A method and system for tracing the source of offshore wind power harmonics based on Bayesian regression

Abstract

The present invention discloses an offshore wind power harmonic source tracing method and system based on Bayesian regression. The method calculates the voltage fluctuation value and the current fluctuation value of each time node by acquiring the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated; constructs a harmonic impedance estimation mathematical model according to the voltage fluctuation value and the current fluctuation value; solves the harmonic impedance estimation mathematical model based on the Bayesian theorem to obtain the harmonic impedance parameters on the system side, accurately tracks the fluctuation of the harmonic impedance on the system side, and quantifies the harmonic responsibility on the user side and the harmonic responsibility on the system side according to the harmonic impedance parameters on the system side, the harmonic voltage time series data and the harmonic current time series data; finally, determines the position of the harmonic source according to the harmonic responsibility on the user side and the harmonic responsibility on the system side, and can accurately realize harmonic source tracing when the impedance changes.

Description

Technical Field

[0001] The present invention relates to the technical field of harmonic source tracing, and in particular to a method and system for offshore wind power harmonic source tracing based on Bayesian regression. Background Art

[0002] With the rapid development of modern power systems, the proportion of renewable energy has increased dramatically. The randomness of wind power generation and the use of a large number of converters have made the harmonic problem of offshore wind power grid connection points more and more serious. The source of harmonics at offshore wind power grid connection points is often difficult to directly determine. It is necessary to conduct harmonic source tracing analysis in combination with actual harmonic monitoring data to locate the source of harmonics and quantify harmonic responsibility. Existing harmonic source tracing methods can be divided into two categories, namely invasive methods and non-invasive methods. The former requires the use of a disturbance generator to inject harmonic disturbances into the power grid, which will affect the normal operation of the power grid; the latter directly uses the harmonic disturbances existing in the power grid to conduct source tracing analysis.

[0003] The existing non-intrusive methods mainly include regression method, correlation analysis method, independent component analysis method, etc. These methods are difficult to obtain good estimation results when the background harmonics and system-side harmonic impedance fluctuations are large. In fact, the system-side harmonic impedance of the offshore wind power grid connection point is variable in the actual system. The existing harmonic methods will produce large errors and cannot accurately distinguish the location of the harmonic source. Summary of the invention

[0004] Based on this, the present invention provides an offshore wind power harmonic source tracing method and system based on Bayesian regression, which can construct a harmonic impedance estimation mathematical model according to the fluctuations of harmonic voltage and harmonic current, and use Bayesian theorem to solve the system-side harmonic impedance parameters for determining the location of the harmonic source, thereby realizing accurate tracing of the harmonic source.

[0005] To achieve the above object, an embodiment of the present invention provides an offshore wind power harmonic source tracing method based on Bayesian regression, comprising:

[0006] Obtain the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated;

[0007] Calculating the voltage fluctuation value at each time node according to the harmonic voltage time series data, and calculating the current fluctuation value at each time node according to the harmonic current time series data;

[0008] Constructing a harmonic impedance estimation mathematical model according to the voltage fluctuation value and the current fluctuation value;

[0009] Solving the harmonic impedance estimation mathematical model based on Bayesian theorem to obtain system-side harmonic impedance parameters;

[0010] quantifying user-side harmonic responsibility and system-side harmonic responsibility according to the system-side harmonic impedance parameter, the harmonic voltage time series data, and the harmonic current time series data;

[0011] The location of the harmonic source is determined according to the user-side harmonic responsibility and the system-side harmonic responsibility.

[0012] To achieve the above purpose, an embodiment of the present invention further provides an offshore wind power harmonic source tracing system based on Bayesian regression, comprising:

[0013] A data acquisition module, used to acquire the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated;

[0014] A fluctuation calculation module, used to calculate the voltage fluctuation value of each time node according to the harmonic voltage time series data, and calculate the current fluctuation value of each time node according to the harmonic current time series data;

[0015] A model building module, used for building a harmonic impedance estimation mathematical model according to the voltage fluctuation value and the current fluctuation value;

[0016] An impedance calculation module, used for solving the harmonic impedance estimation mathematical model based on the Bayesian theorem to obtain system-side harmonic impedance parameters;

[0017] A responsibility determination module, used for quantifying the user-side harmonic responsibility and the system-side harmonic responsibility according to the system-side harmonic impedance parameter, the harmonic voltage time series data and the harmonic current time series data;

[0018] The harmonic positioning module is used to determine the location of the harmonic source according to the user-side harmonic responsibility and the system-side harmonic responsibility.

