Wind power plant performance analysis and evaluation method based on big data

Abstract

The invention discloses a wind power plant performance analysis and evaluation method based on big data, and the method comprises the steps: evaluating the performance of a fan from multiple dimensions in a performance analysis stage of the fan, reading the operation parameter information of each line in a data matrix, and calculating a stability score and a performance score for power output data, the performance of the fan in the aspect of power output is measured from different angles, meanwhile, power factor data are analyzed to obtain a conversion score, finally, the stability score, the expressive score and the conversion score are subjected to normalization processing, and then weighted calculation is performed to obtain a fan performance index. The multi-dimensional comprehensive analysis overcomes the defect that the performance evaluation index of the wind power plant fan is single in the prior art.

Description

Technical Field

[0001] The present invention relates to the technical field of wind farms, and in particular to a wind farm performance analysis and evaluation method based on big data. Background Art

[0002] As the global demand for clean energy continues to grow, the scale and number of wind farms are expanding. The performance of wind farms directly affects their power generation efficiency, economic benefits and contribution to the stability of the power grid.

[0003] However, the existing technologies for wind farm performance evaluation still have the following deficiencies:

[0004] The performance evaluation is singular. When evaluating the performance of wind turbines in a wind farm, the operating data of each wind turbine in the wind farm is not effectively integrated. Most of them are isolated evaluations of the current performance of the wind turbines, lacking a comprehensive analysis of the historical data of the wind turbines.

[0005] At the same time, the existing evaluation methods can only evaluate the performance of the fan, and cannot further mine and analyze the operating data of the fan when the fan performance is abnormal, so as to determine the cause of the abnormal fan performance.

[0006] Therefore, a wind farm performance analysis and evaluation method based on big data is proposed. Summary of the invention

[0007] In view of this, the present invention provides a wind farm performance analysis and evaluation method based on big data to solve the problems raised by the above background technology.

[0008] The purpose of the present invention can be achieved by the following technical solution: A wind farm performance analysis and evaluation method based on big data, comprising:

[0009] Data collection: preset the time interval for performance evaluation. After the preset time interval is reached, collect the operating parameters of each wind turbine in the wind farm within the set time interval, and pre-process the operating parameters of each wind turbine within the set time interval. After pre-processing, organize the operating parameters of each wind turbine within the set time interval into a data matrix, and each row in the data matrix represents the operating parameter information of the corresponding wind turbine within the set time interval; the operating parameters include wind turbine power output data, wind turbine power factor data, wind turbine speed, fan temperature and fan vibration amplitude;

[0010] Performance analysis: read the operating parameter information of each row in the data matrix, and conduct a comprehensive analysis to obtain the wind turbine performance index paex of each wind turbine within the set time interval; where x represents the number of each wind turbine in the wind farm, x=1, 2,..., y; where y represents the total number of wind turbines in the wind farm;

[0011] Performance evaluation: Based on the wind turbine performance index paex of each wind turbine within the set time interval, the performance level of each wind turbine is determined, and the determined performance level results of each wind turbine are added to the corresponding rows of operating parameter information in the data matrix. The data matrix with the added performance level results is used as the preliminary evaluation result of the wind farm within the set time interval; the performance levels include low efficiency level, normal level, good level and high efficiency level;

[0012] Historical synthesis: Based on the wind turbine performance index paex of each wind turbine at the current time point within the set time interval, H groups of historical data matrices before the current time point are extracted, where H>8. The H groups of historical data matrices are analyzed and evaluated to obtain the performance trend change status of each wind turbine at the current time point. After adding the performance trend change status of each wind turbine to each row of operating parameter information in the data matrix, the data matrix after adding the performance trend change status is used as the final evaluation result of the wind farm within the set time interval.

[0013] In some embodiments, the operating parameter information of each row in the data matrix is ​​read and a comprehensive analysis is performed, specifically:

[0014] Read the power output data in each row of the operating parameter information in the data matrix, extract the power output data at each time point in the set time interval, and form a power output time series of each wind turbine in the set time interval;

[0015] The standard deviation formula is used to calculate the power output time series of each wind turbine within the set time interval, and the calculation result is used as the power dispersion index of each wind turbine within the set time interval;

[0016] A reference threshold value of the power dispersion index corresponding to each wind turbine is preset, and the power dispersion index of each wind turbine within a set time interval is calculated to be proportional to the corresponding reference threshold value to obtain the stability score uqx of each wind turbine; that is, it is calculated by power dispersion index / corresponding reference threshold value;

[0017] For the power output time series of each wind turbine within the set time interval, the average value is calculated as the power average value of each wind turbine within the set time interval;

[0018] Extract the rated power corresponding to each wind turbine, calculate the ratio between the average power of each wind turbine and the corresponding rated power, and obtain the performance score uwx of each wind turbine; that is, it is calculated by the average power / corresponding rated power;

[0019] Read the power factor data in each row of the operating parameter information in the data matrix, extract the power factor data at each time point in the set time interval, and form a power factor time series of each wind turbine in the set time interval;

[0020] For the power factor time series of each wind turbine within the set time interval, the average value is calculated as the power factor average value of each wind turbine within the set time interval;

[0021] Preset each group of standard ranges for the average power factor value corresponding to each wind turbine, set each group of standard range to correspond to an electric energy conversion index; match the average power factor value of each wind turbine with the corresponding standard range to obtain the electric energy conversion index corresponding to each wind turbine, and use the electric energy conversion index corresponding to each wind turbine as the transformation score utx of each wind turbine within the set time interval.

