A wind turbine intelligent control system for new energy centralized control center

Abstract

The present invention provides a wind turbine intelligent control system of a new energy centralized control center, which belongs to the field of new energy technology, and includes: an acquisition module: acquiring historical operation data of the new energy centralized control center under historical operation time, and extracting features of the historical operation data to determine the first feature; an analysis module: analyzing the historical operation data and the first feature, constructing a wind turbine intelligent control model, and determining the fault type and fault handling based on the wind turbine intelligent control model; a management module: acquiring the current operation status of the wind turbine in real time, and generating a fault warning sheet based on the analysis result of the current operation status of the wind turbine intelligent control model, and performing fault handling on the wind turbine based on the fault warning sheet to realize intelligent control. It can improve the efficiency and accuracy of fault diagnosis, take effective fault handling measures, reduce production losses and energy losses, optimize maintenance plans and wind turbine operating parameters, and improve the operating efficiency, reliability and system performance of the wind turbine.

Description

Technical Field

[0001] The present invention relates to the field of new energy technology, and in particular to an intelligent wind turbine control system in a new energy centralized control center. Background Art

[0002] At present, the rapid development of new technologies such as intelligent sensors, communication networks, computer software and hardware, and hierarchical management of large systems has provided technical support for new energy power generation companies to achieve daily centralized monitoring, unified dispatching and management of stations within their jurisdiction. The operation and maintenance model of modern new energy power plants needs to develop towards a more intelligent, centralized, safe and reliable centralized control and operation and maintenance model, using advanced intelligent monitoring platforms to centralize data from various wind farms for analysis and prediction.

[0003] Therefore, the present invention proposes an intelligent wind turbine control system for a new energy centralized control center. Summary of the invention

[0004] The present invention provides a wind turbine intelligent management and control system for a new energy centralized control center. According to a first feature determined by historical operation data of the new energy centralized control center, the historical operation data and the first feature are analyzed to construct an intelligent management and control model for the wind turbine, the fault type is determined, and a fault warning sheet is generated in combination with the current operation status of the wind turbine for fault processing, so as to realize intelligent management and control, which can improve the efficiency and accuracy of fault diagnosis, take effective fault processing measures, reduce production losses and energy losses, optimize maintenance plans and wind turbine operating parameters, and improve the operating efficiency, reliability and system performance of the wind turbine.

[0005] The present invention provides a wind turbine intelligent control system for a new energy centralized control center, comprising:

[0006] Acquisition module: acquiring the historical operation data of the new energy centralized control center under the historical operation time, and performing feature extraction on the historical operation data to determine the first feature;

[0007] Analysis module: Analyze historical operation data and the first feature, build a wind turbine intelligent management and control model, and determine the fault type and fault handling based on the wind turbine intelligent management and control model;

[0008] Management module: obtain the current operating status of the fan in real time, analyze the current operating status based on the fan intelligent management and control model to generate a fault warning sheet, and perform fault processing on the fan based on the fault warning sheet to achieve intelligent management and control.

[0009] Preferably, it includes:

[0010] The first characteristics include: vibration characteristics, sound characteristics, current characteristics and temperature characteristics.

[0011] Preferably, the acquisition module includes:

[0012] The first extraction unit is used to extract the vibration value, sound value, current value and temperature value of the fan under the same unit time and hourly operation time from the historical operation data;

[0013] Characteristic autocorrelation value calculation unit: respectively calculates the vibration autocorrelation value based on the vibration value, the sound autocorrelation value based on the sound value, and the current autocorrelation value based on the current value of the fan between every two same unit times, and at the same time, calculates the temperature autocorrelation value of the fan based on the temperature value between every two same unit times;

[0014]

[0015]

[0016]

[0017]

[0018] Among them, VAFK K SAFK represents the vibration autocorrelation value between two units of time when the lag parameter is K. K Represents the sound autocorrelation value between two units of time when the lag parameter is K, CAFK K It represents the current autocorrelation value between two units of time when the hysteresis parameter is K, TAFK K It represents the temperature autocorrelation value between two units of time when the lag parameter is K. represents the lag parameter, [] is the rounding symbol, N1 represents the number of unit time, N2 represents the length of hours in the same unit time, V j1 , S j1 , C j1 、T j1 They represent the vibration value, sound value, current value, and temperature value at the j1st hour of operation time, respectively. They represent the sum of vibration values, sound values, current values, and temperature values ​​in the i1th unit time, respectively. They represent the sum of vibration values, sound values, current values, and temperature values ​​in the i1-kth unit time, respectively. They represent the average vibration value, average sound value, average current value, and average temperature value in N1 unit time respectively;

