Intelligent testing system and method for performance of motor of extractor hood

Abstract

This invention discloses an intelligent testing system and method for the performance of range hood motors, belonging to the field of performance testing technology. This intelligent testing method for range hood motor performance constructs a performance baseline model in a state free of grease and carbonization. By comparing real-time parameter acquisition with the performance baseline model, it initially determines whether the motor is abnormal. When an abnormality is initially determined, carbonization distribution is detected, and real-time parameters are corrected based on the influence of carbonization. After parameter correction, motor abnormality detection is performed again. This method can accurately distinguish between carbonization interference and inherent motor abnormalities, effectively improving the accuracy and reliability of motor performance testing. It avoids both false positives and false negatives, and can clearly define the degree of motor abnormality when a genuine abnormality exists. This solves the problem of inaccurate abnormality judgment caused by traditional range hood motor performance testing which did not consider the interference of grease carbonization on parameters such as temperature rise, vibration, and noise.

Description

Technical Field

[0001] This invention relates to the field of performance testing technology, specifically to an intelligent testing system and method for the performance of a range hood motor. Background Technology

[0002] As a core component in ensuring product quality reliability and safety, the accuracy of range hood motor performance testing directly determines the precision of motor fault diagnosis, product qualification standards, and user experience. In actual kitchen use scenarios, motors are constantly exposed to high-temperature oil fume environments, facing complex conditions such as grease carbonization accumulation, fluctuations in cooking conditions, and differences in usage time. Existing testing technologies reveal the following shortcomings: First, existing tests lack a quantitative correlation mechanism between grease carbonization and motor parameters, failing to match the dynamic interference scenarios of actual use. This leads to parameter deviations being misjudged as faults, resulting in significant discrepancies between laboratory-measured parameters such as temperature rise and vibration and actual operating data.

[0003] Secondly, the interference elimination and fault diagnosis stages in existing tests are independent of each other, lacking targeted carbonization interference correction logic, which significantly reduces the accuracy of anomaly detection. Current solutions only determine whether parameters are abnormal within a fixed standard range, without designing corresponding interference quantification and correction algorithms. This can easily lead to misjudging carbonization interference as a real fault, or causing missed detections because carbonization may mask the core fault.

[0004] Therefore, there is an urgent need for an intelligent motor performance testing technology that can quantify the degree of interference from grease carbonization, dynamically correct the test baseline, and accurately distinguish between interference and real faults, in order to solve the above-mentioned technical bottlenecks and improve the accuracy and reliability of motor performance testing under complex operating environments. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent testing system and method for the performance of range hood motors. This solves the problem that traditional range hood motor performance testing does not consider the interference of grease carbonization and adhesion on parameters such as temperature rise, vibration, and noise, leading to inaccurate anomaly detection.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent testing method for the performance of a range hood motor, comprising the following steps: constructing a performance baseline model of the range hood motor in a state without grease carbonization.

[0007] Collect temperature rise distribution data, vibration characteristic parameters, and noise characteristic parameters of the range hood motor under set operating conditions.

[0008] The parameters collected in real time are compared with the performance baseline model to determine whether there is any abnormality in the range hood motor.

[0009] If any abnormality is found, the carbonization distribution of the range hood motor will be tested to determine the carbonization distribution characteristics.

[0010] Based on the carbonization distribution characteristics and the influence of carbonization on the temperature rise conduction of the motor, the parameters collected in real time are corrected to obtain the corrected performance parameters.

[0011] The corrected performance parameters are compared again with the performance baseline model. If the range hood motor still has abnormalities, the degree of motor abnormality is determined; if there are no abnormalities, the range hood motor is judged to be in good condition.

[0012] Compared with existing technologies, the present invention has the following beneficial effects: By constructing a performance baseline model of a non-grease carbonization adhesion state, and combining real-time parameter acquisition and baseline comparison, carbonization distribution detection, and parameter correction based on the carbonization influence law, the present invention can accurately distinguish between carbonization interference and motor abnormalities, effectively improving the accuracy and reliability of motor performance testing. It avoids both false positives and false negatives, and can clearly identify the degree of motor abnormality when real abnormalities exist. It solves the problem that traditional range hood motor performance testing does not consider the interference of grease carbonization adhesion on parameters such as temperature rise, vibration, and noise, leading to inaccurate abnormality judgment. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the intelligent testing method for the performance of a range hood motor according to the present invention.

