Single-point breakdown electrical detection method for flat wire stator coil

Abstract

The present application belongs to the technical field of single-point damage of coil, and specifically discloses a single-point damage electrical detection method for flat wire stator coil, which synchronously collects partial discharge characteristic parameters such as discharge quantity and signal rising time during the process of applying test voltage to the flat wire stator coil to be measured, directly reflects the physical characteristics of the discharge channel, and compared with the single criterion method which only relies on voltage decay time, the present application realizes mechanism level recognition of discharge mode through multi-parameter fusion, can significantly reduce missed detection and false positives, and at the same time, when the multi-parameter characteristic recognition based on the stator coil is suspected to have single-point damage, the same test voltage is sequentially applied to the remaining phase and the partial discharge signal is collected, a cross-phase retest verification mechanism of the discharge mode is constructed, the authenticity of the defect signal is identified, the accuracy and reliability of the detection result are improved, and the fault tracing direction is clear, which provides traceable judgment basis for subsequent maintenance.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of single-point damage of coils, and specifically discloses a single-point damage electrical detection method for flat wire stator coils. BACKGROUND

[0002] Flat wire stators are widely used in the field of new energy automobile driving motors and the like due to their high slot fill rate, excellent heat dissipation performance and high power density and the like. However, the insulating paint layer on the surface of the flat wire is extremely easy to be damaged at a small single point due to mechanical stress in the manufacturing, wire embedding and the like processes. Such defects may not immediately cause a fault during the initial operation of the motor, but under the working conditions of high voltage, high heat and vibration for a long time, the damaged point will gradually deteriorate, and finally cause the motor to fail. Therefore, it is particularly necessary to detect the single-point damage of the coils before the stator is assembled.

[0003] There are technical solutions related to the single-point damage detection of flat wire stator coils in the prior art, such as a motor stator coil single-point damage detection method disclosed in Chinese Patent Publication No. CN110514999A, which applies a modulated high voltage between the winding and the shell of the motor, or between the coil and the silicon steel sheet under high vacuum conditions; if there is single-point damage or wire damage, discharge occurs, and the voltage will be rapidly consumed. By judging the time of voltage consumption, it can be judged whether the motor stator is qualified.

[0004] Although this solution realizes the preliminary excitation and response detection of the weak insulating point, it still has the following significant technical defects: (1) This method relies on the voltage consumption time as a judgment basis, and does not analyze the physical mechanism of the discharge process in depth. Since the voltage decay rate is affected by many factors, it is difficult to distinguish between small single-point damage and system noise or other non-defect energy consumption by relying on the voltage consumption time, resulting in insufficient detection sensitivity and criterion stability, and easy occurrence of missed detection or false positives.

[0005] (2) The generation of partial discharge signals is not completely derived from the insulation defects of the stator coil body. When the wiring components in the test system have poor contact, sharp tip protrusion or weak insulation, external corona or contact discharge may also be triggered, which produces similar electrical responses to real single-point damage. The existing technology lacks a retest verification mechanism for the discharge mode after detecting abnormal signals, and cannot effectively distinguish whether the abnormal signals are derived from real defects of the winding body or pseudo signals introduced by the test fixture or connection system, resulting in ambiguous fault troubleshooting direction and difficulty for maintenance personnel to locate the real failure source. SUMMARY

[0006] To this end, one purpose of the embodiments of the present application is to provide a kind of flat wire stator coil single-point damage electric detection method capable of fusing partial discharge multi-dimensional feature analysis, with discharge mode recognition and authenticity discrimination capability, to realize accurate, reliable and traceable determination of flat wire stator coil single-point damage.

[0007] The object of the present application can be achieved by the following technical solutions: a kind of flat wire stator coil single-point damage electric detection method, comprising the following steps: Step1: the flat wire stator coil to be tested is fixed on the detection tooling, and vacuum sealing treatment is carried out, then the sealed detection cavity is vacuumized, and whether the current test condition meets the requirements is determined based on the predetermined vacuum pressure maintenance standard.

[0008] Step2: after determining that the test condition meets the requirements, a predetermined test voltage is applied to the specified initial phase of the stator coil through the wiring component, and the partial discharge signal containing at least two characteristic parameters of discharge quantity and signal rise time is collected by the partial discharge sensor during the power test.

[0009] Step3: the collected partial discharge signal is in the single-point damage suspicious area defined in the discharge feature two-dimensional coordinate system, whether the stator coil has single-point damage suspicion is identified by multi-dimensional analysis.

[0010] Step4: after identifying single-point damage suspicion, the electrical connection of the wiring component is switched, the same test voltage is applied to the remaining phases of the stator coil in turn, and the partial discharge signal is collected.

[0011] Step5: the similarity of the partial discharge signals collected at the initial phase and the remaining phases in the discharge mode is compared and analyzed, and the authenticity of single-point damage is determined accordingly.

