Motor heat dissipation performance evaluation method and system based on temperature rise curve fitting

Abstract

This invention discloses a method and system for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting, relating to the field of motor thermal health management technology. The method includes: monitoring the casing temperature and ambient temperature and calculating the temperature rise sequence; collecting current, speed, and cooling characteristic parameters to determine the evaluation period; generating a detection sequence within the evaluation period and injecting current under electromagnetic torque disturbance constraints to obtain a detection thermal input sequence; simultaneously demodulating the detection thermal input sequence and the temperature rise sequence to obtain thermal frequency response estimation and coherence, thereby selecting effective frequency bands and reconstructing a standardized temperature rise curve; fitting the standardized temperature rise curve to obtain thermal impedance parameters and calculating cooling sensitivity under two cooling levels, decomposing convection link parameters and heat conduction link parameters; calculating the heat dissipation deviation and making multi-level judgments, outputting a pre-diagnostic level. By using constrained injection, identifiable thermal excitation is achieved, improving the observability of the evaluation without significantly affecting the output.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of motor thermal health management, and particularly relates to a motor heat dissipation performance evaluation method and system based on temperature rise curve fitting. BACKGROUND

[0002] In the scene of motor thermal health management and pre-diagnosis, heat dissipation performance evaluation is usually carried out around temperature measurement, working condition recording and thermal model identification. The conventional method can obtain the temperature rise curve by arranging the shell and winding temperature sensors, combine the environmental conditions and load information, estimate the thermal parameters by using the equivalent thermal resistance-thermal capacity model or empirical fitting, and evaluate the heat dissipation capacity and thermal margin accordingly; under the test condition, the temperature rise response can also be obtained by constant load or specified working condition sequence to form the reference curve and maintenance criterion, thereby supporting the motor operation safety and life management.

[0003] However, in the face of in-service evaluation, the above conventional method still has engineering difficulties in two aspects: firstly, the amplitude and frequency domain components of thermal excitation are difficult to control due to the fluctuation of actual running working conditions, and the fitting result of the temperature rise curve is easily affected by noise and non-thermal disturbance, which makes it difficult to stably guarantee the comparability and reliability of the parameters in different time periods and different cooling conditions; secondly, the temperature rise curve often reflects the coupling effect of multiple thermal channels, and it is difficult to attribute the change of heat dissipation to the change of external convection capacity or internal heat conduction path only by a single fitting result, thereby limiting the explainability and maintenance directivity of the pre-diagnosis conclusion. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a motor heat dissipation performance evaluation method based on temperature rise curve fitting, which solves the problems of insufficient reliability and comparability of the fitting result caused by uncontrollable thermal excitation and difficulty in attributing the change of heat dissipation to the convection channel or heat conduction channel in the prior art.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present invention provides a method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting. This method includes: monitoring the casing temperature and ambient temperature; calculating the temperature rise sequence; collecting current, speed, and cooling characteristic quantities; and determining the evaluation period. During the evaluation period, a detection sequence is generated; under the constraint of limiting electromagnetic torque disturbance, a detection injection is applied to the current command to obtain a detection heat input sequence. Based on the detection heat input sequence and the temperature rise sequence, synchronous demodulation is performed to obtain a thermal frequency response estimate and coherence. An effective frequency band is generated based on the coherence, and a standardized temperature rise curve is reconstructed. The standardized temperature rise curve is fitted to obtain thermal impedance parameters. Cooling characteristic quantities are sequentially set to a first cooling level and a second cooling level, and the difference is verified. Cooling sensitivity is calculated, and the thermal impedance parameters are decomposed into convection link parameters and heat conduction link parameters. Based on the thermal frequency response estimate, effective frequency band screening data, convection link parameters, and heat conduction link parameters, the heat dissipation deviation is calculated and multi-level judgment is performed, outputting a pre-diagnosis level.

[0008] As a preferred embodiment of the motor heat dissipation performance evaluation method based on temperature rise curve fitting described in this invention, the determination of the evaluation period includes: calculating the speed stability index and the current stability index respectively within the stability evaluation window; determining that the evaluation period is entered when the speed stability index is not greater than the speed stability threshold and the current stability index is not greater than the current stability threshold; obtaining the maximum allowable casing temperature and taking the maximum value of the casing temperature sequence, calculating the temperature margin, and performing probe injection within the evaluation period based on the temperature margin.

[0009] As a preferred embodiment of the motor heat dissipation performance evaluation method based on temperature rise curve fitting described in this invention, the step of generating a probe sequence during the evaluation period includes generating a binary probe sequence using a linear feedback shift register, wherein the binary probe sequence is updated once at the beginning of each symbol.