[0019] Autocorrelation sequence unit: based on The vibration autocorrelation values ​​determine the vibration autocorrelation sequence of the fan, based on The sound autocorrelation value determines the sound autocorrelation sequence of the fan and the The current autocorrelation values ​​determine the current autocorrelation sequence of the wind turbine. The temperature autocorrelation values ​​determine the temperature autocorrelation sequence of the fan;

[0020] Autocorrelation diagram unit: based on all lag parameters and vibration autocorrelation sequence, sound autocorrelation sequence, current autocorrelation sequence, temperature autocorrelation sequence, draw the corresponding vibration autocorrelation diagram, sound autocorrelation diagram, current autocorrelation diagram and temperature autocorrelation diagram;

[0021] Characteristic unit: analyzing the vibration autocorrelation graph to determine the vibration characteristics of the fan, analyzing the sound autocorrelation graph to determine the sound characteristics of the fan, analyzing the current autocorrelation graph to determine the current characteristics of the fan, and analyzing the temperature autocorrelation graph to determine the temperature characteristics of the fan.

[0022] Preferably, the analysis module comprises:

[0023] Annotation unit: annotate the data of fan failure in historical operation data;

[0024] Division unit: divide the annotated historical running data into training set and test set;

[0025] Feature vector unit: based on the vibration feature, sound feature, current feature and temperature feature, it is combined into corresponding vibration vector, sound vector, current vector and temperature vector respectively;

[0026] Sample distance square value calculation unit: determines the distance square value between every two samples in the training set based on the vibration vector, the sound vector, the current vector and the temperature vector;

[0027]

[0028] Where Xa, Xb represent two samples in the training set, 1≤Xa≤N3, 1≤Xb≤N3, and Xa≠Xb, Vn represents the number of standard sub-features of vibration features, Sn represents the number of standard sub-features of sound features, Cn represents the number of standard sub-features of current features, and Tn represents the number of standard sub-features of temperature features; Xa k1 represents the value of sample Xa based on the k1th sub-feature; Xb k1 represents the value of sample Xb based on the k1th sub-feature; Z k1 Represents the feature conversion coefficient of the k1th sub-feature;

[0029] Arrangement unit: Randomly The squared distance values ​​between the samples are arranged in ascending order, and the sorted The middle value mD(Xa, Xb) among the values ​​is taken as the square value of the interval distance, where [] is the integer symbol, N3 represents the number of samples in the training set,

[0030] Adjustment coefficient calculation unit: calculates adjustment parameter σ based on all distance square values ​​and interval distance square values;

[0031]

[0032]

[0033]

[0034] Among them, α1 and α2 represent the weight coefficients corresponding to the average distance adjustment formula and the interval distance adjustment formula, respectively, |D(Xa, Xb) L1 -mD(Xa, Xb)| represents the absolute difference between the L1th distance square value and the interval distance square value; A01 represents the average distance adjustment formula, and A02 represents the interval distance adjustment formula;

[0035] Sample similarity value calculation unit: determines the similarity value between two corresponding samples based on the adjustment parameter and the squared distance between every two samples in the training set;

[0036]

[0037] Construction unit: construct a similarity matrix based on the similarity values ​​between any two samples, wherein the rows in the similarity matrix are similarity values ​​between the same sample and different samples;

[0038] Training set sample vector unit: a sample vector of the corresponding sample is formed according to the row corresponding to each sample in the similarity matrix and the fault label corresponding to the sample;

[0039] Model building unit: input all sample vectors into the neural network model to build a wind turbine intelligent management and control model;

[0040] Test set sample vector unit: calculates the similarity between each test sample and all training samples and forms the corresponding test sample vector;

[0041] Determining unit: inputting the test sample vector into the wind turbine intelligent management and control model, determining the output labeling result of each test sample, wherein the output labeling result includes: output fault type and fault handling, wherein the fault type includes no fault, resettable fault, and non-resettable fault, and the non-resettable fault includes fault content and fault level;

[0042] Evaluation unit: compares the fault labeling results of the test samples and the output labeling results, accumulates the number of true positive examples, true negative examples, false positive examples, and false negative examples in the test set, and evaluates the model performance.

[0043] Preferably, the analysis module comprises:

[0044] A setting unit: setting an autonomous reset threshold for the resettable fault;

[0045] Comparison unit: compares the number of reset times with the autonomous reset threshold. If the number of reset times is less than the autonomous reset threshold, the resettable fault is autonomously reset, and the number of reset times is increased by 1.

[0046] Reset unit: if the number of resets is equal to the autonomous reset threshold, the fault type is changed from the resettable fault to a non-resettable fault, and at the same time, the number of resets is reset.

