CN119471358A - Wind driven generator motor fault on-line detection method based on current sensing - Google Patents
Wind driven generator motor fault on-line detection method based on current sensing Download PDFInfo
- Publication number
- CN119471358A CN119471358A CN202411456751.4A CN202411456751A CN119471358A CN 119471358 A CN119471358 A CN 119471358A CN 202411456751 A CN202411456751 A CN 202411456751A CN 119471358 A CN119471358 A CN 119471358A
- Authority
- CN
- China
- Prior art keywords
- current
- motor
- value
- data
- normal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Protection Of Generators And Motors (AREA)
Abstract
本发明属于风力发电技术领域，公开了一种基于电流感知的风力发电机电机故障在线检测方法。该方法基于预先搭建的电流采集系统实时采集获取电机电流数据，电流采集系统中的云端服务器首先接收每个电流互感器上传的对应电机的电流数据，并将每个电机的电流数据进行数据清洗，然后将过滤掉电机的启动电流后的电机电流数据进行阈值对比，再进行基于KNN聚类算法进行同类型、多电机横向电流对比，以及单个电机纵向历史电流数据对比，分级判断电机运行状态是否正常。本发明中的风力发电机电机故障在线检测方法解决了振动监测方法只能分析电机机械性故障的弊端，多维度预警大大提高了电机故障判断的准确性。
The present invention belongs to the technical field of wind power generation, and discloses a method for online detection of motor faults of wind turbines based on current sensing. The method is based on real-time acquisition of motor current data by a pre-built current acquisition system. The cloud server in the current acquisition system first receives the current data of the corresponding motor uploaded by each current transformer, and cleans the current data of each motor. Then, the motor current data after filtering out the starting current of the motor is threshold-compared, and then the lateral current comparison of the same type and multiple motors, as well as the longitudinal historical current data comparison of a single motor are performed based on the KNN clustering algorithm to grade and judge whether the motor operation status is normal. The online detection method for motor faults of wind turbines in the present invention solves the drawback that the vibration monitoring method can only analyze the mechanical faults of the motor, and the multi-dimensional early warning greatly improves the accuracy of motor fault judgment.
Description
Technical Field
The invention belongs to the field of wind power generation, and relates to a wind power generator motor fault online detection method based on current sensing.
Background
At present, motor faults of a wind driven generator are mainly represented by two types of motor faults, namely yaw motor faults, the wind driven generator is installed in coastal and hilly areas, turbulence intensity is relatively high, a unit transmission system is required to bear more fatigue loads, a transmission function of the yaw system of the wind driven generator unit is mainly realized by virtue of yaw motors and yaw speed reducers, a unit yaw braking function is realized by virtue of yaw motor electromagnetic braking, yaw motor faults seriously influence yaw direction adjustment of a motor to restrict wind power generation efficiency, secondly, the wind driven generator faults are heat dissipation motor faults, various ventilation heat dissipation motors are more in a heat dissipation system in a cabin of the wind driven generator, the faults of the heat dissipation motor influence mechanical systems and electrical systems of the whole wind driven generator, and the faults of any heat dissipation motor can influence normal operation of the wind driven generator and even seriously influence the service life of the wind driven generator system.
At present, the main problems causing the operation failure of the yaw motor and the heat dissipation motor are the situation of motor bearing damage jam and inter-phase or inter-turn short circuit of the internal winding of the motor. In the current stage, fault monitoring of the yaw motor and the heat dissipation motor is mainly achieved through a vibration monitoring method, as shown in fig. 2, vibration sensors are installed in the horizontal direction and the vertical direction of the yaw motor and the heat dissipation motor, parameters such as vibration peak values, effective values and frequencies are measured, and the running state and the condition of the motor are judged through data such as vibration frequencies during motor running, so that data reference is provided for motor fault monitoring. However vibration monitoring has the following problems:
The method for monitoring the vibration only can judge whether the motor has mechanical problems or not and cannot judge whether the motor has circuit problems or not, and the scheme core monitoring target of the vibration monitoring is to analyze the mechanical problems of a motor bearing, a gear and the like through vibration, but the mechanical problems are only one of the manifestations of motor faults, and the method for monitoring the vibration cannot analyze the circuit problems, so that certain limitations exist. Secondly, the vibration monitoring has the precision problem, the vibration sensor is interfered by the vibration of the field installation environment and other equipment, the misjudgment on the running state of the motor is easy to be caused, the frequency change at the initial stage of the motor failure is slight, and the vibration sensor is difficult to capture. Thirdly, the site construction degree of difficulty is high, the deployment is difficult, yaw motor, heat dissipation motor all have fixed position in wind generating set to the installation space of most motors is narrow and small, and cabin space is also narrow and is made wiring, wiring very difficult, leads to vibration sensor to be difficult to install or can not install, and construction degree of difficulty and cost are relatively higher.
Based on this, a detection scheme of motor faults of the wind driven generator is needed to ensure stable operation of the wind driven generator device.
Disclosure of Invention
The invention aims to provide a wind driven generator motor fault on-line detection method based on current sensing, which is based on a pre-built current acquisition system to remotely acquire current data of each motor in a wind driven generator in real time, and (3) carrying out data cleaning on the current data of each motor, and then carrying out threshold comparison, transverse current comparison of multiple motors of the same type and longitudinal historical current data comparison of a single motor so as to judge whether the running state of the motor is normal in a grading manner.
