Intelligent Refrigeration Equipment Predictive Maintenance System

Abstract

The present invention discloses a predictive maintenance system for intelligent refrigeration equipment, and relates to the technical field of analysis and prediction. The predictive maintenance system for intelligent refrigeration equipment includes a data acquisition module, a data analysis module, and a fault prediction module, which constructs a confidence interval of key parameters, constructs a first mapping relationship between the confidence interval of key parameters and the historical fault type, calculates the average gray value of the internal infrared image after the second preprocessing, selects the fault corresponding map, and establishes a second mapping relationship. The present invention improves the accuracy of fault detection through the analysis of historical fault parameters and the selection of key parameters, forms a knowledge base through the continuous accumulation and updating of historical fault types and parameters, so that future maintenance and fault prediction will be more accurate, and visualizes the equipment status through the acquisition and analysis of internal infrared images, helping maintenance personnel to intuitively understand the operation of the equipment, and self-adjust according to historical data, and further improve the ability of fault prediction and analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of analysis and prediction, and in particular to a predictive maintenance system for intelligent refrigeration equipment. Background Art

[0002] In recent years, with the popularization of Internet of Things technology, the types and accuracy of sensors have been continuously improved. Refrigeration equipment can monitor various operating parameters such as temperature, pressure and vibration in real time, provide rich data support, and use big data analysis technology combined with machine learning algorithms to conduct in-depth mining and analysis of historical operating data. It can identify potential failure modes and trends and enhance the accuracy of fault prediction.

[0003] At present, a Chinese invention patent with publication number CN 109308519 A discloses a refrigeration equipment fault prediction method based on a neural network. The method judges the health status of the refrigeration equipment by comparing the probability density of the actual prediction result with the fault threshold, and predicts the fault of the refrigeration equipment by combining the long-term and short-term neural network model with the fault discrimination model. The health status of the refrigeration equipment can be effectively predicted. However, in the related technology, there is no step of simplifying the numerical data by image data according to the inherent correlation between the image data and the numerical data for analysis and calculation, which easily leads to slow response speed. There is no screening of corresponding effective analysis parameters according to different fault types, which is not conducive to improving the accuracy and calculation speed of the system. Summary of the invention

[0004] The technical problem solved by the present invention is that the related technology does not have the steps of simplifying numerical data through image data for analysis and calculation based on the inherent correlation between image data and numerical data, which easily leads to slow response speed, and does not screen corresponding effective analysis parameters according to different fault types, which is not conducive to improving the accuracy and calculation speed of the system.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, a predictive maintenance system for intelligent refrigeration equipment includes a data acquisition module, a data analysis module and a fault prediction module;

[0006] The data acquisition module obtains the historical fault type of the device, obtains the historical fault parameters corresponding to the historical fault type, and assigns a first number to the historical fault type;

[0007] The data analysis module performs a first preprocessing on the historical fault parameters, analyzes the preprocessed historical fault parameters, selects key parameters according to the analysis results, performs a second numbering on the key parameters, calculates the average of the key parameters, and calculates the standard deviation of the key parameters, constructs a confidence interval of the key parameters, constructs a first mapping relationship between the confidence interval of the key parameters and the historical fault type, obtains an internal infrared image corresponding to the historical fault parameters, performs a third numbering on the internal infrared image, performs a second preprocessing on the internal infrared image, calculates the average gray value of the internal infrared image after the second preprocessing, selects a fault corresponding image according to the average gray value, and establishes a second mapping relationship between the fault corresponding image and the confidence interval of the key parameters;

[0008] The fault prediction module obtains the average gray value of the current fault corresponding map, obtains the corresponding confidence interval of the key parameter according to the second mapping relationship, and obtains the corresponding historical fault type according to the first mapping relationship.