[0010] As a preferred embodiment of the motor heat dissipation performance evaluation method based on temperature rise curve fitting described in this invention, the step of applying probe injection to the current command under the constraint of keeping electromagnetic torque disturbance limited includes: in each symbol period, the product of the probe sequence value and the injection amplitude is used as the d-axis injection amount and superimposed on the reference d-axis current command to generate the d-axis probe injection; the compensation coefficient is determined according to the sensitivity ratio formed by the partial derivatives of electromagnetic torque with respect to d-axis current and q-axis current, and the d-axis probe injection amount is converted into the q-axis compensation injection amount.

[0011] As a preferred embodiment of the motor heat dissipation performance evaluation method based on temperature rise curve fitting described in this invention, the method of applying probe injection to the current command under the constraint of keeping electromagnetic torque disturbance limited further includes: calculating an estimated value of the change in torque command before and after injection based on the internal torque; when the estimated value exceeds the upper limit of electromagnetic torque disturbance, reducing the injection amplitude by a fixed ratio and regenerating the original symbol injection amount; when the current command exceeds the limit, preferentially reducing the d-axis injection amount and recalculating the q-axis compensation injection amount; and suspending probe injection when the temperature margin is insufficient.

[0012] As a preferred embodiment of the motor heat dissipation performance evaluation method based on temperature rise curve fitting described in this invention, the method of obtaining the probe heat input sequence includes: recording the actual d-axis current, q-axis current and reference current during the evaluation period, reading the stator resistance estimate, and calculating the difference between the square of the injected current and the square of the reference current; and using the incremental copper loss caused by the injection as the probe heat input sequence.

[0013] As a preferred embodiment of the motor heat dissipation performance evaluation method based on temperature rise curve fitting described in this invention, the synchronous demodulation includes: performing a discrete Fourier transform on the probed thermal input sequence and temperature rise sequence after mean removal processing to calculate the input autospectrum, output autospectrum, and input-output cross spectrum; using the ratio of the cross spectrum to the input autospectrum as a thermal frequency response estimate, and calculating the coherence through the cross spectrum, input autospectrum, and output autospectrum; comparing the coherence at each frequency point with the coherence threshold to generate an effective frequency band, and using the bandwidth covered by the effective frequency points as the effective frequency band measure; comparing the effective frequency band measure with the effective frequency band threshold to determine the reliability of synchronous demodulation.

[0014] As a preferred embodiment of the motor heat dissipation performance evaluation method based on temperature rise curve fitting described in this invention, the reconstruction of the standardized temperature rise curve includes: constructing a frequency domain transfer function based on the thermal frequency response estimation at the effective frequency point and setting the values ​​at the invalid frequency point to zero; performing an inverse transformation after completion according to the conjugate symmetry rule to obtain the discrete sequence of impulse response, and performing cumulative integration on the impulse response to obtain the standardized temperature rise curve.

[0015] As a preferred embodiment of the motor heat dissipation performance evaluation method based on temperature rise curve fitting described in this invention, the following steps are included: calculating the cooling sensitivity and decomposing the thermal impedance parameters into convection link parameters and heat conduction link parameters: expressing the standardized temperature rise curve as a second-order thermal impedance response form and performing least-squares fitting to obtain a set of thermal impedance parameters; setting the cooling characterization quantities sequentially to a first cooling level and a second cooling level, and verifying that the difference between the first cooling level and the second cooling level reaches a cooling switching threshold, calculating the cooling sensitivity based on the difference in thermal impedance parameters under the first cooling level and the second cooling level; and decomposing the thermal impedance parameters according to the comparison between the cooling sensitivity and the cooling sensitivity threshold to obtain the convection link parameters and heat conduction link parameters.

[0016] Secondly, this invention provides a motor heat dissipation performance evaluation system based on temperature rise curve fitting, comprising: a data acquisition module for monitoring casing temperature and ambient temperature, calculating temperature rise sequence, acquiring current, speed, and cooling characterization parameters, and determining the evaluation period; a detection injection module for generating a detection sequence during the evaluation period, applying detection injection to the current command under the constraint of limiting electromagnetic torque disturbance, and obtaining a detection heat input sequence; a synchronous demodulation module for performing synchronous demodulation based on the detection heat input sequence and temperature rise sequence, obtaining thermal frequency response estimation and coherence, generating an effective frequency band based on the coherence, and reconstructing a standardized temperature rise curve; a fitting decomposition module for fitting the standardized temperature rise curve, obtaining thermal impedance parameters, setting the cooling characterization parameters sequentially to a first cooling level and a second cooling level and verifying the difference, calculating the cooling sensitivity, and decomposing the thermal impedance parameters into convection link parameters and thermal conduction link parameters; and a deviation judgment module for calculating the heat dissipation deviation based on the thermal frequency response estimation, effective frequency band screening data, convection link parameters, and thermal conduction link parameters, performing multi-level judgment, and outputting a pre-diagnosis level.

