A non-intrusive detection method and system for isolating switch motor

Abstract

The present application relates to the field of motor detection technology, and specifically to a non-invasive detection method and system for isolating switch motors, the method comprising: collecting temperature data, current data, voltage data, and power data, obtaining voltage-current correlation functions, and then obtaining electrical response indexes; calculating real-time motor thermal sensitivity indexes at each sampling moment; obtaining temperature data with same and different effects, power data with same and different effects, temperature data with non-different effects, and power data with non-different effects; obtaining motor hot spot power influence factors; obtaining updated contraction factors based on motor hot spot power influence factors, and performing abnormal data detection on power sequences. The present application solves the problem of misjudgment caused by temperature changes and realizes accurate diagnosis of motor status.

Description

Technical Field

[0001] The present application relates to the technical field of motor detection, and in particular to a non-invasive detection method and system for an isolating switch motor. Background Art

[0002] Non-intrusive load monitoring technology (NILM) is a revolutionary power monitoring method that allows the identification and monitoring of the working status and energy consumption patterns of individual power devices by analyzing the total load data of power users. Its core advantage lies in its non-intrusiveness, which means that there is no need to install sensors on each device separately, thus greatly reducing costs and reducing interference with users' lives. With the advancement of technology and the deepening of its application, NILM is expected to promote the intelligence of power systems and the modernization of energy management on a global scale, and provide strong technical support for achieving sustainable energy development goals.

[0003] Non-invasive detection technology allows real-time monitoring of motor performance without disassembly or interruption of the production process, thereby reducing maintenance costs and improving production efficiency. However, the impact of temperature changes on motors cannot be ignored. It will cause changes in the resistance, inductance and capacitance characteristics of the motor, which in turn affects the frequency response of the motor. These changes may interfere with the detection signal, making it difficult for the detection algorithm based on fixed parameters to accurately identify the actual state of the motor. Because the working environment of the motor is complex and changeable, and the design and material properties of different motors are different, the non-invasive detection system must have a high degree of flexibility and adaptability, as well as the ability to respond quickly to various environmental changes. In the existing technology, it is impossible to completely avoid the misjudgment of the motor state caused by temperature changes when the motor is working, and it is necessary to reduce the impact of temperature changes and frequency response on detection accuracy. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a non-intrusive detection method and system for isolating switch motors. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides a non-intrusive detection method for an isolating switch motor, the method comprising the following steps:

[0006] Collect temperature data, current data, voltage data, and power data to construct temperature series, current series, voltage series, and power series; use frequency domain conversion algorithms to obtain voltage-current correlation functions based on current series and voltage series;

[0007] Obtain an electrical response index based on the correlation between the temperature sequence, the voltage sequence, the current sequence, and the voltage-current correlation function; obtain a temperature subsequence, a current subsequence, and a voltage subsequence at each sampling moment; obtain a real-time motor thermal sensitivity index at each sampling moment based on the temperature subsequence, the current subsequence, the voltage subsequence, and the electrical response index;

[0008] According to the real-time motor thermal sensitivity index, obtain the temperature data of the same and different effects, the power data of the same and different effects, the temperature data of the non-different effects and the power data of the non-different effects;

[0009] Based on the temperature data with the same or different effects, the power data with the same or different effects, the temperature data with the same or different effects and the power data with the same or different effects, the power factor with the hot spots of the motor is obtained;

[0010] The updated shrinkage factor is obtained according to the hot spot power influencing factor of the motor, and the updated shrinkage factor is used in combination with the clustering algorithm to detect abnormal data in the power sequence.

[0011] Further, the obtaining of the voltage-current correlation function according to the current sequence and the voltage sequence includes:

[0012] The current sequence and the voltage sequence are processed using a frequency domain conversion algorithm to obtain a current spectrum and a voltage spectrum;

[0013] For the current spectrum and voltage spectrum, the ratio of the amplitude in the current spectrum to the amplitude in the voltage spectrum at the same frequency is calculated as the same-frequency amplitude ratio of each frequency, and the function composed of all frequencies and the same-frequency amplitude ratios of frequencies is taken as the voltage-current correlation function.

[0014] Furthermore, the electrical response index is obtained by:

[0015] The minimum value of the frequency in the current spectrum and the voltage spectrum is recorded as the first minimum frequency, and the maximum value of the frequency is recorded as the first maximum frequency;

[0016] The difference between the first maximum frequency and the first minimum frequency is calculated as the first frequency difference, and the sum of all amplitude ratios of the same frequency in the voltage-current correlation function from the first minimum frequency to the first maximum frequency is calculated as the first frequency sum. The electrical response index is negatively correlated with the first frequency difference, and positively correlated with the first frequency sum.