[0009] As a preferred solution of the intelligent refrigeration equipment predictive maintenance system described in the present invention, the data acquisition module obtains the equipment model, retrieves the fault database, inputs the equipment model into the fault database, matches the historical fault type corresponding to the equipment model, and the historical fault type includes bearing wear, gear wear, winding wear, overload, poor air intake and poor exhaust. The historical fault type is first numbered, and the first number is represented as , is a natural number;

[0010] The historical fault parameters corresponding to the historical fault type are obtained, wherein the historical fault parameters include bearing vibration frequency, gear vibration frequency, motor voltage, motor current, noise intensity, suction pressure, discharge pressure and winding vibration frequency.

[0011] As a preferred solution of the intelligent refrigeration equipment predictive maintenance system described in the present invention, wherein: the data analysis module performs a first preprocessing on the historical fault parameters, the first preprocessing includes a duplicate value removal process and a normalization process, and the normalization process makes the value of the historical fault parameter distributed between [0~1];

[0012] The pre-processed historical fault parameters are analyzed, key parameters are selected according to the analysis results, and the key parameters are second numbered. The second number corresponds to the first number. The second number is represented by , is a natural number, is a natural number, among which, It is indicated as a subscript in the first number, and the second number corresponds to the first number;

[0013] The selection logic of the key parameters includes:

[0014] The historical fault type is set as the dependent variable, and the historical fault parameter is set as the independent variable. Multiple regression analysis is performed to obtain the coefficients corresponding to the historical fault parameters. The first value is set as the correlation threshold, and the coefficients corresponding to the historical fault parameters are sorted in descending order. The coefficients corresponding to the historical fault parameters after descending order are compared with the first value, and historical fault parameters greater than or equal to the first value are selected, and the historical fault parameters greater than or equal to the first value are set as key parameters.

[0015] As a preferred solution of the intelligent refrigeration equipment predictive maintenance system described in the present invention, when the historical fault type is bearing wear, the corresponding key parameters include bearing vibration frequency and noise intensity; when the historical fault type is gear wear, the corresponding key parameters include bearing vibration frequency, noise intensity and gear vibration frequency and noise intensity; when the historical fault type is winding wear, the corresponding key parameters include winding vibration frequency and motor current; when the historical fault type is overload, the corresponding key parameters are motor current and motor voltage; when the historical fault type is poor air intake, the corresponding key parameter is suction pressure; when the historical fault type is poor exhaust, the corresponding key parameter is discharge pressure.

[0016] As a preferred solution of the intelligent refrigeration equipment predictive maintenance system described in the present invention, wherein: the data analysis module obtains any historical fault type and its corresponding key parameters, calculates the average of each key parameter respectively, and calculates the standard deviation of the key parameter through the standard deviation formula to construct the confidence interval of the key parameter;

[0017] The construction logic of the confidence interval of the key parameter includes:

[0018] Obtain the standard deviation and the average of the key parameters, calculate the difference between the average of the key parameters and the standard deviation to obtain a first difference, set the first difference as the lower limit of the confidence interval of the key parameters, calculate the sum of the average of the key parameters and the standard deviation to obtain a first sum, set the first sum as the upper limit of the confidence interval of the key parameters, and construct a confidence interval according to the lower limit and upper limit of the confidence interval.

[0019] As a preferred solution of the intelligent refrigeration equipment predictive maintenance system described in the present invention, wherein: the data analysis module constructs a first mapping relationship between the confidence interval of the key parameter and the historical fault type, obtains the internal infrared image corresponding to the historical fault parameter, and assigns a third number to the internal infrared image, the third number corresponds to the first number, and the third number is represented as , the internal infrared image is subjected to a second preprocessing, the second preprocessing includes graying processing and segmentation processing, the internal infrared image is formed by an infrared imager by monitoring the motor winding temperature, bearing temperature, gear temperature, exhaust port temperature and intake port temperature, and the logic of the segmentation processing includes:

[0020] Recognize the motor winding, bearing, gear, exhaust port and air intake port images through machine vision, set the contour lines of the motor winding, bearing, gear, exhaust port and air intake port, and segment the internal infrared image after grayscale processing according to the contour lines to obtain infrared sub-images;

