A low-frequency injection method high-voltage motor insulation detection device

Abstract

This invention provides an insulation testing device for high-voltage motors using a low-frequency injection method. By employing non-invasive injection of ultra-low frequency multi-frequency signals and identification using a fractional-order equivalent circuit model, it achieves a safe, online, and accurate assessment of the insulation status of high-voltage motors. This method completely avoids the cumbersome manual high-voltage operations of traditional testing, fundamentally eliminating the risk of electric shock, and supports testing of equipment in hot standby mode, improving operation and maintenance efficiency and safety. In particular, the introduced fractional-order model can more accurately fit the actual dielectric relaxation behavior of insulating materials. The identified parameters are extremely sensitive to degradation processes such as insulation aging and moisture absorption. Combined with the rich information obtained from multi-frequency sweeps, multi-dimensional characteristic quantities such as the model-predicted polarization index and dielectric loss spectrum can be calculated, thereby achieving more sensitive capture of early insulation defects and more accurate fault mode identification, providing strong technical support for predictive maintenance of high-voltage motors.

Description

Technical Field

[0001] This invention relates to the field of high-voltage motor insulation testing technology, and in particular to an insulation testing device for high-voltage motors using the low-frequency injection method. Background Technology

[0002] High-voltage motors are core power equipment in industrial production, and their operational reliability directly affects the safety and stability of the entire production system. The quality of the stator winding insulation is a key factor determining the lifespan and reliability of high-voltage motors. During long-term operation, the combined effects of electrical, thermal, mechanical, and environmental stresses can cause the insulation materials to gradually age and deteriorate, potentially leading to insulation breakdown faults, resulting in significant economic losses or even safety accidents. Therefore, regular and effective inspection and evaluation of the insulation condition of high-voltage motors is an important technical means to implement predictive maintenance and avoid sudden downtime.

[0003] Traditional high-voltage motor insulation testing primarily relies on periodic power outages for maintenance, using megohmmeters to measure the insulation resistance and absorption ratio of the windings. This method has several inherent drawbacks: First, the operation is cumbersome and risky. Testing requires manual intervention involving a series of switching operations, such as pulling in and out of the switchgear trolley, opening and closing grounding switches, and opening and closing cabinet doors. This is labor-intensive, and frequent operations can easily lead to wear and tear on mechanical parts. More seriously, operators may face the significant safety risk of electric shock if they accidentally enter a live compartment. Second, the testing information is limited and offline. Traditional methods only obtain limited point information near DC or power frequency (such as resistance values ​​at 1 minute and 10 minutes), failing to reflect the dielectric response characteristics of the insulation material across a wide frequency range and making it difficult to reveal early, deep-seated insulation defects (such as early partial discharge or uneven moisture absorption). Furthermore, offline testing requires equipment shutdown, affecting production continuity, and cannot capture the dynamic impact of changes in operating conditions on the insulation state.

[0004] With the popularization of the condition-based maintenance concept, some online monitoring technologies have been proposed, such as monitoring methods based on injecting specific signals into the neutral point or grounding wire. However, most existing online monitoring solutions still rely on simplified RC circuit models to analyze insulation characteristics. This ideal model is difficult to accurately describe the complex relaxation process and frequency-varying characteristics of actual insulation materials, resulting in insufficient accuracy in assessing insulation condition and inadequate sensitivity to early, subtle degradation. How to achieve a safe, efficient, and more profoundly characterizing online detection method for insulation materials without affecting normal equipment operation has become a pressing technical challenge in this field. Therefore, there is an urgent need for an innovative detection device and method that can completely avoid the safety hazards caused by direct manual operation of high-voltage equipment, and achieve more accurate and earlier intelligent assessment and early warning of the insulation condition of high-voltage motors through more advanced sensing and modeling technologies. Summary of the Invention

[0005] The purpose of this invention is to provide an insulation testing device for a low-frequency injection method high-voltage motor to solve the problems existing in the prior art.

[0006] To achieve the above objectives, the present invention provides the following solution: This invention provides an insulation testing device for a low-frequency injection method high-voltage motor, comprising: The signal injection unit is used to generate and inject a set of low-frequency AC test signals of a preset frequency between the stator winding of the high-voltage motor and ground. The signal acquisition unit is connected to the signal injection unit and the high-voltage motor under test, and is used to synchronously acquire the voltage signal at the low-frequency AC test signal application point and the leakage current signal flowing through the grounding wire. The data processing and analysis unit is connected to the signal acquisition unit and is used to process the acquired voltage signal and leakage current signal, and to perform parameter identification and state assessment based on a preset fractional-order equivalent circuit model of the insulation system. The safety isolation and control unit is connected to the signal injection unit, signal acquisition unit, and data processing and analysis unit, respectively, to achieve electrical isolation between the high-voltage side and the low-voltage side, and to control the entire detection process according to the instructions of the data processing and analysis unit.

