithm or spectral clustering algorithm, and both the k-means algorithm and spectral clustering algorithm are common existing technologies.

[0118] 2) For each time window k, find the frequency range [f with the highest correlation degreek,min , f k,max as the optimal transmission frequency of this window. The correlation degree can adopt the aforementioned correlation coefficient r i,j or mutual information I i,j to measure.

[0119] 3) Merge and optimize the optimal frequency ranges of each time window to minimize the frequency overlap between adjacent windows.

[0120] The purpose of this step is to reasonably divide the transmission time window and determine the optimal frequency range for each window according to the time-frequency feature correlation relationship, providing a basis for subsequent adaptive frequency modulation.

[0121] The specific implementation manner of step S60 is as follows: perform sub-band division on the optimal frequency range [f k,min , f k,max of each transmission time window, and perform channel quality evaluation for each sub-band. The specific steps are as follows:

[0122] 1) Divide the optimal frequency range [f k,min , f k,max of each transmission time window into M sub-bands at equal intervals. The sub-band width Δf can be determined according to the bandwidth requirements of the communication system, that is

[0123] 2) For each sub-band j, collect channel quality evaluation indicators, including signal-to-noise ratio SNR j , bit error rate PER j and channel attenuation PER j . These indicators can be obtained by sending test signals and receiving and analyzing them.

[0124] 3) Record the channel quality evaluation data of each sub-band in a matrix, The rows represent time windows and the columns represent sub-bands.

[0125] The purpose of this step is to evaluate the channel quality status of different sub-bands within each transmission time window, providing a basis for subsequent adaptive modulation and frequency selection.

[0126] The specific implementation manner of step S70 is as follows: According to the channel quality evaluation data Q obtained in step S60, fit the noise feature equation set and solve the channel transmission characteristic curve of each transmission time window. The specific steps are as follows:

[0127] 1) Frequency response equation:

[0128]

[0129] Among them, A0 is the reference gain (default value 1), α is the frequency attenuation coefficient (range 0.001 - 0.1), f k is the k-th resonant frequency point, Q k is the quality factor, δ i (f) is the i-th interference component, K is the number of resonant points, and M is the number of interference sources.

[0130] 2) Channel attenuation equation:

[0131]

[0132] Among them, L(d, f) is the path loss (dB) at distance d and frequency f, L0 is the path loss at the reference distance d0, η is the path loss exponent (range 2 - 4), k is the frequency-dependent attenuation coefficient, W i is the attenuation weight of the i-th material, μ i is the attenuation coefficient of the i-th material, and N is the number of material types on the transmission path.

[0133] 3) Signal integrity equation:

[0134]

[0135] Among them, S(t, f) is the signal integrity index, V(t) is the time-domain voltage amplitude, V0 is the nominal voltage value, P(f) is the frequency-domain power, P0 is the nominal power value, x n is the actual sampling value, is the ideal sampling value, λ is the frequency-domain weight coefficient (range 0.1 - 1), ξ is the distortion weight coefficient (range 0.1 - 1), and Ψ is the number of sampling points.

[0136] 4) By fitting the parameter values of the above noise characteristic equations using the least squares method, the channel transmission characteristic curve of each transmission time window is obtained.

[0137] The purpose of this step is to establish a mathematical model of the power line carrier communication channel, describe the characteristics in aspects such as frequency response, attenuation characteristics, and signal integrity, and provide a theoretical basis for subsequent adaptive modulation and frequency selection.

[0138] The specific implementation of step S80 is as follows: According to the channel transmission characteristic curve obtained in step S70, select the optimal transmission sub-band and perform carrier communication within the corresponding transmission time window. At the same time, monitor the communication quality parameters in real time. When the communication quality is lower than the preset threshold, return to execute steps S20 to S80 to re-perform time-frequency characteristic analysis and optimal sub-band selection. The specific method is as follows:

[0139] 1) Select the sub-band j with the optimal channel transmission characteristics * for carrier communication. The preferred criteria include: signal-to-noise ratio Highest, bit error rate Lowest, channel attenuation Minimum, etc.

