A motor detection system based on multi-source data

Abstract

The application discloses a motor detection system based on multi-source data, and relates to the technical field of motor detection.The motor detection system comprises a motor information acquisition module, an information optimization processing module, an automatic detection model establishing module and a motor detection module.The motor information acquisition module is used for collecting parameter information of motors that have been subjected to back electromotive force detection and have correct detection results and time information required by the motors to run at no load to a rated frequency after the load is disconnected.The information optimization processing module is used for avoiding outlier optimization processing on the time data to obtain time reference optimization data.The automatic detection model establishing module is used for establishing an automatic detection time planning model according to the parameter information of the motors and the time reference optimization data.The motor detection module is used for planning back electromotive force detection time of the motors, detecting the back electromotive force of the motors under no load at the planned time, and giving a demagnetization warning when an abnormality is detected, so that the automatic detection of the back electromotive force of the motors under no load is realized.

Description

Technical Field

[0001] This invention relates to the field of motor testing technology, specifically a motor testing system based on multi-source data. Background Technology

[0002] In the selection of rotor magnetic circuit structure for permanent magnet synchronous motors used in the aviation field, permanent magnets can not only provide a large magnetic flux, but also have a high material utilization rate and a relatively simple structure. The radial rotor structure of permanent magnet synchronous motors is superior to the built-in tangential structure in terms of voltage regulation rate, harmonic distortion rate, leakage flux, and power angle characteristics. In order to ensure the normal operation of permanent magnet synchronous motors, multiple aspects of testing are required.

[0003] Among these, demagnetization detection of permanent magnet synchronous motors is particularly important. Demagnetization can lead to a reduction in the motor's output torque and power, and in severe cases, it can even prevent the motor from operating normally. When performing demagnetization detection on a permanent magnet synchronous motor, the motor can be placed in an unloaded state by disconnecting the motor load. When the motor runs at its rated frequency under no-load conditions, the back electromotive force (EMF) amplitude can be measured to determine the motor's demagnetization. If the back EMF amplitude is lower than the standard value, the motor may have already demagnetized. However, since different motors may require different amounts of time to run to their rated frequency after disconnecting the load, existing technologies generally collect the motor's operating frequency data frequently when the motor is running under no-load conditions. This cannot achieve automated back EMF detection time planning for different motors and increases the amount of invalid data collection. Summary of the Invention

[0004] The purpose of this invention is to provide a motor detection system based on multi-source data to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a motor detection system based on multi-source data, the system comprising a motor information acquisition module, an information optimization and processing module, an automated detection model establishment module, and a motor detection module;

[0006] The motor information acquisition module collects parameter information of motors that have previously undergone back electromotive force testing and whose test results were correct, as well as the time required for the motor to run unloaded to the rated frequency after the load is disconnected.

[0007] The information optimization processing module performs outlier avoidance optimization on the time data required for the motor to run unloaded to the rated frequency after the load is disconnected, and obtains the processed time reference optimization data.

[0008] The automated detection model establishment module establishes an automated detection time planning model based on the motor's parameter information and time reference optimization data, which is used for back EMF detection time planning of the motor.

[0009] The motor detection module plans the back EMF detection time of the motor to be tested, detects the back EMF of the permanent magnet synchronous motor under no-load conditions during the planned time, and performs demagnetization warning processing when an abnormal back EMF is detected.

[0010] Preferably, the motor information acquisition module includes a parameter information acquisition unit and an no-load operation data acquisition unit;

[0011] The parameter information acquisition unit collects parameter information of motors that have previously undergone back electromotive force testing and whose test results were correct. The parameter information includes the number of pole pairs and the rated frequency of the motor.

[0012] The no-load operation data acquisition unit collects the time information required for the corresponding motor to run at its rated frequency under no-load conditions after the load is disconnected.

[0013] Preferably, the information optimization processing module includes a data classification unit with constraints and an outlier avoidance optimization unit;

[0014] The collected data is classified by the data classification unit under the specified conditions. The classification method is as follows: the time required for motors with the same number of pole pairs and rated frequency to run unloaded to the rated frequency after the load is disconnected is divided into the same category of values.

[0015] The outlier avoidance optimization unit analyzes each type of time data after segmentation, performs outlier data finding and segmentation optimization on each type of time data separately, calculates the mean of each type of time data remaining after segmentation optimization, and uses the calculated data as the processed time reference optimization data.

[0016] Preferably, the automated detection model building module includes a data integration unit to be fitted and an automated detection time planning model building unit;

[0017] The data integration unit integrates the motor's parameter information and time reference optimization data into data to be fitted.

