A motor shell integrity monitoring method and system based on impedance characteristic analysis

Abstract

The application relates to the technical field of fault prediction and health management, and discloses a motor shell integrity monitoring method and system based on impedance characteristic analysis, wherein the method comprises the following steps: acquiring an impedance response signal of a motor shell and a synchronous environment signal, collecting impedance and environment synchronous data sets through frequency self-adaption residence, checking impedance consistency and correcting environment normalization of the data sets to obtain a corrected complex impedance spectrum, performing partition self-adaption Tikhonov regularization inversion on the corrected complex impedance spectrum to obtain state relaxation time distribution data, extracting a net corrosion characteristic parameter set through differential spectrum decoupling of physical process constraints, performing corrosion quantization correlation and integrity grading determination according to the parameter set to obtain a motor shell integrity monitoring result. The corresponding system is used for executing the above method. The application can identify hidden corrosion under the interference of coating water absorption response, and output integrity state grades and early warning grades.

Description

Technical Field

[0001] This invention relates to the field of fault prediction and health management technology, and in particular to a method and system for monitoring the integrity of motor housing based on impedance characteristic analysis. Background Technology

[0002] Currently, motor housings operating in harsh environments such as coastal salt spray and high humidity often employ multi-layer organic anti-corrosion coatings for protection. However, filamentary or crevice corrosion can still occur beneath the coating, and such hidden substrate damage is difficult to detect with visual inspection and eddy current testing. Electrochemical impedance spectroscopy (EIS), as a non-destructive testing method, obtains complex impedance information at the coating-metal interface by applying broadband excitation and uses modulus, specific frequency phase angles, or equivalent circuit models to determine the coating degradation and corrosion state, making it an important tool for coating health assessment. To achieve separation of multiple time constants, the relaxation time distribution method does not require a pre-set equivalent circuit and can reflect the relaxation characteristics of different physical processes. Typically, Tikhonov regularization is used to inversely derive the relaxation time distribution function from the impedance spectrum, and the peak positions and intensities indicate processes such as coating capacitance, coating water absorption, and substrate corrosion.

[0003] For example, when the coating on a motor housing is macroscopically intact but filamentous corrosion occurs underneath, the coating capacitance peak and the dielectric relaxation peak caused by coating water absorption still dominate in the impedance spectrum. The modulus only decreases slightly in the low-frequency range, and the phase angle change at a single frequency point is easily confused with coating moisture absorption and fluctuations in ambient temperature and humidity. Traditional impedance analysis relies on fixed thresholds or simple equivalent circuit fitting, which cannot effectively decouple the coating dielectric relaxation from the Faraday corrosion process of the metal substrate. This causes the relaxation behaviors of coating water absorption and corrosion to overlap in the impedance spectrum. Corrosion characteristic peaks hidden in the low-frequency region are incorrectly attributed to coating moisture absorption peaks or suppressed by regularization smoothing operations, resulting in missed or misjudged early hidden corrosion.

[0004] Therefore, there is an urgent need for a method for monitoring the integrity of motor housings that can, under the condition that the coating surface appears intact, invert the relaxation time distribution from a broadband impedance spectrum with high resolution and accurately decouple the relaxation peaks of the coating's moisture absorption and the substrate corrosion characteristics. This method should be able to fully protect the relaxation characteristics of low-frequency weak signals through an adaptive regularization mechanism under the interference of environmental temperature and humidity fluctuations and coating moisture absorption, and accurately remove the interference components generated by coating moisture absorption based on a physical model, extracting the net relaxation peak parameters that are only related to the metal substrate corrosion process. This would enable non-destructive online identification and early warning of hidden corrosion, significantly improving monitoring sensitivity and assessment reliability. Summary of the Invention

[0005] To address the aforementioned technical shortcomings, the present invention aims to propose a method for monitoring the integrity of motor housing based on impedance characteristic analysis. This method addresses the problem that existing technologies, which focus solely on impedance magnitude or a single phase angle, are unable to effectively distinguish between the relaxation behavior of hidden corrosion under the coating and simple water absorption by the coating, especially when the coating surface appears intact. This results in the difficulty of detecting hidden corrosion in its early stages.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a method for monitoring the integrity of motor housing based on impedance characteristic analysis.

[0007] The method for monitoring the integrity of a motor housing based on impedance characteristic analysis includes: Step S10: Acquire the motor housing impedance response signal and the synchronous environment signal, and use the frequency adaptive dwell mechanism to perform the wideband impedance acquisition task, and output the impedance and environment synchronization dataset; Step S20: Based on the impedance and environment synchronization dataset, an impedance consistency verification and environment normalization correction task is performed using a hybrid physical and empirical adaptive correction mechanism, and the corrected complex impedance spectrum is output. Step S30: Based on the corrected complex impedance spectrum, a partitioned adaptive Tikhonov regularization mechanism is used to perform the relaxation time distribution inversion preprocessing task, and output the state relaxation time distribution data; Step S40: Based on the state relaxation time distribution data, the net corrosion feature extraction task is performed using the differential spectrum decoupling mechanism constrained by the physical process, and the net corrosion feature parameter set is output. Step S50: Perform corrosion quantification correlation and integrity grading judgment tasks based on the net corrosion characteristic parameter set, and output the motor housing integrity monitoring results.

[0008] Preferably, step S10, which involves acquiring the motor housing impedance response signal and the synchronization environment signal, performing a wideband impedance acquisition task using a frequency adaptive dwell mechanism, and outputting an impedance and environment synchronization dataset, specifically includes: Step S101: Apply a multi-frequency sinusoidal excitation signal with a frequency range of 10mHz to 100kHz to the motor housing through a monitoring patch set on the motor housing, and collect the complex impedance response corresponding to each excitation frequency point to obtain multi-frequency complex impedance sampling data; wherein, the monitoring patch adopts a three-electrode system or a four-point electrode system, and the excitation frequency points are arranged in a logarithmic interval. Step S102: Calculate the signal-to-noise ratio of the multi-frequency complex impedance sampling data at each excitation frequency point. When the signal-to-noise ratio at the current excitation frequency point is lower than the preset signal-to-noise ratio threshold, extend the excitation dwell time at the current excitation frequency point to complete the integral sampling of an integer number of excitation cycles and obtain adaptive dwell complex impedance data. Step S103: Synchronously collect ambient temperature data and relative humidity data output by the temperature and humidity sensor set on the monitoring patch, and associate the ambient temperature data, the relative humidity data and the adaptive dwelling complex impedance data according to a unified sampling time scale to obtain the impedance and environment synchronous dataset.

