On-line monitoring method and device for demagnetization of permanent magnet of permanent magnet wind driven generator

Abstract

The invention relates to an online monitoring method for demagnetization of a permanent magnet of a permanent magnet wind driven generator, and the method comprises the following steps: S1, installing a voltage transformer, a temperature sensor, a humidity sensor, a communication module and a data collection system on a small parallel end of a generator winding, and guaranteeing the connection to be correct; s2, acquiring an output signal of the voltage transformer and environment temperature and humidity data in real time; s3, according to the monitoring results of the temperature sensor and the humidity sensor, establishing an environment compensation model to compensate the output of the voltage transformer; s4, according to the physical characteristics and the electrical principle of the generator, establishing a mathematical model of the generator, performing online parameter identification by using a genetic algorithm, and identifying and correcting the parameters of the generator in real time; s5, calculating the induced voltage amplitude of each column of permanent magnets by using a flux linkage observation model and an induced voltage calculation algorithm; and S6, whether the permanent magnet has a demagnetization phenomenon or not is judged in combination with demagnetization judgment logic, and a threshold value is set according to the actual situation and historical data of the generator.

Description

Technical Field

[0001] The present invention relates to the technical field of permanent magnet wind turbine generators, and in particular to an online monitoring method and device for demagnetization of permanent magnets in permanent magnet wind turbine generators. Background Art

[0002] Permanent magnet wind turbines have the advantages of high efficiency, low noise, and low maintenance costs, so they have been widely used in the field of wind power generation. They use permanent magnets as magnetic poles and can generate a stable magnetic field without external power supply, thereby simplifying the structure of the generator and improving the power generation efficiency. During the operation of permanent magnet wind turbines, permanent magnets may be affected by various factors such as high temperature, vibration, and electromagnetic fields, resulting in demagnetization of permanent magnets. Permanent magnet demagnetization will reduce the output electromagnetic torque of the generator, increase the stator armature current, increase the motor copper loss, and further increase the temperature of the permanent magnet, thereby aggravating the demagnetization process of the permanent magnet. It is particularly important to develop a technology that can monitor the demagnetization status of permanent magnets in real time. This technology needs to be able to accurately and in real time reflect the changes in the magnetic properties of permanent magnets, so as to promptly detect and handle potential demagnetization faults and ensure the safe and stable operation of the generator.

[0003] In the prior art, the technology used for online monitoring of permanent magnet demagnetization of permanent magnet wind turbines mainly involves methods based on flux observation and voltage transformers. However, the accuracy of this method depends on the accuracy of basic parameters of the generator, magnetic circuit saturation characteristics and material property parameters. When the parameters are inaccurate, the calculation results will be inaccurate, and even mislead the judgment of the generator performance. Therefore, it is particularly important to propose a method and device for online monitoring of permanent magnet demagnetization of permanent magnet wind turbines. Summary of the invention

[0004] The purpose of the present invention is to solve the problem that the accuracy of the method based on flux observation and voltage transformer depends on the accuracy of the basic parameters of the generator, the magnetic circuit saturation characteristics and the material property parameters. When the parameters are inaccurate, the calculation results will be inaccurate, and even the judgment of the generator performance will be misled. A method and device for online monitoring of permanent magnet demagnetization of a permanent magnet wind generator are provided.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for online monitoring of permanent magnet demagnetization of a permanent magnet wind turbine generator comprises the following steps:

[0007] S1: Install the voltage transformer, temperature sensor, humidity sensor, communication module and data acquisition system on the small parallel head of the generator winding, and ensure that the connection is correct;

[0008] S2: collect the output signal of the voltage transformer, ambient temperature and humidity data in real time and perform preprocessing;

[0009] S3: According to the monitoring results of the temperature sensor and the humidity sensor, an environmental compensation model is established to compensate the output of the voltage transformer;

[0010] S4: According to the physical characteristics and electrical principles of the generator, a mathematical model of the generator is established, and the genetic algorithm is used for online parameter identification to identify and correct the parameters of the generator in real time;

[0011] S5: Calculate the induced voltage amplitude of each column of permanent magnets using the flux observation model and the induced voltage calculation algorithm;

[0012] S6: Combine the demagnetization judgment logic to determine whether the permanent magnet has demagnetization, set the threshold according to the actual situation and historical data of the generator, and display the monitoring results in real time on the host computer or remote monitoring system. When demagnetization of the permanent magnet is detected, a fault warning or alarm signal is issued.