[0007] Preferably, the signal injection unit includes: An ultra-low frequency signal generator, wherein the ultra-low frequency signal generator is used to generate a sinusoidal signal with a frequency range of 0.01 Hz to 10 Hz; A power amplifier, wherein the power amplifier is used to linearly amplify the sinusoidal signal; A high-insulation isolation coupler is used to safely couple an amplified test signal to the three-phase short-circuit point of the stator winding of a high-voltage motor and ground through a series current-limiting resistor.

[0008] Preferably, in the data processing and analysis unit, the built-in fractional-order equivalent circuit model of the insulation system consists of an insulation resistance characterizing the overall insulation condition, a distributed capacitance characterizing the geometric structure, and at least one fractional-order polarization branch connected in parallel, consisting of a polarization resistor and a constant-phase element connected in series; the impedance expression of the constant-phase element is: ; in, This is the pseudo capacitance coefficient. For fractional order, 0 < ≤1, Angular frequency, It is the imaginary unit.

[0009] Preferably, the data processing and analysis unit is further configured to calculate derived insulation state indices based on the identified optimal model parameters. The indices include: the model-predicted insulation resistance value under DC conditions, the model-predicted polarization index calculated based on the time-domain response curve, and the spectral curve of the dielectric loss factor as a function of frequency.

[0010] Preferably, the data processing and analysis unit further includes a health status assessment module. The health status assessment module is used to compare the model-predicted insulation resistance value, model-predicted polarization index, low-frequency dielectric loss factor and fractional order into a feature vector with a pre-established fault feature library, and output the insulation status level and preliminary degradation type diagnosis results.

[0011] Preferably, the security isolation and control unit includes: The switch, which is a high-voltage vacuum relay or contactor, is used to completely physically disconnect the detection device from the main circuit of the high-voltage motor during non-test periods. The fiber optic communication module is used to transmit control commands and acquire data between the high-voltage side and the low-voltage side, achieving electrical isolation.

[0012] Preferably, it also includes a human-computer interaction and communication unit, which is connected to the data processing and analysis unit, for locally displaying insulation status information and parameter settings, and uploading detection results, early warning information and historical data to the remote monitoring center via wired or wireless network.

[0013] This invention also provides an insulation testing method for a high-voltage motor using the low-frequency injection method, comprising the following steps: S1. System self-test and safety verification steps: Confirm that the high-voltage motor is in a stopped or hot standby state, and control the safety isolation unit to connect the detection device to the test circuit; S2. Multi-frequency signal injection and data acquisition steps: Control the signal injection unit to inject low-frequency test signals sequentially at at least two different frequencies. The signal injection unit injects low-frequency test signals sequentially according to a preset frequency sequence, and the signal acquisition unit synchronously acquires the fundamental voltage phasor and leakage current phasor at each frequency. S3. Model parameter identification steps: Based on the fractional equivalent circuit model, the model parameter set characterizing the insulation state is identified through optimization algorithms using response data at multiple frequency points. S4. Status Assessment and Early Warning Steps: Calculate multi-dimensional assessment indicators based on the identified parameters, combine historical data and fault models to determine the current insulation health status, and generate early warning information if abnormal.

[0014] Preferably, step S3 includes: S31. Calculate the measured admittance of the insulation system under test at each frequency point. The calculation formula is as follows: ; in, For leakage current phasor, It is the fundamental voltage phasor; S32. Construct a theoretical admittance function with the parameter vector of the fractional-order equivalent circuit model as variables. The function formula is: ; in, For the first Injection frequency, For parameter vectors, For insulation resistance, The imaginary unit, Angular frequency, Geometric capacitance Polarization resistor, The pseudo capacitance coefficient of CPE, For fractional order; S33. By solving the nonlinear least squares optimization problem, the optimal model parameters are identified, and the least squares objective function is: ; in, This represents the total number of frequency points.