[0009] Preferably, step S20, based on the impedance and environment synchronization dataset, employs a hybrid physical and empirical adaptive correction mechanism to perform impedance consistency verification and environmental normalization correction tasks, and outputs the corrected complex impedance spectrum. This step specifically includes: Step S201: Extract the adaptive resident complex impedance data from the impedance and environment synchronization dataset, perform Kramers and Kronig consistency checks on the adaptive resident complex impedance data, calculate the consistency check residuals corresponding to each excitation frequency point, and remove abnormal data points accounting for 3% of all excitation frequency points in descending order of the consistency check residuals to obtain the consistency verification complex impedance data. Step S202: Determine the missing frequency points in the consistency verification complex impedance data, and use a regularized spline interpolation algorithm with embedded resistance and capacitance response physical priors and diffusion wake physical priors to perform complex impedance interpolation on the missing frequency points to obtain complete complex impedance data. Step S203: Extract the ambient temperature data and the relative humidity data from the impedance and environment synchronization dataset, and use the pre-calibrated temperature correction coefficient and temperature-humidity coupling correction coefficient to correct the complete complex impedance data to the impedance state corresponding to the preset reference ambient temperature and preset reference relative humidity, thereby obtaining the corrected complex impedance spectrum.

[0010] Preferably, in step S202, the use of a regularized spline interpolation algorithm with embedded physical priors of impedance and capacitance responses and physical priors of diffusion wakes to perform complex impedance interpolation on the missing frequency points specifically includes: The physical priors characterizing the high-frequency RC response of the motor housing coating and the low-frequency diffusion wake are embedded in the spline interpolation kernel function. The interpolation curve constrained by the complex impedance response path is constructed using the effective complex impedance data points located on both sides of the missing frequency point. The Kramers and Kronig consistency test is performed on the interpolation curve to ensure that the interpolated complex impedance data simultaneously satisfies the complex impedance spectrum continuity constraint and the Kramers and Kronig relationship constraint.

[0011] Preferably, step S30, which involves performing a relaxation time distribution inversion preprocessing task based on the corrected complex impedance spectrum using a partitioned adaptive Tikhonov regularization mechanism to output state relaxation time distribution data, specifically includes: Step S301: Input the corrected complex impedance spectrum into the first type of Fredholm integral relation, discretize the relaxation time axis, and establish the inversion relationship between the complex impedance response and the relaxation time distribution based on the discretized relaxation time axis to obtain the relaxation time distribution inversion model. Step S302: Based on the response positions of different physical processes on the relaxation time axis, the relaxation time axis is divided into short relaxation time intervals, medium relaxation time intervals, and long relaxation time intervals. A first preset regularization parameter is configured for the short relaxation time interval and a second-order difference smoothing constraint is applied. A second preset regularization parameter is configured for the medium relaxation time interval and a second-order difference smoothing constraint is applied. A third preset regularization parameter is configured for the long relaxation time interval and a first-order difference smoothing constraint is applied, thus obtaining the partitioned regularization objective function. Wherein, the first preset regularization parameter is greater than the second preset regularization parameter, and the second preset regularization parameter is greater than the third preset regularization parameter. Step S303: Solve the partition regularization objective function under the non-negative constraint of the relaxation time distribution to obtain the current state relaxation time distribution corresponding to the current monitoring state, and obtain the baseline state relaxation time distribution pre-constructed in the initial complete state of the motor housing. Combine the current state relaxation time distribution and the baseline state relaxation time distribution into the state relaxation time distribution data.

[0012] Preferably, step S40, which involves using the differential spectrum decoupling mechanism of physical process constraints to perform the net corrosion feature extraction task based on the state relaxation time distribution data and outputting the net corrosion feature parameter set, specifically includes: Step S401: Extract the current state relaxation time distribution and the baseline state relaxation time distribution from the state relaxation time distribution data. Perform multi-peak fitting using a mixed function composed of Gaussian function and Lorentz function. According to the preset physical process mapping relationship, associate the characteristic peaks in the short relaxation time interval with the coating capacitance response, associate the characteristic peaks in the medium relaxation time interval with the coating water absorption response and the coating-substrate interface charge transfer response, and associate the characteristic peaks in the long relaxation time interval with the substrate corrosion response to obtain the relaxation characteristic peak parameter group. Step S402: Based on the relaxation characteristic peak parameter group, calculate the difference in characteristic peak area between the current state relaxation time distribution and the baseline state relaxation time distribution in the middle relaxation time interval, determine the difference in characteristic peak area as the coating water absorption change, and input the coating water absorption change into the pre-established coating water absorption and micro battery driving potential evolution matrix to obtain the low-frequency interference relaxation time distribution caused by coating water absorption and interface state migration; Step S403: Calculate the net corrosion difference spectrum based on the current state relaxation time distribution, the baseline state relaxation time distribution, and the low-frequency interference relaxation time distribution.

[0013] in, relaxation time Net corrosion differential spectral value at the location, For relaxation time, The relaxation time of the current state is distributed over the relaxation time. Distribution values ​​at that location, The relaxation time of the baseline state is distributed over the relaxation time. Distribution values ​​at that location, The relaxation time distribution of the low-frequency interference is within the relaxation time. Distribution values ​​at; Within a preset corrosion reaction relaxation time interval, a characteristic peak search is performed on the net corrosion differential spectrum to extract the corrosion characteristic peak time constant and corrosion characteristic peak intensity. The corrosion characteristic peak time constant, corrosion characteristic peak intensity and coating water absorption change are then combined to form the net corrosion characteristic parameter set.