[0013] The above scheme further comprises:

[0014] Further, in S3, the specific steps of establishing an environmental compensation model to compensate the output of the voltage transformer are as follows:

[0015] Data collection: collect the ambient temperature and humidity around the generator monitored by the temperature sensor and humidity sensor, and record the output signal of the voltage transformer;

[0016] Establish an environmental compensation model: Environmental factors include temperature T and humidity H. Analyze the relationship between environmental factors and the output signal V of the voltage transformer and establish an environmental compensation model. The formula of the environmental compensation model is V comp =V raw +f(T,H), where V raw is the original output signal of the voltage transformer, V comp is the signal after environmental compensation, and f(T,H) is the compensation function, which represents the influence of environmental factors on the signal;

[0017] Determine the compensation function: Determine the specific form of the compensation function f(T,H), f(T,H) = aT + bH + c, where a, b, c are coefficients to be determined, by collecting multiple sets of experimental data, the experimental data including temperature T, humidity H, the original output signal V of the voltage transformer raw and the real signal V true , use the least squares method to solve the coefficients a, b, c and obtain the compensation function f(T,H);

[0018] Implementation of compensation: During real-time monitoring, the value of the compensation function f(T,H) is calculated based on the current ambient temperature and humidity, and the compensation value is added to the original output signal of the voltage transformer to obtain a signal after environmental compensation.

[0019] Further, in S4, according to the physical characteristics and electrical principles of the generator, a mathematical model of the generator is established, and a genetic algorithm is used to perform online parameter identification, and the parameters of the generator are identified and corrected in real time, including the following steps:

[0020] Data collection: real-time collection of electrical parameters such as voltage and current of the generator, and recording of the operating status and environmental conditions of the generator;

[0021] Establishing mathematical model: Establishing mathematical model of generator according to its physical characteristics and electrical principles;

[0022] Select the identification algorithm: Select the genetic algorithm;

[0023] Initial parameter estimation: Based on the design parameters or historical data of the generator, the parameters in the model are initially estimated. The accuracy of the initial estimation will affect the convergence speed and accuracy of subsequent identification.

[0024] Online identification: Input the real-time collected data into the genetic algorithm for online parameter identification. During the identification process, the genetic algorithm continuously adjusts the parameter values ​​in the model according to the changes in the data;

[0025] Parameter correction and update: According to the identification results, the parameters in the mathematical model are corrected and updated. The corrected parameters will be used in the subsequent monitoring and diagnosis process.

[0026] Further, the specific steps to establish the mathematical model are:

[0027] Determine key parameters: According to the physical characteristics and electrical principles of the generator, determine the key parameters included in the mathematical model, which include: resistance: stator resistance, rotor resistance, etc., which reflect the loss generated when the current passes through the conductor; inductance: stator inductance, rotor inductance and mutual inductance, etc., which describe the storage and change of the magnetic field in the conductor; permanent magnet flux: represents the magnetic flux generated by the magnetic field generated by the permanent magnet in the conductor; other parameters: such as pole pair number, rotor position angular velocity, etc., are used to describe the operating status of the generator;

[0028] Establish mathematical model: Based on the above key parameters, establish the mathematical model of the generator, whose stator resistance is R s , the rotor resistance is R r , the stator inductance is L s , the rotor inductance is L r , the mutual inductance is M, the permanent magnet flux is λ f , then the voltage equation is expressed as Among them, u s and u r are the voltage vectors of the stator and rotor respectively, i s and i r are the current vectors of the stator and rotor, respectively, ss , sr , rr and λ rs They are stator self-inductance flux, stator and rotor mutual inductance flux, rotor self-inductance flux and rotor and stator mutual inductance flux respectively.

[0029] Further, the specific steps of online parameter identification using genetic algorithm are as follows:

[0030] Encoding: Encode the parameters to be identified into individuals (i.e. chromosomes) in the genetic algorithm. Each individual consists of a series of genes, and each gene corresponds to a parameter to be identified.