[0015] Preferably, in the model parameter identification, the optimization algorithm used is the Levenburg-Marquardt algorithm, which iteratively updates the parameter vector and minimizes the sum of squared residuals between the measured and theoretical admittances until a preset convergence condition is met. Or reach the maximum number of iterations, where, The gradient of the objective function. This is the convergence threshold.

[0016] The present invention achieves the following beneficial technical effects compared to the prior art: This invention provides an insulation testing device for high-voltage motors using a low-frequency injection method. By employing non-invasive injection of ultra-low frequency multi-frequency signals and fractional-order equivalent circuit model identification, it achieves safe, online, and accurate assessment of the insulation status of high-voltage motors. This method completely avoids the cumbersome manual high-voltage operations of traditional testing, fundamentally eliminating the risk of electric shock, and supports testing of equipment in hot standby mode, improving operation and maintenance efficiency and safety. In particular, the introduced fractional-order model can more accurately fit the actual dielectric relaxation behavior of the insulating material. The identified parameters are extremely sensitive to degradation processes such as insulation aging and moisture absorption. Combined with the rich information obtained from multi-frequency sweeps, multi-dimensional characteristic quantities such as the model-predicted polarization index and dielectric loss spectrum can be calculated, thereby achieving more sensitive capture of early insulation defects and more accurate fault mode identification, providing strong technical support for predictive maintenance of high-voltage motors. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the insulation detection device for a low-frequency injection method high-voltage motor provided by the present invention. Figure 2 The flowchart illustrates the insulation testing method for high-voltage motors using the low-frequency injection method provided by this invention. Detailed Implementation

[0019] 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.

[0020] The purpose of this invention is to provide an insulation testing device for a low-frequency injection method high-voltage motor to solve the problems existing in the prior art.

[0021] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] Example 1: This invention provides an insulation testing device for a high-voltage motor using a low-frequency injection method, the core architecture of which is as follows: Figure 1As shown, it mainly includes a signal injection unit, a signal acquisition unit, a data processing and analysis unit, and a safety isolation and control unit. These units work together to form a non-intrusive, intelligent online monitoring system that can operate safely in a high-voltage motor shutdown or hot standby state.

[0023] Specifically, the core of the signal injection unit lies in generating a set of ultra-low frequency, high-safety AC test signals. This unit consists of an ultra-low frequency signal generator, a power amplifier, and a high-insulation isolation coupler connected in sequence. The ultra-low frequency signal generator employs high-precision direct digital frequency synthesis technology, capable of generating precisely adjustable sine wave signals within the frequency range of 0.01Hz to 10Hz. Typical frequency sequences include 0.1Hz, 0.3Hz, 1Hz, 3Hz, and 10Hz. The ultra-low frequency band is chosen to significantly reduce the capacitive reactance of the motor windings to ground, increasing the proportion of resistive current flowing through the insulation resistance, thereby improving detection sensitivity and signal-to-noise ratio. The power amplifier linearly amplifies this weak signal to a safe voltage amplitude, typically between 50V and 250V. Subsequently, the high-insulation isolation coupler, through a series high-precision current-limiting resistor, applies the amplified test signal between the three-phase short-circuit point of the high-voltage motor stator winding and the motor grounding wire. The coupler adopts a high-frequency isolation transformer design, which has extremely high insulation strength and extremely low distributed capacitance between its primary and secondary sides. This ensures that the test signal can be effectively injected while strictly limiting the injected current to a safe range of milliamperes, and achieves complete electrical isolation between the test system and the power grid's high-frequency voltage.

[0024] Furthermore, the signal acquisition unit is responsible for accurately and synchronously capturing the insulation system's response to the injected signal. This unit includes a high-isolation differential voltage sensor and a high-sensitivity AC leakage current sensor. The voltage sensor is connected across the current-limiting resistor of the injection unit to accurately measure the voltage signal at the injection point. The current sensor is precisely fitted onto the motor's main grounding wire to measure the leakage current signal induced by the test signal. Both sensors employ isolation designs based on fiber optic transmission or capacitive voltage division principles to ensure the safe acquisition of high-voltage side signals. Their outputs are sent to a multi-channel synchronous analog-to-digital converter module, which performs strict synchronous sampling of the two signals at a sampling rate of no less than 1 kSPS to eliminate phase errors introduced by inter-channel delay, providing an accurate data basis for subsequent complex admittance calculations.

