A power line communication optimization method and system for new power systems

Abstract

The present invention relates to the field of power system communication technology, and in particular to a power line communication optimization method and system for a new power system, including extracting corresponding harmonic noise signals according to operating parameters of renewable energy power generation equipment, classifying the harmonic noise signals of all renewable energy power generation equipment using a clustering algorithm to obtain typical noise frequency patterns, and quantitatively evaluating the degree of influence of harmonic noise signals on power line carrier communication signals, and adaptively modulating power line carrier communication signals based on noise impact evaluation results to obtain power line carrier modulation signals; setting a reward and punishment mechanism based on the power line carrier modulation signal, and using a deep Q learning algorithm to screen out the optimal communication path for data transmission. The present invention optimizes communication efficiency and noise interference resistance by adaptively selecting communication modulation methods and other technologies, thereby improving the noise interference resistance and communication quality of power line carrier communication in wind farms and photovoltaic power station environments.

Description

Technical Field

[0001] The present invention relates to the technical field of power system communication, and in particular to a power line communication optimization method and system for a new power system. Background Art

[0002] In the process of building the distribution communication network of the new power system, the large-scale access of renewable energy such as wind power, photovoltaic power and tidal power to the power grid provides important support for the sustainability of power supply. However, although the access of renewable energy has enriched the energy structure of the power system, it has also brought about complex and changeable power line noise problems, especially renewable energy such as wind power and photovoltaic power. These noise sources seriously interfere with the normal operation of the distribution communication network.

[0003] Specifically, different types of renewable energy will generate power line noise with specific frequencies and amplitudes due to their different operating mechanisms. Wind power generation generates periodic noise due to changes in wind speed and blade speed; photovoltaic power generation generates high-frequency noise due to the switching frequency of the inverter; and the noise of tidal power generation fluctuates with the tidal cycle. These different noise sources produce mutual coupling and superposition effects on the power lines, causing multi-frequency interference to communication signals, which greatly affects the integrity and reliability of communication signals, resulting in the loss or error of key information in data transmission, and thus threatening the safe and stable operation of the power grid.

[0004] There are many types of existing power line noise, including harmonic noise generated by renewable energy sources such as wind power and photovoltaic power generation. These noises are interference sources that cannot be ignored in the power system. The randomness and suddenness of power line noise further increase the complexity of the problem. Therefore, in the construction of the distribution communication network of the new power system, how to accurately identify and effectively suppress the power line noise generated by the access of renewable energy has become a key issue that needs to be solved urgently, so as to ensure that the power grid can operate safely, stably and efficiently in a complex and changeable operating environment. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a power line communication optimization method and system for a new power system.

[0006] In a first aspect, the present invention provides a method for optimizing power line communication for a new power system, the method comprising the following steps:

[0007] According to the collected operating parameters of renewable energy power generation equipment, the harmonic noise signals of different renewable energy power generation equipment are extracted by using the short-time Fourier transform spectrum analysis method; wherein the operating parameters of renewable energy power generation equipment include the operating parameters of wind power equipment at the collection point of each wind farm and the operating parameters of photovoltaic equipment at the grid connection point of the photovoltaic power station;

[0008] According to the spatial correlation of the harmonic noise signals, a clustering algorithm is used to classify the harmonic noise signals of all renewable energy power generation equipment to obtain a typical noise frequency pattern;

[0009] Quantitatively evaluate the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern to obtain a noise influence evaluation result;

[0010] Performing physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulation signal;

[0011] A reward and punishment mechanism is set based on the power line carrier modulation signal, and a deep Q learning algorithm is used to dynamically optimize the power line carrier communication route to select the optimal communication path for data transmission.

[0012] In a further implementation scheme, the wind power equipment operating parameters include the wind turbine blade rotation frequency, the gearbox meshing frequency and the generator rotor rotation frequency, and the photovoltaic equipment operating parameters include the photovoltaic module switching frequency, the inverter switch tube switching frequency and the inverter output current harmonic frequency; the step of extracting the harmonic noise signals of different renewable energy power generation equipment using the short-time Fourier transform spectrum analysis method based on the collected renewable energy power generation equipment operating parameters includes:

[0013] The principal component analysis method is used to reduce the dimension of the collected operating parameters of renewable energy power generation equipment and extract the key operating characteristics of renewable energy power generation equipment;

[0014] The K-means clustering algorithm is used to group the key operating characteristics of renewable energy power generation equipment and identify the equipment parameter patterns under different operating conditions;

[0015] According to the center point of the equipment parameter pattern, the Mamdani fuzzy inference system is used to calculate the Euclidean distance between the currently collected operating parameters of the renewable energy power generation equipment and the center point of each equipment parameter pattern to obtain the pattern matching distance;

[0016] dynamically determining a time window width and an overlap rate of a short-time Fourier transform according to the pattern matching distance;

[0017] According to the time window width and overlap rate of short-time Fourier transform, short-time Fourier transform spectrum analysis is performed on the operating parameters of renewable energy power generation equipment to obtain the time-frequency spectrum diagram of renewable energy power generation equipment;

[0018] According to the time-frequency spectrum of the renewable energy power generation equipment, the harmonic noise signal is extracted; the harmonic noise signal includes timestamp, type of renewable energy power generation equipment, frequency, amplitude and phase information.

[0019] In a further embodiment, the spatial correlation acquisition process of the harmonic noise signal includes:

[0020] Establishing a three-dimensional spatial coordinate system according to the geographic coordinates of the renewable energy power generation equipment, and mapping the spatial location information of the wind farm collection point and the photovoltaic power station grid connection point to the three-dimensional spatial coordinate system;

[0021] Calculating the Euclidean distances between observation points inside the wind farm, between observation points inside the photovoltaic power station, and between the wind farm and the photovoltaic power station in the three-dimensional spatial coordinate system to obtain a distance matrix, and constructing a spatial weight matrix based on the distance matrix;

[0022] The Pearson correlation coefficient is used to calculate the correlation degree of harmonic noise signals between wind farm collection points, and the wind farm correlation coefficient matrix within and between wind farms is obtained;

[0023] The Pearson correlation coefficient is used to calculate the correlation degree of harmonic noise signals between the photovoltaic power station grid connection points, and the photovoltaic power station correlation coefficient matrix within the photovoltaic power station and between photovoltaic power stations is obtained;

[0024] According to the wind farm correlation coefficient matrix, photovoltaic power station correlation coefficient matrix and spatial weight matrix, the global spatial autocorrelation of the harmonic noise signals of wind farms and photovoltaic power stations is calculated using the global Moran index to obtain the spatial correlation of the harmonic noise signals.

[0025] In a further embodiment, the step of classifying the harmonic noise signals of all renewable energy power generation equipment using a clustering algorithm according to the spatial correlation of the harmonic noise signals to obtain a typical noise frequency pattern comprises:

[0026] Combining the harmonic noise signal and its spatial correlation and duration into a multi-dimensional feature vector;

[0027] The elbow rule is used to determine the optimal clustering number of harmonic noise signals of different renewable energy generation equipment by calculating the intra-cluster sum of squares under different clustering numbers.