[0140] 2) Real-time monitor communication quality parameters, including signal-to-noise ratio SNR, bit error rate BER, and packet loss rate PER. The comprehensive communication quality score Q can be calculated in a weighted integration manner:

[0141] Q(t) = w1SNR(t) + w2BER(t) + w3PER(t);

[0142] where w1, w2, and w3 are the weight coefficients of the corresponding indicators and can be set according to actual requirements.

[0143] 3) Perform statistical analysis on the historical communication quality score Q hist (i) to determine the preset threshold θ th :

[0144]

[0145] where μ is the historical sample mean, σ is the standard deviation, k is the confidence coefficient (default value 3), and Γ is the number of historical samples.

[0146] 4) When the real-time monitored communication quality score Q(t) < θ th it indicates that the current channel no longer meets the requirements, and it is necessary to re-analyze the time-frequency characteristics and select the optimal sub-band. At this time, return to step S20 to continue execution.

[0147] The purpose of this step is to achieve adaptive frequency adjustment of power line carrier communication, dynamically select the optimal transmission sub-band according to the real-time channel conditions, and ensure that the communication quality meets the expected requirements.

[0148] In summary, the adaptive frequency adjustment method for the meter carrier communication module proposed by the present invention realizes the self-adaptive optimization of the power line carrier communication system through technologies such as time-frequency characteristic analysis, noise modeling, and adaptive modulation frequency selection.

[0149] The second aspect of the present invention provides a computer-readable storage medium, wherein program instructions are stored in the computer-readable storage medium, and when the program instructions run on a computer, they are used to execute the above-mentioned adaptive frequency adjustment method for a meter carrier communication module.

[0150] The third aspect of the present invention provides an adaptive frequency adjustment system for a meter carrier communication module, which includes the above-mentioned computer-readable storage medium.

[0151] Specifically, the principle of the present invention is:

[0152] 1. Time-frequency characteristic analysis

[0153] First, by sampling and performing time-domain analysis on the power line carrier communication signals, the communication time distribution characteristics within 24 hours of each day are obtained. Specifically, calculate the number of communication requests N(t) and the actual communication duration L i (t) within each time window, and obtain the communication density of each time window according to the communication density formula . In this way, the 24 hours can be divided into communication-intensive periods and communication-sparse periods.

[0154] Secondly, perform frequency-domain analysis on the sampled data to extract the power spectrum characteristics P(f) of the power line noise. Specifically, first use the FFT algorithm to calculate the power spectral density, and then consider the background noise P bg (f) and the correction coefficient γ to obtain the final noise power spectrum characteristics.

[0155] Finally, establish a time-frequency feature correlation matrix M to perform correlation mapping on the time distribution characteristics D(t) and the noise spectrum characteristics P(f). The matrix element M i,j represents the correlation degree between the time window i and the frequency band j, and can be quantified using indicators such as the correlation coefficient or mutual information.

[0156] 2. Adaptive frequency modulation optimization

[0157] Based on the time-frequency correlation matrix M, the present invention proposes an optimization strategy for adaptive frequency modulation. First, use the clustering algorithm to divide 24 hours into K transmission time windows, and find the optimal frequency range [f k,min , f k,max for each time window.

[0158] Then, perform sub-band division on the optimal frequency range of each transmission time window, and collect channel quality indicators for each sub-band, including the signal-to-noise ratio SNR j , the bit error rate BER j and the channel attenuation PER j . Record these indicators in the matrix Q.

[0159] Next, according to the channel quality evaluation data in the matrix Q, fit the parameter values of the noise characteristic equations (frequency response equation H(f), channel attenuation equation L(d, f) and signal integrity equation S(t, f)) to obtain the channel transmission characteristic curves of each transmission time window.