In order to achieve the above purpose, the invention adopts the following technical scheme:
The wind driven generator motor fault online detection method based on current sensing is realized based on a current acquisition system which is built in advance, wherein the current acquisition system comprises a cloud server and a plurality of current transformers;
Each current transformer respectively and correspondingly acquires current data of a certain phase in three-phase current of any motor in the wind driven generator, and the acquired current data is uploaded to a cloud server through an intelligent gateway;
The wind driven generator motor fault on-line detection method based on current sensing comprises the following steps:
step 1, a cloud server receives current data of a corresponding motor uploaded by each current transformer in a current acquisition system, and performs data cleaning on the current data so as to filter out starting current of the corresponding motor;
step 2, roughly judging the threshold value of the motor current data after the data in the step 1 are cleaned;
When the current value of the motor collected by the current transformer does not exceed the preset threshold, the step 3 is switched to;
Step 3, comparing the motor currents which do not exceed the threshold value with the same type and multiple motors in the transverse direction based on KNN clustering;
Comparing the motor current with relatively small data difference with the rated current of the motor, and judging whether the motor current is in a preset safety range or not;
if the motor current value is judged to be within the preset safety range, the motor current is further transferred to the step 4 to perform longitudinal historical current data comparison, and if the motor current value is not within the preset safety range, the motor current value is directly judged to be abnormal;
Step 4, further comparing the motor current value with the normal value of the motor current history of the motor current value;
The average value of the historical current normal value of the motor is the average value of the historical current data of the motor under normal operation;
If the difference value between the current value and the average value of the historical normal value of the motor current is within the preset range, the motor is judged to be normal in operation, otherwise, the development trend of the difference value between the current value of the motor and each historical normal value is further judged;
If the difference development trend gradually increases along with the time change, the circuit or internal mechanical problem situation is gradually worsened, and if the difference development trend gradually stabilizes along with the time change, the circuit or internal mechanical is judged to be caused by dust or circuit aging.
In addition, on the basis of the on-line detection method of the motor faults of the wind driven generator based on current sensing, the invention also provides a wind driven generator motor fault on-line detection system based on current sensing, which is adaptive to the on-line detection method of the motor faults of the wind driven generator based on current sensing, and the technical scheme is as follows:
An on-line detection system for motor faults of a wind driven generator based on current sensing, comprising:
the data cleaning module is used for carrying out data cleaning on the current data of the corresponding motor uploaded by each current transformer in the current acquisition system received by the cloud server so as to filter out the starting current of the motor;
The threshold judgment module is used for roughly judging the threshold of the motor current data after the data cleaning;
judging that the motor fails when the current value of the motor collected by the current transformer exceeds a preset threshold value, and switching to a multi-motor transverse current comparison module of the same type when the current value of the motor collected by the current transformer does not exceed the preset threshold value;
The same-type multi-motor transverse current comparison module is used for comparing motor currents which do not exceed a threshold value with the same-type multi-motor transverse currents further based on a KNN clustering algorithm;
the motor current with relatively small data difference is transferred to a longitudinal historical current data comparison module to be judged, and whether the motor current with relatively large data difference is in a preset safety range is judged by comparing the motor current with a rated current value;
If the motor current is within the preset safety range, the longitudinal historical current data comparison module is switched to further judge, and if the motor current is not within the preset safety range, the motor is directly judged to be abnormal;
The longitudinal historical current data comparison module is used for comparing the motor current value with the normal value of the motor current history of the motor current value;
The average value of the historical current normal value of the motor is the average value of the historical current data of the motor under normal operation;
If the difference value between the current value and the average value of the historical normal value of the motor current is within the preset range, the motor is judged to be normal in operation, otherwise, the development trend of the difference value between the current value of the motor and each historical normal value is further judged;
and further distinguishing the specific abnormal situation of the motor according to whether the change of the difference development trend along with the time is stable.
In addition, on the basis of the current sensing-based wind driven generator motor fault online detection method, the invention further provides computer equipment which comprises a memory and one or more processors, wherein executable codes are stored in the memory, and the processor is used for realizing the steps of the current sensing-based wind driven generator motor fault online detection method when executing the executable codes.
In addition, on the basis of the current sensing-based wind driven generator motor fault online detection method, the invention further provides a computer readable storage medium, wherein a program is stored on the computer readable storage medium, and the program is used for realizing the current sensing-based wind driven generator motor fault online detection method when being executed by a processor.
The invention has the following advantages:
As described above, the invention provides a wind driven generator motor fault online detection method based on current sensing, which solves the defect that the traditional vibration monitoring method can only analyze mechanical faults of a motor, and can rapidly position a circuit problem of the motor or the mechanical faults of the motor through motor current data, for example, when A, B, C three-phase current data of the motor run exceed a preset threshold value, the motor is confirmed to have great mechanical faults to cause the integral increase of motor current, when the current value of one phase of A, B, C three-phase current of the motor is 0 or the current value of one phase is increased, the circuit of the motor is judged to have problems, the corresponding current is in a default phase when the current of one phase is 0, and the corresponding circuit connector lug is loosened to cause the increase of resistance when the current value of one phase is increased. After the current value of the motor is subjected to comparison of the same type and multiple motor transverse currents and the longitudinal historical current data of a single motor, the development change of the running health state of the motor can be finely judged, for example, a certain motor current does not exceed a preset threshold value, the motor is indicated to run normally in the current stage, the motor current enters the comparison stage of the same type and multiple motor transverse currents, and if the motor current value deviates from the same type of motor current for transverse comparison, the motor is judged to have a tiny mechanical fault or a loose phase line head. The method adopts multidimensional early warning, the current data of each motor is subjected to data cleaning and filtering to remove the starting current of the motor, and then threshold comparison, transverse current comparison of multiple motors of the same type and longitudinal historical current data comparison of a single motor are carried out, and a grading judgment mode combining rough judgment and refined judgment is adopted, so that the accuracy of motor fault judgment is greatly improved.