[0028] Based on the optimal cluster number, a K-means clustering algorithm is used to perform clustering analysis on the multidimensional feature vector to obtain a clustering result;

[0029] The cluster center features are extracted according to the clustering results to form a typical noise frequency pattern.

[0030] In a further embodiment, the step of quantitatively evaluating the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern to obtain the noise impact evaluation result comprises:

[0031] Using a similarity measurement algorithm to match the typical noise frequency pattern with a pre-built renewable energy noise frequency feature library to obtain a matching similarity;

[0032] If the matching similarity exceeds a preset similarity threshold, selecting the typical noise frequency pattern with the highest matching similarity as the target noise pattern;

[0033] Calculating the frequency overlap between the target noise pattern and the power line carrier communication signal, and calculating the signal-to-noise ratio using the ratio of the harmonic noise amplitude corresponding to the target noise pattern to the power line carrier communication signal amplitude;

[0034] The influence of harmonic noise on the power line carrier communication signal is quantitatively evaluated according to the frequency overlap and the signal-to-noise ratio to obtain a noise influence evaluation result.

[0035] In a further embodiment, the renewable energy noise frequency feature library includes a wind power noise frequency feature library and a photovoltaic noise frequency feature library, and the construction process of the renewable energy noise frequency feature library is specifically as follows:

[0036] Acquire historical harmonic noise signals at wind farm collection points and photovoltaic power station grid connection points, and preprocess the historical harmonic noise signals to obtain preprocessed harmonic noise signals;

[0037] Performing frequency domain analysis on the preprocessed harmonic noise signal using a fast Fourier transform algorithm to extract the amplitude and frequency characteristics of each harmonic component;

[0038] Calculating the amplitude and frequency characteristics of each harmonic component using a statistical analysis method to obtain a renewable energy noise characteristic statistic;

[0039] Using a kernel density estimation algorithm to perform non-parametric estimation on the amplitude and frequency characteristics of each harmonic component, and construct a probability density function;

[0040] A cumulative distribution function is obtained by integrating the probability density function, and the occurrence probability of historical harmonic noise signals of different amplitudes and frequency intervals is quantified according to the cumulative distribution function;

[0041] A renewable energy noise frequency feature library is constructed according to the renewable energy noise feature statistics, harmonic order and the occurrence probability.

[0042] In a further embodiment, the step of performing physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulated signal comprises:

[0043] Filtering the power line carrier communication signal according to the noise impact assessment result to obtain a power line carrier communication filtered signal;

[0044] Using a frequency scanning method to measure the resonant frequency of the power line carrier communication filter signal within a preset frequency range, and determine the potential crosstalk frequency of the power line carrier communication filter signal;

[0045] Selecting an orthogonal frequency division multiplexing modulation mode according to the resonant frequency and the potential crosstalk frequency;

[0046] Acquire a real-time channel state, and calculate a signal-to-noise ratio of each subcarrier based on the real-time channel state to obtain a channel quality indicator;

[0047] Determine a coding rate according to the channel quality indicator at the link layer, and adjust a subcarrier modulation mode in the orthogonal frequency division multiplexing modulation mode according to the coding rate to obtain a target modulation mode;

[0048] The power line carrier communication filtered signal is modulated according to the target modulation mode to generate a power line carrier modulation signal.

[0049] In a further embodiment, the step of filtering the power line carrier communication signal according to the noise impact assessment result to obtain the power line carrier communication filtered signal comprises:

[0050] If the noise impact assessment result is within a preset interference threshold range, the noise characteristics of the harmonic noise signal are identified, and the corresponding interference source is determined according to the noise characteristics of the harmonic noise signal;

[0051] A filtering strategy is determined according to the interference source, and noise is filtered on the power line carrier communication signal according to the filtering strategy to obtain a power line carrier communication filtered signal; the filtering strategy includes a wavelet analysis method and an empirical mode decomposition method.

[0052] In a further implementation scheme, the step of setting a reward and punishment mechanism based on the power line carrier modulation signal, and dynamically optimizing the power line carrier communication route using a deep Q learning algorithm, and selecting the optimal communication path for data transmission includes:

[0053] Different PLC modulation signals are considered as elements of state space, and different communication paths are defined as action space;

[0054] According to the current network status, identify the spatial distribution pattern of the power line carrier modulation signal on different communication paths;

[0055] A reward and punishment mechanism is set up according to the spatial distribution law and key performance indicators of the power line carrier modulation signal; the key performance indicators include the transmission success rate, rate and transmission delay index of the power line carrier modulation signal;

[0056] Randomly initialize the Q value and select the communication path action using the ε-greedy strategy according to the current network state at each time step;

[0057] Each communication path action is evaluated according to the reward and punishment mechanism, and the Q-Learning algorithm is used to learn the spatial distribution law of the power line carrier modulation signal to dynamically adjust the selection of the communication path, update the Q value of each communication path, until the convergence condition is reached, and find the communication path with the highest Q value in each state;

[0058] The communication path with the highest Q value is selected as the optimal communication path for power line carrier modulation signal transmission.

[0059] In a second aspect, the present invention provides a power line communication optimization system for a new power system, the system comprising:

[0060] The spectrum analysis module is used to extract the harmonic noise signals of different renewable energy power generation equipment using the short-time Fourier transform spectrum analysis method according to the collected operating parameters of the renewable energy power generation equipment; wherein the operating parameters of the renewable energy power generation equipment include the operating parameters of the wind power equipment at the collection point of each wind farm and the operating parameters of the photovoltaic equipment at the grid connection point of the photovoltaic power station;

[0061] A signal classification module, used to classify the harmonic noise signals of all renewable energy power generation equipment according to the spatial correlation of the harmonic noise signals by using a clustering algorithm to obtain a typical noise frequency pattern;

[0062] A noise assessment module, used for quantitatively assessing the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern, and obtaining a noise impact assessment result;

[0063] A signal modulation module, used for performing physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulation signal;

[0064] The communication optimization module is used to set a reward and punishment mechanism based on the power line carrier modulation signal, and dynamically optimize the power line carrier communication route using a deep Q learning algorithm to select the optimal communication path for data transmission.

[0065] The present invention provides a power line communication optimization method and system for a new power system. The method uses a short-time Fourier transform spectrum analysis method to extract harmonic noise signals of different renewable energy power generation equipment based on the collected operating parameters of the renewable energy power generation equipment; uses a clustering algorithm to classify the harmonic noise signals of all renewable energy power generation equipment based on the spatial correlation of the harmonic noise signals to obtain a typical noise frequency pattern; quantitatively evaluates the influence of the harmonic noise signal on the power line carrier communication signal based on the typical noise frequency pattern to obtain a noise impact assessment result; performs physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulation signal; sets a reward and punishment mechanism based on the power line carrier modulation signal, and uses a deep Q learning algorithm to dynamically optimize the power line carrier communication route to screen out the optimal communication path for data transmission. Compared with the existing technology, this method can accurately evaluate the impact of harmonic noise generated by wind power and photovoltaic equipment on communication signals, and provide an effective solution for power line carrier communication after renewable energy is connected to the grid through technologies such as adaptive selection of communication modulation methods, alleviating the interference of harmonic noise on the communication system, and improving the power line carrier communication's anti-noise interference capability and communication quality in wind farm and photovoltaic power station environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a schematic diagram of the flow of a power line communication optimization method for a new power system provided by an embodiment of the present invention;

[0067] Figure 2 It is a block diagram of a power line communication optimization system for a new power system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0068] The following specifically illustrates the implementation mode of the present invention in conjunction with the accompanying drawings. The embodiments are provided for illustrative purposes only and cannot be understood as limiting the present invention. The accompanying drawings are provided for reference and illustration only and do not constitute a limitation on the scope of patent protection of the present invention, because many changes can be made to the present invention without departing from the spirit and scope of the present invention.