[0160] Finally, select the sub-band with the optimal channel transmission characteristics for carrier communication, and monitor the communication quality parameter Q(t) in real time. When Q(t) is lower than the preset threshold θ thWhen it indicates that the current channel no longer meets the requirements, it is necessary to return and re - perform time - frequency feature analysis and optimal sub - band selection.

[0161] This adaptive frequency modulation optimization strategy makes full use of the time - frequency characteristics of the power line channel, dynamically selects the optimal transmission sub - band, maximally avoids noise interference, and greatly improves the communication reliability and stability. At the same time, the algorithm design uses mathematical modeling means such as noise feature equation sets and matrix calculations, which can operate efficiently on small chips, and has the characteristics of low power consumption and low cost, and is very suitable for embedded application scenarios such as smart meters.

[0162] The following provides an embodiment of a specific application scenario of the present invention: A power grid enterprise is using power line carrier communication technology to remotely meter and telemetry monitor smart meters at the user end. This power grid belongs to a typical low - voltage distribution network, with a line length of about 50 kilometers and covering more than 30,000 users. Due to the uneven load distribution in this area, different degrees of aging of user equipment, and numerous surrounding industrial enterprises, there are serious noise interference problems on the power line, which poses a great challenge to the operation of the carrier communication system.

[0163] In order to improve the reliability of carrier communication in the low - voltage distribution network, this power grid enterprise decides to apply the adaptive frequency regulation technology proposed by the present invention to smart meters at the user end. The specific implementation steps are as follows:

[0164] Signal acquisition and time - frequency feature analysis

[0165] First, the carrier communication module built in the smart meter samples the power line carrier communication signal using a high - speed ADC. The sampling rate f s is set to 200 kHz, which is greater than twice the communication bandwidth B = 100 kHz, meeting the Nyquist sampling theorem. During the sampling process, the Hanning window function (where N = 1024 is the window length) is used to window the sampling data to suppress spectral leakage.

[0166] After obtaining the sampling data sequence x(n), perform time - domain statistical analysis on it. Specifically, the number of communication requests N(t) is statistically analyzed every 300 seconds (i.e., 5 minutes), and the actual communication duration L i (t) of each type of service is recorded. Taking the smart metering service as an example, the maximum allowable communication duration M i (t) of this service is set to 60 seconds. According to the communication density calculation formula:

[0167]

[0168] Calculate the communication density D(t) of each time window. Figure 2The curve of the communication density change within 24 hours of a certain day is given. The horizontal axis represents time (hours), and the vertical axis represents the communication density value. The figure clearly shows the change characteristics of the communication sparse period (0:00 - 6:00), the communication intensive period (6:00 - 22:00), and the communication sparse period (22:00 - 24:00). It can be seen that 0:00 - 6:00 is the communication sparse period, 6:00 - 22:00 is the communication intensive period, and 22:00 - 24:00 enters the communication sparse period again.

[0169] Table 1 Statistical results of communication density within 24 hours of a certain day

[0170]

[0171]

[0172] Next, perform frequency-domain analysis on the sampled data x(n). First, calculate the power spectral density using the 1024-point FFT algorithm Then, collect 100 groups of noise data during the communication idle period (22:00 - 24:00), perform FFT analysis on each group of data and take the average to obtain the background noise power spectrum P bg (f). After comparative testing, the correction coefficient γ = 1.2 is calculated. Finally, according to the corrected power spectral density formula Draw the spectral characteristics of the power line noise, as Figure 3 shown, which shows the power spectral characteristics of the power line noise in the frequency range of 50 kHz - 150 kHz. The horizontal axis is the frequency (kHz), and the vertical axis is the power spectral density (dBm / Hz). It can be seen from the figure that the noise power is relatively concentrated around 100 kHz, and there are obvious narrowband interferences.