[0012] Step6: the final detection conclusion is output based on the authenticity discrimination result.

[0013] With all the above technical solutions, the positive effects of the present application are: 1. The present application synchronously collects partial discharge characteristic parameters such as discharge quantity and signal rise time during the process of applying test voltage to the flat wire stator coil to be tested, which directly reflects the physical characteristics of the discharge channel. Compared with the single criterion method relying only on voltage decay time, the present application realizes mechanism-level recognition of discharge mode through multi-parameter fusion, greatly reducing the problems of insufficient sensitivity and poor stability caused by single criterion in traditional methods, and significantly reducing missed detection and false positives.

[0014] 2、The application is based on the multi-parameter characteristics such as discharge capacity and signal rising time to identify the stator coil with single point damage suspicion, through applying the same test voltage to the remaining phase in turn and collecting the partial discharge signal, a cross-phase retest verification mechanism of discharge mode is constructed, the authenticity of the defect signal is identified, not only the accuracy and reliability of the detection result are improved, but also the fault tracing direction is clear, which provides traceable judgment basis for subsequent maintenance.

[0015] 3、The application performs vacuumizing operation on the sealed detection cavity before applying test voltage to the measured flat wire stator coil, and performs vacuum maintenance ability judgment after reaching the target vacuum degree, to verify the cavity airtightness and the stability of the environmental conditions, ensure that the test environment meets the requirements of high sensitivity partial discharge detection, only in the premise of qualified vacuum maintenance judgment and controlled test conditions, high voltage excitation is implemented, so as to ensure the stability of the detection process and the reliability of the signal, and improve the recognition accuracy of the tiny single point damage. BRIEF DESCRIPTION OF DRAWINGS

[0016] The application will be further described by using the drawings, but the embodiments in the drawings do not constitute any limitation on the application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of the following drawings.

[0017] Figure 1 The method embodiment steps of the application are shown in the figure.

[0018] Figure 2 The implementation flowchart of the application in the single point damage suspicion judgment area is shown in the figure.

[0019] Figure 3 The implementation schematic diagram of the application in the single point damage authenticity judgment based on the similarity of the partial discharge signals collected from the initial phase and the remaining phase in the discharge mode is shown in the figure.

[0020] Figure 4 The feature point distribution diagram of the effective suspicious discharge event falling into the single point damage suspicion judgment area in the application is shown in the figure. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the application will be described clearly and completely in the following combined with the drawings in the embodiments of the application, obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor belong to the protection scope of the application.

[0022] Reference Figure 1As shown, the present application proposes a kind of flat wire stator coil single-point damage electric detection method, comprising the following steps:Step1: the flat wire stator coil to be measured is fixed in detection tooling, and vacuum sealing treatment is carried out, then vacuumizing operation is carried out to sealed detection cavity, and whether the current test condition meets the requirement is judged based on predetermined vacuum pressure maintenance standard.

[0023] In the implementation process of the present application, the pre-test clamping of the flat wire stator coil is completed by an automatic system: a manipulator accurately carries and positions the stator coil to be tested to the reference structure of the tooling base; then, the automatic wire clamping mechanism is driven by a cylinder or a servo motor to perform an action, and the integrated conductive clamping jaw or probe accurately contacts the lead terminal of the stator coil to establish a reliable electrical connection; then, the vacuum cover structure is automatically closed by the actuator to seal the stator coil and the connecting parts in the detection cavity; the built-in vacuum sensor monitors the cavity pressure in real time, and when the vacuum degree is reached, the system determines that the environmental conditions meet the test requirements, controls the wiring component to apply a preset test voltage to the stator coil, and starts the partial discharge detection process.

[0024] It should be noted that the reason for using vacuum environment in the single-point damage detection of flat wire stator coil is that the discharge amount generated by single-point damage is extremely small, which is easily covered by environmental noise or air discharge under normal pressure. The vacuum environment reduces the background noise level, making it easier for the weak internal discharge signal to be captured by the partial discharge sensor, improving the signal-to-noise ratio and detection sensitivity.

[0025] If the vacuumization is not sufficient, the residual air pressure in the cavity will be too high, which will result in environmental discharge signals generated by air medium ionization with a much higher amplitude than the intrinsic partial discharge caused by small damage to the paint film, causing the weak defect signal to be overwhelmed by strong noise, resulting in a decrease in detection sensitivity and an increase in the risk of missed detection. Therefore, after the vacuumization reaches the target vacuum degree, the vacuum maintenance capability detection is still needed to verify the air tightness and pressure stability of the sealed cavity, to ensure that the test environment continuously meets the low background noise requirement, to ensure the authenticity and reliability of the partial discharge signal collection, and to avoid false positives or missed detection caused by environmental control.