[0014] Preferably, step S50, which involves performing corrosion quantification correlation and integrity grading determination based on the net corrosion characteristic parameter set and outputting the motor housing integrity monitoring results, specifically includes: Step S501: Extract the corrosion characteristic peak time constant, the corrosion characteristic peak intensity, and the coating water absorption change from the net corrosion characteristic parameter set to obtain the corrosion judgment parameter set; Step S502: Multiply the change in water absorption of the coating by a pre-calibrated low-frequency baseline drift correction coefficient, and add the product to a pre-calibrated base peak intensity threshold to obtain a dynamic corrosion peak intensity threshold; when the corrosion characteristic peak time constant is within the preset corrosion reaction relaxation time interval, and the corrosion characteristic peak intensity is not less than the dynamic corrosion peak intensity threshold, it is determined that there is hidden corrosion on the motor housing; otherwise, it is determined that there is no hidden corrosion on the motor housing, and a corrosion determination result is obtained. Step S503: Based on the corrosion determination result, when it is determined that there is hidden corrosion on the motor housing, the pre-calibrated mapping relationship between the intensity of corrosion characteristic peaks and the effective active area of ​​corrosion is invoked to determine the effective active area of ​​corrosion. The effective active area of ​​corrosion is then compared with a preset corrosion area grading threshold to determine the integrity status level of the motor housing and the corresponding warning level. When it is determined that there is no hidden corrosion on the motor housing, the motor housing is determined to be in an intact state, and the integrity monitoring result of the motor housing, which includes the integrity status level and the warning level, is output.

[0015] The present invention also provides a motor housing integrity monitoring system based on impedance characteristic analysis, comprising: The wideband impedance acquisition module is used to acquire the motor housing impedance response signal and the synchronous environment signal. It uses a frequency adaptive dwell mechanism to perform the wideband impedance acquisition task and outputs an impedance and environment synchronization dataset. The impedance correction module is used to perform impedance consistency verification and environmental normalization correction tasks based on the impedance and environment synchronization dataset, using a hybrid physical and empirical adaptive correction mechanism, and outputs a corrected complex impedance spectrum. The relaxation time distribution inversion module is used to perform the relaxation time distribution inversion preprocessing task based on the corrected complex impedance spectrum and adopts a partitioned adaptive Tikhonov regularization mechanism to output state relaxation time distribution data. The net corrosion feature extraction module is used to perform the net corrosion feature extraction task based on the state relaxation time distribution data and using the differential spectrum decoupling mechanism constrained by physical process, and output the net corrosion feature parameter set. The integrity assessment module is used to perform corrosion quantification correlation and integrity grading judgment tasks based on the net corrosion characteristic parameter set, and output the integrity monitoring results of the motor housing.

[0016] The present invention also provides a motor housing integrity monitoring device based on impedance characteristic analysis. The motor housing integrity monitoring device based on impedance characteristic analysis includes: a memory, a processor, and a motor housing integrity monitoring program based on impedance characteristic analysis stored in the memory and executable on the processor. When the motor housing integrity monitoring program based on impedance characteristic analysis is executed by the processor, it implements the above-described method.

[0017] The present invention also provides a computer program product, which includes a motor housing integrity monitoring program based on impedance characteristic analysis, wherein the motor housing integrity monitoring program based on impedance characteristic analysis implements the above method when executed by a processor.

[0018] The beneficial effects of this invention are as follows: This invention improves the acquisition quality at low signal-to-noise ratio frequency points through a frequency adaptive dwell mechanism, and reduces the impact of noise, abnormal data and environmental changes on the complex impedance spectrum by combining impedance consistency verification and temperature and humidity normalization correction, so that data from different monitoring periods have a consistent comparison benchmark.

[0019] This invention extracts net corrosion characteristic parameters corresponding to the substrate corrosion process by partitioning adaptive Tikhonov regularization to invert the state relaxation time distribution and using a differential spectrum decoupling method constrained by physical processes to deduct low-frequency interference caused by coating water absorption and interface state migration, thereby improving the accuracy of hidden corrosion identification.

[0020] This invention quantifies and correlates corrosion and classifies integrity based on net corrosion characteristic parameters, and dynamically adjusts the corrosion judgment threshold in conjunction with changes in coating water absorption. This reduces misjudgments caused by coating water absorption and outputs integrity status level and warning level, providing a basis for the inspection and maintenance of motor housing. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the first embodiment of a motor housing integrity monitoring method based on impedance characteristic analysis according to the present invention. Detailed Implementation

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0023] 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.

[0024] Example 1: As Figure 1 The diagram shown is a flowchart of the first embodiment of a motor housing integrity monitoring method based on impedance characteristic analysis according to the present invention. The first embodiment of the motor housing integrity monitoring method based on impedance characteristic analysis according to the present invention is presented.

[0025] In the first embodiment, the method for monitoring the integrity of a motor housing based on impedance characteristic analysis includes: Step S10: Acquire the motor housing impedance response signal and the synchronous environment signal, and use the frequency adaptive dwell mechanism to perform the wideband impedance acquisition task, and output the impedance and environment synchronization dataset; The impedance and environment synchronization dataset refers to adaptive dwell-time complex impedance data, ambient temperature data, and relative humidity data associated with a unified sampling timescale. A monitoring patch is installed on the casing of the motor under test, employing a three-electrode or four-point electrode system. Excitation frequency points are set at logarithmic intervals within the range of 10 mHz to 100 kHz, and multi-frequency sinusoidal excitation is applied at each frequency point. The acquisition device calculates the signal-to-noise ratio (SNR) of the complex impedance response at the current frequency point. When this SNR is lower than a preset SNR threshold, the excitation dwell time at the current frequency point is extended until the sampling window covers an integer number of excitation cycles, and then adaptive dwell-time complex impedance data for that frequency point is generated. Temperature and humidity sensors on the patch record ambient temperature and relative humidity at the same sampling timescale, ensuring that each impedance response has traceable environmental conditions, and the obtained data are collectively output to step S20.