[0017] Furthermore, the method for obtaining the temperature subsequence, current subsequence and voltage subsequence at each sampling moment is:

[0018] For the temperature sequence, starting from each sampling moment, all temperature data before each sampling moment are selected and sorted in chronological order to form a temperature subsequence at each sampling moment;

[0019] For the current sequence and the voltage sequence, the current subsequence and the voltage subsequence at each sampling moment are obtained using the same method as the temperature sequence.

[0020] Further, the real-time motor thermal sensitivity index at each sampling moment is obtained, including:

[0021] For each sampling moment, the correlation coefficient between the temperature subsequence and the voltage subsequence is calculated as the first correlation coefficient, and the correlation coefficient between the temperature subsequence and the current subsequence is calculated as the second correlation coefficient;

[0022] Calculating the mean of the absolute value of the first correlation coefficient and the absolute value of the second correlation coefficient as a first correlation mean, and calculating the ratio of the first correlation coefficient to the second correlation coefficient as a first correlation ratio;

[0023] The real-time motor thermal sensitivity index at each sampling moment is obtained according to the first correlation mean, the first correlation ratio and the electrical response coefficient. The real-time motor thermal sensitivity index at each sampling moment is positively correlated with the first correlation mean, negatively correlated with the first correlation ratio, and positively correlated with the electrical response coefficient.

[0024] Furthermore, the method for acquiring the temperature data with the same or different effects, the power data with the same or different effects, the temperature data with the same or different effects, and the power data with the same or different effects is as follows:

[0025] Arrange the real-time motor thermal sensitivity indexes at all sampling moments in the order of the sampling moments to form a motor thermal sensitivity sequence;

[0026] The motor thermal sensor sequence is used as the input of the abnormal data detection algorithm, and the output is the abnormal value in the motor thermal sensor sequence; the power sequence is used as the input of the abnormal data detection algorithm, and the output is the abnormal value in the power sequence;

[0027] If the sampling time of the abnormal value in the motor thermal sensor sequence and the abnormal value in the power sequence is the same, the sampling time is recorded as the same-different sampling time, and the sampling time of other abnormal values ​​in the motor thermal sensor sequence is recorded as each non-different sampling time;

[0028] For the temperature sequence and power sequence, the temperature data at the same and different sampling moments in the temperature sequence are recorded as the same and different affected temperature data, the power data at the same and different sampling moments in the power sequence are recorded as the same and different affected power data, the temperature data at the non-different sampling moments in the temperature sequence are recorded as the non-different affected temperature data, and the power data at the non-different sampling moments in the power sequence are recorded as the non-different affected power data.

[0029] Furthermore, the method for obtaining the motor hot spot power influence factor is:

[0030] Obtain a first similarity-different ratio, a second similarity-different ratio, a first non-similar ratio, and a second non-similar ratio according to the similarity-different influence temperature data, the similarity-different influence power data, the non-similar influence temperature data, and the non-similar influence power data;

[0031] Calculate the mean between the first similarity and difference ratio and the second similarity and difference ratio as the first similarity and difference mean, and calculate the arithmetic mean of all first similarity and difference means as the similarity and difference mean; calculate the mean between the first non-differential ratio and the second non-differential ratio as the first non-differential mean, and calculate the arithmetic mean of all first non-differential means as the non-differential mean;

[0032] The influence factor of the motor thermoelectric power is obtained according to the average number of similarities and differences and the average number of non-similarities. The influence factor of the motor thermoelectric power is positively correlated with the average number of similarities and differences and negatively correlated with the average number of non-similarities.

[0033] Furthermore, the first similarity-difference ratio, the second similarity-difference ratio, the first non-differential ratio and the second non-differential ratio are obtained by:

[0034] For each temperature data with similar or different influence in the temperature sequence, the absolute value of the difference between each temperature data with similar or different influence and its adjacent previous temperature data is taken as the first temperature difference with similar or different influence, and the absolute value of the difference between each temperature data with its adjacent next temperature data is taken as the second temperature difference with similar or different influence; for each temperature data with no similar influence in the temperature sequence, the absolute value of the difference between each temperature data with similar or different influence and its adjacent previous temperature data is taken as the first temperature difference with no similar influence, and the absolute value of the difference between each temperature data with its adjacent next temperature data is taken as the second temperature difference with no similar influence;