[0026] As a preferred implementation of the above scheme, whether the current test condition meets the requirement is judged based on the predetermined vacuum pressure maintenance standard, as follows: after completing the vacuumizing operation of the detection cavity, a preset stable observation period is started, which can be set to 2 minutes for example, and the actual vacuum pressure value in the cavity is continuously collected within this period to form a pressure time sequence.

[0027] The difference between the maximum pressure and the minimum pressure in the pressure time sequence is calculated as the pressure fluctuation amplitude.

[0028] The pressure fluctuation amplitude is compared with the pressure fluctuation critical value in the test standard. If the fluctuation amplitude is less than or equal to the pressure fluctuation critical value, it is determined that the test conditions meet the requirements, otherwise, it is determined that the test conditions do not meet the requirements, and the subsequent process is terminated.

[0029] The pressure fluctuation critical value can be set according to the detection sensitivity requirement, and is usually not greater than ± 50 Pa. The basis for setting the pressure fluctuation critical value according to the detection sensitivity requirement is that when the pressure fluctuation in the vacuum cavity exceeds a certain range, the residual gas density changes accordingly, causing the gas ionization field strength to fluctuate, and then causing measurable false partial discharge signals. This phenomenon constitutes the environmental induced noise boundary of partial discharge detection. In order to ensure that such small defects are not missed, the industry standard and the technical specification of the host manufacturer generally set the sensitivity requirement of the partial discharge detection system to be less than or equal to 5 pC.

[0030] Under the sensitivity requirement of partial discharge detection, the specific method for obtaining the pressure fluctuation critical value is as follows: under the premise that the measured sub-coil does not apply mechanical stress, a controllable micro-leak source is introduced into the detection cavity, the cavity pressure is gradually increased, and after the pressure at each point is kept stable, a gradient voltage is applied until partial discharge is triggered. The corresponding discharge level is recorded, and the pressure fluctuation amplitude is monitored, and then the cavity pressure fluctuation amplitude corresponding to the discharge level reaching the sensitivity requirement of partial discharge detection is recorded. At this time, the gas ionization has been sufficient to produce a false signal that can be recognized by the system, indicating that the test environment has been inaccurate. Take 80% of the cavity pressure fluctuation amplitude in this case as the pressure fluctuation critical value to leave a proper safety margin, and ensure that the background interference is always below the detectable level under normal detection conditions.

[0031] The above-mentioned vacuum maintenance capability detection is based on the following principle: after the vacuum is extracted to the target pressure, if there is a small leak in the detection cavity, material outgassing or incomplete sealing interface, residual gas will continue to escape or external gas will slowly penetrate, causing the pressure in the cavity to change with time. Non-steady state, manifested as pressure drift or periodic fluctuation. By setting a preset stable observation period and continuously collecting pressure data to capture the pressure fluctuation amplitude. Larger fluctuation amplitude directly reflects the insufficient air tightness of the system or the imbalance of thermal-gas dynamics, indicating that the environmental conditions are uncontrollable. Based on this, by comparing the measured fluctuation amplitude with the preset critical threshold, it is determined whether the vacuum environment has entered a steady state, so as to ensure that the partial discharge test is carried out under low noise and high stability conditions, and to ensure the authenticity, repeatability and criterion reliability of the discharge signal acquisition.

[0032] Step 2: After determining that the test conditions meet the requirements, a preset test voltage is applied to the specified initial phase of the stator coil through the wiring member. During the power-on test, the partial discharge sensor is used to collect partial discharge signals, which at least include discharge quantity and signal rise time two types of characteristic parameters.

[0033] It should be noted that drive motors in new energy vehicles generally use three-phase AC motors. Their stator windings consist of three independent phase windings (U, V, and W) that are spatially separated by 120° electrical angles. Each phase winding forms an electrically isolated conductive path through its lead-out terminals. The aforementioned initial phase refers to the first independent winding channel to which the test voltage is applied during the testing process. Due to the electrical isolation between the phases, it is impossible to simultaneously apply excitation to all terminals during testing. Therefore, it is necessary to selectively connect a specific phase winding as the test channel using wiring components to achieve independent phase-by-phase excitation and signal acquisition.

[0034] By conducting independent tests on each phase in sequence, not only can the physical topology of the stator winding be matched, but the data foundation can also be provided for subsequent comparative analysis of multiphase discharge characteristics, and support the pattern identification of defect signals and external interference.

[0035] The initial phase can be chosen arbitrarily in principle, either U, V, or W phase; however, in automated testing systems, a preset sequence is usually followed, such as starting with phase U by default, to ensure the standardization of the testing process, consistency of operation, and traceability of the test data.

[0036] In an optional implementation of the above scheme, a preset test voltage is applied to a specified initial phase of the stator coil through the wiring component as follows: to obtain the rated operating voltage of the stator winding to be tested.

[0037] Understandably, the rated operating voltage is the nominal line voltage under normal operating conditions, determined based on the motor design parameters or application system specifications.

[0038] Based on the insulation class of the enameled wire used in the stator winding, refer to the reference table to find the level of additional electrical stress between turns corresponding to that insulation class.

[0039] Understandably, the level of additional electrical stress in the inter-turn insulation reflects the overvoltage component distributed between the winding turns due to transient voltage surges caused by the switching action of power devices during actual operation of the motor. This component is mainly determined by the inductance-capacitance distribution parameters of the winding and the rate of voltage rise. Because enameled wires of different insulation grades exhibit systematic differences in manufacturing processes, insulation layer thickness, dielectric strength, and heat resistance, the electric field distribution characteristics and withstand voltage capability of their corresponding inter-turn insulation structures under high-frequency voltage stress also show regular variations.

[0040] Therefore, for a flat wire winding of a specific insulation class, a database of turn-to-turn overvoltage amplitudes under typical operating conditions can be established through experimental calibration or electromagnetic transient simulation, and a standardized insulation class-additional electrical stress reference table can be formed. The reference table essentially reflects the response characteristics of the insulation material system and structural parameters to high-frequency voltage components, so that in the test voltage design stage, without modeling each motor individually, the equivalent additional electrical stress value can be quickly obtained according to the insulation class.

[0041] The product of the rated operating voltage and the additional electrical stress borne by the turn-to-turn insulation is calculated, and the alternating current effective value voltage of the addition result is taken as the applied test voltage.

[0042] An example expression applied to the above operation wherein represents the rated operating voltage, represents the rated voltage multiple, preferably 2, to take into account the operating electrical stress simulation and defect excitation capability of the insulation system, represents the additional electrical stress borne by the turn-to-turn insulation.

[0043] It should be noted that the motor insulation system bears two main types of electrical stress in actual operation, namely steady-state operating voltage and transient additional voltage. This method superimposes the equivalent alternating voltage component obtained by multiplying the rated operating voltage by a predetermined multiple, and the turn-to-turn additional electrical stress obtained based on the insulation class reference table, to form a comprehensive test voltage. This design aims to reproduce the composite electrical stress environment experienced by the motor under variable frequency drive operating conditions under static detection conditions, ensuring that the applied voltage is not lower than the maximum electric field strength that the weak point of the insulation may bear in actual service.

[0044] By increasing the test voltage to close to or exceeding the partial discharge inception voltage, the partial discharge activity at single-point defects such as paint film micropores and scratches can be effectively excited, so that characteristic signals can be captured by high-sensitivity sensors.

[0045] Further, the partial discharge signals collected in the energization test include discharge quantity and signal rise time, wherein the discharge quantity represents the equivalent charge quantity released by the partial discharge pulse, reflecting the energy release intensity at the defect, and the signal rise time represents the time from 10% to 90% amplitude of the discharge pulse, reflecting the steepness of the discharge source.

[0046] The above two parameters are core characteristic quantities describing the physical process of partial discharge, and can effectively depict the dynamic behavior of compact discharge sources induced by single-point insulation defects. Since they have clear physical meaning, good repeatability, and can be captured with high precision by discharge sensors, they also have strong robustness to environmental electromagnetic noise.

[0047] Step 3: In the single-point damage suspected area defined in the discharge characteristic two-dimensional coordinate system, it is determined whether the stator coil has single-point damage suspicion through multi-dimensional analysis.

[0048] As the above-mentioned step can be realized, the single-point damage suspected area is defined as follows: a two-dimensional characteristic coordinate system is created with discharge quantity as X axis and signal rise time as Y axis.

[0049] In the coordinate system, a rectangular closed area formed by discharge quantity threshold interval and signal rise time threshold interval is defined as a single-point damage suspected area.

[0050] The discharge quantity threshold interval and the signal rise time threshold interval in the above-mentioned are set according to the typical discharge characteristics of single-point insulation damage, combined with historical sample data and insulation material characteristics, wherein the discharge quantity threshold interval is composed of lower limit and lower limit , wherein is set according to the background electromagnetic noise level, which is used to exclude low-amplitude interference signals and ensure that the detection signal has sufficient signal-to-noise ratio, is determined according to the maximum releasable charge quantity of the paint film insulation layer in the micro-damage state, which is used to distinguish between micro-discharge and large-energy discharge before the adjacent main insulation breakdown, and to avoid misjudging serious defects as single-point damage.

[0051] The signal rise time threshold interval is also composed of lower limit and upper limit , wherein corresponds to the minimum rise time of the compact discharge source, which is used to identify the rapid discharge process caused by micro pinholes or scratches, and embodies the spatial limitation of single-point damage, is set according to the typical rise time of large-volume air gap or cavity discharge, which is used to exclude non-single-point discharge caused by poor winding impregnation or interlayer air bubbles, and improve the identification specificity of the target defect.

[0052] The application of the above-mentioned operation is explained as follows: discharge quantity and signal rise time are used as two independent and complementary physical parameters to construct a two-dimensional characteristic plane coordinate system. In this coordinate system, the closed area surrounded by the threshold interval is essentially the distribution clustering domain of typical single-point defects in the characteristic space, which provides clear geometric criteria for subsequent discharge mode recognition and suspected event screening, realizes the rapid classification and labeling of discharge pulse events falling into the area, and supports the quantitative statistics of effective suspected discharge events.

[0053] Referring to Figure 2As shown, as the further implementation of the above step, whether the stator coil has single-point damage suspicion is identified by multi-dimensional analysis as follows: in the process of applying a preset test voltage to the stator coil, the discharge pulse events are continuously collected by local discharge detection, if no effective discharge signal is detected in the whole test period, the subsequent judgment process is terminated; if at least one discharge pulse event is detected, the discharge quantity and signal rise time corresponding to each event are further extracted.

[0054] The discharge quantity and signal rise time of each discharge pulse event are mapped into the two-dimensional feature coordinate system constructed above as parameters to form a feature point, if the feature point falls within the single-point damage suspicion judgment area, it is determined that the discharge event conforms to the single-point damage discharge mode, and is recorded as an effective suspicious discharge event.

[0055] The proportion of the effective suspicious discharge event in the total number of collected discharge events is calculated and recorded as the effective suspicious proportion, the time interval sequence between adjacent effective suspicious discharge events is calculated, and the extreme value parameter is extracted, and the discharge interval fluctuation degree is defined as the ratio of the maximum time interval to the minimum time interval in the sequence.

[0056] The greater the difference between the maximum time interval and the minimum time interval in the above, the more uneven the distribution of discharge events in the time sequence, the poorer the repeatability, and the more likely the generation mechanism to be affected by random factors rather than a fixed position stable defect. Therefore, as a quantitative indicator for measuring discharge period consistency, the greater the value of the discharge interval fluctuation degree, the more intense the time fluctuation, and the more unstable the discharge behavior; on the contrary, the fluctuation degree is close to 1, which means that the discharge has high periodicity and spatial fixity, which meets the physical characteristics of single-point insulation damage under alternating electric field.

[0057] Understandably, the effective suspicious proportion represents the concentration of the local discharge signal in the feature space. The discharge interval fluctuation degree is used to quantify the time stability and periodic consistency of the discharge behavior.

[0058] The effective suspicious proportion and the discharge interval fluctuation degree are compared with the preset judgment reference value respectively, if any of the following conditions is met, it is determined that the stator coil has single-point damage suspicion; a) the effective suspicious proportion reaches or exceeds the proportion judgment value.

[0059] For example, the proportion judgment value can be set to 0.6, i.e. 60%. This setting is based on the principle of dominant criterion in statistics, which means that when the proportion of discharge events falling into the single-point damage suspicion judgment area is not less than 60% of all discharge signals, it is considered that this type of discharge mode occupies a dominant position in the overall discharge behavior, which is significantly different from random sporadic interference, and meets the mode recognition logic of majority compliance, supporting the preliminary judgment of the existence of typical single-point defects.

[0060] Condition (a) focuses on the spatial aggregation of discharge mode, reflecting the dominant position of defect characteristics in the multi-parameter characteristic space.

[0061] b) The discharge interval fluctuation is less than or equal to the allowable fluctuation determination value.

[0062] For example, the allowable fluctuation determination value can be set to 1.5. This value corresponds to a ratio of the maximum to the minimum discharge time interval of no more than 1.5 times, indicating that the occurrence of discharge events has high periodic consistency. Real single-point breakdowns usually exhibit stable and repetitive local breakdown behavior under alternating voltage, and their discharge time distribution is relatively concentrated, so the discharge interval fluctuation is small. Setting this threshold can effectively distinguish between stable periodic discharges caused by fixed defects and non-steady-state random discharges caused by floating contamination, poor contact, etc., improving the selectivity and reliability of the criterion.

[0063] Condition (b) focuses on the time regularity of the discharge signal, identifying stable and high-frequency repetitive discharge behavior.

[0064] The present application realizes the spatial distribution attribution analysis of partial discharge signals by constructing a discharge quantity-signal rise time two-dimensional characteristic space, and introduces effective suspect proportion and discharge interval fluctuation as multi-dimensional criteria to form a dual-index collaborative determination mechanism combining spatial distribution characteristics and time sequence regularity.

[0065] This design is based on the physical mechanism that single-point insulation breakdown tends to produce spatial position fixed and discharge timing periodic under constant alternating electric field: such defects repeatedly break down near each voltage zero crossing or peak, exhibiting high repetition rate and low dispersion of discharge behavior. When any index reaches the preset benchmark, it indicates that the discharge mode has typical statistical characteristics of single-point breakdown, and it can be determined that there is an insulation defect suspect. This method significantly improves the discrimination sensitivity and anti-interference ability of detection through multi-dimensional feature fusion, reducing the risk of missed detection and false positives.

[0066] Step 4: Switch the electrical connection of the wiring component after identifying the single-point breakdown suspect, and sequentially apply the same test voltage to the remaining phases of the stator coil and collect partial discharge signals.

[0067] Specifically, the electrical connection of the wiring component is as follows: the wiring component is connected to the electrical output end of a multi-way switch.

[0068] The electrical input end of the multi-way switch is connected to the test power supply.

[0069] By controlling the conduction path of the multi-way switch, the test power supply is sequentially connected to the remaining phases of the stator coil.

[0070] Step5: Comparing the similarity of the partial discharge signals collected in the initial phase and the rest of the phases in the discharge mode, and making a single-point damage authenticity judgment accordingly.

[0071] In a specific embodiment of the above scheme, comparing the similarity of the partial discharge signals collected in the initial phase and the rest of the phases in the discharge mode includes the following: mapping the effective suspected discharge events identified in the collected partial discharge signals of each phase to a unified two-dimensional feature coordinate system.

[0072] Taking the discharge feature point distribution of the initial phase as a benchmark template, the feature point distribution of each of the remaining phases in the single-point damage suspected judgment area is compared with the distribution overlap in the feature space calculated by the histogram intersection.

[0073] Compare the distribution overlap of all the remaining phases with the initial phase, and extract the minimum distribution overlap.

[0074] It can be understood that the feature point distribution overlap of the other phases and the initial phase in the single-point damage suspected judgment area reflects the geometric similarity of the two in the feature space, and taking the minimum distribution overlap is the case where the discharge mode of the remaining phase is most inconsistent with the initial phase, representing the degree of pattern matching under the most unfavorable conditions.

[0075] Calculate the effective suspect proportion, discharge interval fluctuation, and relative deviation of the initial phase of each phase, and then take the maximum value of all relative deviations as the maximum suspect proportion deviation and the maximum discharge interval fluctuation deviation.

[0076] It can be understood that the effective suspect proportion, discharge interval fluctuation, and relative deviation of the other phases and the initial phase reflect the consistency of the remaining phases with the initial phase in the two dimensions of spatial dominance and temporal stability of the discharge mode, and taking the maximum value of all relative deviations aims to identify the most significant difference phase, which represents the maximum deviation under the most unfavorable conditions.

[0077] The minimum distribution overlap, the maximum suspect proportion deviation, and the maximum discharge interval fluctuation deviation are all used as indicators of the similarity of the partial discharge signals collected in the initial phase and the rest of the phases in the discharge mode.

[0078] Referring to Figure 3 In further specific embodiments, the single-point damage authenticity judgment is implemented as follows: comparing the minimum distribution overlap of the initial phase and the rest of the phases in the discharge mode , the maximum suspect proportion deviation , and the maximum discharge interval fluctuation deviation with the similarity judgment limit value, respectively.

[0079] The detected discharge signal mode is determined to be highly similar in all phases, originating from interference of the wiring member, not a real single-point breakage, when all the following conditions are met, otherwise it is determined that the stator coil body has a real single-point breakage.

[0080] wherein , , respectively represent the distribution overlap limit value, the suspected proportion deviation limit value, and the discharge interval fluctuation deviation limit value.

[0081] The distribution overlap limit value can be determined by statistically analyzing the minimum overlap distribution of normal interference signals from a large number of known good samples and typical defect samples in multiple phase tests, and taking the lower quartile or empirical stable value as the determination reference.

[0082] The suspected proportion deviation limit value can be determined according to the measurement dispersion of healthy windings in different phases, to ensure that normal fluctuations do not trigger false positives.

[0083] The discharge interval fluctuation deviation limit value can be set in combination with the time jitter level of stable discharge pulses, to exclude the influence of system clock errors and acquisition delays.

[0084] Based on the above-mentioned operation, the real single-point insulation breakage is located at a fixed position of a specific physical phase winding, and its discharge behavior is only excited when the phase is applied with a test voltage, having clear phase dependence and spatial non-migration. Once switched to other unpowered phases, the defect cannot be excited, so the discharge signal with the same characteristics should not be reproduced. On the contrary, when the wiring member has poor contact, a sharp tip, or weak insulation, it is usually caused by the common mode link of the test system, and will be shown in multiple phase tests, and its discharge mode is highly consistent, showing significant reproducibility and cross-phase consistency. Based on this physical difference, after identifying the single-point breakage suspect, the same test voltage is applied to the remaining phases of the stator coil in sequence and the partial discharge signals are collected, and cross-phase discharge mode similarity analysis is performed to realize retest verification of the authenticity of the defect signal.