[0017] The beneficial effects of this invention are as follows: by generating a detection sequence during the evaluation period and applying detection injection to the current command under the constraint of electromagnetic torque disturbance, an identifiable thermal excitation is constructed without significantly affecting the normal output of the motor; by applying two cooling levels sequentially during the same evaluation period and calculating the cooling sensitivity after the switching is achieved, and decomposing the thermal impedance parameter into convection link parameters and heat conduction link parameters, attributable diagnosis of heat dissipation degradation is realized. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Fig. 1 This is a flowchart of a method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting.

[0020] Fig. 2 This is a schematic diagram of a motor heat dissipation performance evaluation system based on temperature rise curve fitting.

[0021] Fig. 3 A flowchart for reconstructing the standardized temperature rise curve.

[0022] Fig. 4 This is a flowchart for fitting decomposition of two cooling levels and determining the deviation of heat dissipation. Detailed Implementation

[0023] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0024] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0025] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0026] Reference Figs. 1-4 As one embodiment of the present invention, this embodiment provides a method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting, including the following steps:

[0027] S1. Monitor the shell temperature and ambient temperature, calculate the temperature rise sequence, collect current, rotation speed and cooling characteristics, and determine the evaluation period.

[0028] Acquire and time-align the following signals at a fixed sampling period (e.g., 1 s): housing temperature, ambient temperature, current, speed, and cooling characteristics (read fan PWM duty cycle).

[0029] At each sampling time, the difference between the shell temperature and the ambient temperature is calculated to obtain the temperature rise sequence.

[0030] To ensure the identifiability of thermal detection and synchronous demodulation, a fixed-time stability evaluation window (e.g., 60 s) is set, and the rotational speed stability index is calculated based on the sampled data. shaft current and The shaft current is calculated at each sampling time as a current amplitude sequence current stability index, expressed as:

[0031] ;

[0032] ;

[0033] in, This indicates the speed stability index. Indicates current stability index, Indicates the duration of the stability assessment window. Indicates the current determination time. Indicates time The rotational speed sample value at that location, Represents the integral variable time. Indicates interval [ , The arithmetic mean of the internal rotational speeds. Indicates time Current amplitude sequence values ​​at the location, Indicates interval [ , The arithmetic mean of the internal current amplitude.

[0034] when Not greater than the speed stability threshold and If the current is not greater than the current stability threshold, the stable operation condition is deemed met; otherwise, the evaluation period is not entered, and the stability index continues to be updated on a rolling basis.

[0035] It should be noted that the speed stability threshold and current stability threshold are determined by continuously collecting multiple sets of stability index samples under the condition that the motor is confirmed to be healthy and operating stably. and The corresponding quantiles are taken as the speed stability threshold and current stability threshold, respectively, such as the 95th quantile, with a value range greater than or equal to 0.

[0036] Read the maximum allowable case temperature. The most recent Within the time period, the maximum value of the shell temperature sequence is taken and compared with the highest allowable shell temperature to calculate the temperature margin. When the temperature margin is not less than the margin threshold, the temperature margin condition is deemed to be met; otherwise, the evaluation is postponed.

[0037] It should be noted that the margin threshold is determined by the maximum allowable housing temperature of the motor. The statistical data on the maximum increase in shell temperature caused by temperature rise detection during the historical assessment period is used. The fixed quantile of the maximum increase is taken as the margin threshold, such as the 95th quantile, and the value range is usually from 8℃ to 12℃.

[0038] Will As the starting point of the evaluation period, The evaluation period is obtained as the end point of the evaluation period.

[0039] S2. Generate a detection sequence during the evaluation period, and apply detection injection to the current command under the constraint of keeping the electromagnetic torque disturbance limited, to obtain the detection thermal input sequence.

[0040] At the start of the assessment period To the finish line An internal probe sequence generator is used to generate a binary sequence output, which employs a linear feedback shift register. The output value is taken as... or The sequence value is updated once at the beginning of each symbol and written to the probe reference register.

[0041] A current command probe injection limited by electromagnetic torque disturbance is used to generate identifiable thermal excitation from motor losses. Specifically, within each symbol period, the probe sequence value and the injection amplitude are... The product is used as the d-axis injection amount and superimposed on the baseline. Shaft current command, generated Shaft detection injection; based on the torque sensitivity of the current operating point, the torque disturbance constraint is transformed into a constraint on... The calculation rules for shaft injection amount are based on electromagnetic torque. shaft current and The partial derivative of the axis current constitutes the sensitivity ratio. This sensitivity ratio is used as a compensation coefficient to adjust the sensitivity of this symbol. Axis probe injection volume conversion The direction of the shaft compensation injection is chosen so that the torque changes caused by the two cancel each other out under the first-order approximation, making the first-order torque disturbance close to zero, thereby achieving disturbance limitation.