[0035] For each power data of the same and different influences in the power sequence, the first same and different power difference, the second same and different power difference, the first non-different power difference and the second non-different power difference are obtained by the same method as the temperature sequence;

[0036] Calculate the sum of the first same-different power difference and the preset constant as the first sum, calculate the sum of the second same-different power difference and the preset constant as the second sum, calculate the sum of the first non-different power difference and the preset constant as the third sum, and calculate the sum of the second non-different power difference and the preset constant as the fourth sum;

[0037] The ratio of the first same-different temperature difference to the first sum is calculated as the first same-different ratio, the ratio of the second same-different temperature difference to the second sum is calculated as the second same-different ratio, the ratio of the first non-different temperature difference to the third sum is calculated as the first non-different ratio, and the ratio of the second non-different temperature difference to the fourth sum is calculated as the second non-different ratio.

[0038] Further, the obtaining of the updated shrinkage factor and using the updated shrinkage factor in combination with a clustering algorithm to perform abnormal data detection on the power sequence include:

[0039] Calculate the result of an exponential function with a natural constant as the base and a negative value of the motor hot spot power influence factor as the exponent, and use the product of the calculation result and a preset initial contraction factor as an updated contraction factor;

[0040] The power data in the power sequence is used as the input of the clustering algorithm. The shrinkage factor of the clustering algorithm is the updated shrinkage factor. The power data in the power sequence is clustered, and the distance between each power data point and the center of the cluster where it is located is calculated. The power data with a distance greater than a preset normal distance is regarded as abnormal data, otherwise, it is normal data.

[0041] When abnormal data appears, it is determined that the motor is faulty; otherwise, the motor is not faulty.

[0042] In a second aspect, an embodiment of the present application also provides a non-invasive detection system for an isolating switch motor, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.

[0043] This application has at least the following beneficial effects:

[0044] The present application proposes a non-intrusive detection method and system for isolating switch motors. Aiming at the problem of interference of temperature changes on monitoring signals during motor operation, the current, voltage and temperature data are deeply analyzed based on partial correlation analysis and frequency domain signal processing methods, and the real-time motor thermal sensitivity index is obtained. This index can accurately reflect the temperature change characteristics of the motor under different operating conditions. Furthermore, aiming at the influence of temperature changes and frequency response on the accuracy of motor power detection, the power data of the motor is analyzed by using the COF abnormality detection algorithm, and the motor thermoelectric power influence factor is constructed, so that it can quantify the influence of electrical abnormalities and fault abnormalities on motor performance, and provide an index that comprehensively considers the temperature and power changes of the motor. Finally, the contraction factor of the CURE algorithm is improved to solve the misjudgment problem caused by temperature changes, and realize accurate diagnosis of the motor state. Through this algorithm optimization, the non-intrusive detection technology can maintain the continuity of motor operation while eliminating the change of motor power caused by temperature changes caused by motor operation, providing strong technical support for motor maintenance and fault prevention. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 A flowchart of the steps of a non-invasive detection method for an isolating switch motor provided in one embodiment of the present application.

[0047] Figure 2 Get the flow chart for updating the shrink factor. DETAILED DESCRIPTION

[0048] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following is a detailed description of the non-invasive detection method and system for the isolating switch motor proposed in accordance with the present application, its specific implementation, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0049] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0050] The following is a detailed description of a non-invasive detection method and system for an isolating switch motor provided by the present application in conjunction with the accompanying drawings.

[0051] See also Figure 1 , which shows a flowchart of a non-intrusive detection method for an isolating switch motor provided by an embodiment of the present application, the method comprising the following steps:

[0052] Step S1, collecting relevant internal data of the disconnector motor during operation, and pre-processing it to form a relevant data sequence for subsequent analysis.

[0053] A thermistor is installed at the winding of the motor to collect real-time temperature data of the motor. The thermistor selected in this embodiment is an NTC thermistor. A current transformer (CT) is installed in the power supply line of the motor to collect current data of the motor. A voltage sensor is installed at the wiring terminal of the motor to collect voltage data of the motor. A power analyzer is installed at the power input end of the motor or a current transformer and a voltage sensor are used in combination to calculate the power factor to collect power data of the motor.

[0054] In the data collection process, the time when the isolating switch motor starts to run is taken as the initial sampling time, and the data is uniformly collected with t as the time interval. In this embodiment, the value of t is 0.5s. Further, the collected data is normalized to eliminate the influence of the dimension. The normalization method includes the average value normalization method and the maximum and minimum value normalization method. This embodiment uses the average value normalization method for processing, and the implementer can choose other methods according to the actual situation; all the normalized data are sorted in chronological order to obtain the temperature sequence, current sequence, voltage sequence and power sequence.