[0047] Preferably, the analysis module comprises:

[0048] Alarm unit: setting a fault alarm for the non-resettable fault based on the corresponding fault content and fault level;

[0049] The fault alarm consists of a fault level alarm and a fault content alarm.

[0050] Preferably, the analysis module comprises:

[0051] Identification unit: performing in-depth analysis on the test set samples of the false negative examples to identify the focus area of ​​the wind turbine;

[0052] Prevention unit: Strengthen monitoring of the concerned areas to prevent potential failures.

[0053] Preferably, the management module includes:

[0054] Input unit: input the current operating status of the fan into the intelligent control model of the fan to determine the fault type and fault handling corresponding to the current operating status;

[0055] Generation unit: Generates a fault warning sheet based on the fault type and fault handling corresponding to the current operating status.

[0056] Compared with the prior art, the beneficial effects of the present application are as follows: according to the first feature determined by the historical operation data of the new energy integrated control center, the historical operation data and the first feature are analyzed to construct an intelligent management and control model for the wind turbine, the fault type is determined, and a fault warning sheet is generated in combination with the current operating status of the wind turbine for fault handling to achieve intelligent management and control, which can improve the efficiency and accuracy of fault diagnosis, take effective fault handling measures, reduce production losses and energy losses, optimize maintenance plans and wind turbine operating parameters, and improve the operating efficiency, reliability and system performance of the wind turbine.

[0057] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or may be understood through the implementation of the present invention. The purposes and other advantages of the present invention may be realized and obtained through the structures particularly pointed out in the written description and the drawings.

[0058] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0060] Figure 1 This is a structural diagram of a wind turbine intelligent management and control system in a new energy centralized control center in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0062] Embodiment 1:

[0063] The embodiment of the present invention provides a wind turbine intelligent control system for a new energy centralized control center, such as Figure 1 As shown, including:

[0064] Acquisition module: acquiring the historical operation data of the new energy centralized control center under the historical operation time, and performing feature extraction on the historical operation data to determine the first feature;

[0065] Analysis module: Analyze historical operation data and the first feature, build a wind turbine intelligent management and control model, and determine the fault type and fault handling based on the wind turbine intelligent management and control model;

[0066] Management module: obtain the current operating status of the fan in real time, analyze the current operating status based on the fan intelligent management and control model to generate a fault warning sheet, and perform fault processing on the fan based on the fault warning sheet to achieve intelligent management and control.

[0067] In this embodiment, feature extraction means extracting feature values ​​of the wind turbine operating time per hour in the same unit time in the historical operating data.

[0068] In this embodiment, the first feature is determined based on the corresponding feature autocorrelation graph, including vibration features, sound features, current features, and temperature features.

[0069] In this embodiment, the intelligent wind turbine management and control model is constructed based on the sample vectors of all training set samples in the historical operation data.

[0070] In this embodiment, the fault types include no fault, resettable fault, and non-resettable fault.

[0071] The beneficial effects of the above technical solution are: according to the first feature determined by the historical operation data of the new energy integrated control center, the historical operation data and the first feature are analyzed to build an intelligent management and control model for the wind turbine, the fault type is determined, and a fault warning sheet is generated in combination with the current operating status of the wind turbine for fault handling to achieve intelligent management and control, which can improve the efficiency and accuracy of fault diagnosis, take effective fault handling measures, reduce production losses and energy losses, optimize maintenance plans and wind turbine operating parameters, and improve the operating efficiency, reliability and system performance of the wind turbine.

[0072] Embodiment 2:

[0073] The embodiment of the present invention provides a wind turbine intelligent control system of a new energy centralized control center, including:

[0074] The first characteristics include: vibration characteristics, sound characteristics, current characteristics, and temperature characteristics.

[0075] In this embodiment, the vibration characteristics calculate the corresponding vibration autocorrelation value based on the vibration value of the fan under the same unit time and hourly operating time. A vibration autocorrelation diagram is drawn based on all vibration autocorrelation values ​​and analyzed and determined to detect fault types such as fan blade imbalance and bearing failure.

[0076] In this embodiment, the sound feature calculates the corresponding sound autocorrelation value according to the sound value of the fan at the operating time per hour in the same unit time, and a sound autocorrelation graph is drawn based on all the sound autocorrelation values ​​and analyzed to detect abnormal noise of the fan, such as blade damage.

[0077] In this embodiment, the current characteristic calculates the corresponding current autocorrelation value according to the current value of the fan under the same unit time and hourly running time, and draws the vibration current autocorrelation diagram based on all current autocorrelation values ​​and analyzes and determines it for detecting motor failure and power supply failure.