[0031] Initialize the population: randomly generate a certain number of individuals as the initial population, and the gene values ​​(i.e. parameter values) of these individuals are randomly selected within a certain range;

[0032] Fitness evaluation: Substitute each individual (i.e. a set of parameter values) into the mathematical model, calculate the difference between the output of the mathematical model and the actual data, and use this difference as the fitness value of the individual. The smaller the fitness value, the better the individual.

[0033] Selection: Select a certain number of excellent individuals as parents according to their fitness values ​​to generate the next generation of individuals;

[0034] Crossover: Perform a crossover operation on the parent individuals to generate new offspring individuals;

[0035] Mutation: Perform mutation operations on offspring individuals to introduce new gene combinations;

[0036] Iteration: Repeat fitness evaluation, selection, crossover, and mutation until a predetermined number of iterations is reached or the fitness value converges to a certain range.

[0037] Output optimal parameters: Select the individual with the smallest fitness value from the final population as the optimal parameter combination.

[0038] Further, in S5, the induced voltage amplitude of each column of permanent magnets is calculated using the flux observation model and the induced voltage calculation algorithm, including the following steps:

[0039] Establishment of flux observation model: The flux observation model comprehensively considers the permanent magnet flux and the inductor flux; permanent magnet flux: this is the flux generated by the permanent magnet itself, which is usually a constant, but will change during the demagnetization process; inductor flux: this is the flux generated by the stator current through the inductor, which is proportional to the stator current; the flux observation model is expressed as Φ = ΦPM + L·I, where Φ represents the total flux, ΦPM represents the permanent magnet flux, L represents the inductance, and I represents the stator current;

[0040] Calculation of induced voltage: According to Faraday's law of electromagnetic induction, the induced electromotive force is proportional to the rate of change of magnetic flux. Therefore, the induced voltage is obtained by calculating the rate of change of magnetic flux. The calculation formula of the induced voltage is E=-dΦ / dt. Substituting the magnetic flux observation model into the calculation formula of the induced voltage, we get: E=-d(ΦPM+L·I)dt=-dtdΦPM-Ld I dt.

[0041] Further, in S6, the demagnetization judgment logic is combined to determine whether the permanent magnet has demagnetization phenomenon, and the threshold is set according to the actual situation of the generator and historical data. The specific steps are:

[0042] Collect data: collect basic parameters of the generator, permanent magnet characteristics, operating conditions and environmental conditions;

[0043] Statistical analysis: Statistical analysis of the induced voltage in historical data to determine its distribution and change trend;

[0044] Setting an initial threshold: Based on the statistical analysis results, combined with the actual situation and operation experience of the generator, a threshold of the induced voltage is preliminarily set, and the preliminarily set threshold of the induced voltage reflects the initial signs of permanent magnet demagnetization;

[0045] Real-time monitoring: Real-time monitoring of the generator's operating data, especially the changes in induced voltage;

[0046] Support vector machine training: After collecting enough data, a model is trained using a support vector machine to predict the demagnetization of the permanent magnet. The input of the initial setting of an induced voltage threshold is the induced voltage, stator current and temperature, and the output is the demagnetization probability or demagnetization level;

[0047] Threshold adjustment: According to the prediction results of the support vector machine, the initial threshold is adjusted. If the prediction accuracy of the model is high, the threshold can be adjusted more strictly to improve the accuracy of demagnetization judgment;

[0048] Validation model: After adjusting the threshold, the support vector machine is validated and tested to ensure the accuracy and reliability of its prediction results;

[0049] Test threshold: Apply the adjusted threshold to actual operation to see whether it accurately determines the demagnetization of the permanent magnet;

[0050] Feedback and adjustment: Adjust and optimize the thresholds based on the results of verification and testing.

[0051] A permanent magnet wind turbine generator permanent magnet demagnetization online monitoring device used in a permanent magnet wind turbine generator permanent magnet demagnetization online monitoring method comprises:

[0052] Voltage transformer: used to monitor the voltage signal of the small parallel head of the generator winding to ensure that the measurement range, accuracy and response time of the voltage transformer meet the monitoring requirements.

[0053] Data acquisition system: Design a data acquisition system to collect the output signal of the voltage transformer in real time and convert it into a digital signal for processing. The data acquisition system should have the characteristics of high speed, high precision and low noise;

[0054] Communication module: Design a communication module to transmit the collected data to the host computer or remote monitoring system in real time to ensure the reliability and real-time performance of data transmission;

[0055] Sensor integration: Temperature sensors and humidity sensors are integrated on the voltage transformer to monitor ambient temperature and humidity in real time and provide data support for the environmental compensation model.

