A remote monitoring method and system for micro-wind generators based on cloud platform

Abstract

The present invention relates to the field of generator operation status monitoring, and specifically to a remote monitoring method and system for a micro-wind generator based on a cloud platform. The method includes: obtaining historical parameter data of the micro-wind generator; calculating the coordination degree between all dimensional parameters in each parameter data; and obtaining the coordination degree sequence corresponding to all moments; using a period to divide the coordination degree sequence into multiple coordination segments, and obtaining the label of each coordination segment; each coordination segment and the label constitute a set of training sets, and using the obtained multiple sets of training sets to respectively train multiple monitoring models to obtain the accuracy of each monitoring model; using each monitoring model to monitor the current parameter data collected in real time to obtain each abnormal probability; using each accuracy rate to perform weighted summation on all abnormal probabilities to obtain the final abnormal probability. That is, the scheme of the present invention can accurately monitor the real-time operation status of the micro-wind generator.

Description

Technical Field

[0001] The present invention relates to the field of generator operation status monitoring, and more specifically, to a remote monitoring method and system for a micro-wind generator based on a cloud platform. Background Art

[0002] Micro wind generators mainly refer to wind power generation equipment with low power and suitable for use in families or small communities. Compared with traditional large wind turbines, micro wind generators have low power and flexible application. They are suitable for remote areas, rural areas or small-scale application scenarios, and can provide power support for single families or small buildings. However, they are greatly affected by the environment. For example, the output power of micro wind generators is greatly affected by climate, wind speed and environmental conditions. Therefore, it is necessary to monitor its operating status in real time to ensure its efficient operation.

[0003] Therefore, how to achieve remote, real-time and accurate monitoring and management to ensure the stable and long-term operation of the wind turbine has become an urgent problem to be solved.

[0004] In the related art, for example, a remote monitoring system and method for wind power generation disclosed in publication number CN118148831A obtains monitoring report data, and performs real-time operation simulation of wind turbines based on the uploaded data through a virtual model module in the operation and maintenance system. The fault management module analyzes the monitoring report data to determine the fault of the wind turbine generator set and sends an alarm message. When performing the analysis, the data of the monitoring report is compared with the preset risk threshold to determine whether there is a fault.

[0005] Although the above method can be used to diagnose faults of generator sets, the output power of the breeze generator is greatly affected by climate, wind speed and environmental conditions. Therefore, if only multiple dimensional data are compared with the thresholds respectively, it is difficult to determine the actual fault condition of the breeze generator, resulting in the problem of being unable to accurately monitor the breeze generator. Summary of the invention

[0006] The purpose of the present invention is to propose a remote monitoring method and system for a micro-wind generator based on a cloud platform, so as to solve the problem that the micro-wind generator cannot be accurately monitored in the prior art; to this end, the present invention provides solutions in the following two aspects.

[0007] In a first aspect, the present invention provides a remote monitoring method for a micro-wind generator based on a cloud platform, comprising:

[0008] Acquire historical parameter data of the micro-wind generator, wherein the historical parameter data includes parameter data at multiple moments, and each parameter data includes multi-dimensional parameters;

[0009] Calculate the coordination degree between all dimension parameters in each parameter data. for: ; , are the number of differential values ​​greater than or equal to 0 and the number of differential values ​​less than 0 in the feature vector of the parameter data at time t, respectively; K is the number of dimensions; the feature vector is composed of the differential value of the target parameter in the parameter data at time t and the differential value of any other dimension at the corresponding moment after the delay time compared with the target parameter; the target parameter is a parameter of any dimension; and the synergy degree sequence corresponding to all moments is obtained;

[0010] The synergy degree sequence is divided into multiple synergy segments using a period, and labels of each synergy segment are obtained; each synergy segment and the label constitute a training set, and the period is the time difference between any two parameter data corresponding to the maximum autocorrelation coefficient;

[0011] Use the obtained multiple training sets to train multiple monitoring models respectively to obtain the accuracy of each monitoring model; during training, one set of training sets corresponds to one monitoring model;

[0012] Each monitoring model is used to monitor the current parameter data collected in real time to obtain each abnormality probability; all abnormality probabilities are weighted and summed using each accuracy rate to obtain the final abnormality probability.