[0018] The automated detection time planning model establishment unit fits the data to be fitted and then establishes the automated detection time planning model.

[0019] Preferably, the motor detection module includes a motor parameter acquisition unit, a detection time planning unit, and a motor anomaly detection unit;

[0020] The number of pole pairs and rated frequency parameters of the motor to be tested are obtained through the motor parameter acquisition unit.

[0021] The detection time planning unit inputs the parameters of the motor to be detected into the automated detection time planning model, and plans the back EMF detection time for the motor to be detected based on the model output results.

[0022] The motor anomaly detection unit uses the time planned in the frequency converter power analyzer to detect the back electromotive force (EMF) amplitude of the motor currently under test. Considering that high or low temperature environments or environments with fluctuating temperatures may interfere with the back EMF detection results, the back EMF of the motor is detected in a normal and constant temperature environment to obtain the normal value of the back EMF of the motor under no-load conditions. This value is then compared with the detected back EMF amplitude. A difference threshold is set. If the back EMF amplitude is lower than the normal back EMF value and the difference between the amplitude and the normal value is greater than the difference threshold, it is determined that the current back EMF of the motor is abnormal and the motor is at risk of demagnetization. A demagnetization warning is then issued. The normal back EMF value refers to the back EMF value on the motor nameplate.

[0023] Preferably, the motor information acquisition module collects the number of pole pairs of motors that have previously undergone back EMF testing and whose test results are correct, which is {p1,p2,...,pn}, and the corresponding rated frequency of the motors is {F1,F2,...,Fn}, where n represents the number of motors that have previously undergone back EMF testing and whose test results are correct. The module also collects the time required for the corresponding motor to run unloaded to its rated frequency after disconnecting the load during back EMF testing preparation, which is T={T1,T2,...,Tn}. Back EMF testing preparation refers to the process of disconnecting the motor load and controlling the motor to run to its rated frequency.

[0024] Preferably, the information optimization processing module is used to group motors with the same number of pole pairs and the same rated frequency among the n motors into the same category, resulting in a total of v categories of motors. The time required for each category of motors to run unloaded to its rated frequency after the load is disconnected is also grouped into the same category of time values, resulting in a random set of time values ​​t={t1,t2,...,tm}, where m represents the number of motors in the random category, Pi represents the number of pole pairs, and fi represents the rated frequency. The time values ​​in set t are then sorted in ascending order from smallest to largest, and the median of the m time values ​​is located as K2. Using K2 as the dividing point, the time values ​​are divided into a lower half and an upper half, with the upper half of the time values ​​arranged in sequence. The time values ​​before K2 are considered, and the time values ​​in the lower half of the time range are those after K2. The median of the time values ​​in the upper half is identified as K1, and the median of the time values ​​in the lower half as K3. The anti-outlier data range is set to [K1-1.5L, K3+1.5L], where L is the interquartile range (L=K3-K1). The time values ​​within set t are compared one by one with the anti-outlier data range: Time values ​​outside the range are considered outliers and removed. This yields a random class of remaining time values. The mean of the remaining time values ​​for each class is calculated, and this mean is taken as the time reference optimization value ti for the random class of motors. ’ The time reference optimization value for the V-type motor is obtained as {t1}. ’ ,t2 ’ ,...ti ’ ,...,tv ’};

[0025] Considering that even permanent magnet synchronous motors with the same rated frequency and number of pole pairs may take different times to reach their rated frequency under no-load conditions—for example, differences in rotational inertia due to different rotor structure designs—this invention collects multi-source data: rated frequency, number of pole pairs, and time required to reach the rated frequency under no-load conditions for different motors. Motors with the same rated frequency and number of pole pairs are grouped together. The time required for no-load operation of the same type of motor to reach the rated frequency is analyzed. An anti-outlier data range is set in a specific way, and time values ​​outside the anti-outlier data range are removed. Since the removed time values ​​differ too much from the remaining data, if they are not removed, they will cause data interference to the subsequent prediction of the time required for the motor to reach the rated frequency under no-load conditions through the model. Therefore, outlier avoidance optimization is prioritized for the time data, reducing the interference of outliers in the collected data on the prediction results and the back EMF detection time planning.

