Fan fault supervision system and method based on intelligent algorithm

Abstract

The invention discloses a fan fault supervision system and method based on an intelligent algorithm, and relates to the technical field of fan fault supervision, and the method comprises the steps: analyzing the fault correlation degree between a monitoring parameter in a historical equipment monitoring record and a reference fan, and obtaining key monitoring data; evaluating the fault monitoring adaptation degree between key monitoring parameters in the key monitoring data and the fan to obtain abnormal monitoring data; acquiring abnormal monitoring data, acquiring a reference historical equipment monitoring record of a reference fan, analyzing abnormal key monitoring parameters in the abnormal monitoring data and a data change state in the reference historical equipment monitoring record, and correcting the operation range of the abnormal key monitoring parameters in the parameter operation range to obtain a target parameter operation range; and performing equipment fault evaluation on the fan in the current period to obtain fault data, and sending the fault data to a worker through the fault supervision platform to manage the fan fault.

Description

Technical Field

[0001] The present invention relates to the technical field of fan fault supervision, and specifically to a fan fault supervision system and method based on intelligent algorithms. Background Art

[0002] Power generation fans are usually referred to as wind turbines or wind power generator sets, which are devices that can convert wind energy into electrical energy. They are usually installed in wind farms to utilize large-area wind resources to provide clean and renewable energy for the power grid. Traditional methods for supervising fan faults usually rely on regular inspections and manual monitoring, and cannot monitor the operating status of fans in real time. In addition, because it mainly relies on manual monitoring, it is impossible to continuously analyze the data collected by sensors in real time, making it difficult to detect potential problems of fans in a timely manner. And even if the data collected by sensors is uploaded to the server through the Internet of Things, only a single parameter is used to set the threshold. When a certain parameter exceeds the set range of the threshold, the system will issue a fault alarm. However, the operating states of different fans in different environments are different, and the preset threshold of the system is easily inaccurate, and it is also impossible to accurately identify fan faults according to the actual situation. This will not only cause false fault alarms and affect the power generation efficiency of the fans, but may even lead to undetected faults, resulting in equipment damage. Summary of the Invention

[0003] The purpose of the present invention is to provide a fan fault supervision system and method based on intelligent algorithms to solve the problems raised in the prior art.

[0004] To achieve the above purpose, the present invention provides the following technical solution: A fan fault supervision method based on intelligent algorithms, the method includes:

[0005] Step S100: Obtain the historical equipment monitoring records of the reference fan of the fan, analyze the degree of fault correlation between the monitoring parameters in the historical equipment monitoring records and the reference fan, and obtain key monitoring data;

[0006] Step S200: Obtain the key monitoring data, obtain the parameter operating range of the fan from the fault supervision platform, evaluate the degree of fault monitoring adaptation between the key monitoring parameters in the key monitoring data and the fan, and obtain abnormal monitoring data;

[0007] Step S300: Obtain the abnormal monitoring data, obtain the reference historical equipment monitoring records of the reference fan, analyze the abnormal key monitoring parameters in the abnormal detection data, the data change state in the reference historical equipment monitoring records, and correct the operating range of the abnormal key monitoring parameters in the parameter operating range to obtain the target parameter operating range;

[0008] Step S400: Obtain the operating range of the target parameters, evaluate the equipment failures of the fans within the current cycle, obtain the failure data, and send the failure data to the staff through the failure supervision platform to manage the fan failures.

[0009] Further, step S100 includes:

[0010] Step S101: Obtain the area where the fan is installed, obtain the equipment model of the fan, obtain other fans with the same equipment model as the fan within the area, and record them as the reference fans of the fan;

[0011] Step S102: Obtain the historical equipment monitoring records of the reference fans. From the historical equipment monitoring records, obtain the data corresponding to the monitoring parameters of the reference fans collected by the sensors, and analyze the degree of fault correlation between the monitoring parameters in the historical equipment monitoring records and the reference fans. The specific analysis process is as follows:

[0012] Preprocess the monitoring parameters in each historical equipment monitoring record of the reference fans, and standardize the monitoring parameters in each historical equipment monitoring record;

[0013] Step S103: And construct the data matrix Z of the reference fan according to the average value of the monitoring parameters in each historical equipment monitoring record;

[0014] Construct the covariance matrix C of the data matrix Z. The specific formula is: C = [1 / n - 1]·Z T ·Z, where n is the total number of each historical equipment monitoring record;

[0015] Perform eigenvalue decomposition on the covariance matrix C to obtain the eigenvalues λ and eigenvectors υ of the covariance matrix C. The specific formula is: C·υ = λ·υ;

[0016] Sort the eigenvalues of the monitoring parameters of the reference fan from small to large and accumulate them. The accumulated value L of the eigenvalues of the current k monitoring parameters k is greater than or equal to the preset accumulated threshold L′, and the accumulated value L of the eigenvalues of the first k - 1 monitoring parameters k is less than the preset accumulated threshold L′, and retain the corresponding several monitoring parameters of the first k monitoring parameters;

[0017] Step S104: When a failure occurs in the reference fan in the historical equipment monitoring record, mark the historical equipment monitoring record;

[0018] Obtain the preset value k > 0. When the historical equipment monitoring record is marked, the failure value of the historical equipment monitoring record is k. Otherwise, the failure value of the historical equipment monitoring record is 0;