[0019] Compared with the prior art, the offshore wind power harmonic source tracing method and system based on Bayesian regression disclosed in the embodiment of the present invention calculates the voltage fluctuation value and current fluctuation value of each time node by obtaining the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated; constructs a harmonic impedance estimation mathematical model based on the voltage fluctuation value and the current fluctuation value; solves the harmonic impedance estimation mathematical model based on the Bayesian theorem to obtain the system-side harmonic impedance parameters, accurately tracks the fluctuation of the system-side harmonic impedance, and quantifies the user-side harmonic responsibility and the system-side harmonic responsibility according to the system-side harmonic impedance parameters, the harmonic voltage time series data and the harmonic current time series data; finally, determines the location of the harmonic source according to the user-side harmonic responsibility and the system-side harmonic responsibility, and can accurately realize harmonic source tracing when the impedance changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solution of the present invention, the drawings required for use in the implementation mode will be briefly introduced below. Obviously, the drawings described below are only some implementation modes of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 It is a flow chart of a method for tracing the source of offshore wind power harmonics based on Bayesian regression provided by one embodiment of the present invention;

[0022] Figure 2 It is a structural schematic diagram of a harmonic source circuit provided by an embodiment of the present invention;

[0023] Figure 3 is a parameter annotation schematic diagram provided by an embodiment of the present invention;

[0024] Figure 4 It is a structural schematic diagram of an offshore wind power harmonic tracing system based on Bayesian regression provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0026] See also Figure 1 , is a flow chart of a method for tracing the source of offshore wind power harmonics based on Bayesian regression provided by an embodiment of the present invention. Specifically, the method for tracing the source of offshore wind power harmonics based on Bayesian regression includes steps S1 to S6:

[0027] S1. Obtaining the harmonic voltage time series data and harmonic current time series data of the node to be evaluated;

[0028] S2. Calculate the voltage fluctuation value at each time node according to the harmonic voltage time series data, and calculate the current fluctuation value at each time node according to the harmonic current time series data;

[0029] S3, constructing a harmonic impedance estimation mathematical model according to the voltage fluctuation value and the current fluctuation value;

[0030] S4. Solving the harmonic impedance estimation mathematical model based on Bayesian theorem to obtain system-side harmonic impedance parameters;

[0031] S5. quantifying the user-side harmonic responsibility and the system-side harmonic responsibility according to the system-side harmonic impedance parameter, the harmonic voltage time series data, and the harmonic current time series data;

[0032] S6. Determine the location of the harmonic source according to the user-side harmonic responsibility and the system-side harmonic responsibility.

[0033] Specifically, the method is applied to an offshore wind farm, and the node to be evaluated is a common connection point. In step S1, the harmonic voltage of the node to be evaluated is continuously monitored, and the monitored harmonic voltage forms harmonic voltage time series data in chronological order, and the harmonic current of the node to be evaluated is continuously monitored, and the monitored harmonic current forms harmonic current time series data in chronological order. In step S2, the voltage fluctuation value between adjacent time nodes in the harmonic voltage series data is determined, and the current fluctuation value between adjacent time nodes in the harmonic current series data is determined. According to the voltage fluctuation value and current fluctuation value of step S2, based on Ohm's law, a harmonic impedance estimation mathematical model in step S3 is constructed. It can be understood that the mathematical model includes the system-side harmonic impedance parameters to be solved. In step S4, the model is solved using Bayes' theorem to predict the system-side harmonic impedance parameters of each time node. In steps S5 and S6, the user-side harmonic responsibility and the system-side harmonic responsibility are quantified according to the system-side harmonic impedance parameters at each time node, wherein the user-side harmonic responsibility characterizes the probability of the harmonic source appearing on the user side to a certain extent, and the system-side harmonic responsibility characterizes the probability of the harmonic source appearing on the system side to a certain extent. By analyzing the user-side harmonic responsibility and the system-side harmonic responsibility, it is determined whether the harmonic source is located on the user side or the system side, thereby achieving accurate tracing of the harmonic source.