[0055] Step S2, by conducting an in-depth analysis of the current, voltage and temperature data of the motor during operation, the impact of scene characteristics such as electrical load and ambient temperature on the motor performance is comprehensively considered, thereby quantifying the impact on the motor's thermoelectric power.

[0056] During the operation of the motor, changes in current and voltage will directly affect its temperature. According to the Joule thermal effect, when current passes through the motor winding, it will generate heat due to resistance. The greater the current, the more heat is generated, which causes the motor temperature to rise. Furthermore, changes in the motor load will also cause changes in current demand. Under heavy load conditions, the motor requires a larger current to provide the necessary torque, thereby generating more heat.

[0057] Analyzing the current sequence and voltage sequence, the relationship between the motor current and voltage depends on many factors, such as motor type, load and electrical characteristics. In practical applications, motor startup, load changes and power supply voltage fluctuations will affect the relationship between the two, making the relationship between current and voltage more complicated.

[0058] With the current sequence and voltage sequence as input, the data in the current sequence and voltage sequence are converted from time domain to frequency domain signals using fast Fourier transform (FFT) to obtain the current spectrum and voltage spectrum. The minimum value of the frequency in the current spectrum and voltage spectrum is recorded as the first minimum frequency, and the maximum value of the frequency is recorded as the first maximum frequency.

[0059] Furthermore, for the current spectrum and the voltage spectrum, the ratio of the amplitude in the current spectrum to the amplitude in the voltage spectrum at the same frequency is calculated as the same-frequency amplitude ratio of each frequency, and the function composed of the same-frequency amplitude ratios of all frequencies is used as the voltage-current correlation function. The voltage-current correlation function reflects the response relationship between the current signal and the voltage signal, and characterizes the interaction and influence between the current and voltage at different frequencies.

[0060] For temperature sequence, current sequence and voltage sequence, taking temperature sequence as an example, starting from each sampling moment, all temperature data before each sampling moment are selected and sorted in chronological order to form a temperature subsequence at each sampling moment; similarly, the current subsequence and voltage subsequence at each sampling moment are obtained.

[0061] Furthermore, for each sampling moment, the correlation coefficient between the temperature subsequence and the voltage subsequence is calculated as the first correlation coefficient, and the correlation coefficient between the temperature subsequence and the current subsequence is calculated as the second correlation coefficient, wherein the correlation coefficient includes the Pearson correlation coefficient, the Spearman rank correlation coefficient and the Kendall rank correlation coefficient. In this embodiment, the Pearson correlation coefficient is selected for calculation, and the implementer may select other correlation coefficients according to actual conditions.

[0062] Construct the electrical response index, which is obtained by:

[0063] The difference between the first maximum frequency and the first minimum frequency is calculated as the first frequency difference, and the sum of all amplitude ratios of the same frequency in the voltage-current correlation function from the first minimum frequency to the first maximum frequency is calculated as the first frequency sum. The electrical response index is negatively correlated with the first frequency difference, and positively correlated with the first frequency sum.

[0064] It should be noted that the positive correlation relationship can be a multiplication, addition, or idempotent relationship of an exponential function, which can be determined based on actual conditions and will not be elaborated in this application; the negative correlation relationship can be a subtraction relationship, a division relationship, etc., which can be determined based on actual conditions and will not be elaborated in this application.

[0065] Specifically, the present application proposes an embodiment in which the ratio of the first frequency sum value to the first frequency difference value is used as the electrical response index.

[0066] The electrical response index reflects the electrical characteristic response of the motor within the frequency range. The larger the value, the stronger the motor's response to voltage changes within the frequency range, and vice versa.

[0067] Furthermore, the real-time motor thermal sensitivity index at each sampling moment is obtained according to the electrical response index, and the acquisition method is:

[0068] The average of the absolute value of the first correlation coefficient and the absolute value of the second correlation coefficient is calculated as the first correlation mean, and the ratio of the first correlation coefficient to the second correlation coefficient is calculated as the first correlation ratio. The real-time motor thermal sensitivity index at each sampling moment is positively correlated with the first correlation mean, negatively correlated with the first correlation ratio, and positively correlated with the electrical response coefficient.

[0069] It should be noted that the positive correlation relationship can be an idempotent relationship of multiplication, addition, or exponential function, which can be determined based on actual conditions and will not be elaborated in this application.