[0021] Obtaining the grayscale value of the second preprocessed internal infrared image, calculating the average grayscale value of the second preprocessed internal infrared image, selecting a fault corresponding map according to the average grayscale value, and establishing a second mapping relationship between the fault corresponding map and the confidence interval of the key parameter, wherein the selection logic of the fault corresponding map includes:

[0022] The average grayscale values ​​of the internal infrared images are traversed, the average grayscale values ​​of the internal infrared images are sorted in descending order, the average grayscale value of the internal infrared image with the largest value is selected, and the infrared sub-image corresponding to the average grayscale value of the internal infrared image with the largest value is set as the fault corresponding image.

[0023] As a preferred solution of the intelligent refrigeration equipment predictive maintenance system described in the present invention, the recognition logic of the machine vision includes:

[0024] Acquire images of standard motor windings, standard bearings, standard gears, standard exhaust ports, and standard air inlets, respectively extract graphic feature quantities of the images of standard motor windings, standard bearings, standard gears, standard exhaust ports, and standard air inlets, record them as first feature quantities, acquire an internal infrared image after grayscale processing, extract graphic feature quantities of the internal infrared image after grayscale processing, record them as second feature quantities, calculate the similarity between the first feature quantity and the second feature quantity by a cosine similarity formula, traverse the similarities, sort the similarities in descending order, select the similarity with the largest numerical value, set the standard image corresponding to the similarity with the largest numerical value as the current image, the standard image includes images of standard motor windings, standard bearings, standard gears, standard exhaust ports, and standard air inlets, and the current image includes images of motor windings, bearings, gears, exhaust ports, and air inlets.

[0025] As a preferred solution of the intelligent refrigeration equipment predictive maintenance system described in the present invention, wherein: the fault prediction module obtains the current infrared sub-image, grays the current infrared sub-image, calculates the average gray value of the current infrared sub-image, performs a difference calculation between the average gray value of the current infrared sub-image and the average gray value of the corresponding fault corresponding image to obtain a second difference value, sets a difference threshold value for a third value, and compares the second difference with the third value;

[0026] When the second difference is less than or equal to the difference threshold, obtaining the corresponding confidence interval of the key parameter according to the second mapping relationship, and obtaining the corresponding historical fault type according to the first mapping relationship;

[0027] When the second difference is greater than the difference threshold, jump to the next current infrared sub-image, and stop the jump operation when the second difference is less than or equal to the difference threshold;

[0028] The fault prediction module outputs and displays the historical fault type and the current infrared sub-image on the OLED display.

[0029] The beneficial effects of the present invention are as follows: through the analysis of historical fault parameters and the selection of key parameters, the main factors affecting the normal operation of refrigeration equipment can be effectively identified, thereby improving the accuracy of fault detection, relying on data collection and analysis; using historical data to predict possible future faults, helping maintenance personnel make more scientific decisions, reducing the subjectivity of human judgment, and being able to intervene before a fault occurs, reducing equipment downtime, improving equipment operating efficiency, and reducing maintenance costs; through the continuous accumulation and updating of historical fault types and parameters, a knowledge base is formed, and future maintenance and fault prediction will be more accurate; through the acquisition and analysis of internal infrared images, the equipment status is visualized, helping maintenance personnel to intuitively understand the equipment operation status, promptly discover potential problems, and make self-adjustments based on historical data, further improving the ability of fault prediction and analysis, and adapting to the operating characteristics of different refrigeration equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A basic flow chart of a smart refrigeration equipment predictive maintenance system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0032] Example 1, reference Figure 1, as an embodiment of the present invention, provides an intelligent refrigeration equipment predictive maintenance system, including a data acquisition module, a data analysis module and a fault prediction module.

[0033] The data acquisition module obtains historical fault types of the device, obtains historical fault parameters corresponding to the historical fault types, and assigns a first number to the historical fault types.