[0014] Figure 2 This is a flowchart showing the corrected performance parameters obtained in the intelligent testing method for the motor performance of a range hood according to the present invention.

[0015] Figure 3 This is a flowchart for determining the abnormality of a motor in the intelligent testing method for the performance of a range hood motor according to the present invention.

[0016] Figure 4 This is a flowchart showing the module connection of the intelligent testing system for the performance of a range hood motor according to the present invention. Detailed Implementation

[0017] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. Please refer to the accompanying drawings. Figure 1 The present invention provides a technical solution: an intelligent testing method for the performance of a range hood motor, comprising the following steps: S1, constructing a performance baseline model of the range hood motor in a state without grease carbonization.

[0018] To ensure that the pure performance baseline in the non-carbonized state can truly reflect the original normal performance of the motor and support subsequent efficient and accurate parameter comparison, and considering that the temperature rise, vibration and noise characteristics of different parts of the motor are significantly different, a single-point sensor cannot reflect the global state.

[0019] In the initial state without grease or carbonization, temperature, vibration, and noise data of various parts of the motor are detected by temperature sensor array, vibration sensor array, and sound sensor array arranged on the motor after startup, and temperature curves, vibration curves, and noise curves of various parts of the motor are obtained.

[0020] It should be noted that the temperature curve reflects the temperature change trend over time; the vibration curve reflects the vibration amplitude and frequency changes over time; and the noise curve reflects the noise level in decibels and the frequency spectrum changes over time.

[0021] For example, the temperature sensor array includes eight thermocouple sensors, which are respectively arranged at the upper end of the motor stator winding, the lower end of the stator winding, the rotor shoulder, the impeller hub, the air inlet side of the casing, the air outlet side of the casing, the inner side of the bearing end cover, and the outer side of the bearing end cover, with a sampling frequency of 10Hz; the vibration sensor array includes four piezoelectric accelerometer sensors, which are respectively arranged at the impeller hub, the middle of the casing, and the two bearing end covers, with a sampling frequency of 1kHz; the sound sensor array includes two capacitive microphones, which are respectively arranged 10cm from the air inlet side and 10cm from the air outlet side of the motor, with a sampling frequency of 44.1kHz.

[0022] By adopting an array-style deployment, parameters of all critical areas of the motor can be fully covered, avoiding the baseline model from missing the normal range due to incomplete data acquisition. This ensures that during subsequent real-time comparisons, abnormalities can be comprehensively judged from temperature rise, vibration, and noise.

[0023] Considering the differences in performance parameters of motors at different operating stages, directly using a unified standard for the entire cycle would lead to serious distortion of the baseline range. Therefore, it is necessary to divide the curves by time windows to establish standards in stages and avoid using a single standard to cover the entire operating cycle. According to the set time windows, such as 10 seconds per time window, the temperature curve, vibration curve, and noise curve are divided into temperature curve segments, vibration curve segments, and noise curve segments respectively.

[0024] By segmenting the time window, a standard range can be established separately for the operating characteristics of each time period, ensuring that the baseline model can accurately match the phased normal fluctuations of the motor's dynamic operation and avoid deviations in subsequent abnormal judgments due to phase mismatch.

[0025] Since the original temperature, vibration, and noise curves are continuous time-domain data containing a large amount of redundant information, directly comparing the curves with the baseline is not only computationally intensive and inefficient, but also prone to misjudging anomalies due to random fluctuations. Therefore, it is necessary to extract feature parameters, performing parameter analysis on the temperature, vibration, and noise curve segments separately to obtain temperature, vibration, and noise feature parameters. Furthermore, feature parameters of the same type and location are arranged according to the time window sequence to obtain the corresponding temperature, vibration, and noise feature parameter sequences for each location.

[0026] Continuous dynamic data is transformed into a set of feature parameters arranged in an ordered manner along the time dimension, providing data support for the subsequent establishment of time-segmented standard ranges. For example, temperature feature parameters include, but are not limited to, the mean, peak value, rate of rise, rate of fall, and steady-state value of the curve segment; vibration feature parameters include, but are not limited to, the peak amplitude, root mean square value, and characteristic frequency components of the curve segment; noise feature parameters include, but are not limited to, the equivalent A-weighted sound level, peak decibel, and noise energy in a specific frequency range of the curve segment.