[0085] The analysis covers the key characteristics of the discharge signal in spatial distribution, mode intensity, and time dynamics from three orthogonal dimensions, ensuring comprehensive criteria. Since the remaining phases are usually two phases, two values are generated for each deviation index. To improve the robustness of the criteria and focus on the most suspicious scenarios, the method extracts the similarity indicators under each dimension as the most unfavorable condition. If all indicators still meet the pre-set similarity threshold in this most unfavorable condition, it is determined that the discharge behavior of each phase is highly consistent, and the discharge signal is not a real defect; otherwise, it supports the existence of a real single-point breakage in the initial phase. Moreover, this extraction strategy not only enhances the recognition sensitivity of real defects, but also only needs to compare a limited number of parameters, significantly reducing the algorithm complexity.

[0086] Step 6: output the final detection conclusion based on the authenticity discrimination result.

[0087] Specific detection conclusions include: for the case of judging as a real single-point damage, identifying the specific phase where the damage occurs in the detection report, and attaching the feature point distribution diagram of the effective suspected discharge event in the two-dimensional feature coordinate system, which falls within the single-point damage suspected judgment area, to visually present the spatial aggregation of the discharge mode as traceable evidence for defect judgment.

[0088] When it is determined that there is no real single-point damage, an inspection alarm for the connecting member is triggered, realizing the forward positioning of the fault source and the self-diagnosis of the test system.

[0089] As an example of the specific implementation of the present application, when it is determined that there is a real single-point damage in phase U of the stator coil, the detection report will output the following content: (1) detection conclusion: phase U confirms the existence of single-point damage.

[0090] (2) Feature point distribution diagram (as shown in the attached Figure 4 ): in the two-dimensional coordinate system with discharge quantity as X-axis and signal rise time as Y-axis, 7 feature points are shown to fall within the single-point damage suspected judgment area (i.e. the rectangular area defined by the discharge quantity threshold interval 300-400 pC and the signal rise time threshold interval 50-75 ns). These points are concentrated, indicating that the discharge mode is highly consistent, supporting the single-point damage judgment.

[0091] The above embodiments can be realized in whole or in part by software, hardware, firmware or any other combination. When realized by software, the above embodiments can be realized in whole or in part in the form of a computer program product.

[0092] Those of ordinary skill in the art can realize that the modules and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solutions. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0093] In addition, the functional modules in each embodiment of the present application can be integrated in one processing module, or each module can exist physically alone, or two or more modules can be integrated in one module.

[0094] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0095] Finally, the above merely provides the preferred embodiments of the present application, but is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for detecting a single-point breakage of a flat wire stator coil by electrical inspection, characterized by, The method comprises the following steps: Step 1: fix the flat wire stator coil to be tested on a detection tool, and perform vacuum sealing treatment on the flat wire stator coil to be tested, and then perform vacuumizing operation on the sealed detection cavity, and determine whether the current test condition meets the requirements based on a predetermined vacuum pressure maintenance standard; Step 2: after determining that the test condition meets the requirements, apply a predetermined test voltage to the specified initial phase of the stator coil through the wiring member, and collect the partial discharge signal containing at least the discharge quantity and the signal rise time during the power test process by using the partial discharge sensor; Step 3: if the collected partial discharge signal is in a single-point damage suspicion judgment area defined in a two-dimensional discharge characteristic coordinate system, identify whether the stator coil has single-point damage suspicion through multi-dimensional analysis; Step 4: after identifying the single-point damage suspicion, switch the electrical connection of the wiring member, apply the same test voltage to the remaining phases of the stator coil in turn, and collect the partial discharge signal; Step 5: compare and analyze the similarity of the partial discharge signals collected from the initial phase and the remaining phases in the discharge mode, and determine the authenticity of the single-point damage; Step 6: output the final detection conclusion based on the authenticity determination result.

2. A method for detecting a single point breakage of a flat wire stator coil by electrical inspection, according to claim 1, characterized in that: The determination of whether the current test condition meets the requirements based on the predetermined vacuum pressure maintenance standard is as follows: After completing the vacuumizing operation of the detection cavity, start a preset stable observation period, and continuously collect the actual vacuum pressure value in the cavity within the period to form a pressure time sequence; Calculate the difference between the maximum pressure and the minimum pressure from the pressure time sequence as the pressure fluctuation amplitude; Compare the pressure fluctuation amplitude with the pressure fluctuation critical value in the test standard. If the fluctuation amplitude is less than or equal to the pressure fluctuation critical value, it is determined that the test condition meets the requirements, otherwise it is determined that the test condition does not meet the requirements, and the subsequent process is terminated.

3. A method for detecting a single point breakage of a flat wire stator coil by electrical method according to claim 1, characterized in that: The preset test voltage is set as follows: Obtain the rated working voltage of the measured stator winding; According to the insulation grade of the enameled wire used in the stator winding, and combined with the reference table, the turn-to-turn insulation additional electric stress level corresponding to the insulation grade is obtained; Determine the test voltage to be applied based on the rated working voltage and the turn-to-turn insulation additional electric stress level.