[0022] In some embodiments, the fan performance index paex of each fan within a set time interval is obtained, specifically:

[0023] After normalizing the stability score uqx, performance score uwx and conversion score utx of each wind turbine within the set time interval, substitute them into the formula A weighted calculation is performed to obtain the fan performance index paex of each wind turbine within the set time interval; wherein a1, a2 and a3 are the influencing weight factors of the stability score uqx, the performance score uwx and the conversion score utx respectively.

[0024] In some embodiments, the performance level of each wind turbine is determined based on the wind turbine performance index paex of each wind turbine within a set time interval, specifically:

[0025] A reference threshold index of the fan performance index paex corresponding to each fan is preset, and the fan performance index paex of each fan within a set time interval is compared with the preset reference threshold index. If the fan performance index paex of a certain fan within the set time interval is lower than the corresponding preset reference threshold index, the performance level of the fan is determined to be a low-efficiency level and a retrieval signaling is triggered;

[0026] If the fan performance index paex of a certain wind turbine in the set time interval is higher than the preset reference threshold index, the difference between the risk performance index paex and the reference threshold index is further calculated as the performance difference, and three groups of difference value ranges corresponding to the preset performance difference are set. The three groups of difference value ranges are set to correspond to the normal level, good level and high-efficiency level respectively. The calculated performance difference is matched with the corresponding preset difference value range to obtain the performance level of the wind turbine in the set time interval.

[0027] In some embodiments, the H groups of historical data matrices are parsed and evaluated, specifically:

[0028] Analyze the H group historical data matrix, extract the H group historical wind turbine performance index of each wind turbine, and integrate it with the wind turbine performance index paex of each wind turbine at the current time point, so as to obtain the performance index sequence of each wind turbine;

[0029] A moving average window size is preset, and the performance index sequence of each wind turbine is calculated according to the preset window size to obtain each group of moving average values ​​corresponding to each wind turbine;

[0030] For each group of moving average values ​​of each wind turbine, the change rate between two adjacent groups of moving average values ​​is calculated, that is, by the formula Calculate and get and Represents two adjacent groups of moving averages;

[0031] If the change rate of the moving averages of two adjacent groups is positive, it is marked as a performance improvement value. If the change rate of the moving averages of two adjacent groups is negative, the absolute value is taken and marked as a performance degradation value.

[0032] After arranging the change rates of each group of wind turbines in chronological order, they are divided from the middle point of the arrangement to obtain the front time zone change rate set and the rear time zone change rate set of each wind turbine.

[0033] In some embodiments, the performance trend change state of each wind turbine at the current time point is obtained, specifically:

[0034] The performance increase value and performance decrease value of each group in the previous time zone change rate set are summed up respectively to obtain the total increase value and the total decrease value of the previous time zone, and the total increase value of the previous time zone is subtracted from the total decrease value of the previous time zone to obtain the previous time zone trend evaluation value. If the previous time zone trend evaluation value is a positive value, it means that the performance state of the fan in the previous time zone is an upward trend. If the previous time zone trend evaluation value is a negative value, it means that the performance state of the fan in the previous time zone is a downward trend.

[0035] The performance increase value and performance decrease value of each group in the rear time zone change rate set are summed up respectively to obtain the rear time zone increase total value and the rear time zone decrease total value, and the rear time zone trend evaluation value is obtained by subtracting the rear time zone decrease total value from the rear time zone increase total value. If the rear time zone trend evaluation value is a positive value, it means that the performance state of the fan in the rear time zone is an upward trend. If the rear time zone trend evaluation value is a negative value, it means that the performance state of the fan in the rear time zone is a downward trend.

[0036] If the performance status of a certain wind turbine in the previous time zone and the next time zone is an upward trend, then the performance trend change status of the wind turbine at the current time point is an overall performance upward trend;

[0037] If the performance status of a certain wind turbine in both the previous time zone and the next time zone is a downward trend, then the performance trend change status of the wind turbine at the current time point is an overall performance downward trend;

[0038] If the performance status of a certain wind turbine in the previous time zone and the next time zone is an upward trend and a downward trend respectively, then the performance trend change status of the wind turbine at the current time point is an overall performance fluctuation trend, and the upward trend changes to a downward trend;

[0039] If the performance status of a wind turbine in the previous time zone and the next time zone is a downward trend and an upward trend respectively, then the performance trend change status of the wind turbine at the current time point is an overall performance fluctuation trend, and the downward trend changes to an upward trend.