[0056] The present invention has the following beneficial effects:

[0057] In the present invention, the online parameter identification technology is introduced. By real-time monitoring of key parameters in the operation process of the generator, the genetic algorithm is used for online parameter identification to improve the monitoring accuracy. The environmental compensation technology is adopted. A temperature sensor and a humidity sensor are installed on the voltage transformer to monitor the ambient temperature and humidity in real time. According to the monitoring results of the temperature sensor and the humidity sensor, an environmental compensation model is established to compensate the output of the voltage transformer to reduce the influence of environmental factors on the measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 A method and step diagram of an online monitoring method and device for permanent magnet demagnetization of a permanent magnet wind turbine generator proposed by the present invention. DETAILED DESCRIPTION

[0059] 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.

[0060] See also Figure 1 As shown, the present invention is an online monitoring method for demagnetization of permanent magnets of a permanent magnet wind turbine generator, comprising the following steps:

[0061] S1: Install the voltage transformer, temperature sensor, humidity sensor, communication module and data acquisition system on the small parallel head of the generator winding, and ensure that the connection is correct;

[0062] S2: collect the output signal of the voltage transformer, ambient temperature and humidity data in real time and perform preprocessing;

[0063] S3: According to the monitoring results of the temperature sensor and the humidity sensor, an environmental compensation model is established to compensate the output of the voltage transformer;

[0064] S4: According to the physical characteristics and electrical principles of the generator, a mathematical model of the generator is established, and the genetic algorithm is used for online parameter identification to identify and correct the parameters of the generator in real time;

[0065] S5: Calculate the induced voltage amplitude of each column of permanent magnets using the flux observation model and the induced voltage calculation algorithm;

[0066] S6: Combine the demagnetization judgment logic to determine whether the permanent magnet has demagnetization, set the threshold according to the actual situation and historical data of the generator, and display the monitoring results in real time on the host computer or remote monitoring system. When demagnetization of the permanent magnet is detected, a fault warning or alarm signal is issued.

[0067] In one embodiment, for the above S3, in S3, the specific steps of establishing an environmental compensation model to compensate the output of the voltage transformer are:

[0068] Data collection: collect the ambient temperature and humidity around the generator monitored by the temperature sensor and humidity sensor, and record the output signal of the voltage transformer;

[0069] Establish an environmental compensation model: Environmental factors include temperature T and humidity H. Analyze the relationship between environmental factors and the output signal V of the voltage transformer and establish an environmental compensation model. The formula of the environmental compensation model is V comp =V raw +f(T,H), where V raw is the original output signal of the voltage transformer, V comp is the signal after environmental compensation, and f(T,H) is the compensation function, which represents the influence of environmental factors on the signal;

[0070] Determine the compensation function: Determine the specific form of the compensation function f(T,H), f(T,H) = aT + bH + c, where a, b, c are coefficients to be determined, by collecting multiple sets of experimental data, the experimental data including temperature T, humidity H, the original output signal V of the voltage transformer raw and the real signal V true , use the least squares method to solve the coefficients a, b, c and obtain the compensation function f(T,H);

[0071] Implementation of compensation: During real-time monitoring, the value of the compensation function f(T,H) is calculated based on the current ambient temperature and humidity, and the compensation value is added to the original output signal of the voltage transformer to obtain a signal after environmental compensation.

[0072] Suppose in an experiment, we collected the following data:

[0073]

[0074] Using the least squares method to solve for the coefficients a, b, c, we get:

[0075] f(T,H)=-0.001T-0.0005H+1.020Therefore, the signal after environmental compensation is expressed as:

[0076] Vcomp=Vraw-0.001T-0.0005H+1.020;

[0077] In the actual monitoring process, the value of the compensation function f(T,H) is calculated according to the current ambient temperature and humidity, and added to the original output signal of the voltage transformer to obtain the signal after environmental compensation.