[0070] Specifically, the present application proposes an embodiment, calculating the absolute value of the difference between the first correlation ratio and the number 1 as the first difference absolute value, calculating the ratio of the first correlation mean and the first difference absolute value as the first ratio, and taking the product of the first ratio and the electrical response coefficient as the real-time motor thermal sensitivity index.

[0071] Specifically, the first correlation coefficient and the second correlation coefficient respectively reflect the influence of voltage and current on motor temperature. The first correlation mean represents the overall influence of current and voltage changes on motor temperature during motor operation. The larger the first correlation mean, the greater the overall influence of current and voltage on motor temperature, and the larger the value of real-time motor thermal sensitivity index. The absolute value of the first difference reflects the degree of balance of the motor's sensitivity to current and voltage changes. The larger its value, the greater the difference in the motor's sensitivity to current and voltage changes, and the smaller the real-time motor thermal sensitivity index.

[0072] When the motor is running, the increase in internal temperature of the motor will cause the resistance to increase, thereby increasing the loss when current passes through, reducing the efficiency and effective output power of the motor. At the same time, high temperature may affect the properties of the magnetic material inside the motor, such as magnetic permeability and magnetic saturation, thereby affecting the magnetic field strength, further causing the efficiency of the motor to change.

[0073] The real-time motor thermal sensitivity indexes at all sampling moments are arranged in the order of the sampling moments to form a motor thermal sensitivity sequence.

[0074] Furthermore, the motor thermal sensor sequence is used as the input of the abnormal data detection algorithm, and the output is the abnormal value in the motor thermal sensor sequence; similarly, the power sequence is used as the input of the abnormal data detection algorithm, and the output is the abnormal value in the power sequence.

[0075] Specifically, the abnormal data detection algorithm includes K nearest neighbor algorithm, LOF abnormal data detection algorithm, isolation forest and COF abnormal data detection algorithm, etc. This embodiment adopts the COF abnormal data detection algorithm, and the implementer can select it according to the actual situation.

[0076] The abnormal values ​​in the motor thermal sensor sequence and the abnormal values ​​in the power sequence are analyzed. If the sampling time of the abnormal value in the motor thermal sensor sequence and the abnormal value in the power sequence is the same, the sampling time is recorded as the same sampling time, and the sampling time of other abnormal values ​​in the motor thermal sensor sequence is recorded as each non-different sampling time.

[0077] For the temperature sequence and power sequence, the temperature data at the same and different sampling moments in the temperature sequence are recorded as the same and different affected temperature data, the power data at the same and different sampling moments in the power sequence are recorded as the same and different affected power data, the temperature data at the non-different sampling moments in the temperature sequence are recorded as the non-different affected temperature data, and the power data at the non-different sampling moments in the power sequence are recorded as the non-different affected power data.

[0078] For each temperature data with similar or different influences in the temperature sequence, the absolute value of the difference between each temperature data with similar or different influences and its adjacent previous temperature data is taken as the first temperature difference with similar or different influences, and the absolute value of the difference between each temperature data with its adjacent next temperature data is taken as the second temperature difference with similar or different influences; for each temperature data with non-different influences in the temperature sequence, the absolute value of the difference between each temperature data with similar or different influences and its adjacent previous temperature data is taken as the first temperature difference with similar or different influences, and the absolute value of the difference between each temperature data with its adjacent next temperature data is taken as the second temperature difference with similar or different influences.

[0079] Similarly, for each power data with similar or different influences in the power sequence, the absolute value of the difference between each power data with similar or different influences and its previous adjacent power data is taken as the first similar or different power difference, and the absolute value of the difference between each power data with its next adjacent power data is taken as the second similar or different power difference; for each power data with non-different influences in the power sequence, the absolute value of the difference between each power data with non-different influences and its previous adjacent power data is taken as the first non-different power difference, and the absolute value of the difference between each power data with its next adjacent power data is taken as the second non-different power difference.

[0080] Furthermore, based on the temperature series and power series, the motor thermoelectric power influence factor is constructed, and the acquisition method is as follows:

[0081] The sum of the first same-different power difference and the preset constant is calculated as the first sum, the sum of the second same-different power difference and the preset constant is calculated as the second sum, the sum of the first non-different power difference and the preset constant is calculated as the third sum, and the sum of the second non-different power difference and the preset constant is calculated as the fourth sum.

[0082] The ratio of the first same-different temperature difference to the first sum is calculated as the first same-different ratio, the ratio of the second same-different temperature difference to the second sum is calculated as the second same-different ratio, the ratio of the first non-different temperature difference to the third sum is calculated as the first non-different ratio, and the ratio of the second non-different temperature difference to the fourth sum is calculated as the second non-different ratio.