[0040] In some embodiments, it also includes:

[0041] Trigger retrieval: When the retrieval signal is triggered, the corresponding wind turbine that triggers the retrieval signal is marked as an abnormal wind turbine, and corresponding steps are executed to determine the triggering reason of the abnormal wind triggering the retrieval signal.

[0042] In some embodiments, corresponding steps are performed to determine the triggering cause of the abnormal wind turbine triggering retrieval signaling, specifically:

[0043] M1: Read the operating parameter information of the row corresponding to the abnormal fan in the data matrix, extract the fan speed of the abnormal fan within the set time interval, and compare the fan speed of the abnormal fan within the set time interval with the set minimum speed value. If the fan speed of the abnormal fan at a certain time point within the set time interval is lower than the set minimum speed and the duration reaches the set time, it is determined that the triggering reason for triggering the retrieval signaling is the abnormal low speed. Otherwise, execute step M2;

[0044] M2: Read the operating parameter information of the row corresponding to the abnormal fan in the data matrix, extract the fan temperature and fan vibration amplitude of the abnormal fan within the set time interval, and perform comprehensive analysis to obtain the health assessment index jkex of the abnormal fan within the set time interval, set the reference threshold index of the health assessment index jkex corresponding to each fan, if the health assessment index jkex of the abnormal fan within the set time interval is higher than the set reference threshold index, then determine that the triggering cause of the triggering retrieval signaling is the abnormal health of the fan, otherwise execute step M3;

[0045] M2-201: extract the fan temperature of the abnormal fan at each time point within the set time interval and calculate the average value, and use the calculated average value as the fan average temperature value of the abnormal fan; set the reference temperature value of the fan corresponding to the fan temperature, calculate the ratio of the fan average temperature value to the set reference temperature value, and obtain the temperature evaluation value wcrx of the abnormal fan; that is, calculate it by dividing the fan average temperature value by the reference temperature value;

[0046] M2-202: Set the standard of vibration events, i.e., the reference threshold corresponding to the vibration amplitude and the reference duration higher than the reference threshold, extract the vibration amplitude of the abnormal fan within the set time interval, and based on the set vibration event standard, count the number of vibration events that occur in the abnormal fan within the set time interval as the number of vibrations of the abnormal fan, set the allowable number of vibrations corresponding to the number of vibrations of the abnormal fan, calculate the ratio of the number of vibrations to the set allowable number of vibrations, and obtain the vibration evaluation value wfrx of the abnormal level; i.e., the number of vibrations / the set allowable number of vibrations;

[0047] M2-203: preset the weight coefficients corresponding to the temperature evaluation value wcrx and the vibration evaluation value wfrx, multiply the abnormal level temperature evaluation value wcrx and the vibration evaluation value wfrx by the corresponding weight coefficients, and then sum them up to obtain the health evaluation index jkex of the abnormal fan within the set time interval;

[0048] M3: Taking the location of the abnormal fan as the target point, select the maintenance personnel with the shortest distance to the target point as the search personnel, and send the location of the abnormal fan to the search personnel. The search personnel go to the location of the abnormal fan and determine the cause of triggering the search signal, and store the determination result.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The present invention evaluates the performance of the wind turbine from multiple dimensions during the performance analysis phase of the wind turbine, reads the operating parameter information of each row in the data matrix, calculates the stability score and the performance score for the power output data, and measures the performance of the wind turbine power output from different angles. At the same time, the power factor data is analyzed to obtain the conversion score. Finally, the stability score, the performance score and the conversion score are normalized and weighted to obtain the wind turbine performance index. This multi-dimensional comprehensive analysis overcomes the deficiency of the prior art in that the wind farm wind turbine performance evaluation index is single.

[0051] The present invention extracts H groups of historical data matrices before the current time point based on the fan performance index of each fan at the current time point, analyzes these historical data matrices, extracts the historical fan performance index of each fan, and integrates it with the current performance index to obtain a performance index sequence, thereby combining historical data with current data, providing a data basis for analyzing the trend change of fan performance, realizing dynamic trend analysis of fan performance based on historical data, and making up for the defect of the prior art lacking comprehensive analysis of historical data;

[0052] When the fan performance is abnormal, the present invention triggers a retrieval signal, marks the corresponding fan as an abnormal fan, and executes different steps to further mine and analyze the operating parameters of the fan, and determines the cause of the abnormal performance of the abnormal fan based on the analysis results, thereby solving the problem in the prior art that the performance of the fan can only be evaluated, but the operating data of the fan cannot be further mined and analyzed when the fan performance is abnormal. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Further details, features and advantages of the present application are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which:

[0054] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0055] Several embodiments of the present application will be described in more detail below with reference to the accompanying drawings so that those skilled in the art can implement the present application. The present application can be embodied in many different forms and purposes and should not be limited to the embodiments described herein. These embodiments are provided to make the present application comprehensive and complete, and to fully convey the scope of the present application to those skilled in the art. The embodiments do not limit the present application.