[0042] It should be noted that the injection amplitude is set as a fixed percentage of the motor's rated current, for example, 1% to 3% of the rated current; the reference value... The shaft current command indicates the current loop output before probe injection. Shaft current setpoint.

[0043] Disturbance limitation verification and amplitude limiting are performed. Specifically, the estimated value of the change in torque command before and after injection is calculated (calculated from the controller's internal torque). If the estimated value of the change exceeds the upper limit of electromagnetic torque disturbance, the injection amplitude is reduced by a fixed proportion (e.g., 50%) and the current symbol injection amount is regenerated. If the current command exceeds the current limit, it is preferentially reduced. Shaft injection volume, recalculate The amount of axial compensation injection is determined; if the temperature margin is insufficient, the current detection injection is paused and the evaluation period is terminated.

[0044] It should be noted that the upper limit of electromagnetic torque disturbance is a fixed torque ratio of the motor's rated torque, such as 1%; the current limit is the controller's rated current; insufficient temperature margin is determined by whether the current housing temperature is close to the maximum allowable housing temperature.

[0045] The actual execution was recorded synchronously during the evaluation period. axis, The shaft current and reference current are read, and the stator resistance is estimated. The incremental copper loss caused by injection is used as the probe thermal input sequence. Specifically, in each sampling period, the sum of squares of the injected current and the sum of squares of the reference current are calculated respectively. The difference between the two is taken and multiplied by the stator resistance estimate and the phase coefficient to obtain the probe thermal input sequence value.

[0046] It should be noted that the reference current is the current command given before the probe injection is superimposed; the stator resistance is estimated to be the factory calibration; the phase number coefficient is obtained from the fixed structure of the motor, such as the phase number coefficient of a three-phase motor being 3.

[0047] S3. Based on the detected thermal input sequence and temperature rise sequence, perform synchronous demodulation to obtain thermal frequency response estimation and coherence. Generate effective frequency bands based on coherence and reconstruct standardized temperature rise curves.

[0048] Read the thermal input sequence and temperature rise sequence, align them with the same timestamp, and then perform mean-removal processing on the thermal input sequence and temperature rise sequence respectively. Specifically, subtract the average value of the thermal input sequence and temperature rise sequence during the evaluation period from the thermal input sequence and temperature rise sequence to obtain the mean-removed input sequence and output sequence.

[0049] The mean-reduced input and output sequences are subjected to Discrete Fourier Transform during the evaluation period to obtain the input frequency domain representation and the output frequency domain representation. The input autospectrum and the input-output cross spectrum are calculated within the frequency band, and the ratio of the cross spectrum to the autospectrum is used as the thermal frequency response estimate. The thermal frequency response estimate is used as the frequency domain representation of the heat dissipation channel within the frequency band.

[0050] Within the same frequency band, the output autospectrum is further calculated, and the coherence is constructed using the cross-spectrum and autospectrum. A higher coherence indicates that the temperature rise response at the current frequency point is more likely to be caused by the probed thermal input. The expression is:

[0051] ;

[0052] in, Represents angular frequency Coherence at that point This indicates the detection of the thermal input sequence and the temperature rise sequence at angular frequencies. The mutual spectrum at the location, Indicates the probe thermal input sequence at angular frequency The self-spectrum of the place, Indicates the temperature rise sequence at angular frequency The self-spectrum of the place.

[0053] The coherence of each frequency point is determined using a binary method. Specifically, frequency points with coherence not lower than the coherence threshold are considered valid, while those with coherence not lower are considered invalid, thus filtering out valid frequency bands. The bandwidth covered by the valid frequency points is used as the effective frequency band measure and compared with the effective frequency band threshold. If the effective frequency band is less than the effective frequency band threshold, the current synchronization demodulation result is deemed unreliable. When the effective frequency band is greater than or equal to the effective frequency band threshold, the effective frequency band is deemed to meet the requirements.

[0054] It should be noted that when the motor is healthy and operating stably, detection and synchronous demodulation are repeatedly performed. The coherence and effective bandwidth measures obtained in each iteration are statistically analyzed. The 95th percentile of the coherence is taken as the coherence threshold, and the 95th percentile of the effective bandwidth measure is taken as the effective bandwidth threshold. If the coherence threshold and effective bandwidth threshold are lower than the 95th percentile, tail samples that should be considered low coherence in the healthy baseline will be allowed to pass. This may lead to the gating still passing when actual operating disturbances, noise, or non-detection heat sources dominate, thus mistaking unreliable demodulation results as reliable and causing false alarms. If the coherence threshold and effective bandwidth threshold are higher than the 95th percentile, normal samples that should be accepted in the healthy baseline will also be frequently rejected, making it difficult for the assessment to pass the gating under normal conditions, resulting in a significant decrease in the number of effective assessments and sparse trend data.