[0078] In this embodiment, the temperature feature calculates the corresponding temperature autocorrelation value based on the temperature value of the fan at each hour of operation in the same unit time, and draws a temperature autocorrelation diagram based on all temperature autocorrelation values ​​and analyzes and determines it to detect overheating problems of the fan, such as motor overload, cooling system failure, etc.

[0079] The beneficial effects of the above technical solution are as follows: by extracting the features of the historical operation data to determine the first feature, different failure modes of the wind turbine can be identified, providing a feature basis for building an intelligent management and control model for the wind turbine.

[0080] Embodiment 3:

[0081] The embodiment of the present invention provides a wind turbine intelligent management and control system of a new energy centralized control center, wherein the acquisition module comprises:

[0082] The first extraction unit is used to extract the vibration value, sound value, current value and temperature value of the fan under the same unit time and hourly operation time from the historical operation data;

[0083] Characteristic autocorrelation value calculation unit: respectively calculates the vibration autocorrelation value based on the vibration value, the sound autocorrelation value based on the sound value, and the current autocorrelation value based on the current value of the fan between every two same unit times, and at the same time, calculates the temperature autocorrelation value of the fan based on the temperature value between every two same unit times;

[0084]

[0085]

[0086]

[0087]

[0088] Among them, VAFK K SAFK represents the vibration autocorrelation value between two units of time when the lag parameter is K. K Represents the sound autocorrelation value between two units of time when the lag parameter is K, CAFK K It represents the current autocorrelation value between two units of time when the hysteresis parameter is K, TAFK K It represents the temperature autocorrelation value between two units of time when the lag parameter is K. represents the lag parameter, [] is the rounding symbol, N1 represents the number of unit time, N2 represents the length of hours in the same unit time, V j1 , S j1 , C j1 , T j1 They represent the vibration value, sound value, current value, and temperature value at the j1st hour of operation time, respectively. They represent the sum of vibration values, sound values, current values, and temperature values ​​in the i1th unit time, respectively. They represent the sum of vibration values, sound values, current values, and temperature values ​​in the i1-kth unit time, respectively. They represent the average vibration value, average sound value, average current value, and average temperature value in N1 unit time respectively;

[0089] Autocorrelation sequence unit: based on The vibration autocorrelation values ​​determine the vibration autocorrelation sequence of the fan, based on The sound autocorrelation value determines the sound autocorrelation sequence of the fan and the The current autocorrelation values ​​determine the current autocorrelation sequence of the wind turbine. The temperature autocorrelation values ​​determine the temperature autocorrelation sequence of the fan;

[0090] Autocorrelation diagram unit: based on all lag parameters and vibration autocorrelation sequence, sound autocorrelation sequence, current autocorrelation sequence, temperature autocorrelation sequence, draw the corresponding vibration autocorrelation diagram, sound autocorrelation diagram, current autocorrelation diagram and temperature autocorrelation diagram;

[0091] Characteristic unit: analyzing the vibration autocorrelation graph to determine the vibration characteristics of the fan, analyzing the sound autocorrelation graph to determine the sound characteristics of the fan, analyzing the current autocorrelation graph to determine the current characteristics of the fan, and analyzing the temperature autocorrelation graph to determine the temperature characteristics of the fan.

[0092] In this embodiment, the vibration autocorrelation value represents the correlation between the vibration value at each hour of operation time in the i1th unit time and the vibration value at each hour of operation time in the i1-Kth unit time corresponding to the lag.

[0093] In this embodiment, the sound autocorrelation value represents the correlation between the sound value at each hour of operation time in the i1th unit time and the sound value at each hour of operation time in the i1-Kth unit time corresponding to the lag.

[0094] In this embodiment, the current autocorrelation value represents the correlation between the current value at each hour of operation time in the i1th unit time and the current value at each hour of operation time in the i1-Kth unit time corresponding to the lag.

[0095] In this embodiment, the temperature autocorrelation value represents the correlation between the temperature value at each hour of operation time in the i1th unit time and the temperature value at each hour of operation time in the i1-Kth unit time corresponding to the lag.

[0096] In this embodiment, the lag parameter represents the interval between the i1th unit time and the previous K unit times.

[0097] In this embodiment, in the vibration autocorrelation diagram, the horizontal axis is the lag value and the vertical axis is the vibration autocorrelation sequence, which reflects the correlation strength, frequency component, periodic distribution, etc. of the vibration value under the historical operating time.

[0098] In this embodiment, in the sound autocorrelation diagram, the horizontal axis is the lag value and the vertical axis is the sound autocorrelation sequence, which reflects the correlation strength, frequency components, periodic distribution, etc. of the sound value under the historical running time.