[0056] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. It will be further understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the relevant art and / or the context of this specification, and will not be interpreted in an idealized or overly formal sense unless explicitly defined as such herein.

[0057] See also Figure 1 As shown in the figure, the wind farm performance analysis and evaluation method based on big data includes:

[0058] Data collection: Utilize various sensors and monitoring equipment in the wind farm, preset the time interval for performance evaluation, collect the operating parameters of each wind turbine in the wind farm within the set time interval after the preset time interval is reached, and pre-process the operating parameters of each wind turbine within the set time interval; pre-processing includes denoising, outlier processing, etc. to ensure the accuracy and integrity of the data; after pre-processing, organize the operating parameters of each wind turbine within the set time interval into a data matrix, and each row in the data matrix represents the operating parameter information of the corresponding wind turbine within the set time interval; the operating parameters include wind turbine power output data, wind turbine power factor data, wind turbine speed, wind turbine temperature and wind turbine vibration amplitude;

[0059] The digital matrix facilitates the subsequent unified analysis and processing of the operation status of each fan;

[0060] Performance analysis: read the operating parameter information of each row in the data matrix, and conduct a comprehensive analysis to obtain the wind turbine performance index paex of each wind turbine within the set time interval; where x represents the number of each wind turbine in the wind farm, x=1, 2,..., y; where y represents the total number of wind turbines in the wind farm;

[0061] Specifically:

[0062] Read the power output data in each row of the operating parameter information in the data matrix, extract the power output data at each time point in the set time interval, and form a power output time series of each wind turbine in the set time interval;

[0063] The standard deviation formula is used to calculate the power output time series of each wind turbine within the set time interval, and the calculation result is used as the power dispersion index of each wind turbine within the set time interval;

[0064] The technicians preset the reference threshold of the power dispersion index corresponding to each wind turbine according to the type of each wind turbine, historical operation data and industry standards;

[0065] The power dispersion index of each wind turbine within the set time interval is calculated by ratio with the corresponding reference threshold value to obtain the stability score uqx of each wind turbine; that is, it is calculated by power dispersion index / corresponding reference threshold value;

[0066] It should be noted that the smaller the stability score, the more stable the power output of the fan is, and vice versa, the power output of the fan fluctuates greatly.

[0067] For the power output time series of each wind turbine within the set time interval, the average value is calculated as the power average value of each wind turbine within the set time interval;

[0068] Extract the rated power corresponding to each wind turbine, calculate the ratio between the average power of each wind turbine and the corresponding rated power, and obtain the performance score uwx of each wind turbine; that is, it is calculated by the average power / corresponding rated power;

[0069] It should be noted that the higher the performance score, the better the power output performance of the wind turbine, and vice versa, it means that the power output performance of the wind turbine is poor;

[0070] Read the power factor data in each row of the operating parameter information in the data matrix, extract the power factor data at each time point in the set time interval, and form a power factor time series of each wind turbine in the set time interval;

[0071] For the power factor time series of each wind turbine within the set time interval, the average value is calculated as the power factor average value of each wind turbine within the set time interval;

[0072] Preset each group of standard ranges of the average power factor value corresponding to each wind turbine, and set each group of standard ranges to correspond to an electric energy conversion index; the electric energy conversion index range is set at 0.63-0.98, and the higher the average power factor value, the higher the corresponding matched electric energy conversion index; match the average power factor value of each wind turbine with the corresponding standard range to obtain the electric energy conversion index corresponding to each wind turbine, and use the electric energy conversion index corresponding to each wind turbine as the transformation score utx of each wind turbine within the set time interval;

[0073] After normalizing the stability score uqx, performance score uwx and conversion score utx of each wind turbine within the set time interval, substitute them into the formula A weighted calculation is performed to obtain the fan performance index paex of each fan within the set time interval; a1, a2 and a3 are the influence weight factors of the stability score uqx, the performance score uwx and the conversion score utx respectively;

[0074] It should be noted that, through the above analysis process, a relatively comprehensive and quantitative performance evaluation index of the wind turbine within a set time interval can be obtained, so as to intuitively understand the power generation performance status of the wind turbine.

[0075] Performance evaluation: Based on the wind turbine performance index paex of each wind turbine within the set time interval, the performance level of each wind turbine is determined, and the determined performance level results of each wind turbine are added to the corresponding rows of operating parameter information in the data matrix. The data matrix with the added performance level results is used as the preliminary evaluation result of the wind farm within the set time interval; the performance levels include low efficiency level, normal level, good level and high efficiency level;

[0076] Specifically:

[0077] A reference threshold index of the fan performance index paex corresponding to each fan is preset, and the fan performance index paex of each fan within a set time interval is compared with the preset reference threshold index. If the fan performance index paex of a certain fan within the set time interval is lower than the corresponding preset reference threshold index, the performance level of the fan is determined to be a low-efficiency level and a retrieval signaling is triggered;