[0034] Compared with the prior art, the embodiment of the present invention calculates the voltage fluctuation value and the current fluctuation value of each time node by obtaining the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated; constructs a harmonic impedance estimation mathematical model based on the voltage fluctuation value and the current fluctuation value; solves the harmonic impedance estimation mathematical model based on the Bayesian theorem to obtain the harmonic impedance parameters on the system side, accurately tracks the fluctuation of the harmonic impedance on the system side, and quantifies the harmonic responsibility on the user side and the harmonic responsibility on the system side according to the harmonic impedance parameters on the system side, the harmonic voltage time series data and the harmonic current time series data; finally, determines the location of the harmonic source according to the harmonic responsibility on the user side and the harmonic responsibility on the system side, and can accurately trace the harmonic source when the impedance changes.

[0035] In a preferred embodiment, on the basis of steps S1 to S6, step S2 specifically includes:

[0036] Subtracting the harmonic voltage of the first time node from the harmonic voltage of the first time node at a time node after the first time node to obtain the voltage fluctuation value of the first time node; wherein the harmonic voltage time series data includes the harmonic voltages of several time nodes, and the first time node is any time node among all the time nodes except the last time node;

[0037] The harmonic current at the first time node is subtracted from the harmonic current at the first time node to obtain the current fluctuation value at the first time node; wherein the harmonic current time series data includes the harmonic currents at several of the time nodes.

[0038] Specifically, the specific calculation formulas for the voltage fluctuation value and the current fluctuation value are:

[0039] ;

[0040] ;

[0041] Wherein, n∈[1,N-1], N is the total number of time nodes, which is equal to the total number of samples. It is worth noting that a harmonic voltage in the harmonic voltage time series data and a harmonic current at the corresponding moment in the harmonic current time series data constitute a sample.

[0042] In one implementation, based on steps S1 to S6, step S3 specifically includes:

[0043] Separating the imaginary part and the real part of the voltage fluctuation value, and constructing a voltage fluctuation matrix according to the imaginary part and the real part of all the voltage fluctuation values;

[0044] Separating the imaginary part and the real part of the current fluctuation value, and constructing a current fluctuation matrix according to the imaginary part and the real part of all the current fluctuation values;

[0045] Setting a weight diagonal matrix for the time node to be estimated, wherein the weight values ​​in the weight diagonal matrix are set according to the time difference between the time node to be estimated and other time nodes;

[0046] Using Ohm's law, a harmonic impedance estimation mathematical model is constructed; wherein the harmonic impedance estimation mathematical model is as follows:

[0047] The weight diagonal matrix of the time node to be estimated × the voltage fluctuation matrix = the weight diagonal matrix of the time node to be estimated × the current fluctuation matrix × the unknown model parameters of the time node to be estimated + the residual vector of the time node to be estimated; wherein, the time node to be estimated is any time node among all the time nodes, and the unknown model parameters are composed of the system side harmonic impedance parameters and background harmonic voltage to be solved.

[0048] Specifically, see Figure 2 The harmonic source circuit structure diagram shown in FIG. 1 includes the system side 1 and the user side 2 with the node to be evaluated as the common connection point PCC. The impedance information of the user side 2 is , the current information of user side 2 is The construction process of the harmonic impedance estimation mathematical model is as follows:

[0049] 1. Based on public connection point monitoring Harmonic voltage at time Harmonic current , the circuit equation can be written as:

[0050] ;

[0051] in, is the system side harmonic impedance, is the background harmonic voltage.

[0052] 2. Taking the wave quantity form of the above circuit equation, we have:

[0053] ;

[0054] in, , , , n∈[1,N-1], N is the total number of samples. It is worth noting that a harmonic voltage in the harmonic voltage time series data and a harmonic current at the corresponding moment in the harmonic current time series data constitute a sample.