[0034] The data analysis module performs a first preprocessing on the historical fault parameters, analyzes the preprocessed historical fault parameters, selects key parameters according to the analysis results, assigns a second number to the key parameters, calculates the average of the key parameters, and calculates the standard deviation of the key parameters, constructs a confidence interval of the key parameters, constructs a first mapping relationship between the confidence interval of the key parameters and the historical fault type, obtains an internal infrared image corresponding to the historical fault parameters, assigns a third number to the internal infrared image, performs a second preprocessing on the internal infrared image, calculates the average grayscale value of the internal infrared image after the second preprocessing, selects a fault correspondence image according to the average grayscale value, and establishes a second mapping relationship between the fault correspondence image and the confidence interval of the key parameters.

[0035] The fault prediction module obtains the average gray value of the current fault corresponding map, obtains the corresponding confidence interval of the key parameter according to the second mapping relationship, and obtains the corresponding historical fault type according to the first mapping relationship.

[0036] The present invention effectively identifies the main factors affecting the normal operation of refrigeration equipment through the analysis of historical fault parameters and the selection of key parameters, thereby improving the accuracy of fault detection, relying on data collection and analysis; using historical data to predict possible future faults, helping maintenance personnel make more scientific decisions, reducing the subjectivity of human judgment, and being able to intervene before a fault occurs, reducing equipment downtime, improving equipment operating efficiency, and reducing maintenance costs; through the continuous accumulation and updating of historical fault types and parameters, a knowledge base is formed, and future maintenance and fault prediction will be more accurate; through the acquisition and analysis of internal infrared images, the equipment status is visualized, helping maintenance personnel to intuitively understand the equipment operation status, timely discover potential problems, and self-adjust according to historical data, further improving the ability of fault prediction and analysis, and adapting to the operating characteristics of different equipment.

[0037] The data acquisition module obtains the device model, retrieves the fault database, inputs the device model into the fault database, matches the historical fault type corresponding to the device model, the historical fault type includes bearing wear, gear wear, winding wear, overload, poor air intake and poor exhaust, and assigns a first number to the historical fault type, the first number is represented as , is a natural number;

[0038] The historical fault parameters corresponding to the historical fault type are obtained, wherein the historical fault parameters include bearing vibration frequency, gear vibration frequency, motor voltage, motor current, noise intensity, suction pressure, discharge pressure and winding vibration frequency.

[0039] In specific implementation, by obtaining the equipment model and matching the historical fault type, the possible problems of the equipment can be quickly located, effectively reducing the time for fault inspection and repair. The use of historical fault types and fault parameters is helpful to establish a fault pattern recognition system, which can provide data support for maintenance personnel and help judge the health status of the equipment. By monitoring and analyzing historical fault parameters, potential faults can be identified in advance, predictive maintenance can be implemented, and equipment downtime can be reduced, thereby improving production efficiency. Combining historical fault types with parameters can help conduct more in-depth fault analysis and provide data support for decision-making.

[0040] The data analysis module performs a first preprocessing on the historical fault parameters, wherein the first preprocessing includes removing duplicate values ​​and normalizing, wherein the normalizing makes the values ​​of the historical fault parameters distributed between [0 and 1]; the preprocessed historical fault parameters are analyzed, key parameters are selected according to the analysis results, and the key parameters are second numbered, wherein the second number corresponds to the first number, and the second number is represented as , is a natural number, is a natural number, among which, It is represented as a subscript in the first number, and the second number corresponds to the first number.

[0041] The selection logic of the key parameters includes: setting the historical fault type as a dependent variable, setting the historical fault parameters as independent variables, performing multiple regression analysis, obtaining coefficients corresponding to the historical fault parameters, setting a first value as a correlation threshold, sorting the coefficients corresponding to the historical fault parameters in descending order, comparing the coefficients corresponding to the historical fault parameters after descending order with the first value, selecting historical fault parameters greater than or equal to the first value, and setting the historical fault parameters greater than or equal to the first value as key parameters.