[0026] Frequency adaptive dwell time allows the sampling duration to vary with the actual signal-to-noise ratio (SNR) at each frequency point, rather than applying a uniform delay to the entire frequency sweep process. This compensates for the effective integration period at low frequencies, and the effects of contact disturbances, electromagnetic noise, and slow polarization drift on the real and imaginary parts of the complex impedance can be constrained during period integration. Frequency points where the SNR already meets requirements retain their original dwell time, avoiding unnecessary increases in the duration of a complete frequency sweep. After the environmental signal and impedance response are correlated using a unified timescale, step S20 can perform consistency checks and environmental normalization for the same sampling time, preventing the incorrect use of temperature and humidity conditions from different times for impedance correction.

[0027] Traditional fixed-dwell acquisition typically processes all frequency points according to a pre-set number of cycles. At low frequencies, if sufficient stable response cycles are not experienced, the obtained impedance values ​​may fluctuate significantly with noise and polarization drift. Increasing the dwell time invariably would significantly lengthen the broadband scan. This step uses the signal-to-noise ratio of the current frequency point as the trigger condition for extending the dwell time. While ensuring integer-cycle integration, it preserves the original acquisition rhythm of other frequency points. The obtained data covers both the high-frequency RC response of the coating and the low-frequency response related to interface charge transfer and substrate corrosion, providing a common input for subsequent correction and relaxation time distribution inversion.

[0028] For example, during the monitoring of motor casings in a high-humidity coastal environment, the system sequentially acquires complex impedance responses from high frequency to low frequency. When the signal-to-noise ratio (SNR) at a certain low-frequency point falls below a preset threshold due to electromagnetic interference, the acquisition controller extends the dwell time at that point until the integration window covers an integer number of cycles and the SNR meets the requirements before moving on to the next frequency point; the temperature and humidity sensors simultaneously record environmental changes during this period. The final data record includes the frequency point, complex impedance response, dwell information, ambient temperature, relative humidity, and a unified time identifier. Step S20 can use this data to identify abnormal frequency points and perform normalization correction according to the same environmental time.

[0029] Step S20: Based on the impedance and environment synchronization dataset, an impedance consistency verification and environment normalization correction task is performed using a hybrid physical and empirical adaptive correction mechanism, and the corrected complex impedance spectrum is output. The corrected complex impedance spectrum refers to the complete complex impedance data after consistency verification, missing frequency point completion, and reference environment normalization. During processing, adaptive resident complex impedance data is first extracted from the impedance and environment synchronization dataset. The Kramers and Kronig consistency tests are used to obtain the test residuals for each excitation frequency point, and the residuals are sorted from largest to smallest. Outliers accounting for 3% of all excitation frequency points are removed. For the missing frequency points after removal, a regularized spline interpolation curve constrained by the complex impedance response path is established using the effective complex impedance points on both sides. The interpolation kernel simultaneously embeds the high-frequency RC response prior of the coating and the low-frequency diffusion wake prior. The interpolation result is again verified by Kramers and Kronig consistency. Subsequently, the ambient temperature and relative humidity at the same time scale are read. Based on the pre-calibrated temperature correction coefficient and temperature-humidity coupling correction coefficient, the complete complex impedance data is converted to the impedance state corresponding to the preset reference temperature and reference relative humidity, and the corrected complex impedance spectrum is output to step S30.

[0030] Outliers are eliminated by sorting residuals to remove a fixed proportion, providing a unified basis for excluding non-physical responses such as poor contact and transient interference. Spline interpolation, constrained by physical priors, recovers missing positions using adjacent valid points, while limiting the interpolation curve's deviation from the reasonable trajectory of the coating's resistive-capacitive response and diffusion wake. Environmental normalization uses temperature and humidity data corresponding to the current measurement time, converting impedance spectra from different monitoring periods to the same reference environment. Therefore, the changes compared in subsequent relaxation time distribution inversions primarily originate from the condition of the motor housing coating and substrate, rather than overall drift caused by temperature and humidity variations.

[0031] Traditional correction schemes, when using only ordinary numerical interpolation, may exhibit bends near missing points that do not conform to the complex impedance response path, resulting in additional relaxation peaks after inversion. Furthermore, single-temperature linear compensation is insufficient to reflect the changes in coating dielectric properties and ion conduction states under the combined effects of temperature and humidity. This step first constrains the linear, causal, and stable response characteristics of the data using the Kramers-Kronig relationship, then limits the interpolation form using the RC response and diffusion wake, and finally unifies the environmental benchmark through pre-calibrated temperature correction coefficients and temperature-humidity coupling correction coefficients. This reduces non-state differences in subsequent inversions from both data continuity and environmental comparability perspectives.

[0032] For example, after frequency sweeps are performed on the same motor casing during dry and high-humidity periods, the two sets of original complex impedance spectra may shift due to different environmental conditions. Step S20 first removes the abnormal frequency points in the top 3% of the residual sorting from each set of data, and then fills in the missing positions with adjacent valid points and physical priors; subsequently, the temperature and relative humidity at the corresponding time are read and corrected to the same reference environment. If the two sets of spectra tend to be consistent in the high-frequency and mid-frequency regions after correction, while still retaining stable differences in the low-frequency region, then this difference can be used as state change information in step S30, avoiding the direct interpretation of environmental drift as coating degradation or hidden corrosion.