[0055] When the effective frequency band is satisfied, the frequency domain transfer function is constructed using the thermal frequency response estimate corresponding to the effective frequency point, and the inverse transformation is performed to obtain the discrete sequence of the thermal channel impulse response; the standard step thermal input amplitude is used as a unified input condition, and the impulse response is cumulatively integrated to obtain the standardized temperature rise curve.

[0056] The construction of the frequency domain transfer function involves processing each discrete frequency point. When a frequency point is valid, the estimated thermal frequency response value of that frequency point is directly taken as the value of the frequency domain transfer function at that frequency point. When a frequency point is invalid, the value of the frequency domain transfer function at that frequency point is set to zero. At the same time, negative frequency points are completed according to the conjugate symmetry rule of the discrete transform, so that the obtained frequency domain transfer function satisfies the symmetry required by the inverse transform, forming a complete set of discrete frequency point values ​​as the discrete representation of the frequency domain transfer function.

[0057] It should be noted that the standard step thermal input amplitude is obtained by performing a known constant injection under healthy and stable motor conditions (so that the probe thermal input sequence reaches and remains constant during the evaluation period), and using the steady-state average power value of the probe thermal input sequence corresponding to the evaluation period when it is constant as the standard step thermal input amplitude.

[0058] S4. Fit the standardized temperature rise curve to obtain the thermal impedance parameters. Set the cooling characterization parameters to the first cooling level and the second cooling level in sequence and verify the difference. Calculate the cooling sensitivity and decompose the thermal impedance parameters into convection link parameters and heat conduction link parameters.

[0059] The standardized temperature rise curve can be expressed in the form of a second-order thermal impedance response, as follows:

[0060] ;

[0061] in, Indicates time The standardized temperature rise curve value, Indicates the coefficient of the first thermal resistance component. Indicates the coefficient of the second thermal resistance component. Represents the first time constant. This represents the second time constant.

[0062] Using discrete sampling points within the evaluation period as the dataset, the first thermal impedance component coefficient, the second thermal impedance component coefficient, the first time constant, and the second time constant are solved by minimizing the error in the sense of least squares. The constraint conditions are used as boundary restrictions. If candidate parameters that violate the constraints appear during the iteration process, they are corrected to be within the constraint boundary and the iteration continues until the error converges. The converged first thermal impedance component coefficient, the second thermal impedance component coefficient, the first time constant, and the second time constant are obtained and used as the set of thermal impedance parameters.

[0063] It should be noted that the constraints are that the coefficients of the first and second thermal impedance components should both be greater than or equal to zero, the first and second time constants should both be greater than zero, and the first time constant should be less than the second time constant.

[0064] To achieve attributable decomposition of the airflow and heat conduction links, the cooling characterization parameters are deterministically set twice within the same evaluation period. The evaluation period is divided into two consecutive sub-periods, including cooling level application and cooling switch achievement verification. Specifically, after the start of the evaluation period, the cooling characterization parameter is first set to the first cooling level, and the fan PWM duty cycle is set to a fixed low value (e.g., 30%) and maintained for a sub-period. Then, the cooling characterization parameter is switched to the second cooling level, and the fan PWM duty cycle is set to a fixed high value (e.g., 70%) and maintained for a sub-period. Both sub-periods are within the same evaluation period, and the switching order is fixed as low to high. The time average of the actual collected cooling characterization parameter sequences is calculated for each of the two sub-periods, and the difference between the two average values ​​is compared with the cooling switch achievement threshold. If the difference between the two average values ​​does not reach the cooling switch achievement threshold, the cooling switch is determined to have failed, and abnormal cooling execution data is output. If the difference between the two average values ​​reaches the cooling switch achievement threshold, the cooling switch achievement verification is output as passed.

[0065] It should be noted that the cooling switch achievement threshold is determined by repeatedly performing two cooling switches under healthy and stable operating conditions, recording the difference in the mean feedback of cooling characteristics before and after each switch, forming a historical sample set. The corresponding quantile of the historical sample set is used as the cooling switch achievement threshold, such as the 5th quantile. If the cooling switch achievement threshold is less than the 5th quantile, many small differences caused only by measurement noise, duty cycle quantization error, or short-term jitter will be judged as achieved, resulting in unreliable decomposition conclusions and increasing the risk of misdiagnosis. If the cooling switch achievement threshold is greater than the 5th quantile, cooling switches that are normal in the baseline stage will be frequently judged as not achieved, lacking link attribution data and reducing diagnostic coverage and trend continuity.