[0013] In the above scheme, by analyzing the delay time between different dimensional parameters in the historical parameter data of the micro wind generator, the coordination degree between each dimensional parameter considering the delay time is calculated. Since the coordination degree takes into account the consistency of changes and the consistency of change amplitudes between different dimensional parameters, it can accurately characterize the consistency between different dimensional parameters, and then obtain the periodicity between each parameter data in the historical parameter data to divide the coordination degree sequence, and use the divided coordination segments to train and establish multiple monitoring models; and use the trained multiple monitoring models to monitor the current parameter data of the micro wind generator collected in real time, and integrate the output results of different monitoring models into one according to the accuracy of the model to obtain the abnormal probability of the micro wind generator, thereby improving the accuracy and reliability of the monitoring of the micro wind generator.

[0014] Optionally, the method further comprises the step of correcting the coordination degree to obtain a corrected coordination degree; the corrected coordination degree for:

[0015] ;

[0016] in, Represents the degree of coordination between all dimensional parameters in the parameter data at time t, Indicates natural normal, is the variance of all difference values ​​in the eigenvector.

[0017] In the above scheme, the accurate coordination degree can be obtained by combining the time change and the amplitude change of the parameters.

[0018] Optionally, the multi-dimensional parameters include at least temperature, voltage, current and wind force.

[0019] Optionally, the final abnormal probability is:

[0020] ;

[0021] in, is the final abnormal probability of the current operating state of the micro-wind generator, represents the abnormal probability output by the i-th monitoring model, is the normalized value of the accuracy of the i-th monitoring model, , represents the accuracy of the i-th monitoring model, and I is the number of monitoring models.

[0022] In the above scheme, by weighting multiple abnormality probabilities using the accuracy of the detection model, an accurate final abnormality probability can be obtained.

[0023] Optionally, the process of obtaining the difference value is:

[0024] Get the parameter sequence of the same dimension parameter at different times;

[0025] Take any parameter sequence as the target sequence, calculate the difference between two adjacent parameters in the target sequence, and obtain the difference at each moment in the target sequence.

[0026] Optionally, the delay time is the time difference between any two-dimensional parameters corresponding to the maximum mutual correlation coefficient.

[0027] Introducing the delay time in the above scheme can obtain data of different dimensions with high consistency of changes, which is convenient for subsequent accurate analysis of parameters of different dimensions.

[0028] Optionally, the monitoring of the current parameter data collected in real time by using each monitoring model to obtain each abnormality probability includes:

[0029] Acquire the coordination degree sequence of current parameter data of the micro-wind generator collected in real time;

[0030] Dividing the coordination degree sequence of the current parameter data according to the period to obtain multiple groups of current coordination segments;

[0031] Input each current collaborative segment into the corresponding monitoring model and output the corresponding abnormal probability.

[0032] The above scheme can accurately obtain the abnormal probabilities of different current coordination segments.

[0033] Optionally, the label includes 1 and / or 0, wherein 1 indicates that the operating state of the micro-wind generator corresponding to the cooperative segment is abnormal, and 0 indicates that the operating state of the micro-wind generator corresponding to the cooperative segment is normal.

[0034] In a second aspect, a remote monitoring system for a micro-wind generator based on a cloud platform comprises:

[0035] processor;

[0036] A memory stores computer instructions for remote monitoring of micro-wind generators based on a cloud platform. When the computer instructions are executed by the processor, the system executes the above-mentioned remote monitoring method of micro-wind generators based on a cloud platform.

[0037] The beneficial effects of the present invention are:

[0038] The solution of the present invention takes into account the hysteresis between the various dimensional parameters collected by the micro wind generator and the coordination between the various dimensional parameters, and can obtain a more accurate degree of coordination to describe the operating status of the micro wind generator, and obtain accurate multiple monitoring models to monitor the real-time operating status of the micro wind generator, thereby improving the operating reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0040] Figure 1 A flowchart of a remote monitoring method for a micro-wind generator based on a cloud platform in this embodiment is schematically shown;

[0041] Figure 2 The structural block diagram of a remote monitoring system for a micro-wind generator based on a cloud platform in this embodiment is schematically shown. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0043] The present invention is directed to remote monitoring of the operating state of a micro-wind generator, so as to monitor whether the operating state of the micro-wind generator is abnormal.