[0026] Preferably, the dataset to be fitted is obtained by integrating the automated detection model building module as {(P1,f1,t1)}. ’ ),(P2,f2,t2 ’ ),...(Pi,fi,ti ’ ),...,(Pv,fv,tv ’ )}, where {P1,P2,...Pi,...,Pv} represent the number of pole pairs of a Class V motor, and {f1,f2,...fi,...,fv} represent the rated frequency of a Class V motor. After fitting the integrated data, an automated detection time planning model is established: Z=G1*X+G2*Y+G3, where G1, G2, and G3 represent fitting coefficients, X represents the variable representing the number of pole pairs in the model, Y represents the variable representing the rated frequency in the model, and Z represents the variable representing the time value in the model.

[0027] Preferably, the number of pole pairs of the motor to be tested is P, obtained using the motor detection module. 待检 The rated frequency is f 待检 , will P 待检 and f 待检 Input into the automated detection time planning model: Let X=P 待检 Y=f 待检 The predicted time required for the motor under test to run unloaded to its rated frequency after the load is disconnected is G1*P. 待检 +G2*f 待检 +G3, the planned time for back EMF detection of the motor under test is: after the load is disconnected from the motor under test, the interval is G1*P. 待检 +G2*f 待检 Back electromotive force detection is performed at +G3;

[0028] Considering that the more pole pairs a permanent magnet synchronous motor has and the lower its rated frequency, the shorter the time required for the motor to reach its rated frequency under no-load conditions, an automated detection time planning model is established by acquiring multi-source data: collecting the time required for different motors to reach their rated frequency before back EMF detection, as well as parameter data for different motors. After fitting the multi-source data, an automated detection time planning model is established. Based on the automated detection time planning model, the optimal time for back EMF detection of different motors is predicted. The optimal time is the time it takes for the motor to reach its rated frequency under no-load conditions. Back EMF detection is performed on the motor at the planned time. The back EMF detection results are used to determine whether the motor has demagnetization. When there is a risk of demagnetization, timely warning processing is carried out, which is conducive to timely optimization of motor performance to ensure normal use of the motor. The establishment of the model not only realizes the automated detection time planning of back EMF for different motors, but also effectively reduces the amount of invalid data collection.

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] This invention utilizes multi-source data acquisition: collecting the time required for different motors to reach their rated frequency before back EMF detection, as well as parameter data for different motors. After fitting the multi-source data, an automated detection time planning model is established. Based on the automated detection time planning model, the optimal time for back EMF detection of different motors is predicted. The optimal time is the time it takes for the motor to run at its rated frequency under no-load. Back EMF detection is performed on the motor at the planned time. The back EMF detection results are used to determine whether the motor has demagnetization. If there is a risk of demagnetization, timely warning processing is carried out, which is conducive to timely optimization of motor performance to ensure normal use of the motor. Through model establishment, not only is automated back EMF detection time planning for different motors realized, but also the amount of invalid data collection is effectively reduced.

[0031] Considering that even permanent magnet synchronous motors with the same rated frequency and number of pole pairs may take different times to reach their rated frequency under no-load conditions—for example, differences in rotational inertia due to different rotor structure designs—this invention collects multi-source data: rated frequency, number of pole pairs, and time required to reach the rated frequency under no-load conditions for different motors. Motors with the same rated frequency and number of pole pairs are grouped together. The time required for no-load operation of the same type of motor to reach the rated frequency is analyzed. An anti-outlier data range is set in a specific way, and time values ​​outside the anti-outlier data range are removed. Since the removed time values ​​differ too much from the remaining data, if they are not removed, they will cause data interference to the subsequent prediction of the time required for the motor to reach the rated frequency under no-load conditions through the model. Therefore, outlier avoidance optimization is prioritized for the time data, reducing the interference of outliers in the collected data on the prediction results and the back EMF detection time planning. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the structure of a motor detection system based on multi-source data according to the present invention. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Example 1: As Figure 1As shown, this embodiment provides a motor detection system based on multi-source data. The system includes: a motor information acquisition module, an information optimization processing module, an automated detection model establishment module, and a motor detection module. The motor information acquisition module acquires parameter information of motors that have previously undergone back EMF detection with correct results, as well as the time required for the motor to run unloaded to its rated frequency after the load is disconnected. The information optimization processing module performs outlier avoidance optimization on the time data required for the motor to run unloaded to its rated frequency after the load is disconnected, obtaining processed time reference optimized data. The automated detection model establishment module establishes an automated detection time planning model based on the motor parameter information and the time reference optimized data, which is used for back EMF detection time planning of the motor. The motor detection module plans the back EMF detection time of the motor to be detected, detects the back EMF of the permanent magnet synchronous motor under no-load conditions within the planned time, and performs demagnetization early warning processing when an abnormal back EMF is detected.