[0078] In one embodiment, for the above S4, in S4, according to the physical characteristics and electrical principles of the generator, a mathematical model of the generator is established, and a genetic algorithm is used to perform online parameter identification, and the parameters of the generator are identified and corrected in real time, including the following steps:

[0079] Data collection: real-time collection of electrical parameters such as voltage and current of the generator, and recording of the operating status and environmental conditions of the generator;

[0080] Establishing mathematical model: Establishing mathematical model of generator according to its physical characteristics and electrical principles;

[0081] Select the identification algorithm: Select the genetic algorithm;

[0082] Initial parameter estimation: Based on the design parameters or historical data of the generator, the parameters in the model are initially estimated. The accuracy of the initial estimation will affect the convergence speed and accuracy of subsequent identification.

[0083] Online identification: Input the real-time collected data into the genetic algorithm for online parameter identification. During the identification process, the genetic algorithm continuously adjusts the parameter values ​​in the model according to the changes in the data;

[0084] Parameter correction and update: According to the identification results, the parameters in the mathematical model are corrected and updated. The corrected parameters will be used in the subsequent monitoring and diagnosis process.

[0085] In one embodiment, for the above-mentioned establishment of a mathematical model, the specific steps of establishing the mathematical model are:

[0086] Determine key parameters: According to the physical characteristics and electrical principles of the generator, determine the key parameters included in the mathematical model, which include: resistance: stator resistance, rotor resistance, etc., which reflect the loss generated when the current passes through the conductor; inductance: stator inductance, rotor inductance and mutual inductance, etc., which describe the storage and change of the magnetic field in the conductor; permanent magnet flux: represents the magnetic flux generated by the magnetic field generated by the permanent magnet in the conductor; other parameters: such as pole pair number, rotor position angular velocity, etc., are used to describe the operating status of the generator;

[0087] Establish mathematical model: Based on the above key parameters, establish the mathematical model of the generator, whose stator resistance is R s , the rotor resistance is R r , the stator inductance is L s , the rotor inductance is L r , the mutual inductance is M, the permanent magnet flux is λ f , then the voltage equation is expressed as Among them, u s and u r are the voltage vectors of the stator and rotor respectively, i s and i r are the current vectors of the stator and rotor, respectively, ss , sr , rr and λ rs They are stator self-inductance flux, stator and rotor mutual inductance flux, rotor self-inductance flux and rotor and stator mutual inductance flux respectively.

[0088] In one embodiment, for the above-mentioned online parameter identification using genetic algorithm, the specific steps of online parameter identification using genetic algorithm are:

[0089] Encoding: Encode the parameters to be identified into individuals (i.e. chromosomes) in the genetic algorithm. Each individual consists of a series of genes, and each gene corresponds to a parameter to be identified.

[0090] Initialize the population: randomly generate a certain number of individuals as the initial population, and the gene values ​​(i.e. parameter values) of these individuals are randomly selected within a certain range;

[0091] Fitness evaluation: Substitute each individual (i.e. a set of parameter values) into the mathematical model, calculate the difference between the output of the mathematical model and the actual data, and use this difference as the fitness value of the individual. The smaller the fitness value, the better the individual.

[0092] Selection: Select a certain number of excellent individuals as parents according to their fitness values ​​to generate the next generation of individuals;

[0093] Crossover: Perform a crossover operation on the parent individuals to generate new offspring individuals;

[0094] Mutation: Perform mutation operations on offspring individuals to introduce new gene combinations;

[0095] Iteration: Repeat fitness evaluation, selection, crossover, and mutation until the predetermined number of iterations is reached or the fitness value converges to a certain range;

[0096] Output optimal parameters: Select the individual with the smallest fitness value from the final population as the optimal parameter combination.

[0097] In one embodiment, for the above S5, in S5, using the flux observation model and the induced voltage calculation algorithm to calculate the induced voltage amplitude of each column of permanent magnets includes the following steps:

[0098] Establishment of flux observation model: The flux observation model comprehensively considers the permanent magnet flux and the inductor flux; permanent magnet flux: this is the flux generated by the permanent magnet itself, which is usually a constant, but will change during the demagnetization process; inductor flux: this is the flux generated by the stator current through the inductor, which is proportional to the stator current; the flux observation model is expressed as Φ = ΦPM + L·I, where Φ represents the total flux, ΦPM represents the permanent magnet flux, L represents the inductance, and I represents the stator current;

[0099] Calculation of induced voltage: According to Faraday's law of electromagnetic induction, the induced electromotive force is proportional to the rate of change of magnetic flux. Therefore, the induced voltage is obtained by calculating the rate of change of magnetic flux. The calculation formula of the induced voltage is E=-dΦ / dt. Substituting the magnetic flux observation model into the calculation formula of the induced voltage, we get: E=-d(ΦPM+L·I)dt=-dtdΦPM-Ld I dt.