[0019] Sort the historical device monitoring records according to the time points of each historical device monitoring record, and collect the failure values and monitoring parameters of each historical device monitoring record respectively to obtain the failure value set of the reference fan and the parameter set S of the monitoring parameters;

[0020] Step S105: Obtain the ranks of each element in the failure value set and the parameter set S in order of magnitude, calculate the square of the difference in ranks between each element in the failure value set and the parameter set S, and accumulate them to obtain the squared rank Q′, and calculate the failure correlation value ρ between the monitoring parameter and the reference fan;

[0021] Step S106: When the absolute value of the failure correlation value ρ is greater than the preset failure correlation threshold ρ′, determine that there is a failure correlation between the monitoring parameter and the reference fan, mark the monitoring parameter, and obtain several marked monitoring parameters of the reference fan;

[0022] Step S107: Obtain the marked monitoring parameters among the reference fans of the fan, and calculate the key value of each monitoring parameter in the fan. Among them, the key value P of the a-th monitoring parameter in the fan a = B (a,sum) / B sum where B sum is the total number of each reference fan, and B (a,sum) is the total number of reference fans marked for the a-th monitoring parameter;

[0023] When the key value is greater than the preset key threshold, determine that the a-th monitoring parameter is a key monitoring parameter in the failure monitoring of the fan, and record it as a key monitoring parameter. Obtain and collect the key monitoring parameters of the fan to obtain the key monitoring data of the fan.

[0024] Further, step S200 includes:

[0025] Step S201: Obtain the parameter operating range of the fan from the failure supervision platform. The parameter operating range includes the operating ranges of the monitoring parameters in the fan, where the parameter operating ranges of each reference fan in the failure supervision platform are the same as those of the fan;

[0026] Step S202: Evaluate the adaptability of the key monitoring parameters in the key monitoring data to the failure monitoring between the fan and the reference fan. The specific evaluation process is as follows:

[0027] Obtain the time period of the historical device monitoring records marked in the reference fan. When the reference fan does not detect a failure through the monitoring parameters collected by the sensor during the time period, mark the historical device monitoring record as an abnormal device monitoring record of the reference fan;

[0028] Obtain the maximum value E of the key monitoring parameter from the abnormal device monitoring record max and the minimum value E min , obtain the maximum threshold E′ max and the minimum threshold E′ min from the operating range of the key monitoring parameter, calculate the characteristic mean μ′=(E′ max +E′ min ) / 2, and calculate the absolute values F max and F min of the differences between the maximum value E max and the minimum value E min and the maximum threshold E′ max and the minimum threshold E′ min respectively;

[0029] Step S203: Calculate the parameter anomaly value G of the key monitoring parameter in the abnormal device monitoring record G = max{(μ′ / F max ),(μ′ / F min )};

[0030] When the parameter anomaly value G is greater than the preset parameter anomaly threshold, mark the abnormal device monitoring record as the target abnormal device monitoring record of the key monitoring parameter;

[0031] Step S204: Obtain the total number M sum of the abnormal device monitoring records in each reference fan, obtain the total number M (△,sum) of the target abnormal device monitoring records of the key monitoring parameter in each reference fan, calculate the adaptation anomaly value Y = M (△,sum) / M sum , when the adaptation anomaly value Y is greater than the preset adaptation anomaly threshold, determine that the key monitoring parameter in the key monitoring data is not adapted to the fault monitoring of the fan, mark the key monitoring parameter as the abnormal key monitoring parameter of the fan, obtain each abnormal key monitoring parameter of the fan, and perform aggregation to obtain the abnormal monitoring data of the fan.

[0032] Further, step S300 includes:

[0033] Step S301: Obtain the abnormal monitoring data of the fan, and obtain each abnormal key monitoring parameter of the fan from the abnormal monitoring data;

[0034] Step S302: Obtain the previous historical period of the current period and denote it as the reference historical period, obtain the unmarked historical device monitoring records of the reference fan of the fan in the reference historical period and denote them as the reference historical device monitoring records, and each abnormal key monitoring parameter in the reference historical device monitoring records has been preprocessed;

[0035] Step S303: Analyze the data change status of the abnormal key monitoring parameters in the abnormal detection data. Specifically: Obtain several reference wind turbines containing reference historical equipment monitoring records, obtain several historical equipment monitoring records of the wind turbines that are not marked, and preprocess the data of each abnormal key detection parameter in the several equipment monitoring records;

[0036] Obtain the average value μ of the reference historical equipment monitoring records of several reference wind turbines and the abnormal key monitoring parameters in the several equipment monitoring records △ and the standard deviation σ △ ;

[0037] Step S304: Modify the operating range W of the abnormal key monitoring parameters in the parameter operating range. The specific modification is as follows:

[0038] Obtain a preset coefficient J, and calculate the corrected maximum threshold E′ of the abnormal key monitoring parameters in the reference operating range (t,max) = μ △ + J×σ △ , calculate the corrected minimum threshold E′ of the abnormal key monitoring parameters in the reference operating range (t,max) = μ △ - J×σ △ , obtain the operating range W = [μ △ - J×σ △ , μ △ + J×σ △ of the abnormal key monitoring parameters in the parameter operating range in the current cycle;

[0039] Obtain the operating ranges of the corrected abnormal key monitoring parameters, and replace the operating ranges of the abnormal key monitoring parameters in the parameter operating range to obtain the target parameter operating range of the wind turbine in the current cycle.