[0055] 3. Build Mathematical model for harmonic impedance estimation at time:

[0056] ;

[0057] in, yes The weight diagonal matrix at time, whose weight value is proportional to the time difference| - |Inversely proportional, Except For other time points other than yes The residual vector at time t, yes The unknown model parameters at time , and It is to be solved The real and imaginary parts of the system-side harmonic impedance parameters at time , and It is to be solved The fluctuation value of background harmonic voltage at the moment The real and imaginary parts of .

[0058] Furthermore, step S4 specifically includes:

[0059] Solving the harmonic impedance estimation mathematical model based on Bayesian theorem to derive probability expressions of the unknown model parameters;

[0060] According to the probability expression, the mean and covariance matrix of the unknown model parameters are derived, and the system-side harmonic impedance parameters are calculated according to the mean and covariance matrix of the unknown model parameters.

[0061] Furthermore, solving the harmonic impedance estimation mathematical model based on the Bayesian theorem to derive the probability expression of the unknown model parameters includes:

[0062] According to Bayes' theorem, the posterior estimate of the unknown model parameters is expressed as a candidate expression:

[0063] ;

[0064] in, for and The joint probability distribution of ,for The conditional probability distribution of It means the mean , the variance is The normal distribution of is the identity matrix; , is the prior probability of the unknown model parameters, is the covariance matrix; for The probability distribution of Figure 3 The parameter annotation diagram shown is Used to characterize the unknown model parameters in the harmonic impedance estimation mathematical model, The product of the weight diagonal matrix and the voltage fluctuation matrix used to characterize the mathematical model of harmonic impedance estimation, The product of the weight diagonal matrix and the current fluctuation matrix used to characterize the mathematical model of harmonic impedance estimation; optional, such as Figure 3 As shown, it can be used To characterize the harmonic impedance estimation mathematical model The residual vector at time .

[0065] According to the conjugate prior property in Bayesian statistics, the candidate expression is converted into a probability expression:

[0066] ;

[0067] in, and are the mean and covariance matrices of the unknown model parameters, respectively.

[0068] Specifically, the solution process of the system-side harmonic impedance parameters is as follows:

[0069] 1. Assume that the elements of the residual vector follow a normal distribution , is the variance; the prior probability of the unknown model parameters follows a zero-mean normal distribution , is the covariance matrix.

[0070] 2. According to Bayes’ theorem, the posterior estimates of the unknown model parameters are expressed as candidate expressions.

[0071] 3. According to the conjugate prior characteristics in Bayesian statistics, the candidate expression is converted into a probability expression.

[0072] 4. Expanding the posterior probability distribution (i.e., probability expression), we can derive:

[0073] ;

[0074] Further deduction yields the mean and covariance matrix of the posterior probability distribution:

[0075] ;

[0076] ;

[0077] It can be understood that the mean vector of the probability distribution of the unknown model parameters is , further determining the solved unknown model parameters can be expressed as ,in, and for The real and imaginary parts of the system-side harmonic impedance parameters at time , and for The real and imaginary parts of the background harmonic voltage at the moment. and Sure System side harmonic impedance parameters at time .

[0078] In one implementation, based on steps S1 to S6, step S5 specifically includes:

[0079] The user-side harmonic responsibility and system-side harmonic responsibility are quantified by the following formula:

[0080] ;

[0081] ;

[0082] in, For the user side harmonic responsibility, For the system side harmonic responsibility, for and The angle between for and The angle between is the system side harmonic impedance parameter at the nth time node, is the harmonic voltage at the nth time node, is the harmonic current at the nth time node; N is the total number of the time nodes.

[0083] In one implementation, based on steps S1 to S6, step S6 specifically includes:

[0084] When the user-side harmonic responsibility is greater than the system-side harmonic responsibility, the harmonic source is located at the user side;

[0085] When the user-side harmonic responsibility is smaller than the system-side harmonic responsibility, the harmonic source is located on the system side.