[0025] Furthermore, the core innovation of the data processing and analysis unit lies in its use of a fractional-order equivalent circuit model to model and identify the frequency domain response of the insulation system. In this embodiment, the high-voltage motor stator winding-to-ground insulation system is characterized as an extended fractional-order Debye model. This model consists of insulation resistance characterizing the overall insulation condition, geometric capacitance characterizing the physical structure, and one or more fractional-order polarization branches connected in parallel, each composed of a series polarization resistor and a constant-phase element. The constant-phase element is crucial for describing the non-ideal dielectric relaxation behavior, and its impedance expression is: ; in, This is the pseudo capacitance coefficient. For fractional order, 0 < ≤1, when When = 1, CPE degenerates into an ideal capacitor; when When the value is less than 1, it can more accurately describe the complex polarization and charge diffusion processes inside the insulating material. The further the value deviates from 1, the more severe the aging or non-uniformity of the insulation material. Angular frequency, It is the imaginary unit.

[0026] The workflow of this unit is as follows: First, the control signal injection unit injects test signals sequentially according to a preset frequency sequence. For each frequency, the unit performs a Fast Fourier Transform on a synchronously acquired steady-state waveform, extracts its fundamental component, and obtains the complex form of the voltage and current phasors. Subsequently, the measured admittance at that frequency point is calculated: ; in, For leakage current phasor, This is the fundamental voltage phasor.

[0027] Next, a theoretical admittance function based on the above fractional-order model is constructed, which is a function of the parameter vector to be identified: ; in, For the first Injection frequency, For parameter vectors, For insulation resistance, The imaginary unit, Angular frequency, Geometric capacitance Polarization resistor, The pseudo capacitance coefficient of CPE, The order is fractional; the goal of parameter identification is to find a set of optimal parameters that minimizes the difference between the theoretical admittance and the measured admittance at all test frequencies. This is achieved by solving a nonlinear least squares optimization problem, the objective function of which is: ; in, This represents the total number of frequency points.

[0028] This invention preferably employs the Levenburg-Marquardt algorithm to iteratively solve this optimization problem. This algorithm dynamically adjusts the damping factor, adaptively switching between the Gauss-Newton method and the steepest descent method, exhibiting fast convergence speed and good stability. The iterative process continues until the update amount of the parameter vector is less than a preset threshold or the maximum number of iterations is reached, at which point the optimal parameter estimate is obtained.

[0029] After obtaining the optimal parameters, the data processing and analysis unit can further calculate a series of derived insulation state indices to achieve a multi-dimensional and in-depth evaluation of insulation performance: Model-predicted insulation resistance and polarization index: By substituting the identified parameters into the model, the time-domain current response under DC voltage step excitation is solved numerically, and the model-predicted insulation resistance as a function of time is calculated. Based on this, a model-predicted polarization index that better reflects physical reality can be calculated. This index has better repeatability and discrimination against early moisture absorption than the traditional manually measured polarization index.

[0030] Dielectric loss factor spectrum: The dielectric loss factor at any frequency can be calculated using optimal parameters. In particular, the dielectric loss factor in the low-frequency region (e.g., 0.1 Hz) is extremely sensitive to insulation moisture and contamination.

[0031] Comprehensive Condition Assessment and Early Warning: The health status assessment module within the unit constructs a feature vector from key parameters and compares and analyzes it with a built-in fault feature library trained on a large amount of historical data and machine learning algorithms (such as support vector machines and random forests). Finally, the module outputs four levels of insulation condition assessment results (such as excellent, caution, warning, and alarm), and can preliminarily diagnose the type of degradation, such as "overall dampness," "local aging," or "surface contamination."

[0032] Furthermore, the safety isolation and control unit provides safety assurance for the entire testing process. It mainly consists of a switch group composed of high-voltage vacuum relays or contactors. During non-testing periods, these switches are in the open state, ensuring complete physical disconnection between the testing device and the main circuit of the high-voltage motor. During testing, the data processing and analysis unit is only remotely controlled to close after its safety logic has confirmed its correctness. In addition, all control commands and data communication between the high-voltage and low-voltage sides are transmitted through a fiber optic communication module, achieving complete electrical isolation and preventing high-voltage intrusion into the low-voltage control section.

[0033] Furthermore, the device of this invention also includes a human-machine interaction and communication unit, which typically includes a local touchscreen and multiple communication interfaces (such as Ethernet, 4G / 5G). It is used for parameter configuration, real-time status display, historical trend query, and uploads all insulation status reports, raw data, and early warning information to the factory's monitoring system or cloud platform via standard industrial protocols (such as Modbus TCP / IP, MQTT) to achieve remote centralized monitoring and data analysis.