[0042] In specific implementation, removing duplicate values ​​can improve the accuracy of data and avoid deviations in analysis. Normalization processing enables historical fault parameters of different dimensions to be compared on the same scale, thereby improving the effectiveness of analysis. Through multivariate regression analysis, the relationship between historical fault parameters and fault types is systematically revealed, thereby identifying key parameters that have a significant impact on equipment failures, helping to focus on the most valuable information, sorting the coefficients of historical fault parameters in descending order, and screening out key parameters according to the set correlation threshold, which can provide a data-driven scientific basis for subsequent fault diagnosis and early warning, and improve the accuracy and reliability of decision-making.

[0043] When the historical fault type is bearing wear, the corresponding key parameters include bearing vibration frequency and noise intensity; when the historical fault type is gear wear, the corresponding key parameters include bearing vibration frequency, noise intensity, gear vibration frequency and noise intensity; when the historical fault type is winding wear, the corresponding key parameters include winding vibration frequency and motor current; when the historical fault type is overload, the corresponding key parameters are motor current and motor voltage; when the historical fault type is poor air intake, the corresponding key parameter is suction pressure; when the historical fault type is poor exhaust, the corresponding key parameter is discharge pressure.

[0044] In the specific implementation, corresponding key parameters are set for different historical fault types, so that the monitoring equipment can concentrate resources and energy on the most critical indicators, improve monitoring efficiency, and associate key parameters with fault types. This can provide a basis for subsequent maintenance decisions, promote maintenance management to develop in a data-driven direction, and improve the scientific nature of decision-making.

[0045] The data analysis module obtains any historical fault type and its corresponding key parameters, calculates the average of each key parameter respectively, calculates the standard deviation of the key parameter through the standard deviation formula, and constructs the confidence interval of the key parameter.

[0046] The construction logic of the confidence interval of the key parameter includes: obtaining the standard deviation and the average of the key parameter, performing a difference calculation between the average of the key parameter and the standard deviation to obtain a first difference value, setting the first difference value as the lower limit of the confidence interval of the key parameter, performing a sum calculation between the average of the key parameter and the standard deviation to obtain a first sum value, setting the first sum value as the upper limit of the confidence interval of the key parameter, and constructing a confidence interval according to the lower limit of the confidence interval and the upper limit of the confidence interval.

[0047] In specific implementation, by calculating the mean and standard deviation of key parameters, the central tendency and dispersion of data can be more accurately evaluated, thereby improving the credibility of the analysis results. Constructing a confidence interval can provide engineers with a reasonable reference range to help them determine whether key parameters are within the normal range and make effective fault diagnosis. The confidence interval provides an interval rather than a single point estimate, thereby helping to identify potential risks and abnormal situations and enhance the ability to prevent faults. By analyzing the confidence interval of each key parameter, the operating standards of the equipment can be adjusted according to different operating environments to improve the safety and efficiency of equipment operation.

[0048] The data analysis module constructs a first mapping relationship between the confidence interval of the key parameter and the historical fault type, obtains the internal infrared image corresponding to the historical fault parameter, and assigns a third number to the internal infrared image, wherein the third number corresponds to the first number, and the third number is represented as , the internal infrared image is subjected to a second preprocessing, the second preprocessing includes graying processing and segmentation processing, and the internal infrared image is formed by an infrared imager through monitoring the motor winding temperature, bearing temperature, gear temperature, exhaust port temperature and intake port temperature.

[0049] The logic of the segmentation process includes:

[0050] Recognize the motor winding, bearing, gear, exhaust port and air intake port images through machine vision, set the contour lines of the motor winding, bearing, gear, exhaust port and air intake port, and segment the internal infrared image after grayscale processing according to the contour lines to obtain infrared sub-images;

[0051] The grayscale value of the second preprocessed internal infrared image is obtained, the average grayscale value of the second preprocessed internal infrared image is calculated, a fault corresponding map is selected according to the average grayscale value, and a second mapping relationship between the fault corresponding map and the confidence interval of the key parameter is established.