[0086] It can be understood that the harmonic responsibility on the user side characterizes the probability of the harmonic source appearing on the user side to a certain extent, and the harmonic responsibility on the system side characterizes the probability of the harmonic source appearing on the system side to a certain extent. After calculating the harmonic responsibility on the system side and the harmonic responsibility on the user side, the harmonic source can be located according to the relationship between the two. When the harmonic responsibility on the system side is greater than the harmonic responsibility on the user side, it means that the probability of the harmonic source being located on the system side is greater than the probability of being located on the user side, and the dominant harmonic source is located on the system side. When the harmonic responsibility on the system side is less than the harmonic responsibility on the user side, it means that the probability of the harmonic source being located on the system side is less than the probability of being located on the user side, and the dominant harmonic source is located on the user side.

[0087] To further illustrate the effect of this embodiment, based on Figure 2 The harmonic source circuit in the embodiment is simulated and tested to verify the effectiveness of the present embodiment, and a comparative analysis is performed with the existing methods. Method 1 is a correlation analysis method, method 2 is an independent component analysis method, method 3 is a covariance method, and method 4 is a method provided in an embodiment of the present invention.

[0088] The harmonic source and harmonic impedance are set as follows: the harmonic impedance parameters on the system side and the harmonic impedance parameters on the user side are set to 10+j15 ohms and 40+j60 ohms respectively, and the impedance amplitudes on both sides have a random fluctuation of 2.5%. The harmonic impedance parameters on the system side are also superimposed with a 1-cycle 20% sinusoidal fluctuation. The background harmonic voltage is set to 2500 volts with a random fluctuation of 5%, and the harmonic current parameters on the user side are set to 100A with a random fluctuation of 10%.

[0089] The size of background harmonics is an important factor affecting harmonic source analysis. k is used to reflect the background harmonic level. In the simulation test, the background harmonic voltage is set to k*2500 volts. For different cases of k values ​​considered, 800 sets of sample data are generated according to the above simulation settings. The calculation results are shown in Tables 1 and 2.

[0090] Table 1: Harmonic liability error

[0091]

[0092] Table 2: Standard deviation of harmonic liability error

[0093]

[0094] By comparing the average estimation errors of the methods in Table 1, it can be seen that for different background harmonic levels, the maximum estimation error of the method provided by the embodiment of the present invention does not exceed 17%, and the estimation effect is the best, while the maximum estimation errors of methods 1 to 3 are approximately 20%, 34%, and 100%, and the estimation errors are relatively large. In addition, the standard deviation of the estimation error of the method provided by the embodiment of the present invention in Table 2 is the smallest, and for different background harmonic levels, the maximum standard deviation does not exceed 2.3%, which proves the effectiveness and stability of the method provided by the embodiment of the present invention. In summary, the harmonic responsibility calculated by the method provided by the embodiment of the present invention is more accurate and can more effectively guide the development of harmonic control work at offshore wind power grid connection points.

[0095] Compared with the prior art, the offshore wind power harmonic source tracing method and system based on Bayesian regression disclosed in the embodiment of the present invention does not need to use a disturbance generating device to inject harmonic disturbances into the power grid, and will not affect the normal operation of the power grid. The voltage fluctuation value and current fluctuation value of each time node are calculated by obtaining the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated; a harmonic impedance estimation mathematical model is constructed according to the voltage fluctuation value and the current fluctuation value; the harmonic impedance estimation mathematical model is solved based on the Bayesian theorem to obtain the system-side harmonic impedance parameters, accurately track the fluctuation of the system-side harmonic impedance, and quantify the user-side harmonic responsibility and the system-side harmonic responsibility according to the system-side harmonic impedance parameters, the harmonic voltage time series data and the harmonic current time series data; finally, the location of the harmonic source is determined according to the user-side harmonic responsibility and the system-side harmonic responsibility, so that better impedance estimation results can be obtained when the background harmonics fluctuate and the impedance fluctuates, and the harmonic responsibility quantification result is more accurate, thereby improving the accuracy of harmonic source tracing.