Drawings
FIG. 1 is a block diagram of a method for online detecting motor faults of a wind driven generator based on current sensing in an embodiment of the invention;
FIG. 2 is a schematic diagram of a vibration detection method;
FIG. 3 is a schematic diagram of a current collection system according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a current collection master control box and a current transformer in a current collection system according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a current collection master control box according to an embodiment of the present invention;
FIG. 6 is a schematic diagram of a motor starting current according to an embodiment of the present invention;
FIG. 7 is a flow chart of data cleaning in an embodiment of the invention;
FIG. 8 is a flow chart of rough threshold judgment in an embodiment of the present invention;
FIG. 9 is a flow chart of the method of bisection to obtain the threshold values of rated current and fault current of the motor in an embodiment of the invention;
FIG. 10 is a flow chart of the comparison of transverse currents of the same type and multiple motors in an embodiment of the invention;
FIG. 11 is a flowchart of setting a K value according to an embodiment of the present invention;
FIG. 12 is a flow chart of comparison of longitudinal historical current data in an embodiment of the invention;
FIG. 13 is a flow chart of a cross-comparison cluster analysis of multiple groups of motors to determine motor line problems and potential motor faults in an embodiment of the invention;
Fig. 14 is a flowchart for judging motor abnormality by comparing longitudinal history current data in the embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the attached drawings and detailed description:
Example 1
Before introducing the current sensing-based wind driven generator motor fault online detection method, a current acquisition system shown in fig. 3 is built in advance, and the whole set of current acquisition system comprises bottom sensing layer equipment and a cloud server.
Specifically, the bottom sensing layer device comprises a current transducer and a plurality of current transformers.
Motors in wind power generators are of two types, including yaw motors and heat dissipation motors. The yaw motors and the heat dissipation motors are respectively multiple, and each current transformer is used for correspondingly collecting one-phase current data in three currents of one motor in the wind driven generator.
For example, one motor has A, B, C three-phase currents, and each motor corresponds to three current transformers, wherein each current transformer respectively corresponds to acquisition of phase A, phase B and phase C currents of the motor.
Similarly, assuming that there are M yaw motors and heat dissipation motors in total, it is necessary to refer to 3M current transformers collecting current data. Where M is a natural number greater than 0, which is related to the number of actual yaw motors and heat dissipating motors.
Each current transformer converts the acquired current data from analog quantity data to digital quantity data through a current transmitter, and the acquired motor current data is uploaded to a cloud server in a wireless transmission mode.
Each wind driven generator corresponds to an intelligent gateway and a current transducer, one current transducer is connected with a plurality of current transformers, and current data collected by each current transformer are transmitted to the same cloud server through the corresponding intelligent gateway.
The cloud server is used for analyzing the uploaded current data, and judging whether the motors normally run or not by comparing the current of each motor with threshold values, comparing the currents of multiple motors of the same type transversely and comparing the longitudinal historical current data of the single motor.
The deployment of the current collection master control box and the current transformer in the current collection system is shown in fig. 4.
The current transformer is installed in fan switch board department, near fan switch board installation current collection master control case, and current collection master control case is responsible for gathering current transformer's data, and current transformer's lead wire is drawn forth by in the fan switch board and is inserted on the current collection master control case in order to accomplish data acquisition, and the data of gathering will be converted into the digital quantity by the analog quantity through current transformer.
Specifically, the current collection master control box structure in this embodiment is shown in fig. 5. The current collection master control box comprises an air switch, a 12V power supply, an intelligent gateway, a current transducer and a terminal strip.
The power line of the current collection main control box is connected with a 12V power supply, and the power is supplied to the intelligent gateway and the current transducer through the 12V power supply. Analog quantity collected by the current transformer is transmitted to the current transducer through the terminal strip.
The current transmitter converts the analog quantity of the current into standard modbus485 data quantity data and sends the standard modbus485 data quantity data to the intelligent gateway, and the intelligent gateway uploads the motor current data to the cloud server in a wireless transmission mode.
In the embodiment, the current transformer preferably adopts an AKH-0.66W-7 series current transformer, and has the advantages of small volume, high precision, strong carrying capacity and the like, and can output small current signals of 5mA-20mA and the like.
Besides high acquisition precision, the current transformer is also convenient for field installation, the fan power distribution cabinet supplies power for each motor, and the current transformer is installed in the fan power distribution cabinet to which the motor is connected.
The current transformer is installed in the fan switch board in a centralized way, and is high in space utilization, easy to fix and simple to deploy, so that the problems of difficult installation wiring and insufficient installation space of the vibration sensor are avoided, the installation work difficulty is reduced, and the installation efficiency is high.
As shown in fig. 1, after a current collection system explicitly built in advance is introduced, a detailed description of a method for detecting a failure of a wind turbine motor on line based on current sensing in this embodiment is described below.
The method for detecting the motor faults of the wind driven generator on line based on current sensing in the embodiment comprises the following steps:
step 1, a cloud server receives current data of a corresponding motor uploaded by each current transformer in a current acquisition system, and performs data cleaning on the current data so as to filter out starting current of the motor.
