A wind turbine fault monitoring and diagnosis system

Abstract

The invention relates to the field of wind power generation monitoring, and discloses a wind turbine fault monitoring and diagnosis system, comprising: a parameter acquisition module, used for acquiring various parameters of the wind turbine during operation, the parameters including operating parameters and environmental parameters, the operating parameters including vibration data, winding temperature, power generation and fan speed, and the environmental parameters including wind speed, wind direction, air temperature and air pressure; a foresight analysis module, which acquires ideal values ​​corresponding to various parameters based on historical data, compares the ideal values ​​with real-time parameters, and acquires the fan operation coefficient according to the comparison result; a fault judgment module, which judges the fan operation state according to the result of whether the fan operation coefficient falls into a preset fluctuation range; and a diagnosis module, which marks the wind turbine with abnormal operation state as a fault according to the fan operation state determined by the fault judgment module, inputs various parameters of the wind turbine after the fault is marked into a neural network model, and outputs a predicted fault cause.

Description

Technical Field

[0001] The invention relates to the field of wind power generation monitoring, and in particular to a wind power generator fault monitoring and diagnosis system. Background Art

[0002] A wind turbine is a device that converts wind energy into electrical energy. Its working principle is to use the power of wind to drive the rotation of the generator rotor through the rotation of the wind rotor blades, thereby generating electrical energy. A wind turbine is mainly composed of a wind rotor, a generator, a tower, etc. The wind rotor is the component that receives wind energy and is composed of blades and a hub; the generator is the component that converts wind energy into electrical energy and is composed of a stator and a rotor; and the tower is the structure that supports the entire wind turbine generator set.

[0003] Wind power generation is a green energy source. It does not produce any pollutants during its operation. Wind energy is a renewable energy source. Therefore, wind power generation is becoming more and more popular. Wind power generation relies on wind power, so it is generally built in areas with rich wind resources. Areas with rich wind resources on land are generally at higher altitudes, and the height of the tower is also higher. The above objective conditions limit the frequency of manual maintenance. Therefore, sensors are installed at various nodes of wind turbines to detect various parameters of wind turbines. However, single parameters are affected by too many external factors, and false alarms are prone to occur, which cannot accurately reflect the actual situation of the wind turbine. False alarms will increase useless personnel maintenance costs, which does not meet actual operation needs. Summary of the invention

[0004] The purpose of the present invention is to provide a wind turbine fault monitoring and diagnosis system to solve the technical problem that sensors are installed at various nodes of the wind turbine to detect various parameters of the wind turbine, but single parameters are affected by too many external factors and are prone to false alarms, which cannot accurately reflect the actual situation of the wind turbine. False alarms will increase useless personnel maintenance costs and do not meet actual operating requirements.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A wind turbine fault monitoring and diagnosis system, comprising:

[0007] A parameter acquisition module is used to acquire various parameters of the wind turbine during operation, including operating parameters and environmental parameters. The operating parameters include vibration data, winding temperature, power generation and wind turbine speed, and the environmental parameters include wind speed, wind direction, temperature and air pressure;

[0008] The foresight analysis module obtains the ideal values ​​corresponding to the operating parameters and environmental parameters based on historical data, compares the ideal values ​​with the real-time parameters, and obtains the fan operating coefficient according to the comparison results;

[0009] A fault judgment module judges the operation status of the fan according to whether the fan operation coefficient falls into a preset fluctuation range;

[0010] The diagnosis module marks the wind turbine with abnormal operating status as a fault according to the wind turbine operating status result determined by the fault judgment module, inputs various parameters of the wind turbine with fault marking into the neural network model, and outputs the predicted fault cause.

[0011] As a further technical solution, the method for obtaining the ideal value is:

[0012] Obtain the change curve of each parameter in the operating parameters over time based on historical data;

[0013] Divide into several sub-areas according to each parameter in the environmental parameters;

[0014] Select the time corresponding to the minimum discrete degree of the operating parameters in each partition as the time point;

[0015] By formula:

[0016] ;

[0017] Get the ideal value of each parameter at the i-th time point ;

[0018] in, is the conversion function between the change amount and weight of the j-th influencing factor, obtained based on historical data; For the ideal value The change in the jth influencing factor of the corresponding parameter, m is the number of influencing factors of the parameter corresponding to the ideal value, It is the actual value of each parameter at the previous time point.