[0096] See also Figure 4 , Figure 4 : is a schematic diagram of the structure of an offshore wind power harmonic tracing system based on Bayesian regression provided by an embodiment of the present invention. The offshore wind power harmonic tracing system based on Bayesian regression 20 includes:

[0097] A data acquisition module 21 is used to acquire the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated;

[0098] A fluctuation calculation module 22, used to calculate the voltage fluctuation value of each time node according to the harmonic voltage time series data, and calculate the current fluctuation value of each time node according to the harmonic current time series data;

[0099] A model building module 23, used to build a harmonic impedance estimation mathematical model according to the voltage fluctuation value and the current fluctuation value;

[0100] An impedance calculation module 24, used for solving the harmonic impedance estimation mathematical model based on the Bayesian theorem to obtain system-side harmonic impedance parameters;

[0101] A responsibility determination module 25, for quantifying user-side harmonic responsibility and system-side harmonic responsibility according to the system-side harmonic impedance parameter, the harmonic voltage time series data and the harmonic current time series data;

[0102] The harmonic locating module 26 is used to determine the location of the harmonic source according to the user-side harmonic responsibility and the system-side harmonic responsibility.

[0103] In one implementation, the fluctuation calculation module 22 is specifically used to:

[0104] Subtracting the harmonic voltage of the first time node from the harmonic voltage of the first time node at a time node after the first time node to obtain the voltage fluctuation value of the first time node; wherein the harmonic voltage time series data includes the harmonic voltages of several time nodes, and the first time node is any time node among all the time nodes except the last time node;

[0105] The harmonic current at the first time node is subtracted from the harmonic current at the first time node to obtain the current fluctuation value at the first time node; wherein the harmonic current time series data includes the harmonic currents at several of the time nodes.

[0106] In one embodiment, the model building module 23 is specifically used to:

[0107] Separating the imaginary part and the real part of the voltage fluctuation value, and constructing a voltage fluctuation matrix according to the imaginary part and the real part of all the voltage fluctuation values;

[0108] Separating the imaginary part and the real part of the current fluctuation value, and constructing a current fluctuation matrix according to the imaginary part and the real part of all the current fluctuation values;

[0109] Setting a weight diagonal matrix for the time node to be estimated, wherein the weight values ​​in the weight diagonal matrix are set according to the time difference between the time node to be estimated and other time nodes;

[0110] Using Ohm's law, a harmonic impedance estimation mathematical model is constructed; wherein the harmonic impedance estimation mathematical model is as follows:

[0111] The weight diagonal matrix of the time node to be estimated × the voltage fluctuation matrix = the weight diagonal matrix of the time node to be estimated × the current fluctuation matrix × the unknown model parameters of the time node to be estimated + the residual vector of the time node to be estimated; wherein, the time node to be estimated is any time node among all the time nodes, and the unknown model parameters are composed of the system side harmonic impedance parameters and background harmonic voltage to be solved.

[0112] It is worth noting that the specific working process of the offshore wind power harmonic tracing system based on Bayesian regression can refer to the working process of the offshore wind power harmonic tracing method based on Bayesian regression described in the above embodiment, which will not be repeated here.

[0113] Compared with the prior art, the offshore wind power harmonic source tracing system based on Bayesian regression provided by the embodiment of the present invention calculates the voltage fluctuation value and the current fluctuation value of each time node by obtaining the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated; constructs a harmonic impedance estimation mathematical model based on the voltage fluctuation value and the current fluctuation value; solves the harmonic impedance estimation mathematical model based on the Bayesian theorem to obtain the system-side harmonic impedance parameters, accurately tracks the fluctuation of the system-side harmonic impedance, and quantifies the user-side harmonic responsibility and the system-side harmonic responsibility according to the system-side harmonic impedance parameters, the harmonic voltage time series data and the harmonic current time series data; finally, determines the location of the harmonic source according to the user-side harmonic responsibility and the system-side harmonic responsibility, and can accurately realize harmonic source tracing when the impedance changes.