[0173] Through the above dual analysis in the time domain and frequency domain, the time distribution characteristics and noise spectral characteristics of the power line carrier communication within 24 hours of each day are obtained. Next, correlate and analyze the characteristics in these two dimensions to construct a time-frequency feature correlation matrix.

[0174] Time-frequency feature correlation analysis

[0175] First, divide 24 hours into 12 transmission time windows, each window being 2 hours. Using the k-means clustering algorithm, map the time distribution feature D(t) to 12 rows and the noise power spectrum P(f) to 12 columns to construct a 12×12 time-frequency feature correlation matrix M. The matrix element M i,j represents the correlation degree between the time window i and the frequency band j, and uses the correlation coefficient r i,j to quantify:

[0176]

[0177] Among them, and are the average values of D i and P j respectively. The calculated time-frequency feature correlation matrix M is shown in Table 2.

[0178] Table 2 Time-frequency feature correlation matrix M;

[0179]

[0180]

[0181] It can be seen from the time-frequency feature correlation matrix M that during the communication-intensive period (6:00 - 22:00), the correlation degree between each frequency band and the time window is relatively high, indicating that the noise environment is relatively complex during this period; while during the communication-sparse period (0:00 - 6:00, 22:00 - 24:00), the correlation degree between each frequency band and the time window is relatively low, and the noise environment is relatively good. This provides a basis for the subsequent adaptive frequency modulation strategy.

[0182] Adaptive frequency modulation optimization

[0183] Based on the time-frequency feature correlation matrix M, the adaptive frequency modulation strategy is implemented next. First, each 2-hour transmission time window is equally divided into 5 sub-bands, with center frequencies of 60 kHz, 80 kHz, 100 kHz, 120 kHz, and 140 kHz respectively, and the bandwidth of each is 20 kHz. Then, for each sub-band, channel quality evaluation indicators are collected, including signal-to-noise ratio SNR j , bit error rate BER j and channel attenuation PER j . The specific test results are shown in Table 3.

[0184] Table 3 Channel quality evaluation of each sub-band

[0185]

[0186]

[0187] It can be seen from Table 3 that during the communication-intensive period (6:00 - 22:00), the channel quality of each sub-band is relatively poor, the signal-to-noise ratio is low, and the bit error rate and channel attenuation are high; while during the communication-sparse period (0:00 - 6:00, 22:00 - 24:00), the channel quality of the sub-band is relatively good. This also verifies the results of the previous time-frequency feature correlation analysis.

[0188] Based on the channel quality data of the sub-bands, the next step is to fit the noise feature equations to solve the channel transmission characteristic curve of each transmission time window. Specifically:

[0189] 1) Frequency response equation:

[0190]

[0191] Where, A0 = 1.0 is the reference gain, α = 0.05 is the frequency attenuation coefficient, f1 = 80 kHz, Q1 = 40 are the parameters of the first resonant frequency point, f2 = 120 kHz, Q2 = 30 are the parameters of the second resonant frequency point, and δ(f - 100000) is the interference component at 100 kHz.

[0192] 2) Channel attenuation equation:

[0193]

[0194] Where, L0 = 20 dB is the path loss at the reference distance of 1 m, η = 3 is the path loss exponent, κ = 0.001 is the frequency-dependent attenuation coefficient, W1 = 5, μ1 = 0.05 are the attenuation parameters of one kind of material.

[0195] 3) Signal integrity equation:

[0196]

[0197] Where, V0 = 220 V is the nominal voltage value, P0 = 1 W is the nominal power value, λ = 0.6, ξ = 0.2 are the weight coefficients, and N = 1024 is the number of sampling points.

[0198] Substituting the parameters of the above noise characteristic equations, the channel transmission characteristic curves for each transmission time window can be obtained. As Figure 4 shown, the channel transmission characteristic curves for three typical time periods (communication sparse, transition, and dense periods) are presented. The horizontal axis is the frequency (kHz), and the vertical axis is the frequency response amplitude.