[0069] refer to Figure 1 , the embodiment of the present invention provides a power line communication optimization method for a new power system, such as Figure 1 As shown, the method comprises the following steps:

[0070] S1. Based on the collected operating parameters of renewable energy power generation equipment, the short-time Fourier transform spectrum analysis method is used to extract the harmonic noise signals of different renewable energy power generation equipment; wherein the operating parameters of renewable energy power generation equipment include the operating parameters of wind power equipment at the collection points of each wind farm and the operating parameters of photovoltaic equipment at the grid-connected points of photovoltaic power stations.

[0071] In this embodiment, the wind power equipment operating parameters include the wind turbine blade rotation frequency, the gearbox meshing frequency and the generator rotor rotation frequency, and the photovoltaic equipment operating parameters include the photovoltaic module switching frequency, the inverter switch tube switching frequency and the inverter output current harmonic frequency; the step of extracting the harmonic noise signals of different renewable energy power generation equipment using the short-time Fourier transform spectrum analysis method based on the collected renewable energy power generation equipment operating parameters includes:

[0072] The principal component analysis method is used to reduce the dimension of the collected operating parameters of renewable energy power generation equipment and extract the key operating characteristics of renewable energy power generation equipment;

[0073] The K-means clustering algorithm is used to group the key operating characteristics of renewable energy power generation equipment and identify the equipment parameter patterns under different operating conditions;

[0074] According to the center point of the equipment parameter pattern, the Mamdani fuzzy inference system is used to calculate the Euclidean distance between the currently collected operating parameters of the renewable energy power generation equipment and the center point of each equipment parameter pattern to obtain the pattern matching distance;

[0075] dynamically determining a time window width and an overlap rate of a short-time Fourier transform according to the pattern matching distance;

[0076] According to the time window width and overlap rate of short-time Fourier transform, short-time Fourier transform spectrum analysis is performed on the operating parameters of renewable energy power generation equipment to obtain the time-frequency spectrum diagram of renewable energy power generation equipment;

[0077] According to the time-frequency spectrum of the renewable energy power generation equipment, the harmonic noise signal is extracted; the harmonic noise signal includes timestamp, type of renewable energy power generation equipment, frequency, amplitude and phase.

[0078] Specifically, this embodiment collects operating parameters from the monitoring system of renewable energy power generation equipment (such as wind farms and photovoltaic power stations) in real time to obtain the operating parameters of wind power equipment at the collection points of each wind farm and the operating parameters of photovoltaic equipment at the grid-connected points of photovoltaic power stations, wherein the operating parameters of wind power equipment include but are not limited to the rotation frequency of wind turbine blades, the meshing frequency of gearboxes and the rotation frequency of generator rotors, and the operating parameters of photovoltaic equipment include but are not limited to the switching frequency of photovoltaic components, the switching frequency of inverter switching tubes and the harmonic frequency of inverter output current, etc. Then, this embodiment uses the principal component analysis method (PCA) to reduce the dimension of these data to extract key feature vectors that can characterize the operating state of the equipment and obtain the key features of the operation of the renewable energy power generation equipment. After the PCA dimensionality reduction, this embodiment uses the K-means clustering algorithm to group these key operating features of the renewable energy power generation equipment to identify the equipment parameter patterns under different operating states. Each cluster center represents the typical operating state of the equipment in that state, and the cluster boundary defines the transition area between each state.

[0079] Next, this embodiment uses the Mamdani fuzzy inference system to calculate the Euclidean distance between the currently collected equipment operating parameters and the center point of each equipment parameter mode. The Mamdani fuzzy inference system is a reasoning method based on fuzzy logic, which can process fuzzy input and output. This embodiment uses the currently collected renewable energy power generation equipment operating parameters as the input of the fuzzy inference system, and calculates the fuzzy distance between it and the center point of each equipment parameter mode through preset fuzzy rules and membership functions, and finally converts it into the Euclidean distance. This distance value reflects the similarity between the current equipment state and each equipment parameter mode. According to the calculated Euclidean distance, the time window width and overlap rate of the short-time Fourier transform (STFT) are dynamically adjusted to optimize the accuracy and efficiency of spectrum analysis. Specifically, this embodiment can appropriately increase or decrease the width of the time window and adjust the overlap rate according to the size of the Euclidean distance to ensure the continuity of the signal and the accuracy of the analysis. For example, when the equipment state is close to or in a certain equipment parameter mode, the time window can be adjusted to capture the signal characteristics more finely; otherwise, a wider time window is used to improve processing efficiency.

[0080] Finally, this embodiment uses a determined time window width and overlap rate to perform short-time Fourier transform spectrum analysis (STFT) on the real-time collected operating parameters of the renewable energy power generation equipment. STFT divides the signal into multiple shorter time periods (i.e., time windows), and performs Fourier transform on each time period to generate a time-frequency spectrum diagram of the renewable energy power generation equipment. The time-frequency spectrum diagram not only shows the frequency change of the signal over time, but also reveals the distribution and change of each harmonic. In the obtained time-frequency spectrum diagram, this embodiment can analyze the frequency change characteristics of the signal over time, identify and separate signal components that are integer multiples of the fundamental frequency to extract the harmonic noise signal of the equipment.

[0081] S2. Based on the spatial correlation of the harmonic noise signals, a clustering algorithm is used to classify the harmonic noise signals of all renewable energy power generation equipment to obtain a typical noise frequency pattern.

[0082] In this embodiment, the spatial correlation acquisition process of the harmonic noise signal includes:

[0083] Establishing a three-dimensional spatial coordinate system according to the geographic coordinates of the renewable energy power generation equipment, and mapping the spatial location information of the wind farm collection point and the photovoltaic power station grid connection point to the three-dimensional spatial coordinate system;

[0084] Calculating the Euclidean distances between observation points inside the wind farm, between observation points inside the photovoltaic power station, and between the wind farm and the photovoltaic power station in the three-dimensional spatial coordinate system to obtain a distance matrix, and constructing a spatial weight matrix based on the distance matrix;

[0085] The Pearson correlation coefficient is used to calculate the correlation degree of harmonic noise signals between wind farm collection points, and the wind farm correlation coefficient matrix within and between wind farms is obtained;

[0086] The Pearson correlation coefficient is used to calculate the correlation degree of harmonic noise signals between the photovoltaic power station grid connection points, and the photovoltaic power station correlation coefficient matrix within the photovoltaic power station and between photovoltaic power stations is obtained;

[0087] According to the wind farm correlation coefficient matrix, photovoltaic power station correlation coefficient matrix and spatial weight matrix, the global spatial autocorrelation of the harmonic noise signals of wind farms and photovoltaic power stations is calculated using the global Moran index to obtain the spatial correlation of the harmonic noise signals.