[0019] As a further technical solution, the process of obtaining the fan operation coefficient includes:

[0020] By formula:

[0021] ;

[0022] Calculate the operating coefficient of the Xth fan ;

[0023] in, is the actual value of the jth influencing factor of the parameter at the rth time point; is the duration between two time points, is the total number of time points, is the proportionality coefficient, which is selected and determined based on historical data and empirical data.

[0024] As a further technical solution, the process of judging the operation status of the fan according to whether the fan operation coefficient falls into the preset fluctuation range is as follows:

[0025] The fan operating coefficient will be obtained With the preset fluctuation range Compare;

[0026] like , then it is judged that the fan operation status is normal;

[0027] like , the fan is judged to be in abnormal operating state, the fan is shut down, and a maintenance task order is generated and sent to the system backend to assign maintenance personnel to repair the fan;

[0028] like , it is judged that the fan is in good operating condition.

[0029] As a further technical solution, the minimum discrete degree of the operating parameters within each partition is determined as follows:

[0030] By formula: ;

[0031] Calculate the coefficient of dispersion ;

[0032] in Indicates the actual value of the i-th parameter in the operating parameters. is the weight coefficient corresponding to the i-th parameter, which is determined based on historical data and empirical data. is the total number of operating parameters, is the average value;

[0033] All discrete coefficients calculated within the current partition are Arranged in ascending order, the time corresponding to the first discrete coefficient is the time point of the minimum discrete degree of the operating parameter.

[0034] As a further technical solution, the system also includes a fault trend analysis module, which further analyzes the fan in normal operation to obtain a fault trend coefficient, and adjusts the frequency of random maintenance of the current fan according to the fault trend coefficient.

[0035] As a further technical solution, the method for obtaining the fault trend coefficient is:

[0036] By formula:

[0037] ;

[0038] Calculate the failure trend coefficient of the xth fan ;

[0039] in, is the conversion coefficient, which is determined based on historical data and experimental data. is the deviation coefficient of the power generated by the x-th wind turbine, It is a preset proportional coefficient, which is selected and determined based on historical data. The preset fluctuation range The endpoint value of .

[0040] As a further technical solution, the deviation coefficient of the current wind turbine power generation The method to obtain is:

[0041] Obtain the curve of wind turbine power generation over time based on historical data normalization fitting And the curve of standard power generation over time ;

[0042] By formula:

[0043] ;

[0044] Calculate the deviation coefficient of the current wind turbine power generation ;

[0045] in, are the two end points of the time period, is the attenuation coefficient, which can be obtained by looking up the table based on empirical data.

[0046] As a further technical solution, according to the failure tendency coefficient The process of adjusting the frequency of random maintenance is:

[0047] Failure tendency factor With preset threshold Compare;

[0048] like , then according to Adjust the frequency of random maintenance of the fan, where: The preset random maintenance frequency;

[0049] like , the current fan random maintenance frequency remains unchanged.

[0050] Beneficial effects of the present invention:

[0051] (1) By classifying the operating parameters and environmental parameters of the wind turbine during power generation, and obtaining the ideal values ​​of the operating parameters and environmental parameters according to the foresight analysis module, the wind turbine operating coefficient is obtained by comparing the ideal values ​​with the real-time parameters. The higher the wind turbine operating coefficient, the more unstable the parameters of the wind turbine in an operating cycle. Therefore, the wind turbine fault can be discovered in time within an operating cycle to ensure the normal operation of the wind turbine.

[0052] (2) The fault trend analysis module is used to further predict and analyze the wind turbine in normal operation to reflect the actual state of the wind turbine and the fault trend coefficient. The smaller the value, the higher the failure risk of the wind turbine, so as to foresee the failure of the normally operating wind turbine. At the same time, a comprehensive model is established by integrating the deviation coefficient of the wind turbine power generation and the wind turbine operation coefficient. The change of the wind turbine power generation can be used to assist in verifying the change of the wind turbine operation coefficient, thereby improving the accuracy of the fault trend coefficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The present invention will be further described below in conjunction with the accompanying drawings.

[0054] Figure 1 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0055] 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.

[0056] See also Figure 1 As shown, the present invention is a wind turbine fault monitoring and diagnosis system, comprising:

[0057] A parameter acquisition module is used to acquire various parameters of the wind turbine during operation, including operating parameters and environmental parameters. The operating parameters include vibration data, winding temperature, power generation and wind turbine speed, and the environmental parameters include wind speed, wind direction, temperature and air pressure;

[0058] The foresight analysis module obtains the ideal values ​​corresponding to the operating parameters and environmental parameters based on historical data, compares the ideal values ​​with the real-time parameters, and obtains the fan operating coefficient according to the comparison results;

[0059] A fault judgment module judges the operation status of the fan according to whether the fan operation coefficient falls into a preset fluctuation range;

[0060] The diagnosis module marks the wind turbine with abnormal operating status as a fault according to the wind turbine operating status result determined by the fault judgment module, inputs various parameters of the wind turbine with fault marking into the neural network model, and outputs the predicted fault cause.

[0061] In this embodiment, the operating parameters and environmental parameters of the wind turbine itself during power generation are classified, and the ideal values ​​of the operating parameters and environmental parameters are obtained according to the predictive analysis module, so that the wind turbine operating coefficient is obtained after comparing the ideal value with the real-time parameters. The higher the wind turbine operating coefficient, the more unstable the various parameters of the wind turbine in an operating cycle, so that the wind turbine fault can be discovered in time within an operating cycle to ensure the normal operation of the wind turbine; at the same time, the fault judgment module is used to judge the operating status of the wind turbine, and the wind turbine with abnormal operating status is marked as faulty through the diagnosis module, and the various parameters of the wind turbine after fault marking are input into the neural network model to output the predicted cause of the fault; it should be noted that the neural network model is a prior art and will not be elaborated here.

[0062] The method for obtaining the ideal value is:

[0063] Obtain the change curve of each parameter in the operating parameters over time based on historical data;

[0064] Divide into several sub-areas according to each parameter in the environmental parameters;

[0065] Select the time corresponding to the minimum discrete degree of the operating parameters in each partition as the time point;

[0066] By formula:

[0067] ;

[0068] Get the ideal value of each parameter at the i-th time point ;

[0069] in, is the conversion function between the change amount and weight of the j-th influencing factor, obtained based on historical data; For the ideal value The change in the jth influencing factor of the corresponding parameter, m is the number of influencing factors of the parameter corresponding to the ideal value, It is the actual value of each parameter at the previous time point.

[0070] In this embodiment, a specific method for obtaining an ideal value is provided. Specifically, based on historical data, a time curve of each parameter in the wind turbine's own operating parameters is fitted through normalization, and then each parameter in the environmental parameters is divided into intervals. For example, the environmental parameters can be divided into intervals with rich wind resources and poor wind resources by season or month, and different intervals of wind strength can be divided by day or hour. Other wind directions, temperatures, etc. are similar to wind speeds. The time when the discrete degree of each numerical value in the operating parameters in each interval is the smallest is used as the time point, which can be more representative and reduce the numerical error caused by the fixed selection of time points. The obtained ideal value can predict and analyze the operating status of the wind turbine in advance, and improve the monitoring efficiency and accuracy of the operating status of the wind turbine. At the same time, the rationality of the ideal value can be guaranteed by comprehensive analysis of the weights converted from the influencing factors corresponding to each parameter.

[0071] For example, when the vibration parameters in the operating parameters change, the corresponding influencing factors are different due to the different environmental parameters.

[0072] The process of obtaining the fan operating coefficient includes:

[0073] By formula:

[0074] ;

[0075] Calculate the operating coefficient of the Xth fan ;

[0076] in, is the actual value of the jth influencing factor of the parameter at the rth time point; is the duration between two time points, is the total number of time points, is the proportionality coefficient, which is selected and determined based on historical data and empirical data.

[0077] The process of judging the fan operating status according to whether the fan operating coefficient falls into the preset fluctuation range is as follows:

[0078] The fan operating coefficient will be obtained With the preset fluctuation range Compare;

[0079] like , then it is judged that the fan operation status is normal;

[0080] like , the fan is judged to be in abnormal operating state, the fan is shut down, and a maintenance task order is generated and sent to the system backend to assign maintenance personnel to repair the fan;

[0081] like , it is judged that the fan is in good operating condition.

[0082] In this embodiment, the fan operation coefficient obtained by comparing the ideal value with the actual value of each parameter can reflect the parameter fluctuation of the fan operation state based on the ideal value and the actual value of the influencing factor, thereby improving the accuracy of the fan operation state judgment.

[0083] The minimum discrete degree of the operating parameters within each partition is determined as follows:

[0084] By formula: ;

[0085] Calculate the coefficient of dispersion ;

[0086] in Indicates the actual value of the i-th parameter in the operating parameters. is the weight coefficient corresponding to the i-th parameter, which is determined based on historical data and empirical data. is the total number of operating parameters;

[0087] All discrete coefficients calculated within the current partition are Arranged in ascending order, the time corresponding to the first discrete coefficient is the time point of the minimum discrete degree of the operating parameter.

[0088] In this embodiment, a method for determining the minimum discrete degree of the operating parameters between each partition is provided. Specifically, the formula is: ; Calculate the coefficient of dispersion , the minimum discrete coefficient is obtained by sorting, and the corresponding time is the time point, according to each parameter and the average value The difference between and can more accurately find the minimum discrete coefficient.

[0089] The system also includes a fault trend analysis module, which further analyzes the fan in a normal operating state to obtain a fault trend coefficient, and adjusts the frequency of random maintenance of the current fan according to the fault trend coefficient.

[0090] The method for obtaining the fault trend coefficient is:

[0091] By formula:

[0092] ;

[0093] Calculate the failure trend coefficient of the xth fan ;

[0094] in, is the conversion coefficient, which is determined based on historical data and experimental data. is the deviation coefficient of the power generated by the x-th wind turbine, It is a preset proportional coefficient, which is selected and determined based on historical data. The preset fluctuation range The endpoint value of .

[0095] In this embodiment, the fault trend analysis module is used to further predict and analyze the wind turbine in normal operation to reflect the real state of the wind turbine. The smaller the value, the higher the failure risk of the wind turbine, so as to foresee the failure of the normally operating wind turbine. At the same time, a comprehensive model is established by integrating the deviation coefficient of the wind turbine power generation and the wind turbine operation coefficient. The change of the wind turbine power generation can be used to assist in verifying the change of the wind turbine operation coefficient, thereby improving the accuracy of the fault trend coefficient.

[0096] Deviation coefficient of current wind turbine power generation The method to obtain is:

[0097] Obtain the curve of wind turbine power generation over time based on historical data normalization fitting And the curve of standard power generation over time ;

[0098] By formula:

[0099] ;

[0100] Calculate the deviation coefficient of the current wind turbine power generation ;

[0101] in, are the two end points of the time period, is the attenuation coefficient, which can be obtained by looking up the table based on empirical data.

[0102] In this embodiment, a specific method for obtaining the deviation coefficient of wind turbine power generation is provided. Specifically, the wind turbine power generation variation curve over time obtained through historical data is Curve of standard power generation over time By comparison, the formula You can calculate Obviously, the greater the difference between the real-time power generation and the standard reference power generation, the greater the power generation deviation coefficient, which means the power generation stability of the wind turbine is worse. Therefore, it can help verify that the wind turbine is more likely to fail during subsequent operation.

[0103] According to the failure tendency factor The process of adjusting the frequency of random maintenance is:

[0104] Failure tendency factor With preset threshold Compare;

[0105] like , then according to Adjust the frequency of random maintenance of the fan, where: The preset random maintenance frequency;

[0106] like , the current fan random maintenance frequency remains unchanged.

[0107] In this embodiment, the fault trend coefficient is analyzed to adjust the frequency of random maintenance of existing fans. Obviously, if the fault trend coefficient , it indicates that the possibility of failure of the fan is higher, so it is necessary to follow Increase the frequency of random maintenance of the fan, including: is the preset random maintenance frequency, which is obtained by looking up the table according to empirical data; otherwise, the existing random maintenance frequency is kept unchanged; through the above process, the random maintenance frequency of the fan can be reasonably adjusted, and the focus can always be on the fan with a high possibility of failure, avoiding the risk of fan shutdown.

[0108] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A wind turbine fault monitoring and diagnosis system, characterized in that: include: A parameter acquisition module is used to acquire various parameters of the wind turbine during operation, including operating parameters and environmental parameters. The operating parameters include vibration data, winding temperature, power generation and wind turbine speed, and the environmental parameters include wind speed, wind direction, temperature and air pressure; The foresight analysis module obtains the ideal values ​​corresponding to the operating parameters and environmental parameters based on historical data, compares the ideal values ​​with the real-time parameters, and obtains the fan operating coefficient according to the comparison results; A fault judgment module judges the operation status of the fan according to whether the fan operation coefficient falls into a preset fluctuation range; A diagnosis module, which marks the wind turbine with abnormal operating status as a fault according to the wind turbine operating status result determined by the fault judgment module, inputs various parameters of the wind turbine with fault marking into the neural network model, and outputs a predicted fault cause; The method for obtaining the ideal value is: Obtain the change curve of each parameter in the operating parameters over time based on historical data; Divide into several sub-areas according to each parameter in the environmental parameters; Select the time corresponding to the minimum discrete degree of the operating parameters in each partition as the time point; By formula: ； Get the ideal value of each parameter at the i-th time point ; in, is the conversion function between the change amount and weight of the j-th influencing factor, obtained based on historical data; For the ideal value The change in the jth influencing factor of the corresponding parameter, m is the number of influencing factors of the parameter corresponding to the ideal value, is the actual value of each parameter at the previous time point; The process of obtaining the fan operating coefficient includes: By formula: ； Calculate the operating coefficient of the Xth fan ; in, is the actual value of the jth influencing factor of the parameter at the rth time point; is the duration between two time points, is the total number of time points, is the proportionality coefficient, which is selected and determined based on historical data and empirical data; The minimum discrete degree of the operating parameters within each partition is determined as follows: By formula: ; Calculate the coefficient of dispersion ; in Indicates the actual value of the i-th parameter in the operating parameters. is the weight coefficient corresponding to the i-th parameter, and the weight coefficient is selected and determined based on historical data and empirical data. is the total number of operating parameters, is the average value; All discrete coefficients calculated within the current partition are Arranged in ascending order, the time corresponding to the first discrete coefficient is the time point of the minimum discrete degree of the operating parameter.

2. The wind turbine fault monitoring and diagnosis system according to claim 1, characterized in that: The process of judging the fan operating status according to whether the fan operating coefficient falls into the preset fluctuation range is as follows: The fan operating coefficient will be obtained With the preset fluctuation range Compare; like , then it is judged that the fan operation status is normal; like , the fan is judged to be in abnormal operating state, the fan is shut down, and a maintenance task order is generated and sent to the system backend to assign maintenance personnel to repair the fan; like , then it is judged that the fan is in good operating condition.

3. The wind turbine fault monitoring and diagnosis system according to claim 2, characterized in that: The system also includes a fault trend analysis module, which further analyzes the fan in a normal operating state to obtain a fault trend coefficient, and adjusts the frequency of random maintenance of the current fan according to the fault trend coefficient.

4. The wind turbine fault monitoring and diagnosis system according to claim 3, characterized in that: The method for obtaining the fault trend coefficient is: By formula: ； Calculate the failure trend coefficient of the xth fan ; in, is the conversion coefficient, which is determined based on historical data and experimental data. is the deviation coefficient of the power generated by the x-th wind turbine, It is a preset proportional coefficient, which is selected and determined based on historical data. The preset fluctuation range The endpoint value of .

5. The wind turbine fault monitoring and diagnosis system according to claim 4, characterized in that: Deviation coefficient of current wind turbine power generation The method to obtain is: Obtain the curve of wind turbine power generation over time based on historical data normalization fitting And the curve of standard power generation over time ; By formula: ； Calculate the deviation coefficient of the current wind turbine power generation ; in, are the two end points of the time period, is the attenuation coefficient, which can be obtained by looking up the table based on empirical data.

6. The wind turbine fault monitoring and diagnosis system according to claim 5, characterized in that: According to the failure tendency factor The process of adjusting the frequency of random maintenance is: Failure tendency factor With preset threshold Compare; like , then according to Adjust the frequency of random maintenance of the fan, where: The preset random maintenance frequency; like , the current fan random maintenance frequency remains unchanged.

Images