[0099] In this embodiment, in the current autocorrelation diagram, the horizontal axis is the lag value and the vertical axis is the current autocorrelation sequence, which reflects the correlation strength, frequency component, periodic distribution, etc. of the current value under the historical operating time.

[0100] In this embodiment, in the temperature autocorrelation diagram, the horizontal axis is the lag value and the vertical axis is the temperature autocorrelation sequence, which reflects the correlation strength, frequency component, periodic distribution, etc. of the temperature value under the historical operating time.

[0101] In this embodiment, the first feature is determined according to the corresponding first feature autocorrelation graph, which may include the peak value, valley value, amplitude, width, mean value and variance of the autocorrelation sequence of the first feature autocorrelation graph.

[0102] The beneficial effects of the above technical solution are as follows: by performing a series of analysis and calculations on the vibration values, sound values, current values ​​and temperature values ​​of the fan under the same unit time and hourly operating time to determine the vibration characteristics, sound characteristics, current characteristics and temperature characteristics of the fan, the characteristics of different fault modes in the historical operating data can be identified, and the data basis for building a smart management and control model for the fan can be improved to realize smart management and control.

[0103] Embodiment 4:

[0104] The embodiment of the present invention provides a wind turbine intelligent management and control system of a new energy centralized control center, wherein the analysis module includes:

[0105] Annotation unit: annotate the data of fan failure in historical operation data;

[0106] Division unit: divide the annotated historical running data into training set and test set;

[0107] Feature vector unit: based on the vibration feature, sound feature, current feature and temperature feature, it is combined into corresponding vibration vector, sound vector, current vector and temperature vector respectively;

[0108] Sample distance square value calculation unit: determines the distance square value between every two samples in the training set based on the vibration vector, the sound vector, the current vector and the temperature vector;

[0109]

[0110] Where Xa, Xb represent two samples in the training set, 1≤Xa≤N3, 1≤Xb≤N3, and Xa≠Xb, Vn represents the number of standard sub-features of vibration features, Sn represents the number of standard sub-features of sound features, Cn represents the number of standard sub-features of current features, and Tn represents the number of standard sub-features of temperature features; Xa k1 represents the value of sample Xa based on the k1th sub-feature; Xb k1represents the value of sample Xb based on the k1th sub-feature; Z k1 Represents the feature conversion coefficient of the k1th sub-feature;

[0111] Arrangement unit: Randomly The squared distance values ​​between the samples are arranged in ascending order, and the sorted The middle value mD(Xa, Xb) among the values ​​is taken as the square value of the interval distance, where [] is the integer symbol, N3 represents the number of samples in the training set,

[0112] Adjustment coefficient calculation unit: Calculate adjustment parameters based on all distance square values ​​and interval distance square values

[0113]

[0114]

[0115]

[0116] Among them, α1 and α2 represent the weight coefficients corresponding to the average distance adjustment formula and the interval distance adjustment formula, respectively, |D(Xa, Xb) L1 -mD(Xa, Xb)| represents the absolute difference between the L1th distance square value and the interval distance square value; A01 represents the average distance adjustment formula, and A02 represents the interval distance adjustment formula;

[0117] Sample similarity value calculation unit: determines the similarity value between two corresponding samples based on the adjustment parameter and the squared distance between every two samples in the training set;

[0118]

[0119] Construction unit: construct a similarity matrix based on the similarity values ​​between any two samples, wherein the rows in the similarity matrix are similarity values ​​between the same sample and different samples;

[0120] Training set sample vector unit: a sample vector of the corresponding sample is formed according to the row corresponding to each sample in the similarity matrix and the fault label corresponding to the sample;

[0121] Model building unit: input all sample vectors into the neural network model to build a wind turbine intelligent management and control model;

[0122] Test set sample vector unit: calculates the similarity between each test sample and all training samples and forms the corresponding test sample vector;

[0123] Determining unit: inputting the test sample vector into the wind turbine intelligent management and control model, determining the output labeling result of each test sample, wherein the output labeling result includes: output fault type and fault handling, wherein the fault type includes no fault, resettable fault, and non-resettable fault, and the non-resettable fault includes fault content and fault level;

[0124] Evaluation unit: compares the fault labeling results of the test samples and the output labeling results, accumulates the number of true positive examples, true negative examples, false positive examples, and false negative examples in the test set, and evaluates the model performance.

[0125] In this embodiment, the fault label includes the fault type and the fault handling.

[0126] In this embodiment, the dimension of the vibration vector is determined based on the feature quantity of the vibration feature.

[0127] In this embodiment, the dimension of the sound vector is determined based on the number of features of the sound feature.