[0034] Example 2: Combination Figure 2 The insulation testing method provided by this invention specifically includes the following steps: First, system initialization and safety self-test are performed. After the device is powered on, the data processing and analysis unit obtains the real-time status of the high-voltage motor through the communication interface to confirm that it is in a stopped or safe standby state. Subsequently, the switch in the safety isolation unit is controlled to safely connect the detection device to the test circuit, and a zero-signal injection self-test is performed to collect background noise for subsequent digital filtering and elimination.

[0035] Next, adaptive multi-frequency signal injection and synchronous data acquisition are performed. The data processing and analysis unit intelligently selects the frequency (usually starting from 0.1Hz) and voltage amplitude of the first injected signal based on motor parameters and historical data. The control signal injection unit generates a sinusoidal test signal at that frequency and injects it into the system. The signal acquisition unit simultaneously acquires voltage and leakage current waveforms for several cycles. After completing data acquisition at that frequency point, the device automatically switches to the next preset frequency point (e.g., 0.3Hz) and repeats the injection and acquisition process until the entire frequency sequence sweep test is completed.

[0036] Next, the core model parameter identification and state index calculation are performed. The data processing and analysis unit performs preprocessing such as filtering and FFT transformation on the time-domain signal acquired at each frequency point to obtain a complex admittance sequence. Then, as detailed in the working principle above, the Levenburg-Marquardt algorithm is used to optimize and identify the parameter vector based on the fractional-order equivalent circuit model to obtain the optimal solution. Subsequently, the calculation model is used to predict a series of derived state indices such as polarization index and dielectric loss spectrum.

[0037] Finally, intelligent condition assessment, data archiving, and early warning are performed. The health condition assessment module compares the calculated multi-dimensional characteristic indicators with the fault model library to obtain a comprehensive diagnostic conclusion and level of insulation health condition. If the diagnostic result is "early warning" or "alarm," the system will immediately issue an audible and visual alarm through the local interface of the human-machine interface unit and send alarm information to a preset mobile phone number or monitoring center through the communication unit. At the same time, all raw data, identification parameters, assessment results, and timestamps of this test are completely stored locally and uploaded to the server to generate a standardized test report, providing data support for historical tracking of insulation condition and predictive maintenance decisions.

[0038] In summary, through the above specific embodiments, this invention provides a safe, accurate, and intelligent online insulation detection solution for high-voltage motors, effectively overcoming the drawbacks of traditional methods and possessing significant practical value and innovation.

[0039] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0040] It should be noted that the components mentioned in the above embodiments are all general standard parts or components known to those skilled in the art. Their structures and principles can be learned by those skilled in the art through technical manuals or conventional experimental methods.

[0041] This invention has illustrated its principles and implementation methods using specific examples. The descriptions of these embodiments are merely illustrative of the method and its core ideas; furthermore, those skilled in the art will recognize that modifications may be made to the specific implementation methods and application scope based on the principles of this invention. Therefore, the content of this specification should not be construed as limiting the invention.

Claims

1. An insulation testing device for a high-voltage motor using a low-frequency injection method, characterized in that, include: The signal injection unit is used to generate and inject a set of low-frequency AC test signals of a preset frequency between the stator winding of the high-voltage motor and ground. The signal acquisition unit is connected to the signal injection unit and the high-voltage motor under test, and is used to synchronously acquire the voltage signal at the low-frequency AC test signal application point and the leakage current signal flowing through the grounding wire. The data processing and analysis unit is connected to the signal acquisition unit and is used to process the acquired voltage signal and leakage current signal, and to perform parameter identification and state assessment based on a preset fractional-order equivalent circuit model of the insulation system. The safety isolation and control unit is connected to the signal injection unit, signal acquisition unit, and data processing and analysis unit, respectively, to achieve electrical isolation between the high-voltage side and the low-voltage side, and to control the entire detection process according to the instructions of the data processing and analysis unit.

2. The insulation testing device for a low-frequency injection method high-voltage motor according to claim 1, characterized in that, The signal injection unit includes: An ultra-low frequency signal generator, wherein the ultra-low frequency signal generator is used to generate a sine wave signal with a frequency range of 0.01 Hz to 10 Hz; A power amplifier, wherein the power amplifier is used to linearly amplify the sinusoidal signal; A high-insulation isolation coupler is used to safely couple an amplified test signal to the three-phase short-circuit point of the stator winding of a high-voltage motor and ground through a series current-limiting resistor.