[0040] Furthermore, step S400 includes:

[0041] Step S401: Use the preset sensors in the wind turbine to monitor the wind turbine in the current cycle, obtain the data of each monitoring parameter of the preprocessed wind turbine, and obtain the target parameter operating range of the wind turbine;

[0042] Step S402: Conduct equipment fault assessment on the wind turbine in the current cycle. Specifically, when the maximum or minimum value of a certain monitoring parameter in the wind turbine is not within the operating range of a certain monitoring parameter in the target parameter operating range, obtain several fault types corresponding to the abnormality of a certain monitoring parameter, collect the data of a certain monitoring parameter exceeding the operating range and several fault types to obtain fault data, and send the fault data to the staff through the fault supervision platform to manage the wind turbine fault;

[0043] In the above steps, the fan in the current cycle is monitored by sensors, and the operating state of the fan can be obtained in real time, and the fault condition of the fan can be judged more quickly and timely. Using the corrected target parameter operation data can also greatly improve the accuracy of the platform's judgment of fan faults and effectively ensure the normal operation of the fan.

[0044] In order to better implement the above method, a fan fault supervision system based on an intelligent algorithm is also proposed. The system includes a fault-related analysis module, a parameter adaptation evaluation module, an operating range correction module, and a fan fault management module;

[0045] The fault-related analysis module is used to analyze the degree of fault correlation between the monitoring parameters in the historical equipment monitoring records and the reference fan to obtain key monitoring data;

[0046] The parameter adaptation evaluation module is used to evaluate the degree of work monitoring adaptation between the key monitoring parameters in the key monitoring data and the fan to obtain abnormal monitoring data;

[0047] The operating range correction module is used to obtain the reference historical equipment monitoring records of the reference fan and correct the operating range of the abnormal key monitoring parameters in the parameter operating range to obtain the target parameter operating range;

[0048] The fan fault management module is used to evaluate the faults of the fans in the current cycle, obtain the fault data of the fans, and send the fault data to the staff of the fans through the fault supervision platform to manage the fan faults.

[0049] Furthermore, the fault-related analysis module includes a reference fan acquisition unit and a fault-related analysis unit;

[0050] The reference fan acquisition unit is used to obtain the equipment model of the fan and the area where it is located, obtain other fans with the equipment model installed in the area, and record them as the reference fans of the fan;

[0051] The fault-related analysis unit is used to obtain the historical equipment monitoring records of the reference fan, analyze the degree of fault correlation between the monitoring parameters in the historical equipment monitoring records and the reference fan, and obtain the key monitoring data of the fan.

[0052] Furthermore, the parameter adaptation evaluation module includes a record marking unit and a parameter adaptation evaluation unit;

[0053] The record marking unit is used to obtain the key monitoring parameters in the key monitoring data, calculate the parameter abnormal values of the key monitoring parameters in the abnormal equipment monitoring records, and mark the abnormal equipment monitoring records according to the reference abnormal values to obtain the target abnormal equipment monitoring records of the key monitoring parameters;

[0054] The parameter adaptation evaluation unit is used to calculate the adaptation abnormality value of the key monitoring parameter according to the number of target abnormal equipment monitoring records of the key monitoring parameter, evaluate the fault monitoring adaptation degree between the key monitoring parameter and the wind turbine, and obtain abnormal monitoring data.

[0055] Further, the operating range correction module includes a reference data acquisition unit and an operating range correction unit;

[0056] The reference data acquisition unit is used to acquire the previous historical period of the current period and record it as the reference historical period, and acquire the unmarked historical equipment monitoring record of the reference wind turbine in the reference historical period and record it as the reference historical equipment monitoring record;

[0057] The operating range correction unit is used to correct the operating range of the abnormal key monitoring parameter in the parameter operating range to obtain the target parameter operating range.

[0058] Further, the fan fault management module includes a fan fault management unit;

[0059] The fan fault management unit is used to use the preset sensors in the fan to monitor the fan in the current cycle, and to perform equipment fault assessment on the fan in the current cycle, obtain the fault data of the fan, and send the fault data to the staff through the fault monitoring platform to manage the fan fault.

[0060] Compared with the prior art, the beneficial effects of the present invention are: the present invention realizes the accurate identification of fan faults, because when considering fault monitoring of the fan, it mainly relies on the various monitoring parameters of the fan collected by the sensor to determine the fault conditions during the operation of the fan, but different fans are in different scenarios and the equipment models are also different. The present invention obtains the historical equipment monitoring records of reference fans of the same equipment model and area as the fan, and obtains monitoring parameters with a greater degree of relevance to fan fault monitoring, obtains the key monitoring parameters of the fan, and corrects the operating range set by the fault monitoring platform for the key monitoring parameters, which greatly improves the accuracy of the fault monitoring platform's supervision of fan faults and effectively guarantees the normal operation of the fan. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a method flow chart of a fan fault monitoring method based on an intelligent algorithm according to the present invention;

[0062] Figure 2 It is a module schematic diagram of the fan fault monitoring system based on intelligent algorithm of the present invention. DETAILED DESCRIPTION