The current data uploaded by the intelligent gateway firstly needs to be subjected to data cleaning through the cloud server, and the starting current of the motor is filtered. The data cleaning function is to ensure the authenticity and reliability of the current data.
As shown in fig. 6, wherein the abscissa t represents time, the ordinate I represents a current value, and I e represents a rated current. Because the motor is in a static state at the moment of power on, the rotor in the motor is static, at the moment, the rotating speed of the rotor is 0, the synchronous rotating magnetic field cuts the rotor winding at the maximum cutting speed, so that the rotor winding induces and reaches the highest electromotive force, a large current is generated in the rotor winding, the current counteracts the magnetic flux of the stator magnetic field in the motor, the stator winding automatically increases the current for maintaining the original magnetic flux which is suitable for the power supply voltage, and the stator current in the motor is increased by 5-7 times of the rated current because the current of the rotor is large at the moment, so that the data of the motor starting current is required to be cleaned and filtered, and the obtained motor current data which is the starting current of the motor is filtered after the data is cleaned.
Because the maintenance time of the starting current of the motor is very short and often does not exceed 0.5s, the flow of filtering the current data based on the online detection method of the motor faults of the wind driven generator based on current sensing is shown in fig. 7.
And judging motor current data acquired by the current acquisition system, when the motor current is identified to be changed from 0A to non-0A, defining the data in the current motor current 2s as starting current, filtering the part of data, namely defined starting current data, and using the data after the current motor current 2s for rough judgment of the threshold value in the step 2.
And 2, roughly judging the threshold value of the motor current data after the data in the step 1 are cleaned.
And when the current value of the motor collected by the current transformer does not exceed the preset threshold, turning to step 3.
The threshold judgment is most direct in expression mode, when the motor is subjected to conditions such as bearing damage and jamming, motor internal windings and the like, the resistance of the motor is increased to cause current change, and the motor with faults can be intuitively and effectively screened out by adopting the threshold judgment mode. Specifically, when the motor has conditions of bearing damage and jamming, motor internal winding and the like, the resistance of the motor is increased so as to cause current change, and at the moment, the numerical values of A, B, C three-phase currents are increased. In addition, if the motor phase failure occurs, for example, a certain phase current becomes 0, but another two phases current increases, it is determined that the motor has a phase failure problem.
The flow of the threshold rough judgment is shown in fig. 8.
Firstly, presetting a threshold parameter of motor current data, comparing the motor current data acquired by a current acquisition system with the preset threshold parameter after data cleaning, and directly triggering early warning push to prompt motor faults when the motor current data exceeds the threshold.
The key point of the rough judgment of the threshold value is to set the early warning threshold value, as shown in fig. 9.
According to the method, the rated current of the motor operation and the historical data of the motor current collected by the cloud server are combined, the current value of the equipment in the fault period, namely the fault motor current, is taken out, the threshold value of the rated current of the motor and the threshold value of the fault current of the motor are obtained through a dichotomy method, and the threshold value is set as a current early warning threshold value, namely a preset threshold value in rough judgment of the threshold value.
The process for obtaining the critical values of rated current and fault current of the motor by using the dichotomy is specifically as follows:
and based on the current value collected by the current fault motor as the upper limit of the current early warning threshold, taking the rated current as the reference lower limit of the set current early warning threshold, determining the intermediate value between the upper limit and the rated current by using a dichotomy, taking the intermediate value as the upper limit of the current early warning threshold, judging whether the current value is the fault current or not through experiments, namely, taking the current intermediate value as the fault current if the motor is fault current, continuing to take the intermediate value between the current value and the rated current, carrying out motor fault verification on the new intermediate value until the current value taken out by the dichotomy is the upper limit of the current value of normal motor operation, and taking the current value of the critical value as the current early warning threshold of the motor.
And 3, comparing the motor currents which do not exceed the threshold value with the same type and multiple motor transverse currents based on a KNN clustering algorithm, and further judging the motor currents with relatively small data difference by the step 4.
The motor current with relatively large data difference is compared with the rated current value.
Judging whether the motor current is in a preset safety range, if the motor current is in the preset safety range, turning to step 4 to further judge, and if the motor current is not in the preset safety range, directly judging that the motor is abnormal.
Specifically, the current average value of the same type and multiple motors is calculated first, and the distance value between the current value of each motor in the same type and the calculated current average value of the same type is further calculated:
Wherein n is the identification of the number of times of acquiring current data, the current data comprises the time of acquiring current comparison and a current value, x n represents the time of acquiring current comparison for the nth time, The average value of the time for acquiring the current comparison for n times is represented, 0.2 is the weight coefficient of the time for acquiring the current comparison, and is used for reducing the error of the whole distance judgment caused by the time difference uploaded by different intelligent gateways, y n is represented by the current value acquired for n times,The average value of the current values obtained n times is represented, and 0.8 is the weight coefficient of the current value.
And (5) carrying out ascending sort on each calculated distance value.
And taking K top positions in each distance value as motor setting K values participating in transverse current comparison.
And (3) comparing the longitudinal historical current data by using the step (4) with the first K motor currents obtained by ascending order, and judging the rest motor currents after the first K motor currents are removed after ascending order as follows:
Calculating whether the difference value between the current value and the rated current value of each other motor is within a preset safety range;
And (3) further turning to the step (4) to continuously compare the longitudinal historical current data for the motor current with the difference value within the safety range, and directly judging the motor current with the difference value outside the safety range as abnormal motor.