3. The insulation testing device for a low-frequency injection method high-voltage motor according to claim 1, characterized in that, The data processing and analysis unit includes a built-in fractional-order equivalent circuit model of the insulation system, consisting of an insulation resistance characterizing the overall insulation condition, a distributed capacitance characterizing the geometric structure, and at least one parallel fractional-order polarization branch composed of a polarization resistor and a constant-phase element connected in series; the impedance expression of the constant-phase element is: ； in, This is the pseudo capacitance coefficient. For fractional order, 0 < ≤1, Angular frequency, It is the imaginary unit.

4. The insulation testing device for a low-frequency injection method high-voltage motor according to claim 1, characterized in that, The data processing and analysis unit is also used to calculate derived insulation state indices based on the identified optimal model parameters. These indices include: the model-predicted insulation resistance value under DC conditions, the model-predicted polarization index calculated based on the time-domain response curve, and the spectral curve of the dielectric loss factor as a function of frequency.

5. The insulation testing device for a low-frequency injection method high-voltage motor according to claim 4, characterized in that, The data processing and analysis unit also includes a health status assessment module. The health status assessment module is used to compare the model-predicted insulation resistance value, model-predicted polarization index, low-frequency dielectric loss factor and fractional order into a feature vector with a pre-established fault feature library, and output the insulation status level and preliminary degradation type diagnosis results.

6. The insulation testing device for a low-frequency injection method high-voltage motor according to claim 1, characterized in that, The security isolation and control unit includes: The switch, which is a high-voltage vacuum relay or contactor, is used to completely physically disconnect the detection device from the main circuit of the high-voltage motor during non-test periods. The fiber optic communication module is used to transmit control commands and acquire data between the high-voltage side and the low-voltage side, achieving electrical isolation.

7. The insulation testing device for a low-frequency injection method high-voltage motor according to claim 1, characterized in that, It also includes a human-computer interaction and communication unit, which is connected to the data processing and analysis unit, for displaying insulation status information and parameter settings locally, and uploading detection results, early warning information and historical data to the remote monitoring center via wired or wireless network.

8. An insulation testing method for a low-frequency injection method high-voltage motor, employing the insulation testing device for a low-frequency injection method high-voltage motor as described in any one of claims 1-7, characterized in that, Includes the following steps: S1. System self-test and safety verification steps: Confirm that the high-voltage motor is in a stopped or hot standby state, and control the safety isolation unit to connect the detection device to the test circuit; S2. Multi-frequency signal injection and data acquisition steps: Control the signal injection unit to inject low-frequency test signals sequentially at at least two different frequencies. The signal injection unit injects low-frequency test signals sequentially according to a preset frequency sequence, and the signal acquisition unit synchronously acquires the fundamental voltage phasor and leakage current phasor at each frequency. S3. Model parameter identification steps: Based on the fractional equivalent circuit model, the model parameter set characterizing the insulation state is identified through optimization algorithms using response data at multiple frequency points. S4. Status Assessment and Early Warning Steps: Calculate multi-dimensional assessment indicators based on the identified parameters, combine historical data and fault models to determine the current insulation health status, and generate early warning information if abnormal.

9. The insulation testing method for a high-voltage motor using the low-frequency injection method according to claim 8, characterized in that, Step S3 includes: S31. Calculate the measured admittance of the insulation system under test at each frequency point. The calculation formula is as follows: ； in, For leakage current phasor, It is the fundamental voltage phasor; S32. Construct a theoretical admittance function with the parameter vector of the fractional-order equivalent circuit model as variables. The function formula is: ； in, For the first Injection frequency, For parameter vectors, For insulation resistance, The imaginary unit, Angular frequency, Geometric capacitance Polarization resistor, The pseudo capacitance coefficient of CPE, For fractional order; S33. By solving the nonlinear least squares optimization problem, the optimal model parameters are identified, and the least squares objective function is: ； in, This represents the total number of frequency points.

10. The insulation testing method for a high-voltage motor using the low-frequency injection method according to claim 9, characterized in that, In the model parameter identification, the Levenburg-Marquardt algorithm was used to iteratively update the parameter vector and minimize the sum of squared residuals between the measured and theoretical admittances until the preset convergence condition was met. Or reach the maximum number of iterations, where, The gradient of the objective function. This is the convergence threshold.

Images