[0066] When the cooling switch passes the verification, the probe heat input sequence and temperature rise sequence are extracted from the corresponding sub-time period; synchronous demodulation is performed to obtain the thermal frequency response estimate and coherence, and the effective frequency band is screened; when the effective frequency band meets the requirements, the standardized temperature rise curve is reconstructed; the standardized temperature rise curve is fitted and the thermal impedance parameter set is output, thereby obtaining two sets of thermal impedance parameters corresponding to the first cooling level and the second cooling level.

[0067] Cooling sensitivity is constructed using the difference in thermal resistance components between two cooling levels to show the degree of response of each thermal resistance component to cooling changes. The expression is as follows:

[0068] ;

[0069] in, This represents the first value obtained by fitting under the second cooling level. Each thermal resistance component coefficient This represents the first cooling level obtained by fitting the data. Each thermal resistance component coefficient This indicates the cooling characteristic value setting for the second cooling level. The cooling characteristic parameter setting value represents the first cooling level. Indicates the first The sensitivity of each thermal resistance component to changes in cooling settings.

[0070] When the sensitivity of a component is not less than the cooling sensitivity threshold, the component is determined to be the convection-dominant component; when the sensitivity of a component is less than the cooling sensitivity threshold, the component is determined to be the heat conduction-dominant component; the coefficients of the components determined to be convection-dominant are combined and the arithmetic sum is taken to obtain the convection link parameters; the coefficients of the components determined to be heat conduction-dominant are combined and the arithmetic sum is taken to obtain the heat conduction link parameters.

[0071] It should be noted that the cooling sensitivity threshold is calculated by performing two cooling level assessments multiple times when the motor is healthy and operating stably. The cooling sensitivity sample set obtained each time is calculated, and the corresponding quantile of the cooling sensitivity sample set is used as the cooling sensitivity threshold, such as the 95th quantile. If the cooling sensitivity threshold is less than the 95th quantile, a large number of components that should be dominated by heat conduction will also be judged as dominated by convection, resulting in incorrect attribution when cooling performance changes. It is easy to misjudge internal heat conduction degradation as external convection degradation. If the cooling sensitivity threshold is greater than the 95th quantile, a considerable number of normally dominated convection components will be judged as dominated by heat conduction. Therefore, when the external cooling capacity decreases, it may still be misattributed to internal heat conduction problems, and the sensitivity to external cooling anomalies will be significantly reduced.

[0072] S5. Calculate the heat dissipation deviation based on thermal frequency response estimation, effective frequency band screening data, convection link parameters, and heat conduction link parameters, and make multi-level judgments to output the pre-diagnosis level.

[0073] Within the frequency band, the heat dissipation deviation is calculated using the following expression:

[0074] ;

[0075] in, Indicates the degree of heat dissipation deviation. Indicates the upper limit of angular frequency. Indicates the lower limit of angular frequency. Indicates the effective frequency point. Represents angular frequency Estimated thermal frequency response at the location, Represents angular frequency Reference thermal frequency response at the location.

[0076] It should be noted that, This is achieved by repeatedly performing detection and synchronous demodulation under healthy and stable operating conditions during the baseline establishment phase. Multiple thermal frequency response estimates are obtained, and the arithmetic mean is taken point-by-point at the same discrete frequency to form a baseline reference curve. The upper limit of the angular frequency is achieved through... The ratio of the sampling period to the angular frequency is used to determine the lower limit of the angular frequency. The ratio of the length of the assessment period to the length of the assessment period is used to determine the assessment period.

[0077] When the heat dissipation deviation is greater than or equal to the heat dissipation anomaly threshold, the output is abnormal; when the heat dissipation deviation is less than the heat dissipation anomaly threshold, the output is normal.

[0078] It should be noted that the heat dissipation anomaly threshold is calculated by repeatedly performing heat dissipation deviation calculations under healthy and stable operating conditions to form a baseline deviation sample set. The corresponding quantile of the baseline deviation sample set is used as the heat dissipation anomaly threshold, such as the 95th quantile. If the heat dissipation anomaly threshold is less than the 95th quantile, the high deviation tail samples that should be judged as normal under the baseline healthy state will also be judged as abnormal, thus significantly increasing false alarms, causing frequent triggering of anomalies and reducing availability. If the heat dissipation anomaly threshold is greater than the 95th quantile, the deviation caused by some mild degradation or early heat dissipation degradation will still fall below the threshold and be judged as normal, resulting in missed alarms and weakening the early diagnosis.

[0079] To identify the type of degradation, the heat dissipation deviation was calculated for the first and second cooling sub-periods, and the heat dissipation deviation of the first and second cooling sub-periods was used as the improvement amount.