[0114] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for tracing the source of offshore wind power harmonics based on Bayesian regression, characterized in that: include: Obtain the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated; Calculating the voltage fluctuation value at each time node according to the harmonic voltage time series data, and calculating the current fluctuation value at each time node according to the harmonic current time series data; Constructing a harmonic impedance estimation mathematical model according to the voltage fluctuation value and the current fluctuation value; Solving the harmonic impedance estimation mathematical model based on Bayesian theorem to obtain system-side harmonic impedance parameters; quantifying user-side harmonic responsibility and system-side harmonic responsibility according to the system-side harmonic impedance parameter, the harmonic voltage time series data, and the harmonic current time series data; Determine the location of the harmonic source according to the user-side harmonic responsibility and the system-side harmonic responsibility; The step of constructing a harmonic impedance estimation mathematical model according to the voltage fluctuation value and the current fluctuation value includes: Separating the imaginary part and the real part of the voltage fluctuation value, and constructing a voltage fluctuation matrix according to the imaginary part and the real part of all the voltage fluctuation values; Separating the imaginary part and the real part of the current fluctuation value, and constructing a current fluctuation matrix according to the imaginary part and the real part of all the current fluctuation values; Setting a weight diagonal matrix for the time node to be estimated, wherein the weight values ​​in the weight diagonal matrix are set according to the time difference between the time node to be estimated and other time nodes; Using Ohm's law, a harmonic impedance estimation mathematical model is constructed; wherein the harmonic impedance estimation mathematical model is as follows: The weight diagonal matrix of the time node to be estimated × the voltage fluctuation matrix = the weight diagonal matrix of the time node to be estimated × the current fluctuation matrix × the unknown model parameters of the time node to be estimated + the residual vector of the time node to be estimated; wherein, the time node to be estimated is any one of all the time nodes, and the unknown model parameters are composed of the system side harmonic impedance parameters and background harmonic voltage to be solved; The method of solving the harmonic impedance estimation mathematical model based on the Bayesian theorem to obtain the system-side harmonic impedance parameters includes: Solving the harmonic impedance estimation mathematical model based on Bayesian theorem to derive probability expressions of the unknown model parameters; According to the probability expression, the mean and covariance matrix of the unknown model parameters are derived, and the system-side harmonic impedance parameters are calculated according to the mean and covariance matrix of the unknown model parameters; The method of solving the harmonic impedance estimation mathematical model based on the Bayesian theorem and deriving the probability expression of the unknown model parameters includes: According to Bayes' theorem, the posterior estimate of the unknown model parameters is expressed as a candidate expression: ； in, for and The joint probability distribution of ,for The conditional probability distribution of Indicates the mean , the variance is The normal distribution of is the identity matrix; , is the prior probability of the unknown model parameters; for The probability distribution of are the unknown model parameters, is equal to the product of the weight diagonal matrix and the voltage fluctuation matrix, is the product of the weight diagonal matrix and the current fluctuation matrix; According to the conjugate prior property in Bayesian statistics, the candidate expression is converted into a probability expression: ； in, and are the mean and covariance matrices of the unknown model parameters, respectively.

2. The offshore wind power harmonic source tracing method according to claim 1, characterized in that: The step of calculating the voltage fluctuation value at each time node according to the harmonic voltage time series data, and calculating the current fluctuation value at each time node according to the harmonic current time series data, comprises: Subtracting the harmonic voltage of the first time node from the harmonic voltage of the first time node at a time node after the first time node to obtain the voltage fluctuation value of the first time node; wherein the harmonic voltage time series data includes the harmonic voltages of several time nodes, and the first time node is any time node among all the time nodes except the last time node; The harmonic current at the first time node is subtracted from the harmonic current at the first time node to obtain the current fluctuation value at the first time node; wherein the harmonic current time series data includes the harmonic currents at several of the time nodes.

3. The offshore wind power harmonic source tracing method according to claim 1, characterized in that: The quantifying the user-side harmonic responsibility and the system-side harmonic responsibility according to the system-side harmonic impedance parameter, the harmonic voltage time series data, and the harmonic current time series data includes: The harmonic responsibility on the user side and the harmonic responsibility on the system side are quantified by the following formula: ； ； in, For the user side harmonic responsibility, For the system side harmonic responsibility, for and The angle between for and The angle between is the system side harmonic impedance parameter at the nth time node, is the harmonic voltage at the nth time node, is the harmonic current at the nth time node; N is the total number of the time nodes.