[0078] If the fan performance index paex of a certain fan in a set time interval is higher than the preset reference threshold index, the difference between the risk performance index paex and the reference threshold index is further calculated as the performance difference, and three groups of difference value ranges corresponding to the preset performance difference are set, and the three groups of difference value ranges are set to correspond to the normal level, the good level and the high-efficiency level respectively; the higher the performance difference, the higher the corresponding matching level; the calculated performance difference is matched with the corresponding preset difference value range, so as to obtain the performance level of the fan in the set time interval;

[0079] It should be noted that, assuming that the reference threshold index is set to hg, the following three sets of difference value ranges and corresponding levels are preset;

[0080] Normal level: When the performance is poor, that is, when paex-hg is between 0-0.13, it is judged as normal level;

[0081] Good grade: When the performance is poor, that is, when paex-hg is between 0.13-0.27, it is judged as good grade;

[0082] High efficiency level: When the performance is poor, that is, paex-hg is above 0.27, it is judged as high efficiency level;

[0083] Now the performance difference calculated for a certain fan within the set time interval is 0.12;

[0084] The corresponding match is between 0-0.13, and according to the preset rules, the performance level of the wind turbine is judged to be a good level.

[0085] Historical synthesis: Based on the wind turbine performance index paex of each wind turbine at the current time point within the set time interval, H groups of historical data matrices before the current time point are extracted, where H>8, and the specific value is set by the technical staff; the H groups of historical data matrices are analyzed and evaluated to obtain the performance trend change status of each wind turbine at the current time point, and the performance trend change status of each wind turbine is added to each row of operating parameter information in the data matrix. The data matrix after adding the performance trend change status is used as the final evaluation result of the wind farm within the set time interval;

[0086] Specifically:

[0087] Analyze the H group historical data matrix, extract the H group historical wind turbine performance index of each wind turbine, and integrate it with the wind turbine performance index paex of each wind turbine at the current time point, so as to obtain the performance index sequence of each wind turbine;

[0088] A moving average window size is preset; for example, the window size is 3; and the performance index sequence of each wind turbine is calculated according to the preset window size to obtain each group of moving average values ​​corresponding to each wind turbine;

[0089] For each group of moving average values ​​of each wind turbine, the change rate between two adjacent groups of moving average values ​​is calculated, that is, by the formula Calculate and get and Represents two adjacent groups of moving averages;

[0090] If the change rate of the moving averages of two adjacent groups is positive, it is marked as a performance improvement value. If the change rate of the moving averages of two adjacent groups is negative, the absolute value is taken and marked as a performance degradation value.

[0091] After arranging the change rates of each group of wind turbines in chronological order, they are divided from the middle point of the arrangement to obtain the change rate set of the front time zone and the change rate set of the rear time zone of each wind turbine;

[0092] The performance increase value and performance decrease value of each group in the previous time zone change rate set are summed up respectively to obtain the total increase value and the total decrease value of the previous time zone, and the total increase value of the previous time zone is subtracted from the total decrease value of the previous time zone to obtain the previous time zone trend evaluation value. If the previous time zone trend evaluation value is a positive value, it means that the performance state of the fan in the previous time zone is an upward trend. If the previous time zone trend evaluation value is a negative value, it means that the performance state of the fan in the previous time zone is a downward trend.

[0093] The performance increase value and performance decrease value of each group in the rear time zone change rate set are summed up respectively to obtain the rear time zone increase total value and the rear time zone decrease total value, and the rear time zone trend evaluation value is obtained by subtracting the rear time zone decrease total value from the rear time zone increase total value. If the rear time zone trend evaluation value is a positive value, it means that the performance state of the fan in the rear time zone is an upward trend. If the rear time zone trend evaluation value is a negative value, it means that the performance state of the fan in the rear time zone is a downward trend.

[0094] If the performance status of a certain wind turbine in the previous time zone and the next time zone is an upward trend, then the performance trend change status of the wind turbine at the current time point is an overall performance upward trend;

[0095] If the performance status of a certain wind turbine in both the previous time zone and the next time zone is a downward trend, then the performance trend change status of the wind turbine at the current time point is an overall performance downward trend;

[0096] If the performance status of a certain wind turbine in the previous time zone and the next time zone is an upward trend and a downward trend respectively, then the performance trend change status of the wind turbine at the current time point is an overall performance fluctuation trend, and the upward trend changes to a downward trend;

[0097] If the performance status of a certain wind turbine in the previous time zone and the next time zone is a downward trend and an upward trend respectively, then the performance trend change status of the wind turbine at the current time point is an overall performance fluctuation trend, and the downward trend changes to an upward trend;

[0098] It should be noted that the comprehensive analysis based on historical data mentioned above has achieved accurate judgment on wind turbine performance trends, providing important decision-making basis for wind turbine operation and maintenance, management and the overall operation of wind farms.