[0052] The selection logic of the fault corresponding map includes: traversing the average grayscale values ​​of the internal infrared map, sorting the average grayscale values ​​of the internal infrared map in descending order, selecting the average grayscale value of the internal infrared map with the largest value, and setting the infrared sub-map corresponding to the average grayscale value of the internal infrared map with the largest value as the fault corresponding map.

[0053] The recognition logic of the machine vision includes: acquiring images of standard motor windings, standard bearings, standard gears, standard exhaust ports and standard air intake ports, respectively extracting graphic feature quantities of the images of standard motor windings, standard bearings, standard gears, standard exhaust ports and standard air intake ports, recording them as first feature quantities, acquiring an internal infrared image after grayscale processing, extracting graphic feature quantities of the internal infrared image after grayscale processing, recording them as second feature quantities, calculating the similarity between the first feature quantity and the second feature quantity by a cosine similarity formula, traversing the similarities, sorting the similarities in descending order, selecting the similarity with the largest numerical value, and setting the standard image corresponding to the similarity with the largest numerical value as the current image, the standard image includes images of standard motor windings, standard bearings, standard gears, standard exhaust ports and standard air intake ports, and the current image includes images of motor windings, bearings, gears, exhaust ports and air intake ports.

[0054] In specific implementation, by preprocessing and segmenting the internal infrared image, the thermal status of key components such as motor windings, bearings, gears, etc. can be effectively identified, helping to accurately locate the location of the fault. By utilizing grayscale processing and machine vision technology, a large amount of infrared image data can be quickly processed, which improves the efficiency of data analysis and shortens the time for fault diagnosis. The fault corresponding map selection mechanism based on the average grayscale value makes fault diagnosis more scientific and reasonable, providing strong data support for subsequent decision-making. It is applied to equipment of different types and models, and can adapt to the needs of various working environments by adjusting processing parameters and models.

[0055] The fault prediction module obtains the current infrared sub-image, performs grayscale processing on the current infrared sub-image, calculates the average grayscale value of the current infrared sub-image, performs a difference calculation between the average grayscale value of the current infrared sub-image and the average grayscale value of the corresponding fault corresponding image to obtain a second difference value, sets a difference threshold value for a third value, and compares the second difference with the third value;

[0056] When the second difference is less than or equal to the difference threshold, obtaining the corresponding confidence interval of the key parameter according to the second mapping relationship, and obtaining the corresponding historical fault type according to the first mapping relationship;

[0057] When the second difference is greater than the difference threshold, the next current infrared sub-image is jumped to, and the jump operation is stopped when the second difference is less than or equal to the difference threshold.

[0058] The fault prediction module outputs and displays the historical fault type and current infrared sub-image through the OLED display.

[0059] In specific implementation, it is able to acquire and process the current infrared sub-image in real time, and timely reflect the thermal status of the equipment through grayscale conversion and average grayscale value calculation, which helps to quickly identify potential faults. By comparing the average grayscale value of the current infrared sub-image with the average grayscale value of the fault corresponding image, the key problem area can be intelligently identified, thereby improving the accuracy of fault prediction. Setting the difference threshold allows the system to be flexibly adjusted according to actual conditions, enhancing the adaptability of fault prediction under different working conditions and reducing the false alarm rate. The historical fault type and the current infrared sub-image are output through the OLED display screen, providing intuitive visual feedback, enhancing the interactivity between users and the system, and facilitating the understanding and decision-making of technicians.

[0060] The present invention effectively identifies the main factors affecting the normal operation of refrigeration equipment through the analysis of historical fault parameters and the selection of key parameters, thereby improving the accuracy of fault detection, relying on data collection and analysis; using historical data to predict possible future faults, helping maintenance personnel make more scientific decisions, reducing the subjectivity of human judgment, and being able to intervene before a fault occurs, reducing equipment downtime, improving equipment operating efficiency, and reducing maintenance costs; through the continuous accumulation and updating of historical fault types and parameters, a knowledge base is formed, and future maintenance and fault prediction will be more accurate; through the acquisition and analysis of internal infrared images, the equipment status is visualized, helping maintenance personnel to intuitively understand the