[0083] Calculate the mean between the first similarity and difference ratio and the second similarity and difference ratio as the first similarity and difference mean, and calculate the arithmetic mean of all first similarity and difference means as the similarity and difference mean; calculate the mean between the first non-differential ratio and the second non-differential ratio as the first non-differential mean, and calculate the arithmetic mean of all first non-differential means as the non-differential mean.

[0084] The motor thermoelectric power influence factor is positively correlated with the homogeneous average and negatively correlated with the non-homogeneous average.

[0085] It should be noted that the positive correlation relationship can be a multiplication, addition, or idempotent relationship of an exponential function, which can be determined based on actual conditions and will not be elaborated in this application; the negative correlation relationship can be a subtraction relationship, a division relationship, etc., which can be determined based on actual conditions and will not be elaborated in this application.

[0086] Specifically, the present application proposes an embodiment, calculating the sum of the non-differential average and a preset constant as the fifth sum, and taking the ratio of the same and different average to the fifth sum as the motor thermoelectric power influence factor. In this embodiment, the preset constant is set to 1 to prevent the denominator from being 0.

[0087] In the calculation of the motor's thermal power influence factor, the moment of the electrical anomaly and the moment of the fault anomaly are first determined through the motor's thermal sensor sequence. The average of the temperature change ratio and the power change ratio of the two sequences near the electrical anomaly value is calculated to quantify the relative change degree of the motor temperature and power when the electrical anomaly occurs. The larger the value, the greater the influence of the motor temperature on the power, that is, the electrical anomaly leads to a significant temperature change, which further leads to a greater power change, and the greater the motor's thermal power influence factor; and the calculation process of the non-differential average is similar to the calculation of the average of the electrical influence characteristic value, but focuses on the influence of the fault anomaly value, the purpose is to quantify the influence of the fault anomaly on the motor temperature and power. In the operation of the motor, electrical problems and fault problems are independent of each other. They represent two different aspects of the motor performance problem, and the total number of faults caused is certain, so the two are inversely proportional. The larger the non-differential average, the smaller the average of the same and different, the smaller the thermoelectric power influence, and the smaller the motor's thermal power influence factor.

[0088] Step S3, improving the CURE algorithm to perform abnormality detection on the motor power, using the clustered outliers as a basis for abnormality judgment, and realizing non-intrusive detection of the motor.

[0089] The non-intrusive detection of the isolating switch is mainly to detect potential defects of the isolating switch during operation. This method does not change the complete structure of the isolating switch to determine whether there is a mechanical failure or performance degradation of the isolating switch.

[0090] Furthermore, this embodiment selects the CURE hierarchical clustering algorithm to realize outlier detection of the power sequence. Here, the power sequence is used as input. At the same time, the CURE hierarchical clustering algorithm needs to preset parameters including representative points and initial shrinkage factors. In this embodiment, the representative points are set to 3, and the initial shrinkage factor is set to 0.7. The CURE hierarchical clustering algorithm is used to cluster the power sequence. The CURE hierarchical clustering algorithm takes each power data in the power sequence as a cluster, combines the clusters with the closest distance to form a new cluster, and reduces the influence of outliers on the clustering effect according to the shrinkage factor. However, due to the mixed changes of current, voltage, and temperature on power, it will be difficult for the algorithm to identify the true cluster structure, resulting in outliers interfering with the recognition of these patterns, making the clustering results more unclear. Here, the shrinkage factor in the CURE algorithm is improved, and the formula is: ;in, To update the shrinkage factor, is the influence factor of the motor thermoelectric power, is the initial shrinkage factor. The flowchart for updating the shrinkage factor is as follows: Figure 2 shown.

[0091] The motor thermoelectric power impact factor is an indicator to measure the impact of electrical anomalies on motor efficiency. The larger the value, the greater the electrical impact on the motor power. At this time, the updated shrinkage factor will decrease, making the clustering algorithm more sensitive to outliers and able to identify outliers more accurately.

[0092] Furthermore, this embodiment uses the CURE hierarchical clustering algorithm and uses an updated shrinkage factor to cluster the power series. After obtaining each cluster, the Euclidean distance of each power data point from the center of the cluster to which it belongs is calculated, and the power data whose Euclidean distance is greater than the preset normal distance is regarded as abnormal data, otherwise, it is regarded as normal data.

[0093] The abnormal values ​​identified by the algorithm are used as the basis for judging the motor power fault. When abnormal data appears, the motor is determined to be faulty, which enables early fault diagnosis of the disconnector and improves the operating reliability of the equipment. This makes it possible to eliminate the impact of temperature changes on the power during motor operation and accurately judge whether the motor power is within the normal range.