[0128] In this embodiment, the dimension of the current vector is determined based on the number of features of the current feature.

[0129] In this embodiment, the dimension of the temperature vector is determined based on the feature quantity of the temperature feature.

[0130] In this embodiment, the square distance value D(Xa, Xb) between samples represents the square of the distance between samples Xa and Xb.

[0131] In this embodiment, the similarity value K(Xa, Xb) between samples represents the similarity value between samples Xa and Xb.

[0132] In this embodiment, the size of the similarity matrix is ​​N3×N3, and each element is a similarity value between a corresponding row sample and a column sample.

[0133] In this embodiment, the similarity matrix maps the vibration characteristics, sound characteristics, current characteristics, and temperature characteristics of the sample into an infinite-dimensional high-dimensional space, so that the originally nonlinearly separable samples become linearly separable in this high-dimensional space.

[0134] In this embodiment, the length of the sample vector is N3+1, and is composed of similarity values ​​between the sample and all other samples in the training set and the fault label corresponding to the sample.

[0135] In this embodiment, the length of the test set sample vector is N3, and is composed of similarity values ​​between the test set sample and N3 training set samples.

[0136] In this embodiment, the true example indicates that the result is faulty and consistent with the fault label.

[0137] In this embodiment, a true negative example indicates that the result is no fault and is labeled as no fault.

[0138] In this embodiment, a false positive example indicates that the result is no fault but has a fault label.

[0139] In this embodiment, a false negative example indicates that the result is a fault but there is no fault labeling.

[0140] The beneficial effects of the above technical solution are as follows: a wind turbine intelligent management and control model is constructed based on the sample vectors of all training set samples in the historical operation data, and the model performance is evaluated by comparing the fault types, fault handling and fault labeling corresponding to the sample vectors of all test set samples in the historical operation data. This can improve the efficiency and accuracy of fault diagnosis, better respond to new fault modes and changing operating conditions, optimize fault handling strategies, improve maintenance efficiency and reduce costs.

[0141] Embodiment 5:

[0142] The embodiment of the present invention provides a wind turbine intelligent management and control system of a new energy centralized control center, wherein the analysis module includes:

[0143] A setting unit: setting an autonomous reset threshold for the resettable fault;

[0144] Comparison unit: compares the number of reset times with the autonomous reset threshold. If the number of reset times is less than the autonomous reset threshold, the resettable fault is autonomously reset, and the number of reset times is increased by 1.

[0145] Reset unit: if the number of resets is equal to the autonomous reset threshold, the fault type is changed from the resettable fault to a non-resettable fault, and at the same time, the number of resets is reset.

[0146] In this embodiment, the autonomous reset threshold represents the maximum number of reset times.

[0147] In this embodiment, after the number of resets reaches the autonomous reset threshold, autonomous reset is no longer possible.

[0148] In this embodiment, after the fault is changed to a non-resettable fault and the fault is processed, the reset times are cleared and autonomous reset can be performed.

[0149] The beneficial effects of the above technical solution are as follows: by setting an autonomous reset threshold for the resettable fault of the fan, the fan can be reset autonomously when the number of resets is lower than the autonomous reset threshold. Under the premise of ensuring the stable operation of the fan, measures can be taken quickly to restore the normal operating state, reduce manual intervention and improve the response speed of the fan.

[0150] Embodiment 6:

[0151] The embodiment of the present invention provides a wind turbine intelligent management and control system of a new energy centralized control center, wherein the analysis module includes:

[0152] Alarm unit: setting a fault alarm for the non-resettable fault based on the corresponding fault content and fault level;

[0153] The fault alarm consists of a fault level alarm and a fault content alarm.

[0154] In this embodiment, the fault level alarm is determined based on the fault level of the non-resettable fault. The higher the fault level, the higher and more frequent the fault level alarm.

[0155] In this embodiment, the fault content alarm is determined based on the fault content of the non-resettable fault, and the fault content alarm has different alarm sounds according to different fault contents.

[0156] In this embodiment, the sound of the fault alarm is determined based on the fault content alarm, and the tone and frequency of the fault alarm are determined based on the fault level alarm.

[0157] The beneficial effects of the above technical solution are as follows: by setting a fault alarm for a non-resettable fault, the non-resettable fault can be discovered in time, the fault can be prevented from further expanding, the time for fault handling can be shortened, the fan downtime can be reduced, and the system reliability can be improved.

[0158] Embodiment 7:

[0159] The embodiment of the present invention provides a wind turbine intelligent management and control system of a new energy centralized control center, wherein the analysis module includes:

[0160] Identification unit: performing in-depth analysis on the test set samples of the false negative examples to identify the focus area of ​​the wind turbine;

[0161] Prevention unit: Strengthen monitoring of the concerned areas to prevent potential failures.