[0100] Assume that the permanent magnet flux of a permanent magnet wind turbine is 0.1Wb, the inductance is 0.01H, the stator current is 10A, and the current increases uniformly at a rate of 1A / s. Then, the calculation results are:

[0101] Magnetic flux observation value:

[0102]

[0103] Induced voltage:

[0104]

[0105] In this example, the amplitude of the induced voltage is 0.01 V. If in actual monitoring, the amplitude of the induced voltage is significantly lower than this value and continues to decrease over time, it can be considered that the permanent magnet has been demagnetized.

[0106] In one embodiment, for the above S6, in S6, the demagnetization judgment logic is combined to determine whether the permanent magnet has demagnetization phenomenon, and the threshold is set according to the actual situation of the generator and historical data. The specific steps are:

[0107] Collect data: collect basic parameters of the generator, permanent magnet characteristics, operating conditions and environmental conditions;

[0108] Statistical analysis: Statistical analysis of the induced voltage in historical data to determine its distribution and change trend;

[0109] Setting an initial threshold: Based on the statistical analysis results, combined with the actual situation and operation experience of the generator, a threshold of the induced voltage is preliminarily set, and the preliminarily set threshold of the induced voltage reflects the initial signs of permanent magnet demagnetization;

[0110] Real-time monitoring: Real-time monitoring of the generator's operating data, especially the changes in induced voltage;

[0111] Support vector machine training: After collecting enough data, a model is trained using a support vector machine to predict the demagnetization of the permanent magnet. The input of the initial setting of an induced voltage threshold is the induced voltage, stator current and temperature, and the output is the demagnetization probability or demagnetization level;

[0112] Threshold adjustment: According to the prediction results of the support vector machine, the initial threshold is adjusted. If the prediction accuracy of the model is high, the threshold can be adjusted more strictly to improve the accuracy of demagnetization judgment;

[0113] Validation model: After adjusting the threshold, the support vector machine is validated and tested to ensure the accuracy and reliability of its prediction results;

[0114] Test threshold: Apply the adjusted threshold to actual operation to see whether it accurately determines the demagnetization of the permanent magnet;

[0115] Feedback and adjustment: Adjust and optimize the thresholds based on the results of verification and testing.

[0116] A permanent magnet wind turbine generator permanent magnet demagnetization online monitoring device used in a permanent magnet wind turbine generator permanent magnet demagnetization online monitoring method comprises:

[0117] Voltage transformer: used to monitor the voltage signal of the small parallel head of the generator winding to ensure that the measurement range, accuracy and response time of the voltage transformer meet the monitoring requirements.

[0118] Data acquisition system: Design a data acquisition system to collect the output signal of the voltage transformer in real time and convert it into a digital signal for processing. The data acquisition system should have the characteristics of high speed, high precision and low noise;

[0119] Communication module: Design a communication module to transmit the collected data to the host computer or remote monitoring system in real time to ensure the reliability and real-time performance of data transmission;

[0120] Sensor integration: Temperature sensors and humidity sensors are integrated on the voltage transformer to monitor ambient temperature and humidity in real time and provide data support for the environmental compensation model.

[0121] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for online monitoring of permanent magnet demagnetization of a permanent magnet wind turbine generator, characterized in that: The following steps are involved: S1: Install the voltage transformer, temperature sensor, humidity sensor, communication module and data acquisition system on the small parallel head of the generator winding, and ensure that the connection is correct; S2: collect the output signal of the voltage transformer, ambient temperature and humidity data in real time and perform preprocessing; S3: According to the monitoring results of the temperature sensor and the humidity sensor, an environmental compensation model is established to compensate the output of the voltage transformer; S4: According to the physical characteristics and electrical principles of the generator, a mathematical model of the generator is established, and the genetic algorithm is used for online parameter identification to identify and correct the parameters of the generator in real time; S5: Calculate the induced voltage amplitude of each column of permanent magnets using the flux observation model and the induced voltage calculation algorithm; S6: Combine the demagnetization judgment logic to determine whether the permanent magnet has demagnetization, set the threshold according to the actual situation and historical data of the generator, and display the monitoring results in real time on the host computer or remote monitoring system. When demagnetization of the permanent magnet is detected, a fault warning or alarm signal is issued.

2. The method for online monitoring of permanent magnet demagnetization of a permanent magnet wind turbine generator according to claim 1, characterized in that: In S3, the specific steps of establishing the environmental compensation model to compensate the output of the voltage transformer are: Data collection: collect the ambient temperature and humidity around the generator monitored by the temperature sensor and humidity sensor, and record the output signal of the voltage transformer; Establish an environmental compensation model: Environmental factors include temperature T and humidity H. Analyze the relationship between environmental factors and the output signal V of the voltage transformer and establish an environmental compensation model. The formula of the environmental compensation model is V comp =V raw +f(T,H), where V raw is the original output signal of the voltage transformer, V comp is the signal after environmental compensation, and f(T,H) is the compensation function, which represents the influence of environmental factors on the signal; Determine the compensation function: Determine the specific form of the compensation function f(T,H), f(T,H) = aT + bH + c, where a, b, c are coefficients to be determined, by collecting multiple sets of experimental data, the experimental data including temperature T, humidity H, the original output signal V of the voltage transformer raw and the real signal V true , use the least squares method to solve the coefficients a, b, c and obtain the compensation function f(T,H); Implementation of compensation: During real-time monitoring, the value of the compensation function f(T,H) is calculated based on the current ambient temperature and humidity, and the compensation value is added to the original output signal of the voltage transformer to obtain a signal after environmental compensation.

3. The method for online monitoring of permanent magnet demagnetization of a permanent magnet wind turbine generator according to claim 1, characterized in that: In S4, according to the physical characteristics and electrical principles of the generator, a mathematical model of the generator is established, and a genetic algorithm is used to perform online parameter identification, and the parameters of the generator are identified and corrected in real time, including the following steps: Data collection: real-time collection of electrical parameters of the generator, and recording of the operating status and environmental conditions of the generator; Establishing mathematical model: Establishing mathematical model of generator according to its physical characteristics and electrical principles; Select the identification algorithm: Select the genetic algorithm; Initial parameter estimation: Make an initial estimate of the parameters in the model based on the design parameters or historical data of the generator; Online identification: Input the real-time collected data into the genetic algorithm for online parameter identification. During the identification process, the genetic algorithm continuously adjusts the parameter values ​​in the model according to the changes in the data; Parameter correction and update: According to the identification results, the parameters in the mathematical model are corrected and updated. The corrected parameters will be used in the subsequent monitoring and diagnosis process.

4. The method for online monitoring demagnetization of permanent magnets of a permanent magnet wind turbine generator according to claim 3, characterized in that: Specific steps to establish a mathematical model: Determine key parameters: According to the physical characteristics and electrical principles of the generator, determine the key parameters included in the mathematical model, which include: resistance: reflects the loss generated when the current passes through the conductor; inductance: describes the storage and change of the magnetic field in the conductor; permanent magnet flux: represents the magnetic flux generated by the magnetic field generated by the permanent magnet in the conductor; other parameters: used to describe the operating status of the generator; Establish mathematical model: Based on the above key parameters, establish the mathematical model of the generator, whose stator resistance is R s , the rotor resistance is R r , the stator inductance is L s , the rotor inductance is L r , the mutual inductance is M, the permanent magnet flux is λ f , then the voltage equation is expressed as Among them, u s and u r are the voltage vectors of the stator and rotor respectively, i s and i r are the current vectors of the stator and rotor, respectively, ss , sr , rr and λ rs They are stator self-inductance flux, stator and rotor mutual inductance flux, rotor self-inductance flux and rotor and stator mutual inductance flux respectively.

5. The method for online monitoring demagnetization of permanent magnets of a permanent magnet wind turbine generator according to claim 3, characterized in that: Specific steps for online parameter identification using genetic algorithm: Encoding: Encode the parameters to be identified into individuals (i.e. chromosomes) in the genetic algorithm. Each individual consists of a series of genes, and each gene corresponds to a parameter to be identified. Initialize the population: randomly generate a certain number of individuals as the initial population, and the gene values ​​(i.e. parameter values) of these individuals are randomly selected within a certain range; Fitness evaluation: Substitute each individual (i.e. a set of parameter values) into the mathematical model, calculate the difference between the output of the mathematical model and the actual data, and use this difference as the fitness value of the individual. The smaller the fitness value, the better the individual. Selection: Select a certain number of excellent individuals as parents according to their fitness values ​​to generate the next generation of individuals; Crossover: Perform a crossover operation on the parent individuals to generate new offspring individuals; Mutation: Perform mutation operations on offspring individuals to introduce new gene combinations; Iteration: Repeat fitness evaluation, selection, crossover, and mutation until the predetermined number of iterations is reached or the fitness value converges to a certain range; Output optimal parameters: Select the individual with the smallest fitness value from the final population as the optimal parameter combination.

6. The method for online monitoring of permanent magnet demagnetization of a permanent magnet wind turbine generator according to claim 1, characterized in that: In S5, the induced voltage amplitude of each column of permanent magnets is calculated using the flux observation model and the induced voltage calculation algorithm, including the following steps: Establishment of flux observation model: The flux observation model comprehensively considers the permanent magnet flux and the inductor flux; permanent magnet flux: this is the flux generated by the permanent magnet itself, which is usually a constant, but will change during the demagnetization process; inductor flux: this is the flux generated by the stator current through the inductor, which is proportional to the stator current; the flux observation model is expressed as Φ = ΦPM + L·I, where Φ represents the total flux, ΦPM represents the permanent magnet flux, L represents the inductance, and I represents the stator current; Calculation of induced voltage: According to Faraday's law of electromagnetic induction, the induced electromotive force is proportional to the rate of change of magnetic flux. Therefore, the induced voltage is obtained by calculating the rate of change of magnetic flux. The calculation formula of the induced voltage is E=-dΦ / dt. Substituting the magnetic flux observation model into the calculation formula of the induced voltage, we get: E=-d(ΦPM+L·I)dt=-dtdΦPM-LdIdt.

7. The method for online monitoring demagnetization of permanent magnets of a permanent magnet wind turbine generator according to claim 1, characterized in that: In S6, the demagnetization judgment logic is combined to determine whether the permanent magnet has demagnetization phenomenon, and the threshold is set according to the actual situation of the generator and historical data. The specific steps are: Collect data: collect basic parameters of the generator, permanent magnet characteristics, operating conditions and environmental conditions; Statistical analysis: Statistical analysis of the induced voltage in historical data to determine its distribution and change trend; Setting an initial threshold: Based on the statistical analysis results, combined with the actual situation and operation experience of the generator, a threshold of the induced voltage is preliminarily set, and the preliminarily set threshold of the induced voltage reflects the initial signs of permanent magnet demagnetization; Real-time monitoring: Real-time monitoring of the generator's operating data, especially the changes in induced voltage; Support vector machine training: After collecting enough data, a model is trained using a support vector machine to predict the demagnetization of the permanent magnet. The input of the initial setting of an induced voltage threshold is the induced voltage, stator current and temperature, and the output is the demagnetization probability or demagnetization level; Threshold adjustment: adjust the initial threshold according to the prediction results of the support vector machine; Validation model: After adjusting the threshold, the support vector machine was validated and tested; Test threshold: Apply the adjusted threshold to actual operation to see whether it accurately determines the demagnetization of the permanent magnet; Feedback and adjustment: Adjust and optimize the thresholds based on the results of verification and testing.

8. The on-line monitoring device for permanent magnet demagnetization of a permanent magnet wind turbine generator used in the on-line monitoring method for permanent magnet demagnetization of a permanent magnet wind turbine generator according to claim 1 is characterized in that: include: Voltage transformer: used to monitor the voltage signal of the small parallel head of the generator winding. Data acquisition system: Design a data acquisition system to collect the output signal of the voltage transformer in real time and convert it into a digital signal for processing; Communication module: Design a communication module to transmit the collected data to the host computer or remote monitoring system in real time; Sensor integration: Temperature sensors and humidity sensors are integrated on the voltage transformer to monitor ambient temperature and humidity in real time and provide data support for the environmental compensation model.

Images