[0027] For three types of characteristic parameter sequences at the same location, statistical methods are used to determine the reasonable range of normal parameters within each time window, avoiding baseline bias caused by a single standard. For multiple sets of temperature characteristic parameter sequences corresponding to a certain location, it is verified whether a certain parameter conforms to a normal distribution. If it does, the standard temperature range for the corresponding time window is determined by plus or minus twice the standard deviation of the parameter's mean. If it does not, the standard temperature range for the corresponding time window is determined by the interquartile range method.

[0028] The vibration characteristic parameter sequence and noise characteristic parameter sequence corresponding to the same location were verified by normal distribution to obtain the vibration standard range and noise standard range corresponding to each time window.

[0029] Statistical methods are used to ensure that the standard range of each time window is the normal fluctuation range of the motor in the non-carbonized state during that period, without omitting real normal data or including abnormal data.

[0030] The performance baseline model is obtained by summarizing the temperature standard range, vibration standard range, and noise standard range corresponding to each time window for each part.

[0031] S2. Collect temperature rise distribution data, vibration characteristic parameters and noise characteristic parameters of the range hood motor under set operating conditions.

[0032] Specifically, temperature, vibration, and noise data from various parts of the motor are detected by arrays of temperature, vibration, and sound sensors mounted on the motor after startup. After preprocessing the collected data, feature parameters are extracted to obtain temperature rise distribution data, vibration characteristic parameters, and noise characteristic parameters.

[0033] S3. Compare the real-time collected parameters with the performance baseline model to determine if there are any abnormalities in the range hood motor.

[0034] If real-time data is not divided into the same time windows, time mismatch will occur. Therefore, it is necessary to adopt the same time window rules as when building the baseline model to divide the continuously collected temperature rise distribution data, vibration characteristic parameters, and noise characteristic parameters into discrete real-time data segments. This avoids comparison deviations caused by time window misalignment. The parameters collected in real time are divided according to time windows, and for each real-time data segment, the same parameter parsing method is used to extract real-time feature values ​​that match the baseline feature parameter types, thus obtaining the temperature feature values, vibration feature values, and noise feature values ​​corresponding to each time window.

[0035] Based on data dimension alignment, the temperature characteristic values, vibration characteristic values, and noise characteristic values ​​of each time window are compared with the temperature standard range, vibration standard range, and noise standard range of the corresponding time window in the performance baseline model.

[0036] The normal range of motor parameters exhibits spatial, temporal, and index-specific characteristics. Only through precise point-to-point matching can the location, time period, and index of anomalies be accurately identified, providing precise guidance for subsequent carbonization testing. Each real-time feature value is compared with its corresponding baseline standard range one by one. Through point-to-point comparison, potentially abnormal feature values ​​are initially screened out.

[0037] Considering the sporadic nature of parameter fluctuations, if the traditional testing logic judges anomalies based on a single instance of exceeding the range, it would trigger a large number of unnecessary carbonization tests, wasting resources and misleading judgments. However, parameter anomalies caused by oil carbonization or actual faults are persistent. If the temperature characteristic value, vibration characteristic value, and noise characteristic value are outside the temperature standard range, vibration standard range, and noise standard range of the corresponding time window of the performance baseline model for N consecutive time windows, such as 3 consecutive time windows, then the abnormal temperature value, abnormal vibration value, and abnormal noise value will each be incremented by 1.

[0038] The abnormal values ​​of temperature, vibration, and noise are weighted and summed to obtain a comprehensive abnormal value. If the comprehensive abnormal value is greater than the set abnormal threshold, the range hood motor is abnormal.

[0039] S4. If any abnormality is found, the carbonization distribution of the range hood motor shall be detected to determine the carbonization distribution characteristics.

[0040] To initially identify potentially carbonized areas from visual images and avoid inefficiency due to excessively large detection areas, an industrial area scan camera, paired with a ring LED light source, was used to capture images of key parts of the motor in sections. The visual inspection images of the range hood motor were then processed for grayscale, and areas with grayscale values ​​below a set threshold were filtered out to obtain preliminary suspected areas.

[0041] It should be noted that the grayscale threshold is determined by the sample training method. 100 sets of visual images of normal motors without carbonization are collected, and the normal grayscale range of their key parts is counted. 80% of the minimum value of the normal range is taken as the grayscale threshold. For example, if the normal minimum grayscale is 150, the threshold is set to 120.