It should be noted that a motor failure corresponds to a significant mechanical failure or line failure of the motor, which directly affects the use of the wind turbine system. The abnormal motor corresponds to the tiny fault or circuit problem of the motor, and has little influence on the current use, but the development trend of the abnormal motor is concerned, and the motor is replaced or maintained in time so as not to influence the subsequent use.
The flow of the lateral contrast cluster analysis is shown in fig. 10.
And comparing the current data of motors of the same model, the same type and the same area of the in-service wind driven generator set, wherein the current running curves of all motors tend to be a certain average value when the motors normally run.
And (3) calculating the distance between the motor current value and the current average value which are involved in comparison by using a KNN clustering algorithm, arranging the motor current data which are involved in comparison in an ascending order according to the difference value of the distances, and then setting a K value.
For example, there are 10 sets of motors involved in the comparison, the K value may be set to 7, and then the current values of the first 7 motors closest to the average current (i.e., the first 7 motors arranged in ascending order of distance values) are all considered to be normal motor currents.
The currents of the remaining 3 motors are compared with the rated currents of the motors.
If the difference value is within the preset safety range, the difference value is continuously used for the subsequent longitudinal historical current data comparison, and if the difference value exceeds the safety range, the motor fault is judged, and the alarm information is triggered to prompt the motor to be abnormal.
The accuracy rate of the KNN clustering algorithm for motor fault analysis and judgment is characterized in that the K value is set.
When the K value is too small, the noise component can have a larger influence on motor fault judgment once, for example, when the K value is 1, a larger deviation can occur once the nearest point is noise, the K value is reduced, which means that the whole model becomes complex and easy to be fitted, when the K value is too large, the K value is equivalent to a training example in a larger neighborhood, the learned approximation error can be increased, and the example far away from an input target point can also play a role in motor fault analysis judgment, so that motor fault judgment is wrong. Thus, the setting of the K value requires constant data feeding and optimization.
The specific flow of setting the K value is shown in fig. 11, and the specific process is as follows:
Firstly, based on N motor currents which participate in comparison of transverse currents of the same type and multiple motors, a K value is tentatively set, the first K motor currents are analyzed, whether an abnormal current value exists in the first K motor currents or not is judged, if the abnormal current value exists, the K value is reduced until no abnormal current value exists in the first K motor currents, if the first K motor currents are all normal current values, the rest N-K motor currents are analyzed, and if the normal current value of the rest N-K motor currents is not less than 30%, the K value is increased until the normal current value of the rest N-K motor currents is less than 30%.
The line problems and potential motor faults of the motors can be analyzed through transverse comparison clustering analysis of multiple groups of motors, and fault hidden dangers of the motors which seem to be in normal operation can also be analyzed and judged, such as loosening of wire ends and wiring of the motors, ageing of motor lines and the like. The specific flow is shown in figure 13, the current motor is illustrated to be still in normal operation through the motor roughly judged by the threshold value, and further the line problem and the potential motor faults of the motors are analyzed through the transverse comparison cluster analysis of a plurality of groups of motors. If the current value of a certain motor is discrete from the current values of other motors in the process of transversely comparing the current values of the motors (specifically, the current value of each motor in the same type and the obtained distance value of the current average value of the type are calculated, the current values of the motors are sequenced in ascending order according to each distance value, and the rest motor currents after the first K motor currents are removed) indicate that the motor current may be abnormal, the three-phase current of the motor A, B, C is further analyzed, if the three-phase current is higher than the normal value of the current of the other motors, the problem that the internal mechanical structure of the motor has tiny mechanical faults is further indicated, and if the current of the motor has abnormal current of a certain phase or two phases, the problem that the line head of the motor is loose or the line is aged is indicated in the line of the motor.
And 4, comparing the motor current which is subjected to longitudinal historical current data comparison with the normal value of the motor current history of the motor current. The average value of the historical current normal value of the motor is the average value of the historical current data of the motor under normal operation.
The current history normal value of the motor is the history current data of the motor under normal operation, and the flow of comparing the longitudinal history current data is shown in figure 12. The method comprises the steps of determining that a motor is normal in operation if a difference value between a current value of the motor and a normal value average value of a current history of the motor is within a preset range, determining that the motor is abnormal if the difference value between the current value of the motor and the normal value average value of the current history of the motor is outside the preset range, and determining the specific condition of the motor abnormality by further carrying out trend analysis on the difference values between the current value of the motor and the normal values of the histories of the current of the motor according to whether the trend of the difference values changes steadily along with time.
For example, if the trend is stationary, it may be caused by dust and line aging of the motor, and if the trend is rising or falling, it is an increasing problem to describe a problem circuit or internal machinery. The current longitudinal comparison analysis is to compare the running current of the motor with the historical normal value so as to eliminate misjudgment of motor faults caused by the problem of the motor attribute.
Specifically, as shown in fig. 14, the average value of the historical current data of the motor under normal operation is set as the average value of the historical current value of the motor, and the current value of the motor is compared with the average value of the historical current value of the motor (the average value is the average value of the historical current data of the motor under normal operation, and is a fixed value after the average value is determined).
And then confirming the up-and-down fluctuation range of the motor current, and assuming that the up-and-down fluctuation range of the current is 0.5A, setting the threshold value of the fluctuation range to be 0.8A for preventing interference, judging that the motor is in normal operation when the up-and-down fluctuation of the average value of the motor current history is not more than 0.8A, and judging that the motor is abnormal if the difference value of the current value from the average value of the motor current history is more than 0.8A.