[0094] Based on the same inventive concept as the above method, an embodiment of the present application also provides a non-invasive detection system for an isolating switch motor, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the system implements the steps of any one of the above-mentioned non-invasive detection methods for isolating switch motors.

[0095] It should be noted that the above sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0096] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0097] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present application should be included in the protection scope of the present application.

Claims

1. A non-intrusive detection method for an isolating switch motor, characterized in that: The method comprises the following steps: Collect temperature data, current data, voltage data, and power data to construct temperature series, current series, voltage series, and power series; use frequency domain conversion algorithms to obtain voltage-current correlation functions based on current series and voltage series; Obtain an electrical response index based on the correlation between the temperature sequence, the voltage sequence, the current sequence, and the voltage-current correlation function; obtain a temperature subsequence, a current subsequence, and a voltage subsequence at each sampling moment; obtain a real-time motor thermal sensitivity index at each sampling moment based on the temperature subsequence, the current subsequence, the voltage subsequence, and the electrical response index; According to the real-time motor thermal sensitivity index, obtain the temperature data of the same and different effects, the power data of the same and different effects, the temperature data of the non-different effects and the power data of the non-different effects; Based on the temperature data with the same or different effects, the power data with the same or different effects, the temperature data with the same or different effects and the power data with the same or different effects, the power factor with the hot spots of the motor is obtained; The updated shrinkage factor is obtained according to the hot spot power influencing factor of the motor, and the updated shrinkage factor is used in combination with the clustering algorithm to detect abnormal data in the power sequence; The method for obtaining the motor hot spot power influence factor is: Obtain a first similarity-different ratio, a second similarity-different ratio, a first non-similar ratio, and a second non-similar ratio according to the similarity-different influence temperature data, the similarity-different influence power data, the non-similar influence temperature data, and the non-similar influence power data; Calculate the mean between the first similarity and difference ratio and the second similarity and difference ratio as the first similarity and difference mean, and calculate the arithmetic mean of all first similarity and difference means as the similarity and difference mean; calculate the mean between the first non-differential ratio and the second non-differential ratio as the first non-differential mean, and calculate the arithmetic mean of all first non-differential means as the non-differential mean; Obtaining a motor thermoelectric power influence factor according to the average of the same and different numbers and the average of the non-different numbers, wherein the motor thermoelectric power influence factor is positively correlated with the average of the same and different numbers and negatively correlated with the average of the non-different numbers; The obtaining of the updated shrinkage factor and using the updated shrinkage factor in combination with a clustering algorithm to perform abnormal data detection on the power sequence includes: Calculate the result of an exponential function with a natural constant as the base and a negative value of the motor hot spot power influence factor as the exponent, and use the product of the calculation result and a preset initial contraction factor as an updated contraction factor; The power data in the power sequence is used as the input of the clustering algorithm. The shrinkage factor of the clustering algorithm is the updated shrinkage factor. The power data in the power sequence is clustered, and the distance between each power data point and the center of the cluster where it is located is calculated. The power data with a distance greater than a preset normal distance is regarded as abnormal data, otherwise, it is normal data. When abnormal data appears, it is determined that the motor is faulty; otherwise, the motor is not faulty.

2. A non-intrusive detection method for an isolating switch motor according to claim 1, characterized in that: The step of obtaining a voltage-current correlation function according to a current sequence and a voltage sequence includes: The current sequence and the voltage sequence are processed using a frequency domain conversion algorithm to obtain a current spectrum and a voltage spectrum; For the current spectrum and voltage spectrum, the ratio of the amplitude in the current spectrum to the amplitude in the voltage spectrum at the same frequency is calculated as the same-frequency amplitude ratio of each frequency, and the function composed of all frequencies and the same-frequency amplitude ratios of frequencies is taken as the voltage-current correlation function.

3. A non-intrusive detection method for an isolating switch motor as claimed in claim 2, characterized in that: The electrical response index is obtained by: The minimum value of the frequency in the current spectrum and the voltage spectrum is recorded as the first minimum frequency, and the maximum value of the frequency is recorded as the first maximum frequency; The difference between the first maximum frequency and the first minimum frequency is calculated as the first frequency difference, and the sum of all amplitude ratios of the same frequency in the voltage-current correlation function from the first minimum frequency to the first maximum frequency is calculated as the first frequency sum. The electrical response index is negatively correlated with the first frequency difference, and positively correlated with the first frequency sum.