[0033] Step S30: Based on the corrected complex impedance spectrum, a partitioned adaptive Tikhonov regularization mechanism is used to perform the relaxation time distribution inversion preprocessing task, and output the state relaxation time distribution data; The state relaxation time distribution data refers to the combined data of the relaxation time distribution corresponding to the current monitoring state and the baseline relaxation time distribution pre-constructed under the initial intact state of the motor casing. During processing, the corrected complex impedance spectrum is substituted into the first type of Fredholm integral relation, and the continuous relaxation time axis is discretized into solvable time nodes. Based on this, an inverse relationship between the complex impedance response and the distribution intensity at each time node is established. According to the response position of different physical processes, the relaxation time axis is divided into three intervals: short, medium, and long. The short relaxation time interval uses a first preset regularization parameter and a second-order difference smoothing constraint; the medium relaxation time interval uses a second preset regularization parameter and a second-order difference smoothing constraint; and the long relaxation time interval uses a third preset regularization parameter and a first-order difference smoothing constraint. The first preset regularization parameter is greater than the second preset regularization parameter, and the second preset regularization parameter is greater than the third preset regularization parameter. During the solution process, the distribution intensity at each time node is limited to be no less than zero, resulting in the current state relaxation time distribution. This current state relaxation time distribution, along with the baseline relaxation time distribution obtained using the same discrete time axis and partitioning rules, is output to step S40.

[0034] Partition regularization assigns a smoothing strength appropriate to the physical response of different relaxation time intervals. Short relaxation time intervals use stronger second-order difference constraints to suppress dense oscillations caused by measurement noise in the inversion results; medium relaxation time intervals preserve the peak shape changes of the coating's water absorption and interfacial charge transfer responses; long relaxation time intervals use smaller regularization parameters and first-order difference constraints to reduce over-smoothing of weak corrosion responses. After applying consistent discretization and constraint rules to the current state and the baseline state, the differences between them in step S40 have the same computational basis, reducing spurious differences caused by different inversion configurations.

[0035] Traditional single-parameter Tikhonov inversion uses the same constraint across the entire relaxation time axis. Too large a parameter flattens weak peaks in long relaxation time intervals, creating tails for coating water absorption peaks; too small a parameter amplifies noise in short relaxation time intervals, turning it into spurious peaks. This step sets decreasing regularization parameters for short, medium, and long relaxation time intervals, and uses first-order difference constraints for long intervals. This suppresses oscillations in short intervals while preserving peak position and intensity information related to substrate corrosion in long intervals. Non-negative constraints further eliminate physically meaningless negative intensity distributions, making the output data suitable for subsequent multi-peak fitting and difference spectrum decoupling.

[0036] For example, when inverting the baseline corrected complex impedance spectrum and the current corrected complex impedance spectrum of the same monitoring patch, the same relaxation time discrete nodes are first used, and then the three intervals are divided according to the preset physical process mapping. If the current state only shows a weak response in the long relaxation time interval, a smaller third preset regularization parameter and first-order difference constraint can preserve the peak position and amplitude of the response; the stronger second-order difference constraint in the short interval suppresses the fine peaks formed by high-frequency noise. The two inversion results are combined into state relaxation time distribution data according to the same nodes, and step S40 can directly compare the peak area and peak shape changes in the same interval.

[0037] Step S40: Based on the state relaxation time distribution data, the net corrosion feature extraction task is performed using the differential spectrum decoupling mechanism constrained by the physical process, and the net corrosion feature parameter set is output. The net corrosion characteristic parameter set refers to the combination of corrosion characteristic peak time constant, corrosion characteristic peak intensity, and coating water absorption change. Step S40 reads the current state distribution and baseline state distribution from the state relaxation time distribution data, performs multi-peak fitting using a mixed function composed of Gaussian and Lorentz functions, and, based on a preset physical process mapping relationship, associates peaks in short relaxation time intervals with the coating capacitance response, peaks in medium relaxation time intervals with the coating water absorption response and the coating-substrate interface charge transfer response, and peaks in long relaxation time intervals with the substrate corrosion response. Then, the difference in characteristic peak area between the current state and the baseline state within the medium relaxation time interval is calculated. This difference is used as the coating water absorption change, and a pre-established coating water absorption and micro-cell driving potential evolution matrix is ​​input to obtain the low-frequency interference relaxation time distribution caused by coating water absorption and interface state migration. The net corrosion difference spectrum is processed according to each relaxation time node: the current state distribution value is subtracted from the baseline state distribution value at the same node, and then the low-frequency interference distribution value at the same node is subtracted. Finally, the net spectrum characteristic peaks are searched within the preset corrosion reaction relaxation time interval, and the time constants and peak intensities corresponding to the peaks are read and output together with the change in water absorption of the coating to step S50.

[0038] The physical process mapping gives peaks in different intervals a clear source of response. The peak area change in the mid-relaxation time interval is used separately to estimate the water absorption effect, and the resulting low-frequency interference distribution is then subtracted from the difference between the current state and the baseline state. The resulting net corrosion difference spectrum retains the long relaxation time response that cannot be explained by the baseline state and water absorption migration. The temporal position and intensity of the corrosion characteristic peaks are no longer directly mixed with the peak position drift and broadening caused by normal moisture absorption of the coating. The change in coating water absorption is also retained in the output parameter set and can be used in step S50 to dynamically adjust the corrosion peak intensity threshold.

[0039] Traditional differential analysis directly compares impedance modulus, phase angle, or simply calculates the difference between the current distribution and the baseline distribution. This can lead to the broadening of the relaxation peak caused by coating water absorption, potentially extending into a longer relaxation time interval and forming low-frequency increments similar to corrosion peaks. This new step does not directly identify all low-frequency increments as corrosion. Instead, it first characterizes water absorption changes using the peak area difference in the mid-range, then estimates the interference distribution in the low-frequency region based on the coating water absorption and the microcell driving potential evolution matrix, and performs subtraction at the same time point. This ensures that corrosion feature extraction is simultaneously constrained by the physical attribution of peak regions and the quantification results of water absorption interference.

[0040] For example, the casing of a motor that has been in a humid environment for a long time shows a significant increase in peak area during the medium relaxation time interval, while a gradual increase also appears during the long relaxation time interval. Step S40 first determines the peak area difference in the medium interval as the change in coating water absorption, and generates the low-frequency interference distribution corresponding to this water absorption state using the evolution matrix; after subtracting the baseline distribution and low-frequency interference distribution point by point, if an independent net peak still exists within the preset corrosion reaction interval, its time constant and intensity are extracted. If the gradual increase can be explained by the water absorption interference distribution, and no corrosion peak satisfying the search conditions is formed in the net spectrum, step S50 will not make a corrosion judgment solely based on coating water absorption.