[0042] For each initially suspected region, the total number of pixels it contains is counted using a connected component analysis algorithm and recorded as the pixel area. The correspondence between pixels and actual size is obtained through camera calibration, that is, a calibration image is taken using a standard checkerboard pattern, and the actual area corresponding to 1 pixel is calculated. Then, the actual area of ​​each initially suspected region is obtained by multiplying the pixel area by the actual pixel area. Initially suspected regions with areas smaller than a set area threshold are filtered out to obtain carbonized suspected regions.

[0043] The suspected carbonized areas are correlated and verified to determine the actual carbonized areas and their detection coordinates, and a unique carbonized area ID is assigned to each actual carbonized area.

[0044] The infrared temperature difference, ultrasonic attenuation, and dielectric constant deviation of the actual carbonized region are attached to the corresponding carbonized region ID in the form of key-value pairs to obtain the carbonization distribution characteristics.

[0045] It should be noted that the infrared temperature difference acquisition process is as follows: An infrared thermal imager is used to collect the average temperature T1 of the suspected carbonized area, and simultaneously, the average temperature T2 of the normal area within a 5mm radius of this area is collected. The difference between T1 and T2 is calculated, with the unit being °C. The ultrasonic attenuation acquisition process is as follows: The probe of the ultrasonic flaw detector is attached to the surface of the suspected carbonized area, ultrasonic waves are emitted and reflected waves are received, and the propagation attenuation A1 of the ultrasonic waves within the area is recorded. This process is repeated in adjacent normal areas, and the attenuation A2 is recorded. The difference between A1 and A2 is taken as the ultrasonic attenuation, which is dimensionless. The propagation attenuation A1 and attenuation A2 are defined as the logarithmic ratio of the incident ultrasonic energy to the transmitted or reflected ultrasonic energy. The dielectric constant deviation acquisition process is as follows: The probe of a portable dielectric constant tester is attached to the suspected carbonized area, and its dielectric constant M1 is measured. The dielectric constant M0 is measured in the normal area, and the absolute value of the difference between M1 and M0 is taken as the dielectric constant deviation, which is dimensionless.

[0046] The process of correlating and verifying suspected carbonization areas to determine the actual carbonization areas and their detection coordinates is as follows: Temperature data is collected from the suspected carbonization areas, and the temperature difference between the collected data and the temperatures of non-suspected areas is calculated to obtain the temperature difference value. The acoustic attenuation and dielectric constant of the suspected carbonization areas are recorded.

[0047] If the temperature difference, acoustic attenuation, and dielectric constant of a suspected carbonized region meet at least two of the set conditions, it is determined to be a real carbonized region. Meeting at least two of the above conditions will result in a real carbonized region, thus avoiding misjudgment based on a single parameter. If only 0-1 conditions are met, it is determined to be an interference region and will be eliminated.

[0048] Using the lens optical center of dynamic vision as the origin of the detection coordinates (0, 0, 0), the intrinsic and extrinsic parameters of the lens are obtained through the camera calibration algorithm. The pixel coordinates of the real carbonized area are converted into three-dimensional coordinates in the world coordinate system to obtain the detection coordinates of the real carbonized area.

[0049] Specifically: A world coordinate system is established, with the X-axis representing the motor axis, the Y-axis representing the motor radial direction, and the Z-axis perpendicular to the motor surface; a checkerboard calibration method is used to capture 10 sets of checkerboard images from different angles, and the camera intrinsic and extrinsic parameters are calculated using calibration algorithms, such as the Zhang Zhengyou calibration method; the pixel coordinates of the real carbonized area are extracted from the visual images, and combined with the camera intrinsic and extrinsic parameters and the depth information of the motor surface, the three-dimensional coordinates of the real carbonized area in the world coordinate system are calculated using the perspective projection formula.

[0050] In order to transform the differences in the physical properties of carbonization into quantifiable and actionable criteria, it should be noted that the set conditions include: the temperature difference is greater than the set temperature difference threshold; the ultrasonic attenuation is greater than the set attenuation threshold; and the deviation of the dielectric constant from the reference value is greater than the set deviation.