[0080] When the improvement is greater than or equal to the improvement threshold, the convection link parameter is greater than the baseline convection link parameter, and the heat conduction link parameter is less than or equal to the baseline heat conduction link parameter, external convection link degradation is output; when the improvement is less than the improvement threshold, the convection link parameter is less than or equal to the baseline convection link parameter, and the heat conduction link parameter is greater than the baseline heat conduction link parameter, internal heat conduction link degradation is output; if the convection link parameter is greater than the baseline convection link parameter and the heat conduction link parameter is greater than the baseline heat conduction link parameter, composite degradation is output and retesting is performed.

[0081] It should be noted that the improvement threshold is determined by repeatedly performing two cooling level assessments during the baseline establishment phase under healthy and stable operating conditions. The improvement in heat dissipation deviation before and after each cooling switch is calculated and a sample set is formed. The corresponding quantile of the sample set is used as the improvement threshold, such as the 95th quantile, with a value range of [0,2). If the value is lower than the 95th quantile, many moderate improvements caused only by noise, minor operating disturbances, or cooling feedback fluctuations will be considered significant improvements, leading to an increase in false alarms of external convection link degradation. If the value is higher than the 95th quantile, even if the cooling enhancement does bring significant improvement, it may be judged as insignificant improvement due to an overly strict threshold, causing situations attributable to external convection problems to be suppressed into internal heat conduction or compound degradation branches.

[0082] This embodiment also provides a motor heat dissipation performance evaluation system based on temperature rise curve fitting, including: a data acquisition module, which monitors the casing temperature and ambient temperature, calculates the temperature rise sequence, and simultaneously acquires current, speed and cooling characterization parameters to determine the evaluation period;

[0083] The probe injection module generates a probe sequence during the evaluation period and applies probe injection to the current command under the constraint of keeping the electromagnetic torque disturbance limited, thereby obtaining the probe thermal input sequence.

[0084] The synchronous demodulation module performs synchronous demodulation based on the detected thermal input sequence and temperature rise sequence to obtain thermal frequency response estimation and coherence. It generates an effective frequency band based on the coherence and reconstructs the standardized temperature rise curve.

[0085] The fitting and decomposition module fits the standardized temperature rise curve, obtains the thermal impedance parameters, sets the cooling characterization parameters to the first and second cooling levels in sequence and verifies the difference, calculates the cooling sensitivity, and decomposes the thermal impedance parameters into convection link parameters and heat conduction link parameters.

[0086] The deviation determination module calculates the heat dissipation deviation based on thermal frequency response estimation, effective frequency band screening data, convection link parameters, and heat conduction link parameters, and performs multi-level determination to output the pre-diagnosis level.

[0087] In summary, this invention generates a detection sequence during the evaluation period and applies a detection injection to the current command under the constraint of electromagnetic torque disturbance, thereby constructing an identifiable thermal excitation without significantly affecting the normal output of the motor. By applying two cooling levels sequentially during the same evaluation period and calculating the cooling sensitivity after the switching is achieved, and by decomposing the thermal impedance parameter into convection link parameters and heat conduction link parameters, attributable diagnosis of heat dissipation degradation is realized.

[0088] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting, characterized in that: include, Monitor the shell temperature and ambient temperature, calculate the temperature rise sequence, collect current, rotation speed and cooling characterization parameters, and determine the evaluation period; During the evaluation period, a detection sequence is generated, and a detection injection is applied to the current command under the constraint of keeping the electromagnetic torque disturbance limited, so as to obtain the detection thermal input sequence. Synchronous demodulation is performed based on the detected thermal input sequence and temperature rise sequence to obtain thermal frequency response estimation and coherence. Effective frequency bands are generated based on coherence, and standardized temperature rise curves are reconstructed. The standardized temperature rise curve is fitted to obtain the thermal impedance parameter. The cooling characterization quantity is set to the first cooling level and the second cooling level in sequence and the difference is verified. The cooling sensitivity is calculated and the thermal impedance parameter is decomposed into convection link parameter and heat conduction link parameter. Based on thermal frequency response estimation, effective frequency band screening data, convection link parameters, and heat conduction link parameters, the heat dissipation deviation is calculated and multi-level judgment is performed to output the pre-diagnosis level.

2. The method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting as described in claim 1, characterized in that: The determination of the evaluation period includes calculating the rotational speed stability index and the current stability index respectively within the stability evaluation window; When the speed stability index is not greater than the speed stability threshold and the current stability index is not greater than the current stability threshold, the evaluation period is determined. Obtain the highest permissible shell temperature and take the maximum value of the shell temperature sequence, calculate the temperature margin, and perform probe injection within the evaluation period based on the temperature margin.

3. The method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting as described in claim 2, characterized in that: The generation of the probe sequence during the evaluation period includes generating a binary probe sequence using a linear feedback shift register, wherein the binary probe sequence is updated once at the beginning of each symbol.