[0041] Step S50: Perform corrosion quantification correlation and integrity grading judgment tasks based on the net corrosion characteristic parameter set, and output the motor housing integrity monitoring results.

[0042] The motor housing integrity monitoring result refers to the output result including the integrity status level and warning level of the motor housing. Step S50 extracts the corrosion characteristic peak time constant, corrosion characteristic peak intensity, and coating water absorption change from the net corrosion characteristic parameter set. The dynamic corrosion peak intensity threshold is determined according to a textual rule: first, the coating water absorption change is multiplied by a pre-calibrated low-frequency baseline drift correction coefficient, and then the product is added to the pre-calibrated base peak intensity threshold. Only when the corrosion characteristic peak time constant is within the preset corrosion reaction relaxation time interval and the corrosion characteristic peak intensity is not less than the dynamic threshold, is it determined that there is hidden corrosion in the motor housing; if either condition is not met, it is determined that there is no hidden corrosion. For cases where hidden corrosion exists, the pre-calibrated mapping relationship between corrosion characteristic peak intensity and effective corrosion active area is called to obtain the effective corrosion active area, and then compared with the preset corrosion area grading threshold level by level to determine the integrity status level and the corresponding warning level; for cases where there is no hidden corrosion, the motor housing is determined to be in an intact state.

[0043] The dynamic threshold incorporates the impact of coating moisture absorption changes on the low-frequency baseline into the judgment criteria for the same monitoring period. When the moisture absorption change increases, the threshold is adjusted accordingly based on the low-frequency baseline drift correction coefficient, which can prevent the rise in low-frequency background caused by normal moisture absorption from directly triggering corrosion alarms; the time constant interval condition excludes net peaks located outside the corrosion reaction range. After corrosion is determined, the peak intensity is mapped to the effective active area, and a status level and warning level are formed based on the area threshold, so that the monitoring results not only indicate whether there is hidden corrosion, but also give the maintenance priority corresponding to the corrosion range.

[0044] Traditional fixed peak intensity thresholds cannot adapt to changes in the coating's water absorption state, and under long-term high humidity conditions, a natural rise in the baseline can easily be mistaken for corrosion. Furthermore, simply outputting a binary result indicating the presence or absence of corrosion fails to reflect whether the effective active area for corrosion crosses the maintenance level boundary. This step sequentially combines the time constant range, the peak intensity threshold adjusted according to water absorption, and the classification of the effective active area for corrosion. The result of the previous judgment serves as the activation condition for the next processing step, avoiding the calculation of the corrosion level when corrosion conditions are not met, and ensuring that the warning level can be traced back to specific parameters in the net corrosion characteristic parameter set.

[0045] For example, during a round of motor casing inspection, the corrosion characteristic peak time constant output in step S40 falls within the preset corrosion reaction relaxation time range. The system calculates the dynamic threshold for this cycle based on the coating water absorption change, low-frequency baseline drift correction coefficient, and basic peak intensity threshold from the same parameter set. If the peak intensity reaches the threshold, it is determined that there is hidden corrosion, and the effective active area of ​​corrosion is obtained according to the pre-calibrated mapping relationship. This area is compared with the preset corrosion area thresholds at each level to form the corresponding integrity status level and warning level, which maintenance personnel can use to arrange retesting, local inspection, or shutdown. If the peak intensity does not reach the dynamic threshold, the integrity status is output to avoid corrosion warnings being triggered solely by water absorption response in a high-humidity environment.

[0046] Example 2: Furthermore, the present invention provides a motor housing integrity monitoring system based on impedance characteristic analysis, employing a motor housing integrity monitoring method based on impedance characteristic analysis as described in the above embodiments, which can solve the technical problem of motor housing integrity monitoring based on impedance characteristic analysis. The beneficial effects of the motor housing integrity monitoring system based on impedance characteristic analysis provided by the present invention are the same as those of the motor housing integrity monitoring method based on impedance characteristic analysis provided in the above embodiments, and other technical features of the motor housing integrity monitoring system based on impedance characteristic analysis are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0047] Example 3: This invention provides a motor housing integrity monitoring device based on impedance characteristic analysis. The device includes at least one processor and a memory communicatively connected to the processor. The memory stores instructions executable by the processor, which in turn execute the methods described in Example 1 for monitoring motor housing integrity based on impedance characteristic analysis. The motor housing integrity monitoring device based on impedance characteristic analysis in this invention can include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. This motor housing integrity monitoring device based on impedance characteristic analysis is merely an example and should not limit the functionality or scope of the invention. An impedance-based motor housing integrity monitoring device may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in a read-only memory or a program loaded from a storage device into a random access memory. The random access memory also stores various programs and data required for the operation of the impedance-based motor housing integrity monitoring device. The processing unit, read-only memory, and random access memory are interconnected via a bus. An I / O interface is also connected to the bus. Typically, the following systems can be connected to the I / O interface: input devices including touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices including liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices including magnetic tapes, hard disks, etc.; and communication devices. The communication device allows the impedance-based motor housing integrity monitoring device to communicate wirelessly or wiredly with other devices to exchange data. While an impedance-based motor housing integrity monitoring device with various systems has been described, it should be understood that it is not required to implement or possess all of the described systems. More or fewer systems may be implemented alternatively.

[0048] Example 4: This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method for monitoring the integrity of a motor housing based on impedance characteristic analysis. The computer program product provided by this invention can solve the technical problem of monitoring the integrity of a motor housing based on impedance characteristic analysis. Compared with the prior art, the beneficial effects of the computer program product provided by this invention are the same as those of the method for monitoring the integrity of a motor housing based on impedance characteristic analysis provided in the above embodiments, and will not be repeated here.

[0049] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a read-only memory. When the computer program is executed by a processing device, it performs the functions defined in the methods of the embodiments disclosed in this invention.

[0050] It should be understood that the various parts disclosed in this invention can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0051] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the present invention and its equivalents, the present invention also intends to include these modifications and variations.