[0051] The reason for setting a temperature difference value greater than the set temperature difference threshold is that when the motor is running, the carbonized area will continuously accumulate heat because heat cannot be quickly conducted, forming a stable temperature difference with the surrounding normal area. However, the presence of a temperature difference does not necessarily mean carbonization. Therefore, a threshold needs to be set to determine what temperature difference is consistent with the heat accumulation characteristics of carbonization. The temperature difference threshold is obtained by statistically analyzing real carbonized samples and non-carbonized interference samples to ensure that the heat accumulation of carbonization and the temperature difference of non-carbonization can be effectively distinguished.

[0052] The reason for setting the condition that the ultrasonic attenuation is greater than a set attenuation threshold is that carbonization is a porous solid formed after the combustion of grease. When ultrasonic waves propagate in it, the energy is greatly attenuated due to reflection and scattering through the pores. In contrast, normal metals and insulating materials have a dense structure, and the ultrasonic attenuation is minimal. However, attenuation does not necessarily mean carbonization. Therefore, a threshold needs to be set to determine how much attenuation is considered sufficient for the porous structure of carbonization.

[0053] The reason for setting the dielectric constant to deviate from the reference value to be greater than the set deviation condition is that the dielectric constant of carbonization is significantly different from that of motor insulation paint and metal parts. However, deviation does not necessarily mean carbonization. Therefore, it is necessary to set a threshold to determine how much deviation is required to meet the dielectric characteristics of carbonization.

[0054] S5. Based on the carbonization distribution characteristics and the influence of carbonization on motor temperature rise conduction, the real-time collected parameters are corrected to obtain the corrected performance parameters, as detailed below. Figure 2 As shown: A parameter influence model is constructed based on the influence law of carbonization on the temperature rise conduction of motor. Specifically: In order to extract the direct correspondence between carbonization characteristic parameters and motor parameter deviations from massive historical carbonization influence data, the input and output logic of the model is clarified, and irrelevant data is avoided from interfering with the accuracy of the model.

[0055] From historical carbonization impact data, we screened out the correlations between the combined characteristics of infrared temperature difference and ultrasonic attenuation and temperature deviation, the combined characteristics of dielectric constant deviation and ultrasonic attenuation and vibration deviation, the combined characteristics of infrared temperature difference and ultrasonic attenuation and aerodynamic noise deviation, and the combined characteristics of dielectric constant deviation and ultrasonic attenuation and vibration transmission noise deviation.

[0056] Using the combined characteristics of infrared temperature difference and ultrasonic attenuation, and the combined characteristics of dielectric constant deviation and ultrasonic attenuation as input variables, and the corresponding temperature deviation and vibration deviation as output variables, the least squares method is used to perform nonlinear curve fitting to establish temperature deviation law models and vibration deviation law models.

[0057] Taking the combined characteristics of infrared temperature difference and ultrasonic attenuation as an example to obtain a temperature deviation law model, the fitting process is as follows: the filtered temperature deviation mapping data is divided into training set and test set in an 8:2 ratio.

[0058] Define a quadratic nonlinear function: ,in, For temperature deviation, Infrared temperature difference, This refers to the attenuation of the ultrasonic wave. For interactive items, , and They are coefficients, This is a constant term.

[0059] Substitute the training set data into the function and minimize the prediction using the least squares method. With reality The residual sum of squares is obtained by solving for the coefficients using MATLAB's lsqcurvefit or Python's scipy.optimize.curve_fit.

[0060] Substituting the infrared temperature difference and ultrasonic attenuation of the test set into the model, the prediction is calculated. With reality The error must be such that the mean absolute error is less than a set threshold. In this embodiment, a = 0.02℃ -1 , b=0.03℃, c=0.01, d=0.1℃.

[0061] Similarly, the method for establishing the vibration deviation law model is the same as above.

[0062] Using infrared temperature difference, ultrasonic attenuation, and dielectric constant deviation as input variables, and the weighted sum of aerodynamic noise deviation and vibration transmission noise deviation as output variables, a neural network fitting method is used to establish a noise influence law model.

[0063] It should be noted that the process of establishing the noise influence law model is as follows: Since the noise deviation is a weighted sum of aerodynamic noise and vibration transmission noise, it is affected by the interaction of infrared temperature difference, ultrasonic attenuation, and dielectric constant deviation. Traditional linear fitting cannot capture the implicit correlation, while BP neural network is good at handling complex nonlinear relationships with multiple inputs and strong coupling.