[0099] Trigger retrieval: When the retrieval signal is triggered, the corresponding fan that triggers the retrieval signal is marked as an abnormal fan, and corresponding steps are performed to determine the triggering reason of the abnormal fan triggering the retrieval signal, and the triggering reason is sent to the management personnel;

[0100] Specifically:

[0101] M1: Read the operating parameter information of the row corresponding to the abnormal fan in the data matrix, extract the fan speed of the abnormal fan within the set time interval, and compare the fan speed of the abnormal fan within the set time interval with the set minimum speed value. If the fan speed of the abnormal fan at a certain time point within the set time interval is lower than the set minimum speed and the duration reaches the set time, it is determined that the triggering reason for triggering the retrieval signaling is the abnormal low speed. Otherwise, execute step M2;

[0102] The minimum speed value is set by the technicians according to the ambient wind speed and direction conditions;

[0103] M2: Read the operating parameter information of the row corresponding to the abnormal fan in the data matrix, extract the fan temperature and fan vibration amplitude of the abnormal fan within the set time interval, and perform comprehensive analysis to obtain the health assessment index jkex of the abnormal fan within the set time interval, set the reference threshold index of the health assessment index jkex corresponding to each fan, if the health assessment index jkex of the abnormal fan within the set time interval is higher than the set reference threshold index, then determine that the triggering cause of the triggering retrieval signaling is the abnormal health of the fan, otherwise execute step M3;

[0104] The fan temperature is the temperature value of the motor;

[0105] M2-201: extract the fan temperature of the abnormal fan at each time point within the set time interval and calculate the average value, and use the calculated average value as the fan average temperature value of the abnormal fan; set the reference temperature value of the fan corresponding to the fan temperature, calculate the ratio of the fan average temperature value to the set reference temperature value, and obtain the temperature evaluation value wcrx of the abnormal fan; that is, calculate it by dividing the fan average temperature value by the reference temperature value;

[0106] M2-202: Set the standard of vibration events, i.e., the reference threshold corresponding to the vibration amplitude and the reference duration higher than the reference threshold, extract the vibration amplitude of the abnormal fan within the set time interval, and based on the set vibration event standard, count the number of vibration events that occur in the abnormal fan within the set time interval as the number of vibrations of the abnormal fan, set the allowable number of vibrations corresponding to the number of vibrations of the abnormal fan, calculate the ratio of the number of vibrations to the set allowable number of vibrations, and obtain the vibration evaluation value wfrx of the abnormal level; i.e., the number of vibrations / the set allowable number of vibrations;

[0107] M2-203: preset the weight coefficients corresponding to the temperature evaluation value wcrx and the vibration evaluation value wfrx, multiply the abnormal level temperature evaluation value wcrx and the vibration evaluation value wfrx by the corresponding weight coefficients, and then sum them up to obtain the health evaluation index jkex of the abnormal fan within the set time interval;

[0108] It should be noted that by combining the two key parameters of fan temperature and vibration amplitude for comprehensive analysis, the limitation of relying on a single parameter to judge the health status of the fan is avoided. The health status of the fan is the result of the combined effect of multiple factors. Excessive temperature may indicate heat dissipation problems or overload of components such as the motor. Abnormal vibration amplitude may indicate imbalance, looseness or wear of mechanical components. The combination of these two parameters can more comprehensively and accurately reflect the actual health level of the fan.

[0109] M3: Taking the location of the abnormal wind turbine as the target point, selecting the maintenance personnel with the shortest distance to the target point as the search personnel, and sending the location of the abnormal wind turbine to the search personnel. The search personnel go to the location of the abnormal wind turbine and determine the cause of triggering the search signaling, and store the determination result;

[0110] It should be noted that, in summary, the entire trigger retrieval and subsequent judgment process has clear and detailed step regulations, from the initial trigger retrieval signaling mark, to the analysis of different parameters through different steps to determine the trigger cause, to the selection of retrieval personnel and the storage of judgment results, etc., each link has a clear operation process and standard. This enables different operation and maintenance personnel to operate according to unified standards and processes when facing abnormal conditions of wind turbines, reducing the interference and uncertainty of human factors, and ensuring the quality and efficiency of operation and maintenance work.

[0111] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A wind farm performance analysis and evaluation method based on big data, characterized in that: include: Data collection: preset the time interval for performance evaluation. After the preset time interval is reached, collect the operating parameters of each wind turbine in the wind farm within the set time interval, and pre-process the operating parameters of each wind turbine within the set time interval. After pre-processing, organize the operating parameters of each wind turbine within the set time interval into a data matrix, and each row in the data matrix represents the operating parameter information of the corresponding wind turbine within the set time interval; the operating parameters include wind turbine power output data, wind turbine power factor data, wind turbine speed, fan temperature and fan vibration amplitude; Performance analysis: read the operating parameter information of each row in the data matrix, and conduct a comprehensive analysis to obtain the fan performance index paex of each fan within the set time interval; Where x represents the number of each wind turbine in the wind farm, x=1, 2, ..., y; where y represents the total number of wind turbines in the wind farm; Performance evaluation: Based on the wind turbine performance index paex of each wind turbine within the set time interval, the performance level of each wind turbine is determined, and the determined performance level results of each wind turbine are added to the corresponding rows of operating parameter information in the data matrix. The data matrix with the added performance level results is used as the preliminary evaluation result of the wind farm within the set time interval; the performance levels include low efficiency level, normal level, good level and high efficiency level; Historical synthesis: Based on the wind turbine performance index paex of each wind turbine at the current time point within the set time interval, H groups of historical data matrices before the current time point are extracted, where H>8. The H groups of historical data matrices are analyzed and evaluated to obtain the performance trend change status of each wind turbine at the current time point. After adding the performance trend change status of each wind turbine to each row of operating parameter information in the data matrix, the data matrix after adding the performance trend change status is used as the final evaluation result of the wind farm within the set time interval.