[0035] The motor information acquisition module includes a parameter information acquisition unit and an no-load operation data acquisition unit. The parameter information acquisition unit acquires parameter information of motors that have previously undergone back EMF testing and whose test results are correct. The parameter information includes the number of pole pairs and the rated frequency of the motor. The no-load operation data acquisition unit acquires the time information required for the corresponding motor to run at the rated frequency under no-load after the load is disconnected.

[0036] The information optimization processing module includes a constraint-based data classification unit and an outlier avoidance optimization unit. The constraint-based data classification unit classifies the collected data by grouping motors with the same number of pole pairs and rated frequency into the same category based on the time required for them to run unloaded to their rated frequency after the load is disconnected. The outlier avoidance optimization unit analyzes each category of time data, performs outlier finding and segmentation optimization on each category, and calculates the mean for each remaining category of time data after segmentation optimization. The calculated data is then used as the processed time reference optimization data.

[0037] The automated detection model establishment module includes a data integration unit and an automated detection time planning model establishment unit. The data integration unit integrates the motor parameter information and time reference optimization data into data to be fitted. The automated detection time planning model establishment unit fits the data to be fitted and then establishes the automated detection time planning model.

[0038] The motor testing module includes a motor parameter acquisition unit, a testing time planning unit, and a motor anomaly detection unit. The motor parameter acquisition unit acquires the pole pair number and rated frequency parameters of the motor to be tested. The testing time planning unit inputs the motor parameters into an automated testing time planning model and plans the back EMF testing time based on the model's output. The motor anomaly detection unit uses the planned time in the frequency converter power analyzer to detect the back EMF amplitude of the motor to be tested, acquiring the normal back EMF value under no-load conditions and comparing it with the detected back EMF amplitude. A difference threshold is set; if the back EMF amplitude is lower than the normal back EMF value and the difference between the amplitude and the normal value is greater than the difference threshold, it is determined that the motor's back EMF is abnormal and the motor is at risk of demagnetization, triggering a demagnetization warning. The normal back EMF value refers to the back EMF value on the motor nameplate.

[0039] The motor information acquisition module collects the number of pole pairs of motors that have previously undergone back EMF testing and whose test results are correct. The corresponding rated frequencies of these motors are {p1, p2, ..., pn}, where n represents the number of motors that have previously undergone back EMF testing and whose test results are correct. The module also collects the time required for the corresponding motor to run unloaded to its rated frequency after disconnecting the load during back EMF testing preparation. This time is T = {T1, T2, ..., Tn}. Back EMF testing preparation refers to the process of disconnecting the motor load and controlling the motor to run to its rated frequency.

[0040] Using an information optimization processing module, motors with the same number of pole pairs and rated frequency among n motors are grouped into the same category, resulting in v categories. The time required for each category of motors to reach its rated frequency under no-load conditions after disconnection is also grouped into a category of time values, resulting in a random set of time values ​​t = {t1, t2, ..., tm}, where m represents the number of motors in the random category, Pi represents the number of pole pairs, and fi represents the rated frequency. The time values ​​in set t are sorted in ascending order, and the median of the m time values ​​is located as K2. Using K2 as the dividing point, the time values ​​are divided into a lower half and an upper half, with the upper half of the time values ​​being those listed at K2. The previous time values, the lower half of the time values ​​are those arranged after K2. The median of the upper half of the time values ​​is located as K1, and the median of the lower half of the time values ​​is located as K3. The anti-outlier data range is set to [K1-1.5L, K3+1.5L], where L is the interquartile range, L=K3-K1. The time values ​​within set t are compared one by one with the anti-outlier data range: It is determined whether the time value is within the anti-outlier data range. Time values ​​outside the range are considered outliers. After removing the outliers, a random class of remaining time values ​​is obtained. The mean of the remaining time values ​​for the corresponding class is calculated, and the calculated mean is taken as the time reference optimization value ti for the random class of motors. ’ The time reference optimization value for the V-type motor is obtained as {t1}. ’ ,t2 ’ ,...ti ’ ,...,tv ’};

[0041] For example: If we obtain a random set of time values ​​t={2, 1.2, 1.5, 1.3}, which has an even number of time values, and sort the time values ​​in set t in ascending order from smallest to largest, we get a set {1.2, 1.3, 1.5, 2}. If we locate the median K2=(1.3+1.5) / 2=1.4, then the time values ​​in the upper half are the time values ​​arranged before K2: 1.2 and 1.3, and the time values ​​in the lower half are the time values ​​arranged after K2: 1.5 and 2.