[0162] In this embodiment, the test set samples of false negative examples indicate that the result of the intelligent wind turbine management and control model is a faulty but not faulty label, and the possibility of a fault occurring in subsequent operation is high.

[0163] In this embodiment, the test set samples of false negative examples are deeply analyzed to determine the probability of failure in subsequent operation, and the test set samples of false negative examples with high probability are determined as key areas of sub-health concern.

[0164] In this embodiment, when the current operating status of the sub-health key focus area changes, potential fault prevention is performed before the fault alarm.

[0165] The beneficial effects of the above technical solution are as follows: by preventing potential faults in sub-healthy key areas, potential faults can be discovered in advance and preventive measures and regular maintenance can be taken, thereby reducing the failure rate in sub-healthy areas, extending the service life of the fan, and reducing maintenance costs.

[0166] Embodiment 8:

[0167] The present invention provides a wind turbine intelligent control system of a new energy centralized control center, wherein the control module comprises:

[0168] Input unit: input the current operating status of the fan into the intelligent control model of the fan to determine the fault type and fault handling corresponding to the current operating status;

[0169] Generation unit: Generates a fault warning sheet based on the fault type and fault handling corresponding to the current operating status.

[0170] In this embodiment, the fault warning sheet includes the fault location, fault time, fault type and fault alarm.

[0171] In this embodiment, intelligent management and control is achieved by performing fault processing on the wind turbine based on the fault processing corresponding to the current operating status and the fault warning sheet.

[0172] The beneficial effects of the above technical solution are as follows: by real-time monitoring of the current operating status of the fan and generating a fault warning sheet for fault handling based on the analysis results of the fan intelligent management and control model, the current operating status of the fan can be diagnosed in a timely and accurate manner, so that timely and effective fault handling measures can be taken when a fault warning sheet is issued, thereby reducing production losses and improving operating efficiency and reliability.

[0173] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A wind turbine intelligent control system for a new energy centralized control center, characterized in that: include: Acquisition module: acquiring the historical operation data of the new energy centralized control center under the historical operation time, and performing feature extraction on the historical operation data to determine the first feature; Analysis module: Analyze historical operation data and the first feature, build a wind turbine intelligent management and control model, and determine the fault type and fault handling based on the wind turbine intelligent management and control model; Management module: obtains the current operating status of the fan in real time, analyzes the current operating status based on the fan intelligent management and control model to generate a fault warning sheet, and performs fault handling on the fan based on the fault warning sheet to achieve intelligent management and control; Wherein, the acquisition module includes: The first extraction unit is used to extract the vibration value, sound value, current value and temperature value of the fan under the same unit time and hourly operation time from the historical operation data; Characteristic autocorrelation value calculation unit: respectively calculates the vibration autocorrelation value based on the vibration value, the sound autocorrelation value based on the sound value, and the current autocorrelation value based on the current value of the fan between every two same unit times, and at the same time, calculates the temperature autocorrelation value of the fan based on the temperature value between every two same unit times; in, It represents the vibration autocorrelation value between two units of time when the lag parameter is K. It represents the sound autocorrelation value between two units of time when the lag parameter is K. It represents the current autocorrelation value between two units of time when the hysteresis parameter is K. It represents the temperature autocorrelation value between two units of time when the lag parameter is K. , represents the lag parameter, [] is the integer symbol, N1 represents the number of unit time, N2 represents the length of hours in the same unit time, They represent the vibration value, sound value, current value, and temperature value at the j1st hour of operation time, respectively. , , They represent the sum of vibration values, sound values, current values, and temperature values ​​in the i1th unit time, respectively. , , Respectively represent The sum of vibration values, sound values, current values, and temperature values ​​within a unit of time. , , They represent the average vibration value, average sound value, average current value, and average temperature value in N1 unit time respectively; Autocorrelation sequence unit: based on The vibration autocorrelation values ​​determine the vibration autocorrelation sequence of the fan, based on The sound autocorrelation value determines the sound autocorrelation sequence of the fan and the The current autocorrelation values ​​determine the current autocorrelation sequence of the wind turbine. The temperature autocorrelation values ​​determine the temperature autocorrelation sequence of the fan; Autocorrelation diagram unit: based on all lag parameters and vibration autocorrelation sequence, sound autocorrelation sequence, current autocorrelation sequence, temperature autocorrelation sequence, draw the corresponding vibration autocorrelation diagram, sound autocorrelation diagram, current autocorrelation diagram and temperature autocorrelation diagram; Characteristic unit: analyzing the vibration autocorrelation graph to determine the vibration characteristics of the fan, analyzing the sound autocorrelation graph to determine the sound characteristics of the fan, analyzing the current autocorrelation graph to determine the current characteristics of the fan, and analyzing the temperature autocorrelation graph to determine the temperature characteristics of the fan.