4. A non-intrusive detection method for an isolating switch motor according to claim 1, characterized in that: The method for obtaining the temperature subsequence, current subsequence and voltage subsequence at each sampling moment is: For the temperature sequence, starting from each sampling moment, all temperature data before each sampling moment are selected and sorted in chronological order to form a temperature subsequence at each sampling moment; For the current sequence and the voltage sequence, the current subsequence and the voltage subsequence at each sampling moment are obtained using the same method as the temperature sequence.

5. A non-intrusive detection method for an isolating switch motor according to claim 1, characterized in that: The step of obtaining the real-time motor thermal sensitivity index at each sampling moment includes: For each sampling moment, the correlation coefficient between the temperature subsequence and the voltage subsequence is calculated as the first correlation coefficient, and the correlation coefficient between the temperature subsequence and the current subsequence is calculated as the second correlation coefficient; Calculating the mean of the absolute value of the first correlation coefficient and the absolute value of the second correlation coefficient as a first correlation mean, and calculating the ratio of the first correlation coefficient to the second correlation coefficient as a first correlation ratio; The real-time motor thermal sensitivity index at each sampling moment is obtained according to the first correlation mean, the first correlation ratio and the electrical response coefficient. The real-time motor thermal sensitivity index at each sampling moment is positively correlated with the first correlation mean, negatively correlated with the first correlation ratio, and positively correlated with the electrical response coefficient.

6. A non-intrusive detection method for an isolating switch motor according to claim 1, characterized in that: The method for obtaining the temperature data with the same or different effects, the power data with the same or different effects, the temperature data with the same or different effects, and the power data with the same or different effects is as follows: Arrange the real-time motor thermal sensitivity indexes at all sampling moments in the order of the sampling moments to form a motor thermal sensitivity sequence; The motor thermal sensor sequence is used as the input of the abnormal data detection algorithm, and the output is the abnormal value in the motor thermal sensor sequence; the power sequence is used as the input of the abnormal data detection algorithm, and the output is the abnormal value in the power sequence; If the sampling time of the abnormal value in the motor thermal sensor sequence and the abnormal value in the power sequence is the same, the sampling time is recorded as the same-different sampling time, and the sampling time of other abnormal values ​​in the motor thermal sensor sequence is recorded as each non-different sampling time; For the temperature sequence and power sequence, the temperature data at the same and different sampling moments in the temperature sequence are recorded as the same and different affected temperature data, the power data at the same and different sampling moments in the power sequence are recorded as the same and different affected power data, the temperature data at the non-different sampling moments in the temperature sequence are recorded as the non-different affected temperature data, and the power data at the non-different sampling moments in the power sequence are recorded as the non-different affected power data.

7. A non-intrusive detection method for an isolating switch motor according to claim 1, characterized in that: The method for obtaining the first similarity-dissimilarity ratio, the second similarity-dissimilarity ratio, the first non-dissimilarity ratio and the second non-dissimilarity ratio is: For each temperature data with similar or different influence in the temperature sequence, the absolute value of the difference between each temperature data with similar or different influence and its adjacent previous temperature data is taken as the first temperature difference with similar or different influence, and the absolute value of the difference between each temperature data with its adjacent next temperature data is taken as the second temperature difference with similar or different influence; for each temperature data with no similar influence in the temperature sequence, the absolute value of the difference between each temperature data with similar or different influence and its adjacent previous temperature data is taken as the first temperature difference with no similar influence, and the absolute value of the difference between each temperature data with its adjacent next temperature data is taken as the second temperature difference with no similar influence; For each power data of the same and different influences in the power sequence, the first same and different power difference, the second same and different power difference, the first non-different power difference and the second non-different power difference are obtained by the same method as the temperature sequence; Calculate the sum of the first same-different power difference and the preset constant as the first sum, calculate the sum of the second same-different power difference and the preset constant as the second sum, calculate the sum of the first non-different power difference and the preset constant as the third sum, and calculate the sum of the second non-different power difference and the preset constant as the fourth sum; The ratio of the first same-different temperature difference to the first sum is calculated as the first same-different ratio, the ratio of the second same-different temperature difference to the second sum is calculated as the second same-different ratio, the ratio of the first non-different temperature difference to the third sum is calculated as the first non-different ratio, and the ratio of the second non-different temperature difference to the fourth sum is calculated as the second non-different ratio.

8. A non-intrusive detection system for an isolating switch motor, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of a non-intrusive detection method for an isolating switch motor as claimed in any one of claims 1 to 7 are implemented.

Images