[0042] If we obtain a random set of time values ​​t={2, 1.2, 1.5, 1.3, 1.6}, which contains an odd number of time values, and sort the time values ​​in set t in ascending order, we get a set {1.2, 1.3, 1.5, 1.6, 2}. If we locate the median K2=1.5, then the time values ​​in the upper half of the set are the time values ​​that come before K2: 1.2 and 1.3, and the time values ​​in the lower half of the set are the time values ​​that come after K2: 1.6 and 2.

[0043] The dataset to be fitted is obtained by integrating modules using an automated detection model: {(P1,f1,t1)} ’ ),(P2,f2,t2 ’ ),...(Pi,fi,ti ’ ),...,(Pv,fv,tv ’ )}, where {P1,P2,...Pi,...,Pv} represent the number of pole pairs of a Class V motor, and {f1,f2,...fi,...,fv} represent the rated frequency of a Class V motor. After fitting the integrated data, an automated detection time planning model is established: Z=G1*X+G2*Y+G3, where G1, G2, and G3 represent fitting coefficients, X represents the variable representing the number of pole pairs in the model, Y represents the variable representing the rated frequency in the model, and Z represents the variable representing the time value in the model.

[0044] The number of pole pairs of the motor to be tested is P, obtained using the motor detection module. 待检 The rated frequency is f 待检 , will P 待检 and f 待检 Input into the automated detection time planning model: Let X=P 待检 Y=f 待检 The predicted time required for the motor under test to run unloaded to its rated frequency after the load is disconnected is G1*P. 待检 +G2*f 待检 +G3, the planned time for back EMF detection of the motor under test is: after the load is disconnected from the motor under test, the interval is G1*P. 待检 +G2*f 待检 Back electromotive force detection is performed at +G3;

[0045] For example: after disconnecting the load from the motor currently being tested, at interval G1*P 待检 +G2*f 待检 When +G3 is applied, the back EMF of the current motor is detected using a frequency converter power analyzer. The difference threshold is set to 50V. If the back EMF amplitude of the current motor is detected to be lower than the normal back EMF value and the difference between the back EMF and the normal back EMF value is 60V, and 60V>50V, it is determined that the back EMF of the current motor is abnormal and the motor is at risk of demagnetization, and a demagnetization warning is issued.

[0046] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A motor detection system based on multi-source data, characterized in that: The system includes a motor information acquisition module, an information optimization and processing module, an automated detection model establishment module, and a motor detection module; The motor information acquisition module collects parameter information of motors that have previously undergone back electromotive force testing and whose test results were correct, as well as the time required for the motor to run unloaded to the rated frequency after the load is disconnected. The information optimization processing module performs outlier avoidance optimization on the time data required for the motor to run unloaded to the rated frequency after the load is disconnected, and obtains the processed time reference optimization data. The automated detection model establishment module establishes an automated detection time planning model based on the motor's parameter information and time reference optimization data, which is used for back EMF detection time planning of the motor. The motor detection module plans the back EMF detection time of the motor to be tested, detects the back EMF of the permanent magnet synchronous motor under no-load conditions during the planned time, and performs demagnetization warning processing when an abnormal back EMF is detected.

2. The motor detection system based on multi-source data according to claim 1, characterized in that: The motor information acquisition module includes a parameter information acquisition unit and an no-load operation data acquisition unit; The parameter information acquisition unit collects parameter information of motors that have previously undergone back electromotive force testing and whose test results were correct. The parameter information includes the number of pole pairs and the rated frequency of the motor. The no-load operation data acquisition unit collects the time information required for the corresponding motor to run at its rated frequency under no-load conditions after the load is disconnected.

3. The motor detection system based on multi-source data according to claim 2, characterized in that: The information optimization processing module includes a data classification unit with constraints and an outlier avoidance optimization unit. The collected data is classified by the data classification unit under the specified conditions. The classification method is as follows: the time required for motors with the same number of pole pairs and rated frequency to run unloaded to the rated frequency after the load is disconnected is divided into the same category of values. The outlier avoidance optimization unit analyzes each type of time data after segmentation, performs outlier data finding and segmentation optimization on each type of time data separately, calculates the mean of each type of time data remaining after segmentation optimization, and uses the calculated data as the processed time reference optimization data.

4. The motor detection system based on multi-source data according to claim 3, characterized in that: The automated detection model establishment module includes a data integration unit for data to be fitted and an automated detection time planning model establishment unit. The data integration unit integrates the motor's parameter information and time reference optimization data into data to be fitted. The automated detection time planning model establishment unit fits the data to be fitted and then establishes the automated detection time planning model.