2. According to the intelligent wind turbine control system of the new energy centralized control center of claim 1, it is characterized by: The first characteristics include: vibration characteristics, sound characteristics, current characteristics and temperature characteristics.

3. According to the intelligent wind turbine control system of the new energy centralized control center of claim 1, it is characterized in that: The analysis module comprises: Annotation unit: annotate the data of fan failure in historical operation data; Division unit: divide the annotated historical running data into training set and test set; Feature vector unit: based on the vibration feature, sound feature, current feature and temperature feature, it is combined into corresponding vibration vector, sound vector, current vector and temperature vector respectively; Sample distance square value calculation unit: determines the distance square value between every two samples in the training set based on the vibration vector, the sound vector, the current vector and the temperature vector; in, Represent two samples in the training set, , The number of standard sub-features representing the vibration feature, The number of standard sub-features representing the sound characteristics, represents the number of standard sub-features of the current feature, Represents the number of standard sub-features of the temperature feature; Representation sample Based on the value of the k1th sub-feature; Representation sample Based on the value of the k1th sub-feature; Represents the feature conversion coefficient of the k1th sub-feature; Arrangement unit: Randomly The squared distance values ​​between the samples are arranged in ascending order, and the sorted The middle value among the values As the square value of the interval distance, [] is the integer symbol, N3 represents the number of samples in the training set, Adjustment coefficient calculation unit: Calculate adjustment parameters based on all distance square values ​​and interval distance square values ; in, 2 represent the weight coefficients corresponding to the average distance adjustment formula and the interval distance adjustment formula, It represents the absolute difference between the L1th distance square value and the interval distance square value; A01 represents the average distance adjustment formula, and A02 represents the interval distance adjustment formula; Sample similarity value calculation unit: determines the similarity value between two corresponding samples based on the adjustment parameter and the squared distance between every two samples in the training set; Construction unit: construct a similarity matrix based on the similarity values ​​between any two samples, wherein the rows in the similarity matrix are similarity values ​​between the same sample and different samples; Training set sample vector unit: a sample vector of the corresponding sample is formed according to the row corresponding to each sample in the similarity matrix and the fault label corresponding to the sample; Model building unit: input all sample vectors into the neural network model to build a wind turbine intelligent management and control model; Test set sample vector unit: calculates the similarity between each test sample and all training samples and forms the corresponding test sample vector; Determining unit: inputting the test sample vector into the wind turbine intelligent management and control model, determining the output labeling result of each test sample, wherein the output labeling result includes: output fault type and fault handling, wherein the fault type includes no fault, resettable fault, and non-resettable fault, and the non-resettable fault includes fault content and fault level; Evaluation unit: compares the fault labeling results of the test samples and the output labeling results, accumulates the number of true positive examples, true negative examples, false positive examples, and false negative examples in the test set, and evaluates the model performance.

4. According to claim 3, the intelligent wind turbine control system of the new energy centralized control center is characterized in that: The analysis module further includes: A setting unit: setting an autonomous reset threshold for the resettable fault; Comparison unit: compares the number of reset times with the autonomous reset threshold. If the number of reset times is less than the autonomous reset threshold, the resettable fault is autonomously reset, and the number of reset times is increased by 1. Reset unit: if the number of resets is equal to the autonomous reset threshold, the fault type is changed from the resettable fault to a non-resettable fault, and at the same time, the number of resets is reset.

5. According to the intelligent wind turbine control system of the new energy centralized control center of claim 3, it is characterized in that: The analysis module further includes: Alarm unit: setting a fault alarm for the non-resettable fault based on the corresponding fault content and fault level; The fault alarm consists of a fault level alarm and a fault content alarm.

6. According to the intelligent wind turbine control system of the new energy centralized control center of claim 3, it is characterized in that: The analysis module further includes: Identification unit: performing in-depth analysis on the test set samples of the false negative examples to identify the focus area of ​​the wind turbine; Prevention unit: Strengthen monitoring of the concerned areas to prevent potential failures.

7. According to the wind turbine intelligent control system of the new energy centralized control center of claim 1, it is characterized by: The management module comprises: Input unit: input the current operating status of the fan into the intelligent control model of the fan to determine the fault type and fault handling corresponding to the current operating status; Generation unit: Generates a fault warning sheet based on the fault type and fault handling corresponding to the current operating status.

Images