2. The wind farm performance analysis and evaluation method based on big data according to claim 1 is characterized in that: Read the operating parameter information of each row in the data matrix and perform a comprehensive analysis, specifically: Read the power output data in each row of the operating parameter information in the data matrix, extract the power output data at each time point in the set time interval, and form a power output time series of each wind turbine in the set time interval; The standard deviation formula is used to calculate the power output time series of each wind turbine within the set time interval, and the calculation result is used as the power dispersion index of each wind turbine within the set time interval; A reference threshold value of the power dispersion index corresponding to each wind turbine is preset, and the power dispersion index of each wind turbine within a set time interval is calculated to be proportional to the corresponding reference threshold value to obtain the stability score uqx of each wind turbine; that is, it is calculated by power dispersion index / corresponding reference threshold value; For the power output time series of each wind turbine within the set time interval, the average value is calculated as the power average value of each wind turbine within the set time interval; Extract the rated power corresponding to each wind turbine, calculate the ratio between the average power of each wind turbine and the corresponding rated power, and obtain the performance score uwx of each wind turbine; that is, it is calculated by the average power / corresponding rated power; Read the power factor data in each row of the operating parameter information in the data matrix, extract the power factor data at each time point in the set time interval, and form a power factor time series of each wind turbine in the set time interval; For the power factor time series of each wind turbine within the set time interval, the average value is calculated as the power factor average value of each wind turbine within the set time interval; Preset each group of standard ranges for the average power factor value corresponding to each wind turbine, set each group of standard range to correspond to an electric energy conversion index; match the average power factor value of each wind turbine with the corresponding standard range to obtain the electric energy conversion index corresponding to each wind turbine, and use the electric energy conversion index corresponding to each wind turbine as the transformation score utx of each wind turbine within the set time interval.

3. The wind farm performance analysis and evaluation method based on big data according to claim 2 is characterized in that: The fan performance index paex of each fan within the set time interval is obtained, specifically: After normalizing the stability score uqx, performance score uwx and conversion score utx of each wind turbine within the set time interval, substitute them into the formula A weighted calculation is performed to obtain the fan performance index paex of each wind turbine within the set time interval; wherein a1, a2 and a3 are the influencing weight factors of the stability score uqx, the performance score uwx and the conversion score utx respectively.

4. The wind farm performance analysis and evaluation method based on big data according to claim 3 is characterized in that: Based on the fan performance index paex of each fan within the set time interval, the performance level of each fan is determined as follows: A reference threshold index of the fan performance index paex corresponding to each fan is preset, and the fan performance index paex of each fan within a set time interval is compared with the preset reference threshold index. If the fan performance index paex of a certain fan within the set time interval is lower than the corresponding preset reference threshold index, the performance level of the fan is determined to be a low-efficiency level and a retrieval signaling is triggered; If the fan performance index paex of a certain wind turbine in the set time interval is higher than the preset reference threshold index, the difference between the risk performance index paex and the reference threshold index is further calculated as the performance difference, and three groups of difference value ranges corresponding to the preset performance difference are set. The three groups of difference value ranges are set to correspond to the normal level, good level and high-efficiency level respectively. The calculated performance difference is matched with the corresponding preset difference value range to obtain the performance level of the wind turbine in the set time interval.

5. The wind farm performance analysis and evaluation method based on big data according to claim 4 is characterized in that: The historical data matrix of group H is analyzed and evaluated, specifically: Analyze the H group historical data matrix, extract the H group historical wind turbine performance index of each wind turbine, and integrate it with the wind turbine performance index paex of each wind turbine at the current time point, so as to obtain the performance index sequence of each wind turbine; A moving average window size is preset, and the performance index sequence of each wind turbine is calculated according to the preset window size to obtain each group of moving average values ​​corresponding to each wind turbine; For each group of moving average values ​​of each wind turbine, the change rate between two adjacent groups of moving average values ​​is calculated, that is, by the formula Calculate and get and Represents two adjacent groups of moving averages; If the change rate of the moving averages of two adjacent groups is positive, it is marked as a performance improvement value. If the change rate of the moving averages of two adjacent groups is negative, the absolute value is taken and marked as a performance degradation value. After arranging the change rates of each group of wind turbines in chronological order, they are divided from the middle point of the arrangement to obtain the front time zone change rate set and the rear time zone change rate set of each wind turbine.