4. The method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting as described in claim 3, characterized in that: The method of applying probe injection to the current command while maintaining the constraint of limiting electromagnetic torque disturbance includes multiplying the probe sequence value and the injection amplitude as the d-axis injection amount in each symbol period and superimposing it on the reference. Shaft current command generation Axis detection injection; According to electromagnetic torque shaft current and The sensitivity ratio, derived from the partial derivatives of the shaft current, determines the compensation coefficient, and... Axis probe injection volume conversion Axis compensation injection volume.

5. The method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting as described in claim 4, characterized in that: The method of applying a probe injection to the current command while keeping the electromagnetic torque disturbance limited also includes calculating an estimated value of the change in torque command before and after the injection based on the internal torque. When the estimated value exceeds the upper limit of electromagnetic torque disturbance, the injection amplitude is reduced by a fixed ratio and the original code injection amount is regenerated. When the current command exceeds the limit, priority is given to reduction. Shaft injection volume and recalculation The shaft compensation injection amount is adjusted, and the probe injection is paused when the temperature margin is insufficient.

6. The method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting as described in claim 5, characterized in that: The acquisition of the probe thermal input sequence includes recording the actual d-axis current, q-axis current, and reference current during the evaluation period, reading the stator resistance estimate, and calculating the difference between the square of the injected current and the square of the reference current. The incremental copper loss caused by injection is used as the probe heat input sequence.

7. The method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting as described in claim 6, characterized in that: The synchronous demodulation includes performing a discrete Fourier transform on the probe heat input sequence and temperature rise sequence after removing the mean, and calculating the input autospectrum, output autospectrum and input-output cross spectrum. The ratio of cross spectrum to input autospectrum is used as the thermal frequency response estimate, and the coherence is calculated using cross spectrum, input autospectrum, and output autospectrum. The coherence and coherence threshold of each frequency point are compared to generate an effective frequency band, and the bandwidth covered by the effective frequency points is used as the effective frequency band metric. The reliability of synchronous demodulation is determined by comparing the effective frequency band measurement and the effective frequency band threshold.

8. The method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting as described in claim 7, characterized in that: The reconstructed standardized temperature rise curve includes constructing a frequency domain transfer function based on the thermal frequency response estimate at the effective frequency points and setting the values ​​at the ineffective frequency points to zero. After completing the sequence according to the conjugate symmetry rule, an inverse transformation is performed to obtain the discrete sequence of the impulse response. Then, the impulse response is accumulated and integrated to obtain the standardized temperature rise curve.

9. The method for evaluating the heat dissipation performance of a motor based on temperature rise curve fitting as described in claim 8, characterized in that: The calculation of cooling sensitivity and decomposition of thermal impedance parameters into convection link parameters and heat conduction link parameters includes expressing the standardized temperature rise curve as a second-order thermal impedance response form and performing least squares fitting to obtain a set of thermal impedance parameters. The cooling characterization parameters are set to the first cooling level and the second cooling level in sequence. After verifying that the difference between the first cooling level and the second cooling level reaches the threshold for cooling switching, the cooling sensitivity is calculated based on the difference in thermal impedance parameters under the first cooling level and the second cooling level. The thermal impedance parameters are decomposed by comparing the cooling sensitivity and the cooling sensitivity threshold to obtain the convection link parameters and the heat conduction link parameters.

10. A motor heat dissipation performance evaluation system based on temperature rise curve fitting, based on the motor heat dissipation performance evaluation method based on temperature rise curve fitting according to any one of claims 1 to 9, characterized in that: This includes a data acquisition module that monitors the casing temperature and ambient temperature, calculates the temperature rise sequence, collects current, rotational speed, and cooling parameters, and determines the evaluation period. The probe injection module generates a probe sequence during the evaluation period and applies probe injection to the current command under the constraint of keeping the electromagnetic torque disturbance limited, thereby obtaining the probe thermal input sequence. The synchronous demodulation module performs synchronous demodulation based on the detected thermal input sequence and temperature rise sequence to obtain thermal frequency response estimation and coherence. It generates an effective frequency band based on the coherence and reconstructs the standardized temperature rise curve. The fitting and decomposition module fits the standardized temperature rise curve, obtains the thermal impedance parameters, sets the cooling characterization parameters to the first and second cooling levels in sequence and verifies the difference, calculates the cooling sensitivity, and decomposes the thermal impedance parameters into convection link parameters and heat conduction link parameters. The deviation determination module calculates the heat dissipation deviation based on thermal frequency response estimation, effective frequency band screening data, convection link parameters, and heat conduction link parameters, and performs multi-level determination to output the pre-diagnosis level.

Images