4. A method for detecting a single point failure of a flat wire stator coil according to claim 1, characterized in that: The single-point damage suspicion judgment area is defined as follows: Create a two-dimensional characteristic coordinate system with discharge quantity as X-axis and signal rise time as Y-axis; In the coordinate system, a rectangular closed area formed by the discharge quantity threshold interval and the signal rise time threshold interval is defined as the single-point damage suspicion judgment area.

5. A method for detecting a single point failure of a flat wire stator coil according to claim 4, characterized in that: The multi-dimensional analysis to identify whether the stator coil has single-point damage suspicion is as follows: During the process of applying the preset test voltage to the stator coil, continuously collect the discharge pulse events through the partial discharge detection. If no effective discharge signal is detected during the entire test period, the subsequent determination process is terminated; If at least one discharge pulse event is detected, the discharge quantity and signal rise time corresponding to each event are further extracted; The discharge amount and signal rising time of each discharge pulse event are mapped to a two-dimensional characteristic coordinate system to form a characteristic point, and if the characteristic point falls within a single-point damage suspected determination area, it is determined that the discharge event conforms to a single-point damage discharge mode, and is recorded as an effective suspected discharge event; The proportion of the effective suspected discharge events in the total number of collected discharge events is calculated and recorded as the effective suspected proportion, and the time interval sequence between adjacent effective suspected discharge events is calculated, and the extreme value parameters are extracted, and the discharge interval fluctuation degree is defined as the ratio of the maximum time interval to the minimum time interval in the sequence; The effective suspected proportion and the discharge interval fluctuation degree are compared with the preset determination reference value, and if any of the following conditions is met, it is determined that the stator coil has a single-point damage suspicion; a) The effective suspected proportion reaches or exceeds the proportion determination value; b) The discharge interval fluctuation degree is less than or equal to the allowable fluctuation determination value.

6. A method for detecting a single point failure of a flat wire stator coil according to claim 1, characterized in that: The electrical connection of the switching connection component is as follows: The connection component is connected to the electrical output end of a multi-way switching switch; The electrical input end of the multi-way switching switch is connected to the test power supply; By controlling the conduction path of the multi-way switching switch, the test power supply is connected to the remaining phases of the stator coil in turn.

7. A method for detecting a single point failure of a flat wire stator coil according to claim 5, characterized in that: The comparison and analysis of the similarity of the partial discharge signals collected from the initial phase and the remaining phases in the discharge mode include the following contents: Map the effective suspected discharge events identified in the collected partial discharge signals of each phase to a unified two-dimensional characteristic coordinate system; Take the discharge characteristic point distribution of the initial phase as a reference template, and compare the characteristic point distribution of each of the remaining phases within the single-point damage suspected determination area with the reference template to calculate the distribution overlap degree of the two in the characteristic space through histogram intersection; Compare the distribution overlap degrees of all the remaining phases with the initial phase, and extract the minimum distribution overlap degree; Calculate the relative deviations of the effective suspected proportion and the discharge interval fluctuation degree of each phase from the initial phase, and then take the maximum value of all the relative deviations as the maximum suspected proportion deviation and the maximum discharge interval fluctuation deviation; The minimum distribution overlap degree, the maximum suspected proportion deviation, and the maximum discharge interval fluctuation deviation are all used as indicators of the similarity of the partial discharge signals collected from the initial phase and the remaining phases in the discharge mode.

8. A method for detecting a single point failure of a flat wire stator coil according to claim 7, characterized in that: The single-point damage true-false determination is as follows: the minimum distribution overlap of the initial phase with the partial discharge signals of the remaining phases in the discharge pattern , maximum suspect proportion deviation , maximum discharge interval fluctuation deviation respectively compared with the similarity determination limit value When all the following conditions are met, it is determined that the discharge signal pattern detected at this time is highly similar in all phases, is interference from the connection component, and is not a real single-point damage, otherwise it is determined that the stator coil body has a real single-point damage. wherein , , respectively represent the distribution overlap limit value, the suspicion proportion deviation limit value, and the discharge interval fluctuation deviation limit value.

9. A method for detecting a single point failure of a flat stator coil according to claim 5, wherein: The final detection conclusion based on the true-false discrimination result is as follows: For the case of determining a real single-point damage, the specific phase where the damage occurs is identified in the detection report, and the characteristic point distribution graph of the effective suspected discharge events of the phase falling within the single-point damage suspected determination area in the two-dimensional characteristic coordinate system is attached.

10. A method for detecting a single point failure of a flat wire stator coil according to claim 1, wherein: The final detection conclusion based on the true-false discrimination result also includes the following: When it is determined that there is no real single-point damage, an alarm is triggered for the inspection of the connection component.

Images