[0063] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0064] Embodiment: As Figure 1 - Figure 2 shown, the present invention provides a technical solution, a fan fault supervision method based on an intelligent algorithm, and the method includes:

[0065] Step S100: Obtain the historical equipment monitoring records of the reference fan of the fan, analyze the degree of fault correlation between the monitoring parameters in the historical equipment monitoring records and the reference fan, and obtain key monitoring data;

[0066] Among them, step S100 includes:

[0067] Step S101: Obtain the area where the fan is installed, obtain the equipment model of the fan, obtain other fans with the same equipment model as the fan in the area, and record them as the reference fan of the fan;

[0068] Step S102: Obtain the historical equipment monitoring records of the reference fan, obtain the data corresponding to the monitoring parameters of the reference fan collected by the sensor from the historical equipment monitoring records, and analyze the degree of fault correlation between the monitoring parameters in the historical equipment monitoring records and the reference fan. The specific analysis process is as follows:

[0069] For example, the monitoring parameters include vibration amplitude, temperature, rotational speed, power, voiceprint signal, etc.;

[0070] Preprocess the monitoring parameters in each historical equipment monitoring record of the reference fan, and perform standardization processing on the monitoring parameters in each historical equipment monitoring record;

[0071] For example, the preprocessing includes cleaning the data cleaning of the monitoring parameters in the historical equipment monitoring records, processing the missing value processing, and removing the outliers;

[0072] For example, the standardization includes z-standardization and normalization;

[0073] Step S103: And construct a data matrix Z of the reference fan according to the average value of the monitoring parameters in each historical equipment monitoring record;

[0074] Construct the covariance matrix C of the data matrix Z. The specific formula is: C = [1 / n - 1] · Z T · Z, where n is the total number of each historical equipment monitoring record;

[0075] Perform eigenvalue decomposition on the covariance matrix C to obtain the eigenvalues λ and eigenvectors υ of the covariance matrix C. The specific formula is: C·υ = λ·υ;

[0076] According to the eigenvalues of the monitoring parameters of the reference fan, sort them from smallest to largest and accumulate them. The accumulated value L of the eigenvalues of the current k monitoring parameters k is greater than or equal to the preset accumulated threshold L′, and the accumulated value L of the eigenvalues of the previous k - 1 monitoring parameters k is less than the preset accumulated threshold L′, then retain several monitoring parameters corresponding to the previous k monitoring parameters;

[0077] Step S104: When the reference fan in the historical device monitoring record fails, mark the historical device monitoring record;

[0078] Obtain a preset value k > 0. When the historical device monitoring record is marked, the fault value of the historical device monitoring record is k. Otherwise, the fault value of the historical device monitoring record is 0;

[0079] According to the time points of each historical device monitoring record, sort each historical device monitoring record, and collect the fault values and monitoring parameters of each historical device monitoring record respectively to obtain the fault value set of the reference fan and the parameter set S of the monitoring parameters;

[0080] Step S105: In the order of size, obtain the ranks of each element in the fault value set and the parameter set S respectively, calculate the square of the difference in ranks between each element in the fault value set and the parameter set S, and accumulate them to obtain the squared rank Q′, and calculate the fault correlation value ρ between the monitoring parameter and the reference fan;

[0081] For example, the process of calculating the fault correlation value ρ between the vibration amplitude of the reference fan and the reference fan is as follows:

[0082] Set the preset value k to 1 (fault), 0 (normal);

[0083] Obtain the parameter set S = {20, 25, 30, 35, 40} of the vibration amplitude of the reference fan, the fault value set U = {0, 0, 1, 1, 1} of the reference fan, obtain the ranks S′ = {1, 2, 3, 4, 5} of each element in the parameter set S of the reference fan, and obtain the U′ = {1.5, 1.5, 4, 4, 4} (take the average rank because there are two 0s and three 1s) of each element in the fault value set U of the reference fan;

[0084] Calculate the squared rank Q′ = 0.25 + 0.25 + 1 + 0 + 1 = 2.5;

[0085] Calculate the fault - related value ρ between the vibration amplitude of the reference fan and the reference fan:

[0086]

[0087] Step S106: When the absolute value of the fault - related value ρ is greater than the preset fault - related threshold ρ′, it is determined that there is a fault - related between the monitoring parameter and the reference fan, and the monitoring parameter is marked, and several marked monitoring parameters of the reference fan are obtained;

[0088] Step S107: Obtain the marked monitoring parameters among the respective reference fans of the fan, and calculate the key values of the monitoring parameters in the fan. Among them, the key value P of the ath monitoring parameter in the fan a =B (a,sum) / B sum where B sum is the total number of all reference fans, and B (a,sum) is the total number of reference fans with the ath monitoring parameter marked;

[0089] When the key value is greater than the preset key threshold, it is determined that the ath monitoring parameter is a key monitoring parameter in the fault monitoring of the fan, and it is recorded as a key monitoring parameter. Obtain and collect the key monitoring parameters of the fan to obtain the key monitoring data of the fan;

[0090] Step S200: Obtain the key monitoring data, obtain the parameter operating range of the fan from the fault supervision platform, and evaluate the fault monitoring adaptation degree between the key monitoring parameters in the key monitoring data and the fan to obtain the abnormal monitoring data;

[0091] Among them, step S200 includes:

[0092] Step S201: Obtain the parameter operating range of the fan from the fault supervision platform. The parameter operating range includes the operating ranges of the monitoring parameters in the fan. Among them, the parameter operating ranges of the respective reference fans in the fault supervision platform are the same as those of the fan;

[0093] Step S202: Evaluate the fault monitoring adaptation degree between the key monitoring parameters in the key monitoring data and the fan. The specific evaluation process is as follows:

[0094] Obtain the time period of the marked historical equipment monitoring records of the reference fan. When the reference fan does not detect a fault through the monitoring parameters collected by the sensor during the time period, the marked historical equipment monitoring records are recorded as the abnormal equipment monitoring records of the reference fan;

[0095] Obtain the maximum value E max and the minimum value E min from the abnormal equipment monitoring records., obtain the maximum threshold E′ from the operating range of the key monitoring parameter max and the minimum threshold E′ min , calculate the characteristic mean μ′ of the key monitoring parameter = (E′ max +E′ min ) / 2, and calculate the absolute values F max of the differences between the maximum value E min and the minimum value E max and the maximum threshold E′ min and the minimum threshold E′ max respectively, and F min ;

[0096] Step S203: Calculate the parameter outlier G of the key monitoring parameter in the abnormal device monitoring record = max{(μ′ / F max ), (μ′ / F min )};

[0097] When the parameter outlier G is greater than the preset parameter outlier threshold, mark the abnormal device monitoring record as the target abnormal device monitoring record of the key monitoring parameter;

[0098] Step S204: Obtain the total number M sum of the abnormal device monitoring records in each reference fan, obtain the total number M (△,sum) of the target abnormal device monitoring records of the key monitoring parameter in each reference fan, calculate the adaptation outlier Y of the key monitoring parameter = M (△,sum) / M sum , when the adaptation outlier Y is greater than the preset adaptation outlier threshold, determine that the key monitoring parameter in the key monitoring data is not adapted to the fault monitoring of the fan, mark the key monitoring parameter as the abnormal key monitoring parameter of the fan, obtain each abnormal key monitoring parameter of the fan, and perform aggregation to obtain the abnormal monitoring data of the fan;

[0099] Step S300: Obtain the abnormal monitoring data, obtain the reference historical device monitoring records of the reference fan, analyze the data change status of the abnormal key monitoring parameter in the abnormal detection data in the reference historical device monitoring records, and correct the operating range of the abnormal key monitoring parameter in the parameter operating range to obtain the target parameter operating range;

[0100] Among them, step S300 includes:

[0101] Step S301: Obtain the abnormal monitoring data of the fan, and obtain each abnormal key monitoring parameter of the fan from the abnormal monitoring data;

[0102] Step S302: Obtain the previous historical cycle of the current cycle, denoted as the reference historical cycle. Obtain the unmarked historical equipment monitoring records of the reference fan of the wind turbine in the reference historical cycle, denoted as the reference historical equipment monitoring records. All abnormal key monitoring parameters in the reference historical equipment monitoring records have been preprocessed;

[0103] Step S303: Analyze the data change status of the abnormal key monitoring parameters in the abnormal detection data. Specifically: Obtain several reference fans containing the reference historical equipment monitoring records, obtain several unmarked historical equipment monitoring records of the wind turbine, and preprocess the data of the abnormal key detection parameters in the several equipment monitoring records;

[0104] Obtain the average value μ of the abnormal key monitoring parameters in the reference historical equipment monitoring records of several reference fans and the several equipment monitoring records △ and the standard deviation σ △ ;

[0105] Step S304: Modify the operating range W of the abnormal key monitoring parameters in the parameter operating range. Specifically, the modification is as follows:

[0106] Obtain the preset coefficient J, and calculate the corrected maximum threshold E′ of the abnormal key monitoring parameters in the reference operating range (t,max) = μ △ + J×σ △ , calculate the corrected minimum threshold E′ of the abnormal key monitoring parameters in the reference operating range (t,max) = μ △ - J×σ △ , obtain the operating range W = [μ △ - J×σ △ , μ △ + J×σ △ of the abnormal key monitoring parameters in the parameter operating range of the current cycle;

[0107] Obtain the operating ranges of the corrected abnormal key monitoring parameters, and replace the operating ranges of the abnormal key monitoring parameters in the parameter operating range to obtain the target parameter operating range of the wind turbine in the current cycle;

[0108] Step S400: Obtain the target parameter operating range, evaluate the equipment failure of the wind turbine in the current cycle to obtain failure data, and send the failure data to the staff through the failure supervision platform to manage the wind turbine failure;

[0109] Among them, Step S400 includes:

[0110] Step S401: Use the preset sensors in the fan to monitor the fan in the current cycle, obtain the data of various monitored parameters of the pre-processed fan, and obtain the target parameter operating range of the fan;

[0111] Step S402: Conduct equipment fault assessment on the fan in the current cycle, including, when the maximum or minimum value of a certain monitored parameter in the fan is not within the operating range of a certain monitored parameter in the target parameter operating range, obtain several fault types corresponding to the abnormality of a certain monitored parameter, collect the data where a certain monitored parameter exceeds the operating range and several fault types to obtain fault data, and send the fault data to the staff through the fault supervision platform to manage the fan fault;

[0112] For example, an excessive vibration amplitude may indicate unbalance, alignment problems or damaged bearings of the fan;

[0113] The temperature exceeding the operating threshold may indicate insufficient lubrication or motor overload;

[0114] For example, several fault types include insufficient lubrication, motor overload, cracks in the fan blades, etc.;

[0115] To better implement the above method, a fan fault supervision system based on an intelligent algorithm is also proposed. The system includes a fault correlation analysis module, a parameter adaptation evaluation module, an operating range correction module and a fan fault management module;

[0116] The fault correlation analysis module is used to analyze the fault correlation degree between the monitored parameters in the historical equipment monitoring records and the reference fan to obtain key monitoring data;

[0117] The parameter adaptation evaluation module is used to evaluate the working monitoring adaptation degree between the key monitoring parameters in the key monitoring data and the fan to obtain abnormal monitoring data;

[0118] The operating range correction module is used to obtain the reference historical equipment monitoring records of the reference fan, and correct the operating range of the abnormal key monitoring parameters in the parameter operating range to obtain the target parameter operating range;

[0119] The fan fault management module is used to conduct fault assessment on the fan in the current cycle, obtain the fault data of the fan, and send the fault data to the staff of the fan through the fault supervision platform to manage the fan fault;

[0120] Among them, the fault correlation analysis module includes a reference fan acquisition unit and a fault correlation analysis unit;

[0121] A reference fan acquisition unit is used to acquire the equipment model of the fan and the area where it is located, acquire other fans with the equipment model installed in the area, and record them as the reference fans of the fan;

[0122] A fault-related analysis unit is used to acquire the historical equipment monitoring records of the reference fan, analyze the degree of fault correlation between each monitoring parameter in the historical equipment monitoring records and the reference fan, and obtain the key monitoring data of the fan;

[0123] Among them, the parameter adaptation evaluation module includes a record marking unit and a parameter adaptation evaluation unit;

[0124] The record marking unit is used to acquire the key monitoring parameters in the key monitoring data, calculate the parameter anomaly value of the key monitoring parameters in the abnormal equipment monitoring records, and mark the abnormal equipment monitoring records according to the reference anomaly value to obtain the target abnormal equipment monitoring records of the key monitoring parameters;

[0125] The parameter adaptation evaluation unit is used to calculate the adaptation anomaly value of the key monitoring parameters according to the number of target abnormal equipment monitoring records of the key monitoring parameters, evaluate the fault monitoring adaptation degree between the key monitoring parameters and the fan, and obtain the abnormal monitoring data;

[0126] Among them, the operating range correction module includes a reference data acquisition unit and an operating range correction unit;

[0127] The reference data acquisition unit is used to acquire the previous historical cycle of the current cycle and record it as the reference historical cycle, acquire the unmarked historical equipment monitoring records of the reference fan in the reference historical cycle, and record them as the reference historical equipment monitoring records;

[0128] The operating range correction unit is used to correct the operating range of the abnormal key monitoring parameters in the parameter operating range to obtain the target parameter operating range;

[0129] Among them, the fan fault management module includes a fan fault management unit;

[0130] The fan fault management unit is used to monitor the fan in the current cycle using a preset sensor in the fan, conduct equipment fault assessment on the fan in the current cycle, obtain the fault data of the fan, and send the fault data to the staff through the fault supervision platform to manage the fan fault.

[0131] It is obvious to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Accordingly, all changes that fall within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be construed as limiting the claims concerned.

Claims

1. A fan fault supervision method based on an intelligent algorithm, characterized in that The method includes: Step S100: Obtain the historical equipment monitoring records of the reference fan of the wind turbine, analyze the degree of fault correlation between the monitoring parameters in the historical equipment monitoring records and the reference fan, and obtain key monitoring data; Step S200: Obtain the key monitoring data, obtain the parameter operating range of the wind turbine from the fault supervision platform, evaluate the degree of fault monitoring adaptation between the key monitoring parameters in the key monitoring data and the wind turbine, and obtain abnormal monitoring data; Step S300: Obtain the abnormal monitoring data, obtain the reference historical equipment monitoring records of the reference fan, analyze the abnormal key monitoring parameters in the abnormal detection data, the data change status in the reference historical equipment monitoring records, and correct the operating range of the abnormal key monitoring parameters in the parameter operating range to obtain the target parameter operating range; Step S400: Obtain the target parameter operating range, evaluate the equipment fault of the wind turbine in the current cycle, obtain fault data, and send the fault data to the staff through the fault supervision platform to manage the wind turbine fault.

2. The blower fault supervision method based on intelligent algorithm according to claim 1, characterized in that, The step S100 includes: Step S101: Obtain the area where the wind turbine is installed, obtain the equipment model of the wind turbine, obtain other wind turbines with the same equipment model as the wind turbine in the area, and record them as the reference fan of the wind turbine; Step S102: Obtain the historical equipment monitoring records of the reference fan, obtain the data corresponding to the monitoring parameters of the reference fan collected by the sensor from the historical equipment monitoring records, and analyze the degree of fault correlation between the monitoring parameters in the historical equipment monitoring records and the reference fan. The specific analysis process is as follows: Preprocess the monitoring parameters in each historical equipment monitoring record of the reference fan, and standardize the monitoring parameters in each historical equipment monitoring record; Step S103: And construct the data matrix Z of the reference fan according to the average value of the monitoring parameters in each historical equipment monitoring record; Construct the covariance matrix C of the data matrix Z, and the specific formula is: C = [1 / n - 1] · Z T · Z, where n is the total number of the respective historical device monitoring records; Perform eigenvalue decomposition on the covariance matrix C to obtain the eigenvalues λ and eigenvectors υ of the covariance matrix C. The specific formula is: C·υ = λ·υ; Sort the characteristic values of the monitoring parameters of the reference fan from smallest to largest and accumulate them. The accumulated value L of the characteristic values of the current k monitoring parameters k is greater than or equal to the preset accumulated threshold L′, and the accumulated value L of the characteristic values of the first k - 1 monitoring parameters k is less than the preset accumulated threshold L′, then retain the several monitoring parameters corresponding to the first k monitoring parameters; Step S104: When the reference fan in the historical equipment monitoring record fails, mark the historical equipment monitoring record; Obtain a preset value k>0. When the historical equipment monitoring record is marked, the fault value of the historical equipment monitoring record is k. Otherwise, the fault value of the historical equipment monitoring record is 0; Sort the historical equipment monitoring records according to the time points of each historical equipment monitoring record, and collect the fault values and monitoring parameters of each historical equipment monitoring record respectively to obtain the fault value set of the reference fan and the parameter set S of the monitoring parameters; Step S105: Obtain the ranks of each element in the set of fault values and the set of parameters S in order of magnitude, calculate the square of the difference in ranks between each element in the set of fault values and the set of parameters S, and accumulate them to obtain the squared rank Q′, and calculate the fault correlation value ρ between the monitored parameter and the reference fan; Step S106: When the absolute value of the fault correlation value ρ is greater than the preset fault correlation threshold ρ′, determine that there is a fault correlation between the monitored parameter and the reference fan, mark the monitored parameter, and obtain several monitored parameters marked for the reference fan; Step S107: Obtain the marked monitoring parameters in each reference fan of the fan, and calculate the key values of the monitoring parameters in the fan. Among them, the key value P of the a-th monitoring parameter in the fan a = B (a,sum) / B sum , B sum is the total number of the respective reference fans, and B (a,sum) is the total number of the reference fans in which the a-th monitoring parameter is marked; When the key value is greater than the preset key threshold, determine that the a-th monitored parameter is a key monitored parameter in the fault monitoring of the fan, and record it as a key monitored parameter. Obtain and collect the key monitored parameters of the fan to obtain the key monitoring data of the fan.

3. The fan fault supervision method based on intelligent algorithm according to claim 2, wherein The step S200 includes: Step S201: Obtain the parameter operating range of the fan from the fault supervision platform. The parameter operating range includes the operating ranges of various monitored parameters in the fan. Among them, the parameter operating ranges of the respective reference fans in the fault supervision platform are the same as those of the fan; Step S202: Evaluate the adaptability of the key monitored parameters in the key monitoring data to the fault monitoring of the fan. The specific evaluation process is as follows: Obtain the time period of the historical equipment monitoring record marked for the reference fan. When the reference fan does not detect a fault through various monitored parameters collected by the sensor during this time period, record the marked historical equipment monitoring record as the abnormal equipment monitoring record of the reference fan; Obtain the maximum value E of the key monitoring parameter from the abnormal device monitoring record max and the minimum value E min , obtain the maximum threshold E′ from the operating range of the key monitoring parameter max and the minimum threshold E′ min , calculate the characteristic mean μ′ of the key monitoring parameter = (E′ max +E′ min ) / 2, respectively calculate the absolute values F max of the differences between the maximum value E min and the minimum value E max and the maximum threshold E′ min and the minimum threshold E′ max and F min ; Step S203: Calculate the parameter anomaly value G of the key monitoring parameter in the abnormal device monitoring record as G = max{(μ′ / F max ),(μ′ / F min )}; When the parameter abnormal value G is greater than the preset parameter abnormal threshold, mark the abnormal equipment monitoring record as the target abnormal equipment monitoring record of the key monitored parameter; Step S204: Obtain the total number M of abnormal device monitoring records in each of the reference wind turbines sum , obtain the total number M of target abnormal device monitoring records of the key monitoring parameters in each of the reference wind turbines (△,sum) , calculate the adapted abnormal value Y of the key monitoring parameter as Y = M (△,sum) / M sum , when the adapted abnormal value Y is greater than a preset adapted abnormal threshold, determine that the key monitoring parameter in the key monitoring data is not adapted to the fault monitoring of the wind turbine, record the key monitoring parameter as the abnormal key monitoring parameter of the wind turbine, obtain each abnormal key monitoring parameter of the wind turbine, and perform aggregation to obtain the abnormal monitoring data of the wind turbine 4. The method for monitoring fan faults based on intelligent algorithms according to claim 2, wherein, The step S300 includes: Step S301: Obtain the abnormal monitoring data of the fan, and obtain various abnormal key monitored parameters of the fan from the abnormal monitoring data; Step S302: Obtain the previous historical period of the current period and record it as the reference historical period. Obtain the unmarked historical equipment monitoring record of the reference fan of the fan in the reference historical period and record it as the reference historical equipment monitoring record. The various abnormal key monitored parameters in the reference historical equipment monitoring record have been preprocessed; Step S303: Analyze the data change status of the abnormal key monitored parameters in the abnormal detection data. Specifically: Obtain several reference fans containing the reference historical equipment monitoring record, obtain several unmarked historical equipment monitoring records of the fan, and preprocess the data of the various abnormal key detection parameters in the several equipment monitoring records; Obtain the average value μ of the reference historical device monitoring records of the several reference wind turbines and the abnormal key monitoring parameters in the several device monitoring records △ and the standard deviation σ △ ; Step S304: Modify the operating range W of the abnormal key monitored parameter in the parameter operating range. The specific modification is as follows: Obtain the preset coefficient J, and calculate the corrected maximum threshold E′ of the abnormal key monitoring parameter in the reference operating range (t,max) = μ △ + J×σ △ , calculate the corrected minimum threshold E′ of the abnormal key monitoring parameter in the reference operating range (t,max) = μ △ - J×σ △ , obtain the operating range W = [μ △ - J×σ △ , μ △ + J×σ △ of the abnormal key monitoring parameter in the parameter operating range in the current period; Obtain the operating ranges of the corrected abnormal key monitoring parameters, and replace the operating ranges of the abnormal key monitoring parameters in the parameter operating ranges to obtain the target parameter operating ranges of the fan in the current cycle.

5. The fan fault supervision method based on an intelligent algorithm according to claim 4, characterized in that, The step S400 includes: Step S401: Use the preset sensors in the fan to monitor the fan in the current cycle, obtain the data of the monitoring parameters of the fan after preprocessing, and obtain the target parameter operating ranges of the fan; Step S402: Conduct an equipment failure assessment on the fan in the current cycle, including, when the maximum or minimum value of a certain monitoring parameter in the fan is not within the operating range of the certain monitoring parameter in the target parameter operating ranges, obtain several types of failure corresponding to the abnormality of the certain monitoring parameter, collect the data of the certain monitoring parameter exceeding the operating range and several types of failure, obtain failure data, and send the failure data to the staff through the failure supervision platform to manage the fan failure.

6. A fan fault supervision system based on an intelligent algorithm, which is used to execute the fan fault supervision method based on an intelligent algorithm described in any one of claims 1-5, and is characterized in that The system includes a failure-related analysis module, a parameter adaptation evaluation module, an operating range correction module, and a fan failure management module; The failure-related analysis module is used to analyze the failure-related degree between the monitoring parameters in the historical equipment monitoring records and the reference fan to obtain key monitoring data; The parameter adaptation evaluation module is used to evaluate the working monitoring adaptation degree between the key monitoring parameters in the key monitoring data and the fan to obtain abnormal monitoring data; The operating range correction module is used to obtain the reference historical equipment monitoring records of the reference fan, correct the operating ranges of the abnormal key monitoring parameters in the parameter operating ranges to obtain the target parameter operating ranges; The fan failure management module is used to conduct a failure assessment on the fan in the current cycle, obtain the failure data of the fan, and send the failure data to the staff of the fan through the failure supervision platform to manage the fan failure.

7. The fan fault supervision system based on an intelligent algorithm according to claim 6, wherein The failure-related analysis module includes a reference fan acquisition unit and a failure-related analysis unit; The reference fan acquisition unit is used to obtain the equipment model of the fan and the area where it is located, obtain other fans of the equipment model installed in the area, and record them as the reference fans of the fan; The failure-related analysis unit is used to obtain the historical equipment monitoring records of the reference fan, analyze the failure-related degree between the monitoring parameters in the historical equipment monitoring records and the reference fan, and obtain the key monitoring data of the fan.

8. The fan fault supervision system based on intelligent algorithm according to claim 6, wherein, The parameter adaptation evaluation module includes a record marking unit and a parameter adaptation evaluation unit; The record marking unit is used to obtain the key monitoring parameters in the key monitoring data, calculate the parameter anomaly value of the key monitoring parameters in the abnormal device monitoring record, and mark the abnormal device monitoring record according to the reference anomaly value to obtain the target abnormal device monitoring record of the key monitoring parameters; The parameter adaptation evaluation unit is used to calculate the adaptation anomaly value of the key monitoring parameters according to the number of target abnormal device monitoring records of the key monitoring parameters, evaluate the fault monitoring adaptation degree between the key monitoring parameters and the fan, and obtain the abnormal monitoring data.

9. The fan fault supervision system based on an intelligent algorithm according to claim 6, wherein The operating range correction module includes a reference data acquisition unit and an operating range correction unit; The reference data acquisition unit is used to obtain the previous historical cycle of the current cycle, denoted as the reference historical cycle, and obtain the unmarked historical device monitoring records of the reference fan in the reference historical cycle, denoted as the reference historical device monitoring records; The operating range correction unit is used to correct the operating range of the abnormal key monitoring parameters in the parameter operating range to obtain the target parameter operating range.

10. The fan fault supervision system based on intelligent algorithm according to claim 6, characterized in that, The fan fault management module includes a fan fault management unit; The fan fault management unit is used to monitor the fan in the current cycle using a preset sensor in the fan, evaluate the device fault of the fan in the current cycle, obtain the fault data of the fan, and send the fault data to the staff through the fault supervision platform to manage the fan fault.

Images