Claims

1. A method for monitoring the integrity of a motor housing based on impedance characteristic analysis, characterized in that, The method includes: Step S10: Acquire the motor housing impedance response signal and the synchronous environment signal, and use the frequency adaptive dwell mechanism to perform the wideband impedance acquisition task, and output the impedance and environment synchronization dataset; Step S20: Based on the impedance and environment synchronization dataset, an impedance consistency verification and environment normalization correction task is performed using a hybrid physical and empirical adaptive correction mechanism, and the corrected complex impedance spectrum is output. Step S30: Based on the corrected complex impedance spectrum, a partitioned adaptive Tikhonov regularization mechanism is used to perform the relaxation time distribution inversion preprocessing task, and output the state relaxation time distribution data; Step S40: Based on the state relaxation time distribution data, the net corrosion feature extraction task is performed using the differential spectrum decoupling mechanism constrained by the physical process, and the net corrosion feature parameter set is output. Step S50: Perform corrosion quantification correlation and integrity grading judgment tasks based on the net corrosion characteristic parameter set, and output the motor housing integrity monitoring results.

2. The method for monitoring the integrity of a motor housing based on impedance characteristic analysis as described in claim 1, characterized in that, Step S10, which involves acquiring the motor housing impedance response signal and the synchronization environment signal, performing a wideband impedance acquisition task using a frequency adaptive dwell mechanism, and outputting an impedance and environment synchronization dataset, specifically includes: Step S101: Apply a multi-frequency sinusoidal excitation signal with a frequency range of 10 mHz to 100 kHz to the motor housing through a monitoring patch set on the motor housing, and collect the complex impedance response corresponding to each excitation frequency point to obtain multi-frequency complex impedance sampling data; wherein, the monitoring patch adopts a three-electrode system or a four-point electrode system, and the excitation frequency points are arranged in a logarithmic interval. Step S102: Calculate the signal-to-noise ratio of the multi-frequency complex impedance sampling data at each excitation frequency point. When the signal-to-noise ratio at the current excitation frequency point is lower than the preset signal-to-noise ratio threshold, extend the excitation dwell time at the current excitation frequency point to complete the integral sampling of an integer number of excitation cycles and obtain adaptive dwell complex impedance data. Step S103: Synchronously collect ambient temperature data and relative humidity data output by the temperature and humidity sensor set on the monitoring patch, and associate the ambient temperature data, the relative humidity data and the adaptive dwelling complex impedance data according to a unified sampling time scale to obtain the impedance and environment synchronous dataset.

3. The method for monitoring the integrity of a motor housing based on impedance characteristic analysis as described in claim 2, characterized in that, In step S20, based on the impedance and environment synchronization dataset, an impedance consistency verification and environment normalization correction task is performed using a hybrid physical and empirical adaptive correction mechanism, and the corrected complex impedance spectrum is output. This step specifically includes: Step S201: Extract the adaptive resident complex impedance data from the impedance and environment synchronization dataset, perform Kramers and Kronig consistency checks on the adaptive resident complex impedance data, calculate the consistency check residuals corresponding to each excitation frequency point, and remove abnormal data points accounting for 3% of all excitation frequency points in descending order of the consistency check residuals to obtain the consistency verification complex impedance data. Step S202: Determine the missing frequency points in the consistency verification complex impedance data, and use a regularized spline interpolation algorithm with embedded resistance and capacitance response physical priors and diffusion wake physical priors to perform complex impedance interpolation on the missing frequency points to obtain complete complex impedance data. Step S203: Extract the ambient temperature data and the relative humidity data from the impedance and environment synchronization dataset, and use the pre-calibrated temperature correction coefficient and temperature-humidity coupling correction coefficient to correct the complete complex impedance data to the impedance state corresponding to the preset reference ambient temperature and preset reference relative humidity, thereby obtaining the corrected complex impedance spectrum.

4. The method for monitoring the integrity of a motor housing based on impedance characteristic analysis as described in claim 3, characterized in that, In step S202, the use of a regularized spline interpolation algorithm with embedded physical priors of impedance and capacitance response and physical priors of diffusion wakes to perform complex impedance interpolation on the missing frequency points specifically includes: The physical priors characterizing the high-frequency RC response of the motor housing coating and the low-frequency diffusion wake are embedded in the spline interpolation kernel function. The interpolation curve constrained by the complex impedance response path is constructed using the effective complex impedance data points located on both sides of the missing frequency point. The Kramers and Kronig consistency test is performed on the interpolation curve to ensure that the interpolated complex impedance data simultaneously satisfies the complex impedance spectrum continuity constraint and the Kramers and Kronig relationship constraint.

5. The method for monitoring the integrity of a motor housing based on impedance characteristic analysis as described in claim 3, characterized in that, Step S30, based on the corrected complex impedance spectrum, employs a partitioned adaptive Tikhonov regularization mechanism to perform the relaxation time distribution inversion preprocessing task, and outputs the state relaxation time distribution data. This step specifically includes: Step S301: Input the corrected complex impedance spectrum into the first type of Fredholm integral relation, discretize the relaxation time axis, and establish the inversion relationship between the complex impedance response and the relaxation time distribution based on the discretized relaxation time axis to obtain the relaxation time distribution inversion model. Step S302: Based on the response positions of different physical processes on the relaxation time axis, the relaxation time axis is divided into short relaxation time intervals, medium relaxation time intervals, and long relaxation time intervals. A first preset regularization parameter is configured for the short relaxation time interval and a second-order difference smoothing constraint is applied. A second preset regularization parameter is configured for the medium relaxation time interval and a second-order difference smoothing constraint is applied. A third preset regularization parameter is configured for the long relaxation time interval and a first-order difference smoothing constraint is applied, thus obtaining the partitioned regularization objective function. Wherein, the first preset regularization parameter is greater than the second preset regularization parameter, and the second preset regularization parameter is greater than the third preset regularization parameter. Step S303: Solve the partition regularization objective function under the non-negative constraint of the relaxation time distribution to obtain the current state relaxation time distribution corresponding to the current monitoring state, and obtain the baseline state relaxation time distribution pre-constructed in the initial complete state of the motor housing. Combine the current state relaxation time distribution and the baseline state relaxation time distribution into the state relaxation time distribution data.