And (3) carrying out current difference analysis on the motor with abnormal motor current, wherein when trend analysis is carried out on the abnormal motor, the difference value of the current data of the normal historical motor operated before the motor is stored, and when difference analysis is carried out, the difference value historical data before abnormal conditions are also called out for analysis. The motor current value is gradually different from the normal value of the current history of the motor, if the difference value data gradually increases along with time, the problem circuit or the internal mechanical problem is increasingly serious, if the difference value always tends to be flat and stable, the difference value is possibly caused by dust and line aging, and the wind driven generator needs to be cleaned and maintained in time.
According to the method, the motor current data is acquired based on real-time remote acquisition of the current acquisition system, real-time analysis and processing of the motor current data are achieved through the cloud server, and the current acquisition system is convenient to install and wire on site and has high feasibility. The motor fault early warning method can judge motor operation faults and abnormal conditions based on current data analysis, performs threshold comparison on motor current data after starting current of a motor is filtered, and performs transverse current comparison of the same type and multiple motors and longitudinal historical current data comparison of a single motor based on a KNN clustering algorithm so as to judge whether the motor operation state is normal or not in a grading manner.
The method is designed to use three-level judgment of threshold comparison, transverse current comparison of multiple motors of the same type and longitudinal current comparison of a single motor. Firstly, a threshold value is used for roughly judging a motor with large mechanical faults or circuit problems, the motor has serious influence on a system and needs to be replaced in time, then, the motors entering transverse comparison are all in the threshold value, the motor can normally run at present, the influence on the running of the current system is small, the transverse comparison is carried out on multiple motors of the same type and the same area, the potential problems and faults of the running of a certain motor can be analyzed and found in a finer mode, the motor of the type can not be replaced in time but needs to pay attention continuously, for the final longitudinal comparison, besides the healthy development change of the motor, the false alarm caused by the characteristic problem of the motor can be filtered, if the current data of the running of the certain motor is different from the other motor, the motor is found to be always in the state through the longitudinal comparison of the motor, but the motor state is normal, the motor belongs to a special case, and the motor is filtered during the transverse comparison, so that the reliability and the stability of three-level early warning can be ensured.
For example, in the extreme case, the current data of a motor in normal operation is higher than the current data of other motors, and the difference value between the current value and the rated current is greater than the set safety range, then the motor current is always judged to be abnormal when the motor is subjected to transverse comparison, so that special marking treatment is required to be carried out on the motor, the motor is not involved in the transverse comparison of the motor selected in the area, and the motor is directly involved in the final longitudinal comparison link, thereby reducing the data interference caused by special data on transverse comparison analysis and improving the accuracy of the transverse comparison analysis.
Example 2
Embodiment 2 describes a current sensing-based wind turbine motor fault on-line detection system, which is based on the same inventive concept as the current sensing-based wind turbine motor fault on-line detection method in embodiment 1.
Specifically, the wind driven generator motor fault on-line detection system based on current sensing comprises the following modules:
the data cleaning module is used for carrying out data cleaning on the current data of the corresponding motor uploaded by each current transformer in the current acquisition system received by the cloud server so as to filter out the starting current of the motor;
the threshold value judging module is used for roughly judging the threshold value of the motor current data after the data are cleaned;
judging that the motor fails when the current value of the motor collected by the current transformer exceeds a preset threshold value, and switching to a multi-motor transverse current comparison module of the same type when the current value of the motor collected by the current transformer does not exceed the preset threshold value;
The same-type multi-motor transverse current comparison module is used for comparing motor currents which do not exceed a threshold value with the same-type multi-motor transverse currents further based on a KNN clustering algorithm;
the motor current with relatively small data difference is transferred to a longitudinal historical current data comparison module to be judged, and whether the motor current with relatively large data difference is in a preset safety range is judged by comparing the motor current with a rated current value;
If the motor current is within the preset safety range, the longitudinal historical current data comparison module is switched to further judge, and if the motor current is not within the preset safety range, the motor is directly judged to be abnormal;
the longitudinal historical current data comparison module is used for comparing the current value of the motor with the current historical normal value of the motor, wherein the current historical normal value of the motor is the historical current data of the motor under normal operation;
If the difference value between the current value and the average value of the historical normal value of the motor current is within the preset range, the motor is judged to be normal in operation, otherwise, the development trend of the difference value between the current value of the motor and each historical normal value is further judged;
and further distinguishing the specific abnormal situation of the motor according to whether the change of the difference development trend along with the time is stable.
It should be noted that, in the wind driven generator motor fault online detection system based on current sensing, the implementation process of the functions and roles of each functional module is specifically shown in the implementation process of the corresponding steps in the method in embodiment 1, and will not be described herein again.
Example 3
Embodiment 3 describes a computer device including a memory and one or more processors.
Executable codes are stored in the memory, and when the processor executes the executable codes, the executable codes are used for realizing the steps of the wind driven generator motor fault online detection method based on current sensing in the embodiment 1.
In this embodiment, the computer device is any device or apparatus having data processing capability, which is not described herein.
Example 4
Embodiment 4 describes a computer-readable storage medium having stored thereon a program which, when executed by a processor, is used for the steps of a method for on-line detection of a failure of a wind turbine motor based on current sensing.