[0088] Specifically, for ease of understanding, this embodiment is described in detail by taking the harmonic noise signals of different renewable energy power generation equipment, which mainly include wind farm harmonic noise signals and photovoltaic power station harmonic noise signals, as an example. This embodiment establishes a three-dimensional space coordinate system based on the GPS coordinate system data of the wind farm and the photovoltaic power station, and maps the spatial position information of each collection point of the wind farm and each grid-connected point of the photovoltaic power station to the three-dimensional coordinate system to obtain the spatial coordinates of each point. The Euclidean distance formula in the three-dimensional space is used to calculate the distance between each point (including the observation points inside the wind farm, the observation points inside the photovoltaic power station, and the wind farm and the photovoltaic power station) to form a distance matrix.

[0089] Then, for each confluence point of the wind farm, the Pearson correlation coefficient is used to calculate the correlation of the harmonic noise signals between them, and the fundamental wave, the third harmonic and the fifth harmonic are analyzed to obtain the correlation coefficient matrix within the wind farm and between the wind farms respectively; the Pearson correlation coefficient is used to calculate the correlation degree of the harmonic noise signals between the grid-connected points of the photovoltaic power station, and the photovoltaic power station correlation coefficient matrix within the photovoltaic power station and between the photovoltaic power stations is obtained. Then, the obtained correlation coefficient matrix of the wind farm and the photovoltaic power station and the distance matrix are used to calculate the global Moran's index. Specifically, this embodiment constructs an initial spatial weight matrix based on the distance matrix. The initial spatial weight matrix describes the proximity relationship between each point. The embodiment uses the values ​​in the correlation coefficient matrix as adjustment factors of the weights in the initial spatial weight matrix to adjust the initial spatial weight matrix. This embodiment can multiply the values ​​in the correlation coefficient matrix by the corresponding values ​​in the preliminary spatial weight matrix to obtain a spatial weight matrix that finally incorporates the linear correlation information between the harmonic noise signals. The final spatial weight matrix is ​​used to calculate the global spatial autocorrelation of the harmonic noise signals of the wind farm and the photovoltaic power station to obtain the spatial correlation of the harmonic noise signals. The global spatial autocorrelation simultaneously reflects the spatial distribution pattern, spatial autocorrelation and linear correlation between the harmonic noise signals of the wind farm and the photovoltaic power station in the entire region.

[0090] In this embodiment, the step of classifying the harmonic noise signals of all renewable energy power generation equipment using a clustering algorithm according to the spatial correlation of the harmonic noise signals to obtain a typical noise frequency pattern includes:

[0091] Combining the harmonic noise signal and its spatial correlation and duration into a multi-dimensional feature vector;

[0092] The elbow rule is used to determine the optimal clustering number of harmonic noise signals of different renewable energy generation equipment by calculating the intra-cluster sum of squares under different clustering numbers.

[0093] Based on the optimal cluster number, a K-means clustering algorithm is used to perform clustering analysis on the multidimensional feature vector to obtain a clustering result;

[0094] The cluster center features are extracted according to the clustering results to form a typical noise frequency pattern.

[0095] Specifically, for the feature vector, this embodiment adopts the elbow rule, by calculating the WCSS (Within-Cluster Sum of Squares) under different cluster numbers (k values), drawing a curve graph of WCSS changing with k value, identifying and selecting the cluster number at the inflection point of the curve from the curve graph as the optimal cluster number for wind power noise and photovoltaic harmonic noise signals, and the inflection point indicates that as the number of clusters increases, the reduction rate of WCSS begins to slow down; this embodiment adopts the K-means clustering algorithm for the multidimensional feature vector, sets the random seed to 42, uses the Euclidean distance as the distance metric, sets the maximum number of iterations to 100, and when calculating the distance, not only the differences in amplitude, frequency and duration are considered, but also the spatial location information is introduced to make full use of the spatial correlation of the harmonic noise signal, and iterates the optimization until When the convergence conditions are met (such as the cluster center no longer changes significantly or the maximum number of iterations is reached), the clustering results are obtained. For the obtained clustering results, the coordinates of the central feature points of each cluster are calculated. These central points represent the typical characteristics of each category. The cluster center features corresponding to the central points are extracted to form typical frequency patterns of wind power noise and photovoltaic noise. The cluster center features include amplitude, frequency, duration features and spatial distribution features. Through the above steps, this embodiment makes full use of the spatial correlation of harmonic noise signals, realizes the effective classification of harmonic noise signals of renewable energy power generation equipment through clustering algorithms, and extracts different types of noise frequency patterns.

[0096] S3. Quantitatively evaluate the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern to obtain a noise influence evaluation result.

[0097] In this embodiment, the step of quantitatively evaluating the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern to obtain the noise influence evaluation result includes:

[0098] Using a similarity measurement algorithm to match the typical noise frequency pattern with a pre-built renewable energy noise frequency feature library to obtain a matching similarity;

[0099] If the matching similarity exceeds a preset similarity threshold, selecting the typical noise frequency pattern with the highest matching similarity as the target noise pattern;

[0100] Calculating the frequency overlap between the target noise pattern and the power line carrier communication signal, and calculating the signal-to-noise ratio using the ratio of the harmonic noise amplitude corresponding to the target noise pattern to the power line carrier communication signal amplitude;

[0101] The influence of harmonic noise on the power line carrier communication signal is quantitatively evaluated according to the frequency overlap and the signal-to-noise ratio to obtain a noise influence evaluation result.

[0102] Specifically, in the power line carrier communication system of the wind farm and the photovoltaic power station, harmonic noise is one of the key factors affecting the communication quality. Traditional evaluation methods often find it difficult to accurately quantify the specific impact of harmonic noise on the communication signal. Therefore, this embodiment calculates the similarity between the typical noise frequency pattern and the feature vector in the pre-constructed renewable energy noise frequency feature library through a similarity measurement algorithm (such as Euclidean distance, cosine similarity, etc.), obtains the matching similarity, and screens out the typical noise frequency pattern with the highest matching similarity as the target noise pattern. For the identified target noise pattern, the frequency overlap between the target noise pattern and the power line carrier communication signal and the ratio of the corresponding harmonic noise amplitude to the power line carrier communication signal amplitude (i.e., the signal-to-noise ratio) are calculated, and these two parameters are used to evaluate the degree of influence of harmonic noise on the communication signal. This embodiment calculates the weighted average of the frequency overlap and the signal-to-noise ratio according to the preset first weight and second weight, calculates the degree of influence of harmonic noise on the power line carrier communication signal, and obtains the noise impact evaluation result, thereby realizing the accurate quantification of the harmonic noise impact on the power line carrier communication signal and obtaining the noise impact evaluation result.