6. The wind farm performance analysis and evaluation method based on big data according to claim 5 is characterized in that: Get the performance trend change status of each fan at the current time point, specifically: The performance increase value and performance decrease value of each group in the previous time zone change rate set are summed up respectively to obtain the total increase value and the total decrease value of the previous time zone, and the total increase value of the previous time zone is subtracted from the total decrease value of the previous time zone to obtain the previous time zone trend evaluation value. If the previous time zone trend evaluation value is a positive value, it means that the performance state of the fan in the previous time zone is an upward trend. If the previous time zone trend evaluation value is a negative value, it means that the performance state of the fan in the previous time zone is a downward trend. The performance increase value and performance decrease value of each group in the rear time zone change rate set are summed up respectively to obtain the rear time zone increase total value and the rear time zone decrease total value, and the rear time zone trend evaluation value is obtained by subtracting the rear time zone decrease total value from the rear time zone increase total value. If the rear time zone trend evaluation value is a positive value, it means that the performance state of the fan in the rear time zone is an upward trend. If the rear time zone trend evaluation value is a negative value, it means that the performance state of the fan in the rear time zone is a downward trend. If the performance status of a certain wind turbine in the previous time zone and the next time zone is an upward trend, then the performance trend change status of the wind turbine at the current time point is an overall performance upward trend; If the performance status of a certain wind turbine in both the previous time zone and the next time zone is a downward trend, then the performance trend change status of the wind turbine at the current time point is an overall performance downward trend; If the performance status of a certain wind turbine in the previous time zone and the next time zone is an upward trend and a downward trend respectively, then the performance trend change status of the wind turbine at the current time point is an overall performance fluctuation trend, and the upward trend changes to a downward trend; If the performance status of a wind turbine in the previous time zone and the next time zone is a downward trend and an upward trend respectively, then the performance trend change status of the wind turbine at the current time point is an overall performance fluctuation trend, and the downward trend changes to an upward trend.

7. The wind farm performance analysis and evaluation method based on big data according to claim 1 is characterized in that: Also includes: Trigger retrieval: When the retrieval signal is triggered, the corresponding wind turbine that triggers the retrieval signal is marked as an abnormal wind turbine, and corresponding steps are executed to determine the triggering reason of the abnormal wind triggering the retrieval signal.

8. The wind farm performance analysis and evaluation method based on big data according to claim 7 is characterized in that: Execute corresponding steps to determine the triggering cause of the abnormal wind turbine triggering retrieval signaling, specifically: M1: Read the operating parameter information of the row corresponding to the abnormal fan in the data matrix, extract the fan speed of the abnormal fan within the set time interval, and compare the fan speed of the abnormal fan within the set time interval with the set minimum speed value. If the fan speed of the abnormal fan at a certain time point within the set time interval is lower than the set minimum speed and the duration reaches the set time, it is determined that the triggering reason for triggering the retrieval signaling is the abnormal low speed. Otherwise, execute step M2; M2: Read the operating parameter information of the row corresponding to the abnormal fan in the data matrix, extract the fan temperature and fan vibration amplitude of the abnormal fan within the set time interval, and perform comprehensive analysis to obtain the health assessment index jkex of the abnormal fan within the set time interval, set the reference threshold index of the health assessment index jkex corresponding to each fan, if the health assessment index jkex of the abnormal fan within the set time interval is higher than the set reference threshold index, then determine that the triggering cause of the triggering retrieval signaling is the abnormal health of the fan, otherwise execute step M3; M2-201: extract the fan temperature of the abnormal fan at each time point within the set time interval and calculate the average value, and use the calculated average value as the fan average temperature value of the abnormal fan; set the reference temperature value of the fan corresponding to the fan temperature, calculate the ratio of the fan average temperature value to the set reference temperature value, and obtain the temperature evaluation value wcrx of the abnormal fan; that is, calculate it by dividing the fan average temperature value by the reference temperature value; M2-202: Set the standard of vibration events, i.e., the reference threshold corresponding to the vibration amplitude and the reference duration higher than the reference threshold, extract the vibration amplitude of the abnormal fan within the set time interval, and based on the set vibration event standard, count the number of vibration events that occur in the abnormal fan within the set time interval as the number of vibrations of the abnormal fan, set the allowable number of vibrations corresponding to the number of vibrations of the abnormal fan, calculate the ratio of the number of vibrations to the set allowable number of vibrations, and obtain the vibration evaluation value wfrx of the abnormal level; i.e., the number of vibrations / the set allowable number of vibrations; M2-203: preset the weight coefficients corresponding to the temperature evaluation value wcrx and the vibration evaluation value wfrx, multiply the abnormal level temperature evaluation value wcrx and the vibration evaluation value wfrx by the corresponding weight coefficients, and then sum them up to obtain the health evaluation index jkex of the abnormal fan within the set time interval; M3: Taking the location of the abnormal fan as the target point, select the maintenance personnel with the shortest distance to the target point as the search personnel, and send the location of the abnormal fan to the search personnel. The search personnel go to the location of the abnormal fan and determine the cause of triggering the search signal, and store the determination result.

Images