The computer readable storage medium may be any internal storage unit of a device or apparatus having data processing capability, such as a hard disk or a memory, or may be any external storage device of a device having data processing capability, such as a plug-in hard disk, a smart memory card (SMART MEDIA CARD, SMC), an SD card, a flash memory card (FLASH CARD), or the like, provided on the device.
The foregoing description is, of course, merely illustrative of preferred embodiments of the present invention, and it should be understood that the present invention is not limited to the above-described embodiments, but is intended to cover all modifications, equivalents and alternatives falling within the spirit and scope of the present invention as defined by the appended claims.
Claims (10)
1. The wind driven generator motor fault online detection method based on current sensing is realized based on a current acquisition system built in advance, and is characterized in that the current acquisition system comprises a cloud server and a plurality of current transformers;
Each current transformer respectively and correspondingly acquires current data of a certain phase in three-phase current of any motor in the wind driven generator, and the acquired current data is uploaded to a cloud server through an intelligent gateway;
The wind driven generator motor fault on-line detection method based on current sensing comprises the following steps:
step 1, a cloud server receives current data of a corresponding motor uploaded by each current transformer in a current acquisition system, and performs data cleaning on the current data so as to filter out starting current of the corresponding motor;
step 2, roughly judging the threshold value of the motor current data after the data in the step 1 are cleaned;
When the current value of the motor collected by the current transformer does not exceed the preset threshold, the step 3 is switched to;
Step 3, comparing the motor currents which do not exceed the threshold value with the same type and multiple motors in the transverse direction based on KNN clustering;
Comparing the motor current with relatively small data difference with the rated current of the motor, and judging whether the motor current is in a preset safety range or not;
If the motor current value is judged to be within the preset safety range, the step 4 is further carried out to compare the longitudinal historical current data of the motor current, and if the motor current value is not within the preset safety range, the motor is judged to be abnormal;
Step 4, further comparing the motor current value with the current history normal value of the motor, wherein the average value of the current history normal value of the motor is the average value of the history current data of the motor in normal operation;
If the difference value between the current value and the average value of the historical normal value of the motor current is within the preset range, the motor is judged to be normal in operation, otherwise, the development trend of the difference value between the current value of the motor and each historical normal value is further judged;
and further distinguishing the specific abnormal situation of the motor according to whether the change of the difference development trend along with the time is stable.
2. The online detection method of the motor faults of the wind driven generator based on current sensing according to claim 1, wherein the current acquisition system comprises bottom sensing layer equipment and a cloud server;
The bottom sensing layer equipment comprises a current transducer and a plurality of current transformers, wherein each current transformer is used for correspondingly acquiring one-phase current data of one motor in the wind driven generator, the motor current data is converted into digital data from analog data through the current transducer, and the acquired motor current data is uploaded to the cloud server in a wireless transmission mode.
3. The method for detecting the motor fault of the wind driven generator based on current sensing according to claim 1, wherein in the step 1, motor current data collected by a current collection system is judged, and when the motor current is recognized to be changed from 0A to non-0A, data in the current motor current 2s are defined as starting current;
And (3) filtering the data of the starting current, and using the data of the current motor current for 2s for rough judgment of the threshold value in the step (2).
4. The method for online detecting motor faults of wind driven generators based on current sensing according to claim 1, wherein in the step2, the method for determining the preset threshold value is as follows:
and taking out the current value of the motor fault time period, namely the motor fault current, according to the rated current of the motor operation and the historical data of the motor current collected by the intelligent gateway, obtaining the critical value of the rated current of the motor and the motor fault current by using a dichotomy, and setting the critical value as a current early warning threshold, namely a preset threshold in rough judgment of the threshold.
5. The method for detecting the motor failure of the wind driven generator on line based on current sensing according to claim 1, wherein the step 3 is specifically:
Firstly, calculating the current average value of the same type and multiple motors, further calculating the distance value between the current value of each motor in the same type and the obtained current average value of the type, and carrying out ascending order sequencing on each calculated distance value;
taking K most front positions in each distance value as motor setting K values participating in transverse current comparison;
and (3) comparing the longitudinal historical current data by using the step (4) with the first K motor currents obtained by ascending order, and judging the rest motor currents after the first K motor currents are removed after ascending order as follows:
Calculating whether the difference value between the current value and the rated current value of each other motor is within a preset safety range;
And (3) further turning to the step (4) to continuously compare the longitudinal historical current data for the motor current with the difference value within the safety range, and directly judging the motor current with the difference value outside the safety range as abnormal motor.
6. The method for online detection of motor faults of wind turbines based on current sensing according to claim 5, wherein in the step 3, a formula for calculating a distance value between a current value of each motor in the same type and a current average value of the type is as follows:
Wherein n is the identification of the number of times of acquiring current data, the current data comprises the time of acquiring current comparison and a current value, x n represents the time of acquiring current comparison for the nth time, The average value of the time for acquiring the current comparison for n times is represented, 0.2 is the weight coefficient of the time for acquiring the current comparison, and is used for reducing the error of the whole distance judgment caused by the time difference uploaded by different intelligent gateways, y n is represented by the current value acquired for n times,The average value of the current values obtained n times is represented, and 0.8 is the weight coefficient of the current value.