[0103] In this embodiment, the renewable energy noise frequency feature library includes a wind power noise frequency feature library and a photovoltaic noise frequency feature library. The construction process of the renewable energy noise frequency feature library is specifically as follows:

[0104] Acquire historical harmonic noise signals at wind farm collection points and photovoltaic power station grid connection points, and preprocess the historical harmonic noise signals to obtain preprocessed harmonic noise signals;

[0105] Performing frequency domain analysis on the preprocessed harmonic noise signal using a fast Fourier transform algorithm to extract the amplitude and frequency characteristics of each harmonic component;

[0106] Calculating the amplitude and frequency characteristics of each harmonic component using a statistical analysis method to obtain a renewable energy noise characteristic statistic;

[0107] Using a kernel density estimation algorithm to perform non-parametric estimation on the amplitude and frequency characteristics of each harmonic component, and construct a probability density function;

[0108] A cumulative distribution function is obtained by integrating the probability density function, and the occurrence probability of historical harmonic noise signals of different amplitudes and frequency intervals is quantified according to the cumulative distribution function;

[0109] A renewable energy noise frequency feature library is constructed according to the renewable energy noise feature statistics, harmonic order and the occurrence probability.

[0110] Specifically, this embodiment captures historical harmonic noise signals of different renewable energy sources from wind farm collection points and photovoltaic power station grid connection points. The historical harmonic noise signals include amplitude, frequency and timestamp information. The collected historical harmonic noise signals are preprocessed, and outliers in the historical harmonic noise signal data set are removed to reduce the impact of errors. At the same time, linear interpolation is used to effectively fill in missing values ​​in the data to ensure the continuity and integrity of the data, thereby obtaining high-quality preprocessed harmonic noise signals. This embodiment uses a fast Fourier transform algorithm to perform frequency domain conversion on the preprocessed harmonic noise signal and analyzes its characteristics at different frequencies. This embodiment sets the sampling rate to 1kHz to fully capture signal details, and uses a Hanning window function to reduce spectrum leakage and improve the accuracy of spectrum analysis. The amplitude and frequency characteristics of each harmonic are extracted through fast Fourier transform algorithm analysis.

[0111] This embodiment calculates the statistics of each feature based on the extracted harmonic amplitude and frequency features, for example, the mean, standard deviation, maximum value and minimum value, and quantifies the overall distribution characteristics and mutation of the harmonic noise signal through the statistics. This embodiment adopts a kernel density estimation algorithm (such as a Gaussian kernel function) to perform non-parametric estimation on the amplitude and frequency of the harmonic noise signal to construct its probability density function (PDF), and determines the bandwidth of the kernel function through the Silverman's rule of thumb method to ensure the accuracy and smoothness of the estimation. Then, by integrating the probability density function, the cumulative distribution function (CDF) is calculated to quantify the occurrence probability of the noise signal in different amplitude and frequency ranges, and a relational database for storing the frequency characteristic data of wind power noise and photovoltaic noise is constructed to obtain a renewable energy noise frequency characteristic library. The renewable energy noise frequency characteristic library may include a primary key ID to ensure data uniqueness, as well as harmonic order, amplitude mean, amplitude standard deviation, amplitude maximum value, amplitude minimum value, frequency mean, frequency standard deviation, frequency maximum value, frequency minimum value and occurrence probability, so as to comprehensively record and manage noise characteristic information.

[0112] S4. Based on the noise impact assessment result, perform physical layer adaptive multi-carrier modulation on the power line carrier communication signal to obtain a power line carrier modulation signal.

[0113] In this embodiment, the step of performing physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulation signal includes:

[0114] Filtering the power line carrier communication signal according to the noise impact assessment result to obtain a power line carrier communication filtered signal;

[0115] Using a frequency scanning method to measure the resonant frequency of the power line carrier communication filter signal within a preset frequency range, and determine the potential crosstalk frequency of the power line carrier communication filter signal;

[0116] Selecting an orthogonal frequency division multiplexing modulation mode according to the resonant frequency and the potential crosstalk frequency;

[0117] Acquire a real-time channel state, and calculate a signal-to-noise ratio of each subcarrier based on the real-time channel state to obtain a channel quality indicator;

[0118] Determine a coding rate according to the channel quality indicator at the link layer, and adjust a subcarrier modulation mode in the orthogonal frequency division multiplexing modulation mode according to the coding rate to obtain a target modulation mode;

[0119] The power line carrier communication filtered signal is modulated according to the target modulation mode to generate a power line carrier modulation signal.

[0120] Specifically, this embodiment filters the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier communication filtered signal. The filtered signal can improve the effect of subsequent processing after noise suppression. Then, a frequency scanning method is used to scan the power line carrier communication filtered signal within a preset frequency range covering the power line communication frequency band to obtain a frequency scanning result. The resonant frequency of the power line carrier communication filtered signal is identified according to the frequency scanning result. These resonant frequencies are the optimal transmission frequency bands in the channel. At the same time, according to the frequency scanning result, combined with the power line network characteristics and known crosstalk patterns, the potential crosstalk frequency is predicted and determined, so as to provide frequency characteristic data for the subsequent determination of the modulation scheme.

[0121] Next, this embodiment selects a suitable orthogonal frequency division multiplexing (OFDM) modulation method according to the measured resonant frequency and the distribution of the potential crosstalk frequency, and uses the selected orthogonal frequency division multiplexing modulation method to independently modulate each subcarrier. Initially, a unified orthogonal frequency division multiplexing modulation method can be used. At the same time, this embodiment uses a pilot signal or a channel detection mechanism to monitor and feedback the channel status of each subcarrier in real time, obtain the real-time channel status, and calculate the signal-to-noise ratio (SNR) of each subcarrier based on the real-time channel status to form a channel quality indicator. At the link layer, this embodiment dynamically adjusts the coding rate and modulation method (such as modulation order, bit loading strategy) of each subcarrier according to the channel quality indicator. For subcarriers with a signal-to-noise ratio lower than a first preset value, the coding rate and modulation method (such as modulation order, bit loading strategy) of each subcarrier are adjusted. For subcarriers with a signal-to-noise ratio threshold, a lower coding rate is set to enhance anti-interference capability; for subcarriers with a signal-to-noise ratio higher than a second preset threshold, a higher coding rate is used to improve the data transmission rate, and the modulation mode of each subcarrier in the orthogonal frequency division multiplexing modulation mode is adjusted according to the adjusted coding rate to obtain the target modulation mode, and the power line carrier communication filter signal is modulated by the target modulation mode to finally generate a power line carrier modulated signal, which can adapt to real-time changing channel conditions after dynamic adjustment, and optimizes communication performance by adaptively adjusting the coding rate and subcarrier modulation mode, thereby effectively realizing the physical layer adaptive multi-carrier modulation of the power line carrier communication signal and improving transmission efficiency and stability.

[0122] In this embodiment, the step of filtering the power line carrier communication signal according to the noise impact assessment result to obtain a power line carrier communication filtered signal includes:

[0123] If the noise impact assessment result is within a preset interference threshold range, the noise characteristics of the harmonic noise signal are identified, and the corresponding interference source is determined according to the noise characteristics of the harmonic noise signal;

[0124] A filtering strategy is determined according to the interference source, and noise is filtered on the power line carrier communication signal according to the filtering strategy to obtain a power line carrier communication filtered signal; the filtering strategy includes a wavelet analysis method and an empirical mode decomposition method.

[0125] Specifically, this embodiment determines whether it is necessary to filter the power line carrier communication signal according to the noise impact assessment result and the preset interference threshold range. If the assessment result shows that the noise impact assessment result is within the preset interference threshold range, it is determined that the power line carrier communication signal is seriously interfered with, and the filtering processing mechanism is triggered. For the filtering processing mechanism, this embodiment uses a spectrum feature recognition method to analyze the received power line carrier communication signal, identify the noise characteristics of the harmonic noise signal, distinguish the interference source according to the noise characteristics of the harmonic noise signal, and determine the filtering strategy according to the interference source. For example: this embodiment determines that the noise comes from wind power equipment or photovoltaic equipment according to the interference source. If the interference source is determined to be from wind power equipment, Daubechies wavelet is used. Wavelet) is used as the basis function to perform multi-scale decomposition on the power line carrier communication signal to obtain wavelet coefficients in different frequency bands. The threshold is set by the soft threshold method (the threshold is set to a preset multiple of the noise standard deviation), the wavelet coefficients are screened, and the noise component is removed to obtain the power line carrier communication filtered signal; if the interference source is determined to be from the photovoltaic equipment, the empirical mode decomposition method (EMD) is used to decompose the power line carrier communication signal into several intrinsic mode functions (IMFs). By analyzing the frequency characteristics of each IMF, a preset number of IMF components are selectively retained for signal reconstruction, thereby suppressing the noise component of a specific frequency.

[0126] The present invention achieves accurate identification and suppression of wind power and photovoltaic noise by accurately identifying noise sources and adopting targeted filtering strategies, combining a hybrid signal processing method of wavelet analysis and empirical mode decomposition (EMD), effectively solving the problem of interference of harmonic noise of wind power and photovoltaic equipment on communication signals, and improving the stability and reliability of the power line carrier communication system. Those skilled in the art can set filtering strategies according to specific implementation circumstances, which is not limited to the embodiments of the present invention.

[0127] S5. A reward and punishment mechanism is set based on the power line carrier modulation signal, and a deep Q learning algorithm is used to dynamically optimize the power line carrier communication route to select the optimal communication path for data transmission.

[0128] In this embodiment, the steps of setting a reward and punishment mechanism based on the power line carrier modulation signal, dynamically optimizing the power line carrier communication route using a deep Q learning algorithm, and selecting the optimal communication path for data transmission include:

[0129] Different PLC modulation signals are considered as elements of state space, and different communication paths are defined as action space;

[0130] According to the current network status, identify the spatial distribution pattern of the power line carrier modulation signal on different communication paths;

[0131] A reward and punishment mechanism is set up according to the spatial distribution law and key performance indicators of the power line carrier modulation signal; the key performance indicators include the transmission success rate, rate and transmission delay index of the power line carrier modulation signal;

[0132] Randomly initialize the Q value and select the communication path action using the ε-greedy strategy according to the current network state at each time step;

[0133] Each communication path action is evaluated according to the reward and punishment mechanism, and the Q-Learning algorithm is used to learn the spatial distribution law of the power line carrier modulation signal to dynamically adjust the selection of the communication path, update the Q value of each communication path, until the convergence condition is reached, and find the communication path with the highest Q value in each state;

[0134] The communication path with the highest Q value is selected as the optimal communication path for power line carrier modulation signal transmission.

[0135] Specifically, this embodiment regards different power line carrier modulation signals as elements of a state space, where each state represents signal characteristics such as signal strength and noise level; defines different communication paths as an action space, which contains all available transmission paths. Then, according to the current network state, a signal analysis method is used to identify the spatial distribution laws of harmonic noise, attenuation, etc. of harmonic noise signals on different communication paths. A reward and punishment mechanism is set based on the spatial distribution law of power line carrier modulation signals and key performance indicators (such as transmission success rate, rate, and delay index) to quantify the pros and cons of each communication path, where the transmission success rate refers to the frequency of successful data transmission on a given communication path, the transmission rate refers to the rate at which data is transmitted on a given path, and the transmission delay index refers to the transmission delay time of data from a source node to a destination node. Then, this embodiment randomly initializes Q The Q-value table is used to store the expected cumulative rewards of each action (communication path) in each state. At each time step, an action (i.e., a communication path) is selected according to the current state. The selection strategy can adopt the ε-greedy strategy or other strategies. The selected communication path action is evaluated according to the set reward and punishment mechanism, and the selection of the communication path is dynamically adjusted by comparing the Q values ​​of each path, and the Q value of each communication path is updated. With the continuous updating of the Q-value table, the Q-Learning algorithm can learn the spatial distribution law of the power line carrier modulation signal and the performance differences of different communication paths, thereby dynamically adjusting the selection strategy of the communication path, repeating the above process until the preset convergence condition is reached, finding the communication path with the highest Q value in each state, and taking the communication path with the highest Q value as the optimal communication path for power line carrier modulation signal transmission.

[0136] An embodiment of the present invention provides a power line communication optimization method for a new power system. The method uses a short-time Fourier transform spectrum analysis method to extract harmonic noise signals of different renewable energy power generation equipment according to operating parameters of renewable energy power generation equipment; uses a clustering algorithm to classify the harmonic noise signals of all renewable energy power generation equipment to obtain a typical noise frequency pattern; quantitatively evaluates the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern to obtain a noise impact assessment result; performs physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulation signal; sets a reward and punishment mechanism based on the power line carrier modulation signal, and uses a deep Q learning algorithm to dynamically optimize the power line carrier communication route to screen out the optimal communication path for data transmission. Compared with the prior art, this embodiment can accurately evaluate the impact of harmonic noise generated by wind power and photovoltaic equipment on communication signals, and optimize communication efficiency and noise interference resistance by adaptively selecting communication modulation methods and deep Q learning algorithms and other technologies, so that the communication system can effectively cope with the complex noise environment of different wind farms and photovoltaic power stations, improve the noise interference resistance and communication quality of power line carrier communication in wind farm and photovoltaic power station environments, and provide reliable technical support for power communication under renewable energy grid connection.

[0137] It should be noted that the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0138] In one embodiment, Figure 2 As shown, an embodiment of the present invention provides a power line communication optimization system for a new power system, the system comprising:

[0139] The spectrum analysis module 101 is used to extract the harmonic noise signals of different renewable energy power generation equipment using the short-time Fourier transform spectrum analysis method according to the collected operating parameters of the renewable energy power generation equipment; wherein the operating parameters of the renewable energy power generation equipment include the operating parameters of the wind power equipment at the collection point of each wind farm and the operating parameters of the photovoltaic equipment at the grid connection point of the photovoltaic power station;

[0140] A signal classification module 102, configured to classify the harmonic noise signals of all renewable energy power generation equipment using a clustering algorithm according to the spatial correlation of the harmonic noise signals to obtain a typical noise frequency pattern;

[0141] A noise assessment module 103, configured to quantitatively assess the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern, and obtain a noise impact assessment result;

[0142] A signal modulation module 104, configured to perform physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulation signal;

[0143] The communication optimization module 105 is used to set a reward and punishment mechanism based on the power line carrier modulation signal, and dynamically optimize the power line carrier communication route using a deep Q learning algorithm to select the optimal communication path for data transmission.

[0144] For the specific definition of a power line communication optimization system for a new power system, please refer to the above-mentioned definition of a power line communication optimization method for a new power system, which will not be repeated here. A person of ordinary skill in the art can appreciate that the various modules and steps described in conjunction with the embodiments disclosed in this application can be implemented in hardware, software, or a combination of both. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0145] The embodiment of the present invention provides a power line communication optimization system for a new power system. The system uses a spectrum analysis module to analyze the operating parameters of renewable energy power generation equipment using a short-time Fourier transform spectrum analysis method to extract harmonic noise signals of different renewable energy power generation equipment; the signal classification module uses a clustering algorithm to classify the harmonic noise signals of all renewable energy power generation equipment to obtain a typical noise frequency pattern; the noise evaluation module quantitatively evaluates the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern to obtain a noise impact evaluation result; the signal modulation module performs physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact evaluation result to obtain a power line carrier modulation signal; the communication optimization module sets a reward and punishment mechanism based on the power line carrier modulation signal, and uses a deep Q learning algorithm to dynamically optimize the power line carrier communication route to screen out the optimal communication path for data transmission. Compared with the prior art, this embodiment can accurately evaluate the impact of harmonic noise generated by wind power and photovoltaic equipment on communication signals, and optimize communication efficiency and noise interference resistance by adaptively selecting communication modulation methods and deep Q learning algorithms and other technologies, so that the communication system can effectively cope with the complex noise environment of different wind farms and photovoltaic power stations, improve the noise interference resistance and communication quality of power line carrier communication in wind farm and photovoltaic power station environments, and provide reliable technical support for power communication under renewable energy grid connection.

[0146] The above-mentioned embodiments only express several preferred implementation modes of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in the technical field, several improvements and substitutions can be made without departing from the technical principles of the present invention, and these improvements and substitutions should also be regarded as the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be based on the protection scope of the claims.

Claims

1. A power line communication optimization method for a new power system, characterized in that: The following steps are involved: According to the collected operating parameters of renewable energy power generation equipment, the harmonic noise signals of different renewable energy power generation equipment are extracted by using the short-time Fourier transform spectrum analysis method; wherein the operating parameters of renewable energy power generation equipment include the operating parameters of wind power equipment at the collection point of each wind farm and the operating parameters of photovoltaic equipment at the grid connection point of the photovoltaic power station; According to the spatial correlation of the harmonic noise signals, a clustering algorithm is used to classify the harmonic noise signals of all renewable energy power generation equipment to obtain a typical noise frequency pattern; Quantitatively evaluate the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern to obtain a noise influence evaluation result; Performing physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulation signal; A reward and punishment mechanism is set based on the power line carrier modulation signal, and a deep Q learning algorithm is used to dynamically optimize the power line carrier communication route to select the optimal communication path for data transmission; The step of quantitatively evaluating the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern to obtain the noise influence evaluation result comprises: Using a similarity measurement algorithm to match the typical noise frequency pattern with a pre-built renewable energy noise frequency feature library to obtain a matching similarity; If the matching similarity exceeds a preset similarity threshold, selecting the typical noise frequency pattern with the highest matching similarity as the target noise pattern; Calculating the frequency overlap between the target noise pattern and the power line carrier communication signal, and calculating the signal-to-noise ratio using the ratio of the harmonic noise amplitude corresponding to the target noise pattern to the power line carrier communication signal amplitude; The influence of harmonic noise on the power line carrier communication signal is quantitatively evaluated according to the frequency overlap and the signal-to-noise ratio to obtain a noise influence evaluation result; The step of performing physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulation signal comprises: Filtering the power line carrier communication signal according to the noise impact assessment result to obtain a power line carrier communication filtered signal; Using a frequency scanning method to measure the resonant frequency of the power line carrier communication filter signal within a preset frequency range, and determine the potential crosstalk frequency of the power line carrier communication filter signal; Selecting an orthogonal frequency division multiplexing modulation mode according to the resonant frequency and the potential crosstalk frequency; Acquire a real-time channel state, and calculate a signal-to-noise ratio of each subcarrier based on the real-time channel state to obtain a channel quality indicator; Determine a coding rate according to the channel quality indicator at the link layer, and adjust a subcarrier modulation mode in the orthogonal frequency division multiplexing modulation mode according to the coding rate to obtain a target modulation mode; The power line carrier communication filtered signal is modulated according to the target modulation mode to generate a power line carrier modulation signal.

2. A power line communication optimization method for a new power system according to claim 1, characterized in that: The wind power equipment operating parameters include the wind turbine blade rotation frequency, the gearbox meshing frequency and the generator rotor rotation frequency, and the photovoltaic equipment operating parameters include the photovoltaic module switching frequency, the inverter switch tube switching frequency and the inverter output current harmonic frequency; the step of extracting the harmonic noise signals of different renewable energy power generation equipment using the short-time Fourier transform spectrum analysis method based on the collected renewable energy power generation equipment operating parameters includes: The principal component analysis method is used to reduce the dimension of the collected operating parameters of renewable energy power generation equipment and extract the key operating characteristics of renewable energy power generation equipment; The K-means clustering algorithm is used to group the key operating characteristics of renewable energy power generation equipment and identify the equipment parameter patterns under different operating conditions; According to the center point of the equipment parameter pattern, the Mamdani fuzzy inference system is used to calculate the Euclidean distance between the currently collected operating parameters of the renewable energy power generation equipment and the center point of each equipment parameter pattern to obtain the pattern matching distance; dynamically determining a time window width and an overlap rate of a short-time Fourier transform according to the pattern matching distance; According to the time window width and overlap rate of short-time Fourier transform, short-time Fourier transform spectrum analysis is performed on the operating parameters of renewable energy power generation equipment to obtain the time-frequency spectrum diagram of renewable energy power generation equipment; According to the time-frequency spectrum of the renewable energy power generation equipment, the harmonic noise signal is extracted; the harmonic noise signal includes timestamp, type of renewable energy power generation equipment, frequency, amplitude and phase information.

3. A power line communication optimization method for a new power system according to claim 1, characterized in that: The spatial correlation acquisition process of the harmonic noise signal includes: Establishing a three-dimensional spatial coordinate system according to the geographic coordinates of the renewable energy power generation equipment, and mapping the spatial location information of the wind farm collection point and the photovoltaic power station grid connection point to the three-dimensional spatial coordinate system; Calculating the Euclidean distances between observation points inside the wind farm, between observation points inside the photovoltaic power station, and between the wind farm and the photovoltaic power station in the three-dimensional spatial coordinate system to obtain a distance matrix, and constructing a spatial weight matrix based on the distance matrix; The Pearson correlation coefficient is used to calculate the correlation degree of harmonic noise signals between wind farm collection points, and the wind farm correlation coefficient matrix within and between wind farms is obtained; The Pearson correlation coefficient is used to calculate the correlation degree of harmonic noise signals between the photovoltaic power station grid connection points, and the photovoltaic power station correlation coefficient matrix within the photovoltaic power station and between photovoltaic power stations is obtained; According to the wind farm correlation coefficient matrix, photovoltaic power station correlation coefficient matrix and spatial weight matrix, the global spatial autocorrelation of the harmonic noise signals of wind farms and photovoltaic power stations is calculated using the global Moran index to obtain the spatial correlation of the harmonic noise signals.

4. A power line communication optimization method for a new power system according to claim 1, characterized in that: The step of classifying the harmonic noise signals of all renewable energy power generation equipment using a clustering algorithm according to the spatial correlation of the harmonic noise signals to obtain a typical noise frequency pattern comprises: Combining the harmonic noise signal and its spatial correlation and duration into a multi-dimensional feature vector; The elbow rule is used to determine the optimal clustering number of harmonic noise signals of different renewable energy generation equipment by calculating the intra-cluster sum of squares under different clustering numbers. Based on the optimal cluster number, a K-means clustering algorithm is used to perform clustering analysis on the multidimensional feature vector to obtain a clustering result; The cluster center features are extracted according to the clustering results to form a typical noise frequency pattern.

5. A power line communication optimization method for a new power system according to claim 1, characterized in that: The renewable energy noise frequency feature library includes a wind power noise frequency feature library and a photovoltaic noise frequency feature library. The construction process of the renewable energy noise frequency feature library is specifically as follows: Acquire historical harmonic noise signals at wind farm collection points and photovoltaic power station grid connection points, and preprocess the historical harmonic noise signals to obtain preprocessed harmonic noise signals; Performing frequency domain analysis on the preprocessed harmonic noise signal using a fast Fourier transform algorithm to extract the amplitude and frequency characteristics of each harmonic component; Calculating the amplitude and frequency characteristics of each harmonic component using a statistical analysis method to obtain a renewable energy noise characteristic statistic; Using a kernel density estimation algorithm to perform non-parametric estimation on the amplitude and frequency characteristics of each harmonic component, and construct a probability density function; A cumulative distribution function is obtained by integrating the probability density function, and the occurrence probability of historical harmonic noise signals of different amplitudes and frequency intervals is quantified according to the cumulative distribution function; A renewable energy noise frequency feature library is constructed according to the renewable energy noise feature statistics, harmonic order and the occurrence probability.

6. A power line communication optimization method for a new power system according to claim 1, characterized in that: The step of filtering the power line carrier communication signal according to the noise impact assessment result to obtain a power line carrier communication filtered signal comprises: If the noise impact assessment result is within a preset interference threshold range, the noise characteristics of the harmonic noise signal are identified, and the corresponding interference source is determined according to the noise characteristics of the harmonic noise signal; A filtering strategy is determined according to the interference source, and noise is filtered on the power line carrier communication signal according to the filtering strategy to obtain a power line carrier communication filtered signal; the filtering strategy includes a wavelet analysis method and an empirical mode decomposition method.

7. A power line communication optimization method for a new power system according to claim 1, characterized in that: The steps of setting a reward and punishment mechanism based on the power line carrier modulation signal, dynamically optimizing the power line carrier communication route using a deep Q learning algorithm, and selecting the optimal communication path for data transmission include: Different PLC modulation signals are considered as elements of state space, and different communication paths are defined as action space; According to the current network status, identify the spatial distribution pattern of the power line carrier modulation signal on different communication paths; A reward and punishment mechanism is set up according to the spatial distribution law and key performance indicators of the power line carrier modulation signal; the key performance indicators include the transmission success rate, rate and transmission delay index of the power line carrier modulation signal; Randomly initialize the Q value and select the communication path action using the ε-greedy strategy according to the current network state at each time step; Each communication path action is evaluated according to the reward and punishment mechanism, and the Q-Learning algorithm is used to learn the spatial distribution law of the power line carrier modulation signal to dynamically adjust the selection of the communication path, update the Q value of each communication path, until the convergence condition is reached, and find the communication path with the highest Q value in each state; The communication path with the highest Q value is selected as the optimal communication path for power line carrier modulation signal transmission.

8. A power line communication optimization system for a new power system, characterized in that: The system comprises: The spectrum analysis module is used to extract the harmonic noise signals of different renewable energy power generation equipment using the short-time Fourier transform spectrum analysis method according to the collected operating parameters of the renewable energy power generation equipment; wherein the operating parameters of the renewable energy power generation equipment include the operating parameters of the wind power equipment at the collection point of each wind farm and the operating parameters of the photovoltaic equipment at the grid connection point of the photovoltaic power station; A signal classification module, used to classify the harmonic noise signals of all renewable energy power generation equipment according to the spatial correlation of the harmonic noise signals by using a clustering algorithm to obtain a typical noise frequency pattern; A noise assessment module, used for quantitatively assessing the influence of the harmonic noise signal on the power line carrier communication signal according to the typical noise frequency pattern, and obtaining a noise impact assessment result; A signal modulation module, used for performing physical layer adaptive multi-carrier modulation on the power line carrier communication signal based on the noise impact assessment result to obtain a power line carrier modulation signal; A communication optimization module, used to set a reward and punishment mechanism based on the power line carrier modulation signal, and dynamically optimize the power line carrier communication route using a deep Q learning algorithm to select the optimal communication path for data transmission; The noise assessment module is specifically used for: Using a similarity measurement algorithm to match the typical noise frequency pattern with a pre-built renewable energy noise frequency feature library to obtain a matching similarity; If the matching similarity exceeds a preset similarity threshold, selecting the typical noise frequency pattern with the highest matching similarity as the target noise pattern; Calculating the frequency overlap between the target noise pattern and the power line carrier communication signal, and calculating the signal-to-noise ratio using the ratio of the harmonic noise amplitude corresponding to the target noise pattern to the power line carrier communication signal amplitude; The influence of harmonic noise on the power line carrier communication signal is quantitatively evaluated according to the frequency overlap and the signal-to-noise ratio to obtain a noise influence evaluation result; The signal modulation module is specifically used for: Filtering the power line carrier communication signal according to the noise impact assessment result to obtain a power line carrier communication filtered signal; Using a frequency scanning method to measure the resonant frequency of the power line carrier communication filter signal within a preset frequency range, and determine the potential crosstalk frequency of the power line carrier communication filter signal; Selecting an orthogonal frequency division multiplexing modulation mode according to the resonant frequency and the potential crosstalk frequency; Acquire a real-time channel state, and calculate a signal-to-noise ratio of each subcarrier based on the real-time channel state to obtain a channel quality indicator; Determine a coding rate according to the channel quality indicator at the link layer, and adjust a subcarrier modulation mode in the orthogonal frequency division multiplexing modulation mode according to the coding rate to obtain a target modulation mode; The power line carrier communication filtered signal is modulated according to the target modulation mode to generate a power line carrier modulation signal.

Images