[0064] The specific BP network architecture is as follows: the input layer has three neurons, corresponding to the infrared temperature difference, ultrasonic attenuation, and dielectric constant deviation, respectively; the first hidden layer has 10 neurons, the second hidden layer has 10 neurons, and the output layer has 2 neurons, corresponding to the aerodynamic noise deviation and vibration conduction noise deviation, respectively. The activation function for the hidden layer neurons is the ReLU function, and the activation function for the output layer is the Sigmoid function.

[0065] The filtered noise bias mapping data was divided into training and test sets in a 7:3 ratio. The training set was then further divided into sub-training and validation sets in a 9:1 ratio. The loss function was set to mean squared error. The Adam optimizer was used with a learning rate of 0.001 and 1000 iterations, with validation performed every 100 iterations. To avoid overfitting, an L2 regularization term with a coefficient of 0.001 was introduced. The iteration stopped when the mean squared error of the validation set was less than 0.005.

[0066] The parameter influence model is obtained by storing the temperature deviation law model, vibration deviation law model, and noise influence law model in the database.

[0067] The classification design avoids the coarseness of a single model covering all biases, ensuring that each mapping relationship directly corresponds to the carbonization effect mechanism, and the correction accuracy is greatly improved.

[0068] The carbonization distribution characteristics are input into the parameter influence model to output temperature deviation, vibration deviation, and noise deviation.

[0069] The real-time collected temperature, vibration, and noise data from each part are superimposed with the corresponding temperature, vibration, and noise deviations to offset the parameter distortion caused by carbonization and obtain the corrected performance parameters.

[0070] S6. Compare the corrected performance parameters with the performance baseline model again. If the range hood motor is still abnormal, determine the degree of motor abnormality; if there is no abnormality, determine that the range hood motor is in good condition.

[0071] The process of determining whether the range hood motor is still malfunctioning is the same as the first determination, and will not be repeated here.

[0072] like Figure 3 As shown, the process of determining the motor anomaly degree is as follows: for each time window, determine whether the parameters in the corrected performance parameters are within the standard range corresponding to the performance baseline model.

[0073] If it falls within the corresponding standard range, then the anomaly factor corresponding to this parameter is 0.

[0074] If it is not within the corresponding standard range, calculate the half-width of the corresponding standard range, calculate the absolute value of the deviation between the parameter and the median value of the corresponding standard range, and use the ratio of the absolute value of the deviation to the half-width as the anomaly factor corresponding to the parameter.

[0075] Because the baseline standard ranges for different parameters vary, directly comparing out-of-range values ​​cannot objectively reflect the severity. Abnormal factors are uniformly measured through relative deviation, avoiding judgment imbalances caused by standard differences. Instead of simply judging whether something is out of range, the ratio of the absolute value of the deviation to the half-width of the standard range intuitively reflects the extent to which a single parameter deviates from the normal state, making the severity of the abnormality quantifiable and comparable.

[0076] The motor anomaly degree for the current time window is obtained by weighted summation of all anomaly factors.

[0077] By integrating the effects of all abnormal parameters, rather than drawing conclusions based on just one parameter, the motor condition assessment becomes more comprehensive and more in line with actual operating conditions.

[0078] A smart testing system for the performance of range hood motors, such as Figure 4 As shown, it includes a baseline model building module, a data acquisition module, a motor anomaly judgment module, a carbonization feature determination module, a carbonization feature determination module, and a secondary motor anomaly judgment module. Among them, the baseline model building module is used to build a performance baseline model of the range hood motor in a state without grease carbonization adhesion.

[0079] The data acquisition module is used to collect temperature rise distribution data, vibration characteristic parameters, and noise characteristic parameters of the range hood motor under set operating conditions.

[0080] The motor anomaly detection module is used to compare the various parameters collected in real time with the performance baseline model to determine whether there is an anomaly in the range hood motor.

[0081] The carbonization characteristic determination module is used to detect the carbonization distribution of the range hood motor and determine its carbonization distribution characteristics when there is an abnormality in the motor.

[0082] The parameter correction module is used to correct various parameters collected in real time based on the carbonization distribution characteristics and the influence of carbonization on the temperature rise conduction of the motor, so as to obtain the corrected performance parameters.

[0083] The secondary motor anomaly detection module is used to compare the corrected performance parameters with the performance baseline model again. If the range hood motor still has anomalies, the degree of motor anomaly is determined; if there are no anomalies, the range hood motor is determined to be in good working order.

[0084] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0085] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0086] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0087] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0088] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent testing of the performance of a range hood motor, characterized in that, Includes the following steps: Construct a performance baseline model for the range hood motor under a state without grease carbonization adhesion; Collect temperature rise distribution data, vibration characteristic parameters, and noise characteristic parameters of the range hood motor under set operating conditions; The parameters collected in real time are compared with the performance baseline model to determine whether there is any abnormality in the range hood motor; If any abnormality is found, the carbonization distribution of the range hood motor will be tested to determine the carbonization distribution characteristics. Based on the carbonization distribution characteristics and the influence of carbonization on the temperature rise conduction of the motor, the parameters collected in real time are corrected to obtain the corrected performance parameters. The corrected performance parameters are compared with the performance baseline model again. If the range hood motor is still abnormal, the degree of motor abnormality is determined. If no abnormalities are found, the range hood motor is considered to be in good working order. The process of constructing a performance baseline model for a range hood motor in a state without grease carbonization is as follows: Based on the temperature sensor array, vibration sensor array and sound sensor array arranged on the motor, the temperature data, vibration data and noise data of various parts of the motor are detected after the motor starts, and the temperature curve, vibration curve and noise curve of various parts of the motor are obtained. The temperature curve, vibration curve, and noise curve are divided into temperature curve segments, vibration curve segments, and noise curve segments according to the set time window. The temperature curve segment, vibration curve segment, and noise curve segment are analyzed separately to obtain temperature characteristic parameters, vibration characteristic parameters, and noise characteristic parameters. These parameters are then arranged according to the order of the time windows to obtain the corresponding temperature characteristic parameter sequence, vibration characteristic parameter sequence, and noise characteristic parameter sequence for each part. For a certain part, multiple temperature characteristic parameter sequences are used to verify whether a certain parameter conforms to a normal distribution. If it does, the standard range of temperature for the corresponding time window is taken as the mean plus or minus two standard deviations of the parameter. If it does not, the standard range of temperature for the corresponding time window is determined by the interquartile range method. The vibration characteristic parameter sequence and noise characteristic parameter sequence corresponding to the same location were verified by normal distribution to obtain the vibration standard range and noise standard range corresponding to each time window. The performance baseline model is obtained by summarizing the temperature standard range, vibration standard range and noise standard range corresponding to each time window for each part. The process of detecting carbonization distribution in the motor of a range hood and determining its characteristics is as follows: The visual inspection image of the range hood motor is processed in grayscale, and areas with grayscale values ​​less than a set grayscale threshold are filtered out to obtain preliminary suspected areas; Preliminary suspected areas with an area smaller than a set area threshold are screened out to obtain suspected carbonization areas. The suspected carbonized areas are correlated and verified to determine the actual carbonized areas and their detection coordinates, and a unique carbonized area ID is assigned to each actual carbonized area. The infrared temperature difference, ultrasonic attenuation, and dielectric constant deviation of the actual carbonized region are attached to the corresponding carbonized region ID in the form of key-value pairs to obtain the carbonization distribution characteristics. The process of correcting the real-time collected parameters based on the carbonization distribution characteristics and the influence of carbonization on the temperature rise conduction of the motor, to obtain the corrected performance parameters, is as follows: A parameter influence model was constructed based on the influence of carbonization on the temperature rise conduction of motors. Input the carbonization distribution characteristics into the parameter influence model to output temperature deviation, vibration deviation, and noise deviation; The real-time collected temperature, vibration, and noise data of each part are superimposed with the corresponding temperature deviation, vibration deviation, and noise deviation to obtain the corrected performance parameters. The process of constructing the parameter influence model based on the influence of carbonization on the conduction of motor temperature rise is as follows: From historical carbonization impact data, we screened out the correspondence between the combined characteristics of infrared temperature difference and ultrasonic attenuation and temperature deviation, the combined characteristics of dielectric constant deviation and ultrasonic attenuation and vibration deviation, the combined characteristics of infrared temperature difference and ultrasonic attenuation and aerodynamic noise deviation, and the combined characteristics of dielectric constant deviation and ultrasonic attenuation and vibration transmission noise deviation. Using the combined characteristics of infrared temperature difference and ultrasonic attenuation, and the combined characteristics of dielectric constant deviation and ultrasonic attenuation as input variables, and the corresponding temperature deviation and vibration deviation as output variables, the least squares method is used to perform nonlinear curve fitting to establish temperature deviation law models and vibration deviation law models. Using infrared temperature difference, ultrasonic attenuation, and dielectric constant deviation as input variables, and the weighted sum of aerodynamic noise deviation and vibration transmission noise deviation as output variables, a neural network fitting method is used to establish a noise influence law model. The parameter influence model is obtained by storing the temperature deviation law model, vibration deviation law model, and noise influence law model in the database.

2. The intelligent testing method for the performance of a range hood motor according to claim 1, characterized in that, The process of comparing the real-time collected parameters with the performance baseline model to determine whether there is any abnormality in the range hood motor is as follows: The parameters collected in real time are divided into time windows and analyzed to obtain the temperature characteristic value, vibration characteristic value and noise characteristic value corresponding to each time window; The temperature characteristic values, vibration characteristic values, and noise characteristic values ​​of each time window are compared with the temperature standard range, vibration standard range, and noise standard range of the corresponding time window in the performance baseline model. If the temperature characteristic value, vibration characteristic value, and noise characteristic value are outside the temperature standard range, vibration standard range, and noise standard range of the corresponding time window of the performance baseline model for N consecutive time windows, then the abnormal temperature value, abnormal vibration value, and abnormal noise value are each incremented by 1; The abnormal values ​​of temperature, vibration, and noise are weighted and summed to obtain a comprehensive abnormal value. If the comprehensive abnormal value is greater than the set abnormal threshold, the range hood motor is abnormal.

3. The intelligent testing method for the performance of a range hood motor according to claim 1, characterized in that, The process of correlating and verifying suspected carbonized areas to determine the actual carbonized areas and their detection coordinates is as follows: Temperature data was collected from suspected carbonization areas, and the temperature difference between the temperature data and the temperature of non-suspected areas was calculated to obtain the temperature difference value. Record the acoustic attenuation and dielectric constant of the suspected carbonized region; If the temperature difference, acoustic attenuation, and dielectric constant of a suspected carbonized region meet at least two of the set conditions, it is determined to be a real carbonized region. Using the lens optical center of dynamic vision as the origin of the detection coordinates, the intrinsic and extrinsic parameters of the lens are obtained through the camera calibration algorithm. The pixel coordinates of the real carbonized area are converted into three-dimensional coordinates in the world coordinate system to obtain the detection coordinates of the real carbonized area.

4. The intelligent testing method for the performance of a range hood motor according to claim 3, characterized in that, The setting conditions include: The temperature difference exceeds the set temperature difference threshold. The ultrasonic wave attenuation is greater than the set attenuation threshold. The deviation of the dielectric constant from the reference value is greater than the set deviation.

5. The intelligent testing method for the performance of a range hood motor according to claim 1, characterized in that, If the range hood motor still exhibits abnormalities, the process for determining the degree of motor abnormality is as follows: For each time window, determine whether the parameters in the corrected performance parameters are within the standard range corresponding to the performance baseline model; If it is within the corresponding standard range, then the anomaly factor corresponding to this parameter is 0; If it is not within the corresponding standard range, calculate the half-width of the corresponding standard range, calculate the absolute value of the deviation between the parameter and the median value of the corresponding standard range, and use the ratio of the absolute value of the deviation to the half-width as the outlier factor corresponding to the parameter. The motor anomaly degree for the current time window is obtained by weighted summation of all anomaly factors.

6. A smart testing system for the performance of a range hood motor, used in the smart testing method for the performance of a range hood motor as described in any one of claims 1-5, comprising: The baseline model building module is used to build a performance baseline model of the range hood motor in a state without grease carbonization. The data acquisition module is used to collect temperature rise distribution data, vibration characteristic parameters, and noise characteristic parameters of the range hood motor under set operating conditions; The motor anomaly detection module is used to compare the various parameters collected in real time with the performance baseline model to determine whether there is an anomaly in the range hood motor. The carbonization characteristic determination module is used to detect the carbonization distribution of the range hood motor and determine the carbonization distribution characteristics when there is an abnormality in the range hood motor. The parameter correction module is used to correct various parameters collected in real time based on the carbonization distribution characteristics and the influence of carbonization on the temperature rise conduction of the motor, so as to obtain the corrected performance parameters. The motor anomaly secondary judgment module is used to compare the corrected performance parameters with the performance baseline model again. If the range hood motor is still abnormal, the degree of motor anomaly is determined. If no abnormalities are found, the range hood motor is considered to be in good working order.

Images