5. The motor detection system based on multi-source data according to claim 4, characterized in that: The motor detection module includes a motor parameter acquisition unit, a detection time planning unit, and a motor anomaly detection unit; The number of pole pairs and rated frequency parameters of the motor to be tested are obtained through the motor parameter acquisition unit. The detection time planning unit inputs the parameters of the motor to be detected into the automated detection time planning model, and plans the back EMF detection time for the motor to be detected based on the model output results. The motor anomaly detection unit uses the planned time in the frequency converter power analyzer to detect the back electromotive force amplitude of the motor to be tested, obtains the normal value of the back electromotive force of the motor under no-load conditions, and compares it with the detected back electromotive force amplitude of the motor: a difference threshold is set. If the back electromotive force amplitude is lower than the normal back electromotive force value and the difference between the amplitude and the normal value is greater than the difference threshold, it is determined that the back electromotive force of the current motor is abnormal and the motor has a risk of demagnetization, and a demagnetization warning is issued.

6. The motor detection system based on multi-source data according to claim 5, characterized in that: The motor information acquisition module collects the number of pole pairs of motors that have previously undergone back EMF testing and whose test results are correct, which is {p1,p2,...,pn}, and the corresponding rated frequency of the motors is {F1,F2,...,Fn}, where n represents the number of motors that have previously undergone back EMF testing and whose test results are correct. The module also collects the time required for the corresponding motor to run unloaded to the rated frequency after disconnecting the load during back EMF testing preparation, which is T={T1,T2,...,Tn}.

7. The motor detection system based on multi-source data according to claim 6, characterized in that: The information optimization processing module uses n motors with the same number of pole pairs and the same rated frequency to classify them into the same category, resulting in v categories of motors. The time required for each category of motors to run unloaded to its rated frequency after being disconnected from the load is also classified into the same category of time values, resulting in a random set of time values ​​t={t1,t2,...,tm}, where m represents the number of motors in the random category, Pi represents the number of pole pairs, and fi represents the rated frequency. The time values ​​in set t are sorted in ascending order, and the median of the m time values ​​is located as K2. Using K2 as the dividing point, the time values ​​are divided into a lower half and an upper half, with the upper half of the time values ​​being those ranked at K2. The time values ​​before K2 are considered, and the time values ​​in the lower half of the time range are those following K2. The median of the time values ​​in the upper half is identified as K1, and the median of the time values ​​in the lower half as K3. The anti-outlier data range is set to [K1-1.5L, K3+1.5L], where L is the interquartile range (L=K3-K1). The time values ​​within set t are compared one by one with the anti-outlier data range: Time values ​​outside the range are considered outliers and removed. This yields a random class of remaining time values. The mean of the remaining time values ​​for each class is calculated, and this mean is taken as the time reference optimization value ti for the random class of motors. ’ The time reference optimization value for the V-type motor is obtained as {t1}. ’ ,t2 ’ ,...ti ’ ,...,tv ’ } 8. The motor detection system based on multi-source data according to claim 7, characterized in that: The dataset to be fitted is obtained by integrating the automated detection model building module as {(P1,f1,t1)}. ’ ),(P2,f2,t2 ’ ),...(Pi,fi,ti ’ ),...,(Pv,fv,tv ’ )}, where {P1,P2,...Pi,...,Pv} represent the number of pole pairs of a Class V motor, and {f1,f2,...fi,...,fv} represent the rated frequency of a Class V motor. After fitting the integrated data, an automated detection time planning model is established: Z=G1*X+G2*Y+G3, where G1, G2, and G3 represent fitting coefficients, X represents the variable representing the number of pole pairs in the model, Y represents the variable representing the rated frequency in the model, and Z represents the variable representing the time value in the model.

9. A motor detection system based on multi-source data according to claim 8, characterized in that: The number of pole pairs of the motor to be tested is P, obtained using the motor detection module. 待检 The rated frequency is f 待检 , will P 待检 and f 待检 Input into the automated detection time planning model: Let X=P 待检 Y=f 待检 The predicted time required for the motor under test to run unloaded to its rated frequency after the load is disconnected is G1*P. 待检 +G2*f 待检 +G3, the planned time for back EMF detection of the motor under test is: after the load is disconnected from the motor under test, the interval is G1*P. 待检 +G2*f 待检 Back electromotive force is detected at +G3.

Images