7. The method for online detection of motor faults of wind driven generators based on current sensing according to claim 5, wherein in the step 3, the process of adjusting the K value is as follows:
Firstly, tentatively setting a K value based on N motor currents participating in transverse current comparison of the same type and multiple motors;
Analyzing the first K motor currents, judging whether abnormal current values exist in the first K motor currents, if so, reducing the K value until no abnormal current values exist in the first K motor currents, if the first K motor currents are all normal current values, analyzing the remaining N-K motor currents, and if the normal current value of the remaining N-K motor currents is not less than 30%, increasing the K value until the normal current value of the remaining N-K motor currents is less than 30%.
8. Wind driven generator motor fault on-line measuring system based on current perception, characterized by comprising:
the data cleaning module is used for carrying out data cleaning on the current data of the corresponding motor uploaded by each current transformer in the current acquisition system received by the cloud server so as to filter out the starting current of the motor;
The threshold judgment module is used for roughly judging the threshold of the motor current data after the data cleaning;
judging that the motor fails when the current value of the motor collected by the current transformer exceeds a preset threshold value, and switching to a multi-motor transverse current comparison module of the same type when the current value of the motor collected by the current transformer does not exceed the preset threshold value;
The same-type multi-motor transverse current comparison module is used for comparing motor currents which do not exceed a threshold value with the same-type multi-motor transverse currents further based on a KNN clustering algorithm;
the motor current with relatively small data difference is transferred to a longitudinal historical current data comparison module to be judged, and whether the motor current with relatively large data difference is in a preset safety range is judged by comparing the motor current with a rated current value;
If the motor current is within the preset safety range, the longitudinal historical current data comparison module is switched to further judge, and if the motor current is not within the preset safety range, the motor is directly judged to be abnormal;
the longitudinal historical current data comparison module is used for comparing the current value of the motor with the normal value of the current history of the motor, wherein the average value of the normal value of the current history of the motor is the average value of the historical current data of the motor in normal operation;
If the difference value between the current value and the average value of the historical normal value of the motor current is within the preset range, the motor is judged to be normal in operation, otherwise, the development trend of the difference value between the current value of the motor and each historical normal value is further judged;
and further distinguishing the specific abnormal situation of the motor according to whether the change of the difference development trend along with the time is stable.
9. A computer device comprising a memory and one or more processors, the memory having executable code stored therein, wherein the processor, when executing the executable code, performs the steps of the current sensing based on-line detection method of a wind turbine motor failure of any of claims 1 to 7.
10. A computer-readable storage medium, on which a program is stored, characterized in that the program, when executed by a processor, implements the steps of the current-aware-based on-line detection method of motor faults of a wind turbine as claimed in any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411456751.4A CN119471358A (en) | 2024-10-18 | 2024-10-18 | Wind driven generator motor fault on-line detection method based on current sensing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411456751.4A CN119471358A (en) | 2024-10-18 | 2024-10-18 | Wind driven generator motor fault on-line detection method based on current sensing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN119471358A true CN119471358A (en) | 2025-02-18 |
Family
ID=94595828
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411456751.4A Pending CN119471358A (en) | 2024-10-18 | 2024-10-18 | Wind driven generator motor fault on-line detection method based on current sensing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119471358A (en) |
-
2024
- 2024-10-18 CN CN202411456751.4A patent/CN119471358A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103912448B (en) | A kind of regional wind power power of the assembling unit characteristic monitoring method | |
| JP2003525002A (en) | Shaft voltage and current monitoring system | |
| CN117056764A (en) | Intelligent protection method and system for generator set | |
| CN110749810A (en) | Insulation fault prediction method and system for phase modulator | |
| CN109185018B (en) | Variable-frequency permanent magnet hydroelectric power generation system and control method thereof | |
| CN117786385B (en) | Three-phase asynchronous motor fault monitoring method and system based on twin network | |
| CN114609462B (en) | Intelligent early warning diagnosis model for comprehensive multi-characteristic parameters of electrical equipment state | |
| CN114061743A (en) | Vibration monitoring method, device, equipment and medium for wind turbine | |
| Carroll et al. | Drivetrain availability in offshore wind turbines | |
| CN112385109A (en) | Method for detecting low-frequency oscillations and detection device therefor | |
| CN117627862A (en) | Marine wind turbine generator control system based on IEC61499 framework | |
| Hamatwi et al. | Condition monitoring and fault diagnosis of stator and rotor interturn winding faults in a DFIG-based wind turbine system: A Review | |
| CN113847197A (en) | Wind generating set and rotation avoiding speed control method and device thereof | |
| CN119471358A (en) | Wind driven generator motor fault on-line detection method based on current sensing | |
| EP3846302B1 (en) | System and method for stabilizing weak grids with one or more wind farms connected thereto | |
| CN110809851A (en) | Electrical power system and method using distortion filter | |
| CN118483598A (en) | Testing method and testing system for voltage filter of asynchronous generator for ship | |
| Karaköse et al. | The intelligent fault diagnosis frameworks based on fuzzy integral | |
| CN118270628A (en) | Escalator comprehensive detection system with automatic diagnosis function | |
| CN117349790A (en) | Transformer mechanical stability analysis method based on multi-source heterogeneous data | |
| CN109188185A (en) | A kind of generator rotor interturn short-circuit early stage online test method | |
| CN115913102A (en) | Photovoltaic power station and inverter abnormality detection method, device and monitoring system thereof | |
| CN113408199A (en) | Gearbox oil temperature fault early warning method based on multilayer perception neural network | |
| Skrimpas et al. | Detection of wind turbine power performance abnormalities using eigenvalue analysis | |
| CN111667120A (en) | Insulation fault prediction method and system for phase modulator |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |