Method and device for testing comprehensive performance of refrigerators

Abstract

The application relates to the field of refrigerators, and discloses a refrigerator comprehensive performance test method and device, the method comprising the following steps: S1, electronic tag identification; S2, collecting data and forming time sequence data; S3, calculating a test characteristic vector of the refrigerator to be tested; S4, adaptively controlling a subsequent test process; and S5, sorting after detection is completed; the device comprises a refrigerator, a first conveying belt, a detection conveying belt, a second conveying belt, a detection sensor, an electronic tag and a central control console. The exponential weighted moving average algorithm is used to dynamically correct a reference characteristic vector and a reference covariance matrix, so that the central control console system can self-learn and adapt to different production processes. The accuracy and adaptability of the test are improved, and the rationality and effectiveness of the refrigerator test standard can be maintained for a long time.

Description

Technical Field

[0001] This invention relates to the field of refrigerator technology, specifically to a method and apparatus for comprehensive performance testing of refrigerators. Background Technology

[0002] With the increasing demand for cold chain logistics and household refrigeration, performance testing technology for refrigerated display cases has gradually developed into an important research field. Currently, various performance testing methods and devices for refrigerated display cases have emerged on the market. These technologies typically use specialized sensors to collect data and monitor and analyze multiple performance parameters such as temperature, power, and energy consumption. Furthermore, companies have begun to apply electronic tag technology in the refrigerated display case production process to facilitate product information traceability and simplify testing procedures, thereby improving production efficiency.

[0003] While existing technologies have made some progress in refrigerated display case performance testing, limitations remain in areas such as real-time data analysis, adaptive control strategies, and dynamic updates to testing standards. These limitations not only affect testing efficiency and accuracy but also make it difficult to guarantee the adaptability and stability of refrigerated display cases under different conditions. Therefore, there is an urgent need for improved testing methods and devices to effectively address these technical requirements and industry challenges, and further enhance the comprehensive performance evaluation capabilities of refrigerated display cases. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a comprehensive performance testing method and apparatus for refrigerators, which solves the problems of low data acquisition efficiency, insufficient adaptability of the testing process, and inaccurate updating of the reference model in refrigerator performance testing.

[0005] To achieve the above objectives, the present invention provides a comprehensive performance testing method for refrigerated display cases, comprising the following steps:

[0006] S1. The refrigerator to be tested is transported from the first conveyor belt. The electronic tag attached to the refrigerator is identified by the detection sensor, and then it is transported to one of the multiple detection conveyor belts.

[0007] S2. On the testing conveyor belt, within a preset testing period, collect multiple performance parameters of the refrigerator under test to form time-series data;

[0008] S3. Calculate the test feature vector of the refrigerator under test based on the time series data;

[0009] S4. Compare the test feature vector with the pre-established reference feature vector and reference covariance matrix to obtain a deviation degree, and adaptively control the subsequent test process based on the deviation degree.

[0010] S5. After the inspection is completed, the refrigerated cabinets are transferred to the second conveyor belt for sorting.

[0011] Preferably, in step S1, the information of the electronic tag includes:

[0012] The unique identification code of the refrigerator;

[0013] The product model and specifications of the refrigerator;

[0014] The production batch number of the refrigerated cabinet;

[0015] The model information of the core components used in the refrigerator.

[0016] Preferably, in step S2, the performance parameters of the refrigerator to be tested include:

[0017] The internal temperature of the refrigerator;

[0018] The ambient temperature for monitoring the refrigerated display case;

[0019] The voltage of the refrigerator;

[0020] The current of the refrigerator;

[0021] The power of the refrigerator.

[0022] Preferably, in step S2, the process of forming time-series data includes the following steps:

[0023] S21. Deploy the temperature sensing probe used to collect the internal temperature to a preset position inside the refrigerator to be tested;

[0024] S22. During the preset test period, the instantaneous values ​​of the multiple performance parameters are collected synchronously at a preset sampling frequency.

[0025] S23. Associate the collected instantaneous values ​​with the corresponding timestamps to generate the time series data.

[0026] Preferably, in step S3, the feature vector includes:

[0027] Initial cooling rate;

[0028] Thermodynamic characteristic time constant;

[0029] Cooling efficiency per unit of energy consumption.

[0030] Preferably, the unit energy consumption cooling efficiency is calculated as follows:

[0031] ;

[0032] In the formula, This represents the absolute value of the temperature change inside the refrigerator during a preset test period. This is the integral of the refrigerator's power during the preset test period.

[0033] Preferably, the adaptive control of subsequent testing procedures includes the following steps:

[0034] S41. If the deviation is less than the first preset threshold, the refrigerator under test is determined to be a qualified product and the test is terminated.

[0035] S42. If the deviation is greater than or equal to the first preset threshold and less than the second preset threshold, then perform an extended test on the refrigerator under test.

[0036] S43. If the deviation is greater than or equal to the second preset threshold, the refrigerator under test is determined to be a defective product and the test is terminated.

[0037] S44. Collect the test feature vectors of the refrigerators that have been judged to be defective, and form a failure fingerprint database.

[0038] S45. Perform unsupervised clustering analysis on the test feature vectors in the failure fingerprint database to identify one or more failure modes, and generate associated failure mode codes for the identified failure modes.

[0039] S46. Use the test feature vector of the refrigerator under test that is determined to be qualified to dynamically correct the reference feature vector and the reference covariance moment.

[0040] Preferably, the reference feature vector and the reference covariance matrix are pre-established based on the feature vectors of multiple qualified refrigerated cabinet samples through offline modeling;

[0041] The reference feature vector represents the mean of the characteristics of a qualified refrigerator, and the reference covariance matrix represents the degree of dispersion of the characteristics of a qualified refrigerator.

[0042] In step S46, the dynamic correction of the reference eigenvector and the reference covariance matrix is ​​specifically implemented using an exponentially weighted moving average algorithm, as follows:

[0043] The dynamic correction formula for the reference feature vector is:

[0044] ;

[0045] In the formula, The corrected reference feature vector, The reference feature vector before correction. This is the test feature vector of the qualified product. The preset learning rate;

[0046] The dynamic correction of the reference covariance matrix is ​​based on the test feature vector of each refrigerator that is judged to be qualified, and the covariance relationship between it and the reference feature vector is updated by exponential weighting.

[0047] Preferably, the deviation is obtained by the Mahalanobis distance between the test feature vector and the reference feature vector, and the specific calculation formula is as follows:

[0048] ;

[0049] In the formula, The deviation is mentioned. The test feature vector, The reference feature vector, Let be the reference covariance matrix.

[0050] The comprehensive performance testing device for refrigerated display cases includes:

[0051] Refrigerated cabinets, which are the objects to be tested, are set up in multiples and arranged in sequence;

[0052] The first conveyor belt, which serves as the main conveyor belt, is used to transport the refrigerated cabinets to the corresponding inspection positions;

[0053] The inspection conveyor belt is connected to the first conveyor belt and is used to receive the refrigerated cabinets to be inspected from the first conveyor belt as they enter the inspection area.

[0054] The second conveyor belt, which is connected to the inspection conveyor belt, is used to transport the refrigerated cabinets after inspection to the qualified and defective product areas.

[0055] A detection sensor is installed on one side of the first conveyor belt to identify which detection conveyor belt the corresponding refrigerated cabinet should enter for detection.

[0056] Electronic tags, which are set on the outer wall of the refrigerator, are used by sensors to identify information and facilitate their assignment to different devices.

[0057] The central control unit is located on one side of the first conveyor belt and is electrically connected to the detection sensors and the detection equipment of the conveyor belt. It is used to monitor the detected real-time data and perform intelligent processing.

[0058] This invention provides a method and apparatus for comprehensive performance testing of refrigerated display cases. It possesses at least one of the following beneficial effects:

[0059] 1. This invention dynamically corrects the reference eigenvector and reference covariance matrix using an exponentially weighted moving average algorithm, enabling the central control system to learn and adapt to different production processes. This improves the accuracy and adaptability of testing, and helps maintain the rationality and effectiveness of refrigerated display case testing standards in the long term.

[0060] 2. This invention compares the test feature vector with the reference feature vector, and uses the deviation to adjust the test process in real time, thereby realizing dynamic evaluation and optimization of the performance of the refrigerator. It allows for different processing methods for qualified and unqualified products, thus improving testing efficiency.

[0061] 3. This invention, through integrated detection sensors and electronic tags, enables the real-time acquisition and generation of time-series data for multiple performance parameters of the refrigerator within a preset testing period. This feature makes performance monitoring more flexible and highly real-time, thereby allowing for rapid response and adjustment of the refrigerator's operating status. Attached Figure Description

[0062] Figure 1 This is a flowchart of the method steps of the present invention;

[0063] Figure 2 This is a schematic diagram of time-series data acquisition and feature extraction according to the present invention;

[0064] Figure 3 This is a schematic diagram of the deviation decision and adaptive control of the present invention;

[0065] Figure 4 This is a schematic diagram of the device of the present invention.

[0066] The components include: 1. Refrigerated cabinet; 2. First conveyor belt; 3. Detection conveyor belt; 4. Second conveyor belt; 5. Detection sensor; 6. Electronic tag; 7. Central control panel. Detailed Implementation Example 1

[0067] Please see the appendix Figure 1 - Appendix Figure 3 This invention provides a method for testing the comprehensive performance of a refrigerator, comprising the following steps:

[0068] S1. The refrigerator to be tested is transported from the first conveyor belt. The electronic tag attached to the refrigerator is identified by the detection sensor, and then it is transported to one of the multiple detection conveyor belts.

[0069] S2. On the testing conveyor belt, within a preset testing period, collect multiple performance parameters of the refrigerator under test to form time-series data;

[0070] S3. Calculate the test feature vector of the refrigerator under test based on the time series data;

[0071] S4. Compare the test feature vector with the pre-established reference feature vector and reference covariance matrix to obtain a deviation, and adaptively control the subsequent test process based on the deviation.

[0072] S5. After the inspection is completed, the refrigerated cabinets are transferred to the second conveyor belt for sorting.

[0073] The specific technical solution for step S1 above is as follows:

[0074] Specifically, step S1 aims to identify and automatically transfer the refrigerators to be tested to the designated testing line.

[0075] Among them, an ultra-high frequency RFID electronic tag is fixedly affixed to the upper right front corner of the outer wall of refrigerator 1. The information stored in the tag includes:

[0076] The unique identification code of refrigerator 1 (such as serial number "SN20240728-001");

[0077] Product model and specifications (e.g., model "CRF-280L", specification "220V / 50Hz");

[0078] Production batch number (e.g., "BATCH0725A");

[0079] Model information of core components (such as compressor model).

[0080] Furthermore, the specific operational process is as follows:

[0081] The refrigerated cabinet 1 is conveyed by the main conveyor belt (first conveyor belt 2). When it reaches the area of ​​the detection sensor (fixed UHF RFID reader, reading distance approximately 10cm), the detection sensor 5 automatically reads the electronic tag information. The central control unit 7, based on the identified product model, specifications, and core component model information (e.g., identifying model "CRF-280L"), and combined with preset diversion rules (e.g., different models or batches of key components correspond to different testing lines), controls the diversion mechanism (e.g., electric push rod or turntable) at the end of the first conveyor belt 2 to precisely guide the current refrigerated cabinet 1 onto the corresponding one of the three parallel testing conveyor belts 3, preparing it for subsequent performance testing. This process ensures that refrigerated cabinets with different configurations can enter the appropriate testing environment or process, and achieves automatic binding of testing tasks and product information.

[0082] The specific technical solution for step S2 above is as follows:

[0083] Specifically, step S2 provides essential foundational data for subsequent performance evaluation. This stage is achieved through the coordinated operation of the detection conveyor belt 3, multiple performance sensors, and temperature probe modules, enabling comprehensive and accurate recording of the refrigerator's key performance parameters.

[0084] After the refrigerator 1 is successfully transferred to the testing conveyor belt 3, the conveyor belt 3 will move the refrigerator to the preset testing position at a stable speed (e.g., one refrigerator per minute), ensuring that it remains level for subsequent accurate readings. The testing conveyor belt 3 is equipped with various sensors specifically designed to collect different performance parameters. These sensors include temperature sensors, current sensors, voltage sensors, and power sensors, and their specific organizational structure is as follows:

[0085] Temperature sensor: This sensor consists of a high-precision thermocouple or thermistor and is mainly used for real-time monitoring of the internal and external temperature values ​​of the refrigerator. The circuit design uses an AD converter to convert the analog temperature signal into a digital signal, ensuring an acquisition accuracy of ±0.1℃.

[0086] Voltage sensor: This sensor measures the input voltage of the refrigerator during operation. Its range is designed to be 0-250V, ensuring stable operation under different voltage conditions. Each voltage measurement cycle is typically 1 second, and the data is immediately transmitted to the central control panel 7 for storage and processing after acquisition.

[0087] Current sensor: A current sensor employing the Hall effect principle is used to monitor the power input of the refrigerator in real time. The current sensor has a range of 0-10A, high sensitivity, and can promptly capture transient characteristics of current changes.

[0088] Power sensor: By calculating data from voltage and current sensors, the power sensor's main function is to assess the energy consumption during the operation of the refrigerated display case. The formula for calculating power is:

[0089] ;

[0090] In the formula, P is power, U is voltage, and I is current.

[0091] To ensure data authenticity and accuracy, a stable electrical connection is established between the probe module and each sensor via high-quality cables, reducing errors introduced by external interference. Furthermore, the sensor module is installed in the core operating part of the refrigerated cabinet to ensure data acquisition during normal equipment operation.

[0092] The data acquisition process is automatically managed by the central control console 7. Within the preset test time, the central control console 7 continuously receives data from various performance sensors and performs real-time data acquisition. The acquired instantaneous parameters are correlated with the corresponding timestamps to form a complete time-series dataset.

[0093] To improve the efficiency and accuracy of data analysis, the system also features multiple data acquisition modes, such as timed acquisition and event-triggered acquisition. Timed acquisition automatically records parameters at set time intervals (e.g., once every second), while event-triggered acquisition responds quickly and records relevant performance parameters based on changes in the refrigerator's operating status. Once all data has been collected, the system generates feature vectors containing key performance indicators, preparing for the next feature vector calculation stage.

[0094] This efficient performance parameter acquisition process allows for accurate and real-time recording of the refrigerated display case's operational characteristics, laying a solid foundation for subsequent performance evaluation. This process not only provides a quantitative analysis of the refrigerated display case's actual operating performance but is also a crucial step in improving the overall quality and efficiency of cold chain management.

[0095] The specific technical solution for step S3 above is as follows:

[0096] Specifically, step S3 is a crucial part of the entire refrigerated display case performance evaluation system. This stage involves in-depth analysis of the collected performance parameters to generate high-dimensional feature vectors, providing quantitative evidence for the performance evaluation of the refrigerated display case.

[0097] First, after collecting performance parameters, the system will obtain a time-series dataset containing various performance data of the refrigerator. This dataset covers metrics including internal temperature, ambient temperature, voltage, current, and power. After processing, this data will be used to construct a test feature vector. A feature vector is a mathematically formalized multidimensional data structure containing multiple key metrics related to the refrigerator's performance. Specifically, the feature vector includes, but is not limited to, the following parameters:

[0098] Initial cooling rate: This indicator reflects the rate at which the refrigerator reaches its set temperature. It is calculated by monitoring the internal temperature change over a certain period of time, using the following formula:

[0099] ;

[0100] In the formula, The initial cooling rate, The target temperature set inside the refrigerator. t represents the initial temperature of the refrigerator, and t represents time (seconds).

[0101] Thermodynamic characteristic time constant: This parameter defines the response time of a refrigerator to reach stable operation and is an important basis for evaluating the design and manufacturing quality of a refrigerator. Typically, it is calculated by fitting an exponential curve of temperature change over time, using the following formula:

[0102] ;

[0103] In the formula, It is a time constant. For a period of time, To stabilize the temperature.

[0104] Refrigeration efficiency per unit energy consumption: This indicator is used to evaluate the energy efficiency ratio of a refrigerator under operating conditions, reflecting the relationship between energy consumption and cooling effect. Its calculation formula is:

[0105] ;

[0106] In the formula, This represents the absolute value of the internal temperature change during the preset test period. This is the integral of the power within the preset test period.

[0107] When constructing the feature vector, these performance metrics will be combined into a multidimensional vector. The format is as follows:

[0108] ;

[0109] In the formula, This represents the average temperature inside the refrigerator. Average voltage The average current, This represents the average power.

[0110] During the eigenvector calculation process, the system utilizes data processing algorithms to ensure dynamic adjustment of various eigenvector indicators. Based on the usage of different refrigerated display cases, it selects effective features that represent their performance. Furthermore, to improve the expressive power of the eigenvectors, the system employs methods such as Principal Component Analysis (PCA) to reduce the dimensionality of the eigenvectors, ensuring that the final output eigenvectors possess high information content and representativeness.

[0111] After the feature vector calculation is completed, the generated vector will be stored in the temporary data cache area of ​​the central control panel 7 for use in the subsequent deviation calculation stage. The output of this stage not only provides necessary data support for subsequent performance evaluation, but also provides guidance for optimizing the design of the refrigerated cabinet.

[0112] Through this systematic feature vector calculation process, the key performance indicators of the refrigerated display case are clearly presented, ensuring the scientific rigor and effectiveness of subsequent deviation analysis and performance evaluation. Simultaneously, maintaining high data processing accuracy and real-time performance throughout significantly improves the reliability of the refrigerated display case performance testing.

[0113] The specific technical solution for step S4 above is as follows:

[0114] Specifically, step S4 determines the subsequent testing process by comparing and analyzing the test feature vector and the reference feature vector, ensuring that the test results of the refrigerator are accurate and reliable.

[0115] After completing the eigenvector calculation (S3), the system will generate a test eigenvector. This vector will be compared with a preset reference feature vector. The comparison is performed. The reference feature vector is set based on historical data or standard refrigerated display case performance characteristics, and has high representativeness and stability.

[0116] To facilitate comparison between the two, this invention employs Mahalanobis distance as the method for calculating deviation. This method effectively considers the covariance of the data and provides an accurate measure of deviation. The calculation formula is as follows:

[0117] ;

[0118] In the formula, For deviation, This is the test feature vector of the refrigerated display case. As a reference feature vector, The covariance matrix of the reference eigenvectors.

[0119] In the specific implementation process, the system will perform the following steps:

[0120] Data preparation: First, the calculated test feature vector is extracted from the central control unit 7, and the preset reference feature vector and corresponding reference covariance matrix are loaded simultaneously. Prior to this, the reference covariance matrix needs to be calculated using historical data analysis and sample data to ensure that statistical characteristics are fully reflected.

[0121] Deviation Calculation: Next, the system uses the Mahalanobis distance formula described above to calculate... and Deviation between During this process, all mathematical calculations are performed in real time on the central control console 7, ensuring timely feedback of calculation results and efficient execution.

[0122] Judgment logic: Based on the calculated deviation The system will classify the refrigerators according to preset thresholds:

[0123] Qualified product determination: If If the result is less than the preset threshold of 1, the refrigerator is considered to meet the standard. The system will automatically record the test result to the list of qualified products and terminate the subsequent testing process.

[0124] Extended test implementation: If the preset threshold 1 ≤ If it is less than the preset threshold 2, an extended test will be performed to further analyze the performance of the refrigerator cabinet, collect more data to determine its compliance. This extended test will focus on monitoring non-standard indicators, such as the operating performance under different load conditions.

[0125] Handling of non-conforming products: If it is greater than the preset thresholds 1 and 2, it will be determined as a non-conforming product. The system will record the test data and issue an alarm notification. In addition, the relevant data will be marked for subsequent investigation and analysis.

[0126] Adaptive control: After the judgment result, the system will adaptively adjust the subsequent test process through control instructions. For qualified products, the system will automatically update their performance data to the database for subsequent verification. For non-conforming products, the system will unlock the failure mode analysis program, start recording the failure fingerprint and filling it into the failure fingerprint database.

[0127] Result storage: Finally, the test results and parameters will be stored in the database of the console for future query, comparison and statistical analysis. All data storage uses timestamps and status identifiers, making subsequent data review and equipment tracking more efficient.

[0128] Through this step of deviation comparison and adaptive control, the evaluation process of the refrigerator cabinet performance is ensured to be objective and scientific, capable of timely feedback of test results, and optimizing the test plan through an efficient decision-making mechanism. This process not only improves the automation and intelligence of the test, but also effectively prevents the outflow of non-conforming products, providing reliable technical support for the cold chain industry.

[0129] The specific technical solution content of the above S5 step is as follows:

[0130] Specifically, the S5 step is an important link to ensure the continuous improvement and traceability of the refrigerator cabinet performance. Through this step, the performance data of qualified products collected can inversely affect the update of the reference feature vector, and at the same time, the data of non-conforming products are systematically recorded, providing a scientific basis for subsequent quality improvement.

[0131] After the determination of qualified products, the system will automatically start the dynamic correction function to update the reference feature vector and covariance matrix. This process is as follows:

[0132] Updating the reference feature vector: For the refrigerator cabinet determined to be qualified, its test feature vector will be used to dynamically update the reference feature vector The update formula is:

[0133] ;

[0134] In the formula, A preset learning rate is typically chosen between 0 and 1 (e.g., 0.1) to ensure that the impact of newly collected data on the reference dataset is controllable. During dynamic calibration, if a large number of qualified product tests occur... The newly collected feature vectors will gradually converge, effectively reflecting the latest dynamics of product performance.

[0135] Updated reference covariance matrix: Reference covariance matrix It will also be adjusted based on the new data. Updates can be achieved using the following formula:

[0136] ;

[0137] Here, β is also a preset learning rate to ensure that new measurements effectively affect the update of the covariance matrix.

[0138] Through the above dynamic correction, the system ensures that the reference eigenvector and covariance matrix always reflect the latest performance information of the refrigerator, thereby improving the accuracy and reliability of subsequent tests.

[0139] Formation of the Failure Fingerprint Database: During the processing of non-conforming products, the system will automatically generate a failure fingerprint database to record the test feature vector and other relevant information for each non-conforming refrigerator. This part of the implementation includes the following steps:

[0140] Record the feature vector of non-conforming products: For refrigerated cabinets determined to be non-conforming, the system will save their test feature vector, test date, model, fault type, and other relevant data. All data will be indexed according to time series and frequency of occurrence for subsequent queries.

[0141] Establish a failure mode identification (FMO) mechanism: Data from the failure fingerprint database will be used to build a FMO model. Machine learning or data mining techniques will be used to perform cluster analysis on non-conforming product data to identify common features associated with specific failures. This model will be trained based on past non-conforming data to identify refrigerated display case operating modes with similar symptoms and generate corresponding failure mode codes.

[0142] Storage and Optimization: The failure fingerprint database will be stored in a structured format in the central control panel's database to ensure data integrity and efficient retrieval. The database will be regularly maintained and optimized (e.g., deleting outdated or no longer applicable data) to maintain the effectiveness of the identification model.

[0143] Through the dynamic correction and failure fingerprint database established above, the refrigerated display case performance evaluation system can effectively achieve self-optimization and provide a reliable basis for subsequent fault analysis and quality improvement. This process not only enhances product quality control but also promotes the design and manufacturing level of refrigerated display cases and improves overall cold chain management efficiency. Example 2

[0144] Please see the appendix Figure 4 A comprehensive performance testing device for refrigerated display cases, including:

[0145] Refrigerated cabinet 1, which serves as the object to be tested, is provided in multiple units arranged in sequence;

[0146] The first conveyor belt 2, which serves as the main conveyor belt, is used to transport the refrigerated cabinet 1 to the corresponding inspection position;

[0147] The inspection conveyor belt 3 is connected to the first conveyor belt 2 and is used to receive the refrigerator 1 to be inspected from the first conveyor belt 2 as it enters the inspection area.

[0148] The second conveyor belt 4 is connected to the inspection conveyor belt 3 and is used to transport the refrigerated cabinet 1 after inspection to the qualified and defective product areas.

[0149] The detection sensor 5 is located on one side of the first conveyor belt 2 and is used to identify which detection conveyor belt 3 the corresponding refrigerator 1 should enter for detection.

[0150] Electronic tags 6 are set on the outer wall of the refrigerator 1 and are used by the detection sensor 5 to identify information, so that they can be easily assigned to different tags.

[0151] The central control console 7 is located on one side of the first conveyor belt 2 and is electrically connected to the detection sensor 5 and the detection equipment of the detection conveyor belt 3. It is used to monitor the detected real-time data and perform intelligent processing.

[0152] Specifically, the refrigerator 1 is transported by the first conveyor belt 2 to the designated detection sensor 5, where the electronic tag 6 is identified. After the identification is confirmed, the refrigerator 1 is transported to the detection conveyor belt 3 for detection. Finally, based on the judgment of the detection results by the central control console 7, it is transported to the qualified area or the unqualified area via the second conveyor belt 4.

[0153] The first conveyor belt 2 and the second conveyor belt 4 are arranged in parallel. The first conveyor belt 2 and the second conveyor belt 4 are connected by a detection conveyor belt 3. The connection positions of the detection conveyor belt 3 and the first conveyor belt 2 and the second conveyor belt 4 are all at a 90° right angle. The connection points are equipped with inclined roller groups to assist the material to be transported smoothly at a 90° right angle.

Claims

1. A comprehensive performance testing method for refrigerated display cases, characterized in that, Includes the following steps: S1. The refrigerator to be tested is transported from the first conveyor belt. The electronic tag attached to the refrigerator is identified by the detection sensor, and then it is transported to one of the multiple detection conveyor belts. S2. On the testing conveyor belt, within a preset testing period, collect multiple performance parameters of the refrigerator under test to form time-series data; The performance parameters of the refrigerator under test include: The internal temperature of the refrigerator; The ambient temperature for monitoring the refrigerated display case; The voltage of the refrigerator; The current of the refrigerator; The power of the refrigerator; S3. Calculate the test feature vector of the refrigerator under test based on the time series data; The feature vector includes: Initial cooling rate; Thermodynamic characteristic time constant; Cooling efficiency per unit energy consumption; The calculation method for the unit energy consumption cooling efficiency is as follows: ； In the formula, This represents the absolute value of the temperature change inside the refrigerator during a preset test period. The integral of the refrigerator's power during the preset test period; The thermodynamic characteristic time constant is calculated as follows: ； In the formula, It is a time constant. For a period of time, To stabilize the temperature; The target temperature set inside the refrigerator. The initial temperature of the refrigerator is given by t, which is time in seconds. The initial cooling rate is calculated as follows: ； In the formula, The initial cooling rate, The target temperature set inside the refrigerator. The initial temperature of the refrigerator is given by t, which is time in seconds. S4. Compare the test feature vector with the pre-established reference feature vector and reference covariance matrix to obtain a deviation degree, and adaptively control the subsequent test process based on the deviation degree. S5. After the inspection is completed, the refrigerated cabinets are transferred to the second conveyor belt for sorting.

2. The method for testing the comprehensive performance of a refrigerator according to claim 1, characterized in that, In step S1, the information of the electronic tag includes: The unique identification code of the refrigerator; The product model and specifications of the refrigerator; The production batch number of the refrigerated cabinet; The model information of the core components used in the refrigerator.

3. The method for testing the comprehensive performance of a refrigerator according to claim 1, characterized in that, In step S2, the process of forming time-series data includes the following steps: S21. Deploy the temperature sensing probe used to collect the internal temperature to a preset position inside the refrigerator to be tested; S22. During the preset test period, the instantaneous values ​​of the multiple performance parameters are collected synchronously at a preset sampling frequency. S23. Associate the collected instantaneous values ​​with the corresponding timestamps to generate the time series data.

4. The method for testing the comprehensive performance of a refrigerator according to claim 1, characterized in that, In step S4, the adaptive control of the subsequent test process includes the following steps: S41. If the deviation is less than the first preset threshold, the refrigerator under test is determined to be a qualified product and the test is terminated. S42. If the deviation is greater than or equal to the first preset threshold and less than the second preset threshold, then perform an extended test on the refrigerator under test. S43. If the deviation is greater than or equal to the second preset threshold, the refrigerator under test is determined to be a defective product and the test is terminated. S44. Collect the test feature vectors of the refrigerators that have been judged to be defective, and form a failure fingerprint database. S45. Perform unsupervised clustering analysis on the test feature vectors in the failure fingerprint database to identify one or more failure modes, and generate associated failure mode codes for the identified failure modes. S46. Use the test feature vector of the refrigerator to be tested that is judged to be qualified to dynamically correct the reference feature vector and the reference covariance matrix.

5. The method for testing the comprehensive performance of a refrigerated display case according to claim 4, characterized in that, In step S46, the reference feature vector and the reference covariance matrix are pre-established based on the feature vectors of multiple qualified refrigerated cabinet samples through offline modeling. The reference feature vector represents the mean of the characteristics of a qualified refrigerator, and the reference covariance matrix represents the degree of dispersion of the characteristics of a qualified refrigerator. In step S46, the dynamic correction of the reference eigenvector and the reference covariance matrix is ​​specifically implemented using an exponentially weighted moving average algorithm, as follows: The dynamic correction formula for the reference feature vector is: ； In the formula, The corrected reference feature vector, The reference feature vector before correction. This is the test feature vector of the qualified product. The preset learning rate; The dynamic correction of the reference covariance matrix is ​​based on the test feature vector of each refrigerator that is judged to be qualified, and the covariance relationship between it and the reference feature vector is updated by exponential weighting.

6. The method for testing the comprehensive performance of a refrigerated display case according to claim 1, characterized in that, In step S4, the deviation is obtained by the Mahalanobis distance between the test feature vector and the reference feature vector, and the specific calculation formula is as follows: ； In the formula, The deviation is mentioned. The test feature vector, The reference feature vector, Let be the reference covariance matrix.

7. A comprehensive performance testing device for refrigerated display cases, comprising the comprehensive performance testing method for refrigerated display cases according to any one of claims 1-6, characterized in that, include: Refrigerated cabinets (1), which are the objects to be tested, are provided in multiple and arranged in order; The first conveyor belt (2) serves as the main conveyor belt and is used to transport the refrigerated cabinet (1) to the corresponding inspection position. The detection conveyor belt (3) is connected to the first conveyor belt (2) and is used to receive the refrigerator (1) to be tested transmitted from the first conveyor belt (2) into the detection area; The second conveyor belt (4), which is connected to the inspection conveyor belt (3), is used to transport the refrigerated cabinet (1) after inspection to the qualified and defective product areas; A detection sensor (5) is set on one side of the first conveyor belt (2) to identify which detection conveyor belt (3) the corresponding refrigerator (1) should enter for detection; Electronic tags (6) are set on the outer wall of the refrigerator (1) for the detection sensor (5) to identify information and facilitate their allocation to different; The central control unit (7) is located on one side of the first conveyor belt (2) and is electrically connected to the detection sensor (5) and the detection conveyor belt (3) detection equipment. It is used to monitor the detected real-time data and perform intelligent processing.

Images