4. The offshore wind power harmonic source tracing method according to claim 1, characterized in that: The determining the location of the harmonic source according to the user-side harmonic responsibility and the system-side harmonic responsibility includes: When the user-side harmonic responsibility is greater than the system-side harmonic responsibility, the harmonic source is located at the user side; When the user-side harmonic responsibility is smaller than the system-side harmonic responsibility, the harmonic source is located on the system side.

5. An offshore wind power harmonic source tracing system based on Bayesian regression, characterized in that: include: A data acquisition module, used to acquire the harmonic voltage time series data and the harmonic current time series data of the node to be evaluated; A fluctuation calculation module, used to calculate the voltage fluctuation value of each time node according to the harmonic voltage time series data, and calculate the current fluctuation value of each time node according to the harmonic current time series data; A model building module, used for building a harmonic impedance estimation mathematical model according to the voltage fluctuation value and the current fluctuation value; An impedance calculation module, used for solving the harmonic impedance estimation mathematical model based on the Bayesian theorem to obtain system-side harmonic impedance parameters; A responsibility determination module, used for quantifying the user-side harmonic responsibility and the system-side harmonic responsibility according to the system-side harmonic impedance parameter, the harmonic voltage time series data and the harmonic current time series data; A harmonic positioning module, used to determine the location of the harmonic source according to the user-side harmonic responsibility and the system-side harmonic responsibility; The model building module is specifically used for: Separating the imaginary part and the real part of the voltage fluctuation value, and constructing a voltage fluctuation matrix according to the imaginary part and the real part of all the voltage fluctuation values; Separating the imaginary part and the real part of the current fluctuation value, and constructing a current fluctuation matrix according to the imaginary part and the real part of all the current fluctuation values; Setting a weight diagonal matrix for the time node to be estimated, wherein the weight values ​​in the weight diagonal matrix are set according to the time difference between the time node to be estimated and other time nodes; Using Ohm's law, a harmonic impedance estimation mathematical model is constructed; wherein the harmonic impedance estimation mathematical model is as follows: The weight diagonal matrix of the time node to be estimated × the voltage fluctuation matrix = the weight diagonal matrix of the time node to be estimated × the current fluctuation matrix × the unknown model parameters of the time node to be estimated + the residual vector of the time node to be estimated; wherein, the time node to be estimated is any one of all the time nodes, and the unknown model parameters are composed of the system side harmonic impedance parameters and background harmonic voltage to be solved; The impedance calculation module is specifically used for: Solving the harmonic impedance estimation mathematical model based on Bayesian theorem to derive probability expressions of the unknown model parameters; According to the probability expression, the mean and covariance matrix of the unknown model parameters are derived, and the system-side harmonic impedance parameters are calculated according to the mean and covariance matrix of the unknown model parameters; The method of solving the harmonic impedance estimation mathematical model based on the Bayesian theorem and deriving the probability expression of the unknown model parameters includes: According to Bayes' theorem, the posterior estimate of the unknown model parameters is expressed as a candidate expression: ； in, for and The joint probability distribution of ,for The conditional probability distribution of Indicates the mean , the variance is The normal distribution of is the identity matrix; , is the prior probability of the unknown model parameters; for The probability distribution of are the unknown model parameters, is equal to the product of the weight diagonal matrix and the voltage fluctuation matrix, is the product of the weight diagonal matrix and the current fluctuation matrix; According to the conjugate prior property in Bayesian statistics, the candidate expression is converted into a probability expression: ； in, and are the mean and covariance matrices of the unknown model parameters, respectively.

6. The offshore wind power harmonic source tracing system according to claim 5, characterized in that: The fluctuation calculation module is specifically used for: Subtracting the harmonic voltage of the first time node from the harmonic voltage of the first time node at a time node after the first time node to obtain the voltage fluctuation value of the first time node; wherein the harmonic voltage time series data includes the harmonic voltages of several time nodes, and the first time node is any time node among all the time nodes except the last time node; The harmonic current at the first time node is subtracted from the harmonic current at the first time node to obtain the current fluctuation value at the first time node; wherein the harmonic current time series data includes the harmonic currents at several of the time nodes.

Images