[0199] Finally, according to the channel transmission characteristics of each transmission time window, the optimal sub-bands are dynamically selected for carrier communication. During the communication dense period (6:00 - 22:00), the 120 kHz and 140 kHz sub-bands are preferred; during the communication sparse periods (0:00 - 6:00, 22:00 - 24:00), the 60 kHz and 80 kHz sub-bands can be selected. Meanwhile, the comprehensive communication quality score Q(t) = 0.4·SNR(t) + 0.3·BER(t) + 0.3·PER(t) is monitored in real-time. When Q(t) < 28 (corresponding to an average signal-to-noise ratio of 27 dB, an average bit error rate of 3e-3, and an average packet loss rate of 3%), it indicates that the current channel quality no longer meets the requirements, and it is necessary to return for time-frequency characteristic analysis and optimal sub-band selection again.

[0200] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention.

Claims

1. A method for adaptive frequency adjustment of an electric meter carrier communication module, characterized in that: include: Acquire the power line carrier communication signal and sample it to obtain a sampling data sequence; perform time domain analysis on the sampling data sequence to obtain time distribution characteristics; Perform frequency domain analysis on the sampled data sequence to obtain the spectrum characteristics of power line noise; Establish a time-frequency feature correlation matrix; divide the 24 hours of each day into multiple transmission time windows according to the time-frequency feature correlation matrix; divide the optimal frequency range in each transmission time window into sub-bands and perform channel quality assessment; substitute the channel quality assessment data into the noise characteristic equation group to solve and obtain the channel transmission characteristic curve; select the optimal transmission sub-band for carrier communication according to the channel transmission characteristic curve, and re-analyze and select when the communication quality parameters are lower than the preset threshold.

2. The method for adaptive frequency adjustment of the electric meter carrier communication module according to claim 1, characterized in that: The time distribution characteristics include communication intensive periods and communication sparse periods; the power line noise spectrum characteristics include frequency interference distribution in different periods; the time-frequency characteristic association matrix includes the correspondence between time distribution characteristics and power line noise spectrum characteristics.

3. The method for adaptive frequency adjustment of the electric meter carrier communication module according to claim 2, characterized in that: Each of the transmission time windows has a corresponding optimal frequency range; the channel quality assessment data includes signal-to-noise ratio, bit error rate, and channel attenuation.

4. The method for adaptive frequency adjustment of the electric meter carrier communication module according to claim 3, characterized in that: The noise characteristic equation group includes a frequency response equation, a channel attenuation equation and a signal integrity equation.

5. The method for adaptively adjusting the frequency of the electric meter carrier communication module according to claim 4, characterized in that: The frequency response equation characterizes the amplitude variation characteristics of signals in different frequency bands.

6. The method for adaptively adjusting the frequency of the electric meter carrier communication module according to claim 5, characterized in that: The channel attenuation equation describes the energy loss of a signal during transmission.

7. The method for adaptively adjusting the frequency of the electric meter carrier communication module according to claim 6, characterized in that: The signal integrity equation represents the degree of waveform distortion in signal transmission.

8. The method for adaptively adjusting the frequency of the electric meter carrier communication module according to claim 7, characterized in that: When the communication quality parameter is lower than the preset threshold, the process returns to the steps of executing time domain analysis, frequency domain analysis, establishing a time-frequency feature correlation matrix, dividing the transmission time window, performing sub-band division and channel quality evaluation, solving a set of characteristic equations, and selecting the optimal transmission sub-band, and re-performing the time-frequency feature analysis and the optimal sub-band selection.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program instructions, and when the program instructions are executed in a computer, the method for adaptively adjusting the frequency of an electric meter carrier communication module according to any one of claims 1 to 8 is used to execute the method.

10. An adaptive frequency adjustment system for an electric meter carrier communication module, characterized in that: A computer-readable storage medium comprising the computer-readable storage medium of claim 9.

Images