[0044] The remote monitoring device for a micro-wind generator based on a cloud platform in this embodiment includes a monitoring unit and a cloud platform, and the monitoring unit is connected to the cloud platform.

[0045] The monitoring unit includes a collection module, a sensor module and a transmission module; and is used to collect parameter data of the breeze generator in real time. The collection module collects the parameter data of the sensor module in real time, and sends the parameter data to the cloud platform through the transmission module.

[0046] The sensor modules include anemometer, wind vane, temperature and humidity sensor, voltage sensor, current sensor, etc., which are used to collect wind force, wind direction, temperature, humidity, voltage, and current to form the parameter data of the breeze generator.

[0047] The cloud platform includes an analysis module and an alarm module. The analysis module is used to receive parameter data uploaded by the monitoring system and analyze the parameter data to determine whether the micro-wind generator has a fault. The alarm module issues an alarm based on the received fault result.

[0048] The alarm module may be a voice alarm.

[0049] When analyzing the parameter data, a remote monitoring method of a micro-wind generator based on a cloud platform is adopted to accurately analyze the parameter data collected in real time, realize the monitoring of the micro-wind generator, and improve the reliability of the operation of the micro-wind generator.

[0050] Specifically, Figure 1 As shown, a remote monitoring method for a micro-wind generator based on a cloud platform in this embodiment includes the following steps:

[0051] Step S1, obtaining multiple training sets.

[0052] In this embodiment, the acquisition of multiple training sets includes steps S11 to S13, which are specifically:

[0053] Step S11, obtaining historical parameter data of the micro-wind generator.

[0054] Specifically, the historical parameter data of the micro-wind generator includes parameter data at multiple moments in a historical time period; wherein each parameter data includes multi-dimensional parameters, and the same-dimensional parameters at different moments constitute a parameter sequence.

[0055] The dimensions of the multidimensional parameters in this embodiment include at least two combinations of temperature, voltage, current and wind force. Of course, the dimensions of the multidimensional parameters may also include wind direction.

[0056] It should be noted that the above historical parameter data is the data of the historical operating status of the breeze generator, and the various dimensional parameters thereof are collected by using an anemometer, an anemometer, a temperature and humidity sensor, a voltage sensor, and a current sensor.

[0057] In this embodiment, historical parameter data of multiple micro-wind generators of the same model may be obtained, and historical parameter data of the current micro-wind generator may also be obtained.

[0058] Step S12, calculating the coordination degree between all dimensional parameters in each parameter data, and obtaining the coordination degree sequence corresponding to all moments.

[0059] In the process of wind power generation, when different dimensional parameters are collected at the same time, there will be a certain delay between the parameters. For example, the wind force is detected first, and the voltage and current can only be detected after wind power generation. Therefore, when calculating the synergy between parameters, it is necessary to first calculate the delay time between different dimensional parameters.

[0060] In this embodiment, the cross-correlation function is used to calculate the delay time between parameters of different dimensions. Since the cross-correlation function describes the degree of correlation between the values ​​of different random signals at any two different times, the delay time between parameters of different dimensions in the historical parameter data is determined by the cross-correlation function.

[0061] Specifically, taking any dimensional parameter sequence as the target parameter sequence, respectively calculate the mutual correlation coefficient between each target parameter in the target parameter sequence and the parameters in any other dimensional parameter sequence, select the moment corresponding to the parameter when the mutual correlation coefficient is the largest as the delayed moment, so that the difference between the delayed moment and the moment corresponding to the target parameter can be obtained, and the difference is the delay time.

[0062] The above delay time is a delay of a period of time relative to the time of collecting the target parameter. However, in fact, the parameter corresponding to the time after the delay corresponds to the target parameter.

[0063] For example, the wind speed collected at time t may be delayed due to the collection of different dimensional parameters. The current or voltage can be collected at time t. The current or voltage at each moment is corresponding.

[0064] The purpose of acquiring the above delay time is to obtain data of different dimensions with high consistency of changes, so as to facilitate the subsequent accurate analysis of parameters of different dimensions.

[0065] After the delay time between different dimensional parameters is obtained, the synergy between the different dimensional parameters at the corresponding delay time is calculated.