6. The method for monitoring the integrity of a motor housing based on impedance characteristic analysis as described in claim 5, characterized in that, Step S40, which involves using the state relaxation time distribution data and employing a physical process-constrained differential spectrum decoupling mechanism to perform net corrosion feature extraction and output a set of net corrosion feature parameters, specifically includes: Step S401: Extract the current state relaxation time distribution and the baseline state relaxation time distribution from the state relaxation time distribution data. Perform multi-peak fitting using a mixed function composed of Gaussian function and Lorentz function. According to the preset physical process mapping relationship, associate the characteristic peaks in the short relaxation time interval with the coating capacitance response, associate the characteristic peaks in the medium relaxation time interval with the coating water absorption response and the coating-substrate interface charge transfer response, and associate the characteristic peaks in the long relaxation time interval with the substrate corrosion response to obtain the relaxation characteristic peak parameter group. Step S402: Based on the relaxation characteristic peak parameter group, calculate the difference in characteristic peak area between the current state relaxation time distribution and the baseline state relaxation time distribution in the middle relaxation time interval, determine the difference in characteristic peak area as the coating water absorption change, and input the coating water absorption change into the pre-established coating water absorption and micro battery driving potential evolution matrix to obtain the low-frequency interference relaxation time distribution caused by coating water absorption and interface state migration; Step S403: Calculate the net corrosion difference spectrum based on the current state relaxation time distribution, the baseline state relaxation time distribution, and the low-frequency interference relaxation time distribution. in, relaxation time Net corrosion differential spectral value at the location, For relaxation time, The relaxation time of the current state is distributed over the relaxation time. Distribution values ​​at that location, The relaxation time of the baseline state is distributed over the relaxation time. Distribution values ​​at that location, The relaxation time distribution of the low-frequency interference is within the relaxation time. Distribution values ​​at; Within a preset corrosion reaction relaxation time interval, a characteristic peak search is performed on the net corrosion differential spectrum to extract the corrosion characteristic peak time constant and corrosion characteristic peak intensity. The corrosion characteristic peak time constant, corrosion characteristic peak intensity and coating water absorption change are then combined to form the net corrosion characteristic parameter set.

7. The method for monitoring the integrity of a motor housing based on impedance characteristic analysis as described in claim 6, characterized in that, Step S50, which involves performing corrosion quantification correlation and integrity grading judgment tasks based on the net corrosion characteristic parameter set and outputting the motor housing integrity monitoring results, specifically includes: Step S501: Extract the corrosion characteristic peak time constant, the corrosion characteristic peak intensity, and the coating water absorption change from the net corrosion characteristic parameter set to obtain the corrosion judgment parameter set; Step S502: Multiply the change in water absorption of the coating by a pre-calibrated low-frequency baseline drift correction coefficient, and add the product to a pre-calibrated base peak intensity threshold to obtain a dynamic corrosion peak intensity threshold; when the corrosion characteristic peak time constant is within the preset corrosion reaction relaxation time interval, and the corrosion characteristic peak intensity is not less than the dynamic corrosion peak intensity threshold, it is determined that there is hidden corrosion on the motor housing; otherwise, it is determined that there is no hidden corrosion on the motor housing, and a corrosion determination result is obtained. Step S503: Based on the corrosion determination result, when it is determined that there is hidden corrosion on the motor housing, the pre-calibrated mapping relationship between the intensity of corrosion characteristic peaks and the effective active area of ​​corrosion is invoked to determine the effective active area of ​​corrosion. The effective active area of ​​corrosion is then compared with a preset corrosion area grading threshold to determine the integrity status level of the motor housing and the corresponding warning level. When it is determined that there is no hidden corrosion on the motor housing, the motor housing is determined to be in an intact state, and the integrity monitoring result of the motor housing, which includes the integrity status level and the warning level, is output.

8. A motor housing integrity monitoring system based on impedance characteristic analysis, applied to the motor housing integrity monitoring method based on impedance characteristic analysis as described in any one of claims 1 to 7, characterized in that, The motor housing integrity monitoring system includes: The wideband impedance acquisition module is used to acquire the motor housing impedance response signal and the synchronous environment signal. It uses a frequency adaptive dwell mechanism to perform the wideband impedance acquisition task and outputs an impedance and environment synchronization dataset. The impedance correction module is used to perform impedance consistency verification and environmental normalization correction tasks based on the impedance and environment synchronization dataset, using a hybrid physical and empirical adaptive correction mechanism, and outputs a corrected complex impedance spectrum. The relaxation time distribution inversion module is used to perform the relaxation time distribution inversion preprocessing task based on the corrected complex impedance spectrum and adopts a partitioned adaptive Tikhonov regularization mechanism to output state relaxation time distribution data. The net corrosion feature extraction module is used to perform the net corrosion feature extraction task based on the state relaxation time distribution data and using the differential spectrum decoupling mechanism constrained by physical process, and output the net corrosion feature parameter set. The integrity assessment module is used to perform corrosion quantification correlation and integrity grading judgment tasks based on the net corrosion characteristic parameter set, and output the integrity monitoring results of the motor housing.

9. A motor housing integrity monitoring device based on impedance characteristic analysis, characterized in that, The motor housing integrity monitoring device based on impedance characteristic analysis includes: a memory, a processor, and a motor housing integrity monitoring program based on impedance characteristic analysis stored in the memory and executable on the processor. When the motor housing integrity monitoring program based on impedance characteristic analysis is executed by the processor, it implements a motor housing integrity monitoring method based on impedance characteristic analysis as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes a motor housing integrity monitoring program based on impedance characteristic analysis, which, when executed by a processor, implements a motor housing integrity monitoring method based on impedance characteristic analysis as described in any one of claims 1 to 7.

Images