[0066] Specifically, the method for obtaining the degree of coordination is:

[0067] First, obtain the difference sequence of any parameter sequence. For any parameter sequence, first calculate the difference sequence of the sequence. The difference sequence can reflect the change of parameters over time. The parameter situation at a single moment is difficult to reflect the real situation of the generator.

[0068] Secondly, obtain the characteristic vector of the parameter data at each moment.

[0069] Specifically, the feature vector is composed of the differential value of the target parameter in the parameter data at any moment and the differential value of any other dimension compared with the target parameter at the corresponding moment after a delay time.

[0070] For example, let the feature vector at time t be ,in, represents the difference value of the first dimension parameter at time t, Indicates that the second dimension parameter is The difference value of time, It represents the delay time of the second dimension parameter relative to the first dimension parameter at time t. Indicates that the K-th dimension parameter is The difference value of time, It shows the delay time of the K-th dimension parameter relative to the 1-th dimension parameter at time t.

[0071] The above delay time is due to the fact that when different dimensional parameters are collected, although in theory they are collected at the same time, in reality the data of different dimensions are not collected synchronously, and there is a time delay between them. Therefore, by obtaining the delay time, the corresponding relationship between the real data of different dimensions can be found, which can improve the accuracy of subsequent data analysis.

[0072] Then, the degree of coordination is calculated.

[0073] In one embodiment, the degree of coordination for: ; , They are respectively the number of differential values ​​greater than or equal to 0 and the number of differential values ​​less than 0 in the eigenvector of the parameter data at time t, and K is the number of dimensions.

[0074] Furthermore, since the above-mentioned coordination degree only considers the changes of various dimensional parameters over time, during the power generation process, the large differences in the amplitude of changes between the parameters also indicate that the generator may be abnormal. Therefore, it is also necessary to consider the amplitude of change over time to correct the coordination degree.

[0075] Corrected Coordination for:

[0076] ;

[0077] in, Represents the degree of coordination between all dimensional parameters in the parameter data at time t, represents a natural constant, is the variance of all difference values ​​in the eigenvector.

[0078] in, It represents variance, which reflects the difference in the amplitude of changes among parameters at time t. The larger the variance, the greater the difference in amplitude change, and the lower the coordination between parameters of different dimensions.

[0079] In this embodiment, the coordination degree sequence at all times can be obtained.

[0080] The synergy between the different dimensional parameters in this embodiment reflects the consistency of the changes between the different dimensional parameters, that is, when the generator works normally, the changes between its various dimensional parameters should be consistent. Therefore, the synergy between the calculated parameters can characterize the operating status of the generator.

[0081] Step S13, using the period to divide the synergy degree sequence into multiple synergy segments, and obtaining the label of each synergy segment; each synergy segment and the label constitute a training set, and multiple training sets are obtained.

[0082] In this embodiment, the period between parameter data at different times in the historical parameter data is also obtained, which is specifically obtained by using an autocorrelation function.

[0083] Since the autocorrelation function describes the correlation between the values ​​of a random signal at any two moments, and indicates the correlation between the same sequence at different moments, it is a digital tool for finding repeated patterns. Therefore, in this embodiment, the cycle between parameter data at different moments in the historical parameter data is calculated by the autocorrelation function.

[0084] Specifically, the autocorrelation coefficient between any two historical parameter data is calculated, and the time difference corresponding to the two parameter data when the autocorrelation coefficient is maximum is selected as the period.

[0085] After obtaining the cycle, the coordination degree sequence is divided to obtain coordination segments under multiple cycles.

[0086] After obtaining the coordination segments in each cycle, in this embodiment, a label is given to each coordination segment by combining the historical working status of the micro-wind generator and the actual situation, and the label is normal or abnormal.

[0087] The label in this embodiment includes 1 and / or 0, wherein 1 indicates that the operation state of the micro-wind generator corresponding to the coordination segment is abnormal, and 0 indicates that the operation state of the micro-wind generator corresponding to the coordination segment is normal.

[0088] Each collaborative segment and the corresponding label constitute a training set, and multiple training sets are obtained.

[0089] Step S2: Use the obtained multiple training sets to train multiple monitoring models respectively to obtain the accuracy of each monitoring model.

[0090] In this embodiment, the same number of monitoring models as the number of training sets are constructed. Specifically, the monitoring model may be an LSTM (Long Short-Term Memory) model. Of course, as another implementation, the monitoring model may also be a support vector machine.

[0091] Taking the LSTM model as an example, the specific training process is:

[0092] The collaborative segments in any group of training sets are used as the input of the LSTM model. The output of the model is the operating status of the generator. The cross entropy loss is used to calculate the loss between the output and the label, and the gradient descent algorithm is used to update the model parameters. When the model reaches the set maximum number of training times or the loss of the model is less than the set threshold, the model stops training. After the training is completed, the accuracy corresponding to the group of training sets is obtained.

[0093] The above accuracy is the ratio of the correct number predicted by the monitoring model to the total number.

[0094] Step S3, using each monitoring model to monitor the current parameter data collected in real time to obtain each abnormality probability; using each accuracy rate to perform weighted summation on all abnormality probabilities to obtain the final abnormality probability.

[0095] In this embodiment, each monitoring model is used to monitor the parameter data of the micro-wind generator collected in real time to obtain the abnormal probability output by each monitoring model, including:

[0096] First, the coordination degree sequence of the current parameter data of the micro-wind generator collected in real time is obtained. Specifically, the coordination degree between all dimensional data in the parameter data at different moments in the current set time period is first calculated, and then the coordination degree sequence corresponding to all moments in the current set time period is obtained.

[0097] Since the calculation method of the synergy degree in this embodiment is the same as that in step S1, the only difference is that the values ​​of the parameters of each dimension in the collected parameter data are different. Therefore, the specific acquisition process of the synergy degree sequence is not introduced here.

[0098] Secondly, the coordination degree sequence of the current parameter data is divided according to the period to obtain multiple groups of current coordination segments. In this embodiment, the current parameter data is divided using the period corresponding to the historical parameter data to ensure the correspondence with the monitoring model.

[0099] Then, each current coordination segment is input into the corresponding monitoring model, and the corresponding abnormal probability is output. Specifically, the coordination degree sequences of different lengths are input into the corresponding generator state monitoring model to output the corresponding abnormal probability. It should be noted that when collecting parameter data during the operation of the micro-wind generator in real time, since the amount of data collected at the beginning is small and cannot meet the input requirements of some models, only the monitoring model that meets the input conditions is used to monitor the real-time data. As the amount of data increases, the number of models during integration increases.

[0100] The final abnormal probability in this embodiment is:

[0101] ;

[0102] in, is the final abnormal probability of the current operating state of the micro-wind generator, represents the abnormal probability output by the i-th monitoring model, is the normalized value of the accuracy of the i-th monitoring model.

[0103] In one embodiment, the normalized value for: , represents the accuracy of the i-th monitoring model, and I is the number of monitoring models.

[0104] In another embodiment, the normalized value may also be obtained by using a maximum and minimum value normalization method.

[0105] After obtaining the final abnormal probability, it is compared with the threshold to determine whether the operating state of the microwind generator is abnormal; that is, when the final abnormal probability is greater than or equal to the threshold, the operating state of the microwind generator is abnormal, and the staff is reminded to shut down or repair it in time.

[0106] The solution of the present invention analyzes the delay time between different dimensional parameters in the historical parameter data of the micro-wind generator, calculates the coordination degree between the various dimensional parameters under the consideration of the delay time, and since the coordination degree takes into account the consistency of the changes and the consistency of the change amplitude between the different dimensional parameters, it can accurately characterize the consistency between the different dimensional parameters, and then obtain the periodicity between the various parameter data in the historical parameter data to divide the coordination degree sequence, and use the divided coordination segments to train and establish multiple monitoring models; and use the trained multiple monitoring models to monitor the current parameter data of the micro-wind generator collected in real time, and integrate the output results of different monitoring models into one according to the accuracy of the model, and obtain the abnormal probability of the micro-wind generator to realize the monitoring of the micro-wind generator. Since the present invention takes into account the hysteresis between the various dimensional parameters of the micro-wind generator and the coordination between the various dimensional parameters, it makes it more accurate to describe the characteristics of the micro-wind generator, and thus it is more capable of monitoring the operating status of the micro-wind generator and improving the operating reliability.

[0107] The present invention also provides a remote monitoring system for micro-wind generators based on a cloud platform. Figure 2 As shown, the system includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a remote monitoring method for a micro-wind generator based on a cloud platform according to the present invention is implemented.

[0108] The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface, whose configuration and functions are known in the art and thus will not be described in detail here.

[0109] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented by computer-readable / executable instructions stored or otherwise maintained by such a computer-readable medium.

[0110] In the description of this specification, “plurality” means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0111] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A remote monitoring method for a micro-wind generator based on a cloud platform, characterized in that: include: Acquire historical parameter data of the micro-wind generator, wherein the historical parameter data includes parameter data at multiple moments, and each parameter data includes multi-dimensional parameters; Calculate the coordination degree between all dimension parameters in each parameter data. for: ; , are the number of differential values ​​greater than or equal to 0 and the number of differential values ​​less than 0 in the feature vector of the parameter data at time t, respectively; K is the number of dimensions; the feature vector is composed of the differential value of the target parameter in the parameter data at time t and the differential value of any other dimension at the corresponding moment after the delay time compared with the target parameter; the target parameter is a parameter of any dimension; and the synergy degree sequence corresponding to all moments is obtained; The synergy degree sequence is divided into multiple synergy segments using a period, and labels of each synergy segment are obtained; each synergy segment and the label constitute a training set; the period is the time difference between any two parameter data corresponding to the maximum autocorrelation coefficient; Use the obtained multiple training sets to train multiple monitoring models respectively to obtain the accuracy of each monitoring model; during training, one set of training sets corresponds to one monitoring model; Each monitoring model is used to monitor the current parameter data collected in real time to obtain each abnormality probability; all abnormality probabilities are weighted and summed using each accuracy rate to obtain the final abnormality probability.

2. A remote monitoring method for a micro-wind generator based on a cloud platform according to claim 1, characterized in that: The step of correcting the coordination degree to obtain a corrected coordination degree is also included; the corrected coordination degree for: ； in, Represents the degree of coordination between all dimensional parameters in the parameter data at time t, represents a natural constant, is the variance of all difference values ​​in the eigenvector.

3. A remote monitoring method for a micro-wind generator based on a cloud platform according to claim 2, characterized in that: The multi-dimensional parameters include at least temperature, voltage, current and wind force.

4. A remote monitoring method for a micro-wind generator based on a cloud platform according to claim 1, characterized in that: The final abnormal probability is: ； in, is the final abnormal probability of the current operating state of the micro-wind generator, represents the abnormal probability output by the i-th monitoring model, is the normalized value of the accuracy of the i-th monitoring model, , represents the accuracy of the i-th monitoring model, and I is the number of monitoring models.

5. The remote monitoring method of a micro-wind generator based on a cloud platform according to claim 1 is characterized in that: The process of obtaining the difference value is as follows: Get the parameter sequence of the same dimension parameter at different times; Take any parameter sequence as the target sequence, calculate the difference between two adjacent parameters in the target sequence, and obtain the difference at each moment in the target sequence.

6. A remote monitoring method for a micro-wind generator based on a cloud platform according to claim 1, characterized in that: The delay time is the time difference between any two-dimensional parameters corresponding to the maximum mutual correlation coefficient.

7. A remote monitoring method for a micro-wind generator based on a cloud platform according to claim 5, characterized in that: The monitoring model is used to monitor the current parameter data collected in real time to obtain each abnormal probability, including: Acquire the coordination degree sequence of current parameter data of the micro-wind generator collected in real time; Dividing the coordination degree sequence of the current parameter data according to the period to obtain multiple groups of current coordination segments; Input each current collaborative segment into the corresponding monitoring model and output the corresponding abnormal probability.

8. The remote monitoring method of a micro-wind generator based on a cloud platform according to claim 1 is characterized in that: The label includes 1 and / or 0, wherein 1 indicates that the operation state of the micro-wind generator corresponding to the coordination segment is abnormal, and 0 indicates that the operation state of the micro-wind generator corresponding to the coordination segment is normal.

9. A remote monitoring system for micro-wind generators based on a cloud platform, characterized in that: include: processor; A memory storing computer instructions for remote monitoring of micro-wind generators based on a cloud platform. When the computer instructions are executed by the processor, the system executes a remote monitoring method for micro-wind generators based on a cloud platform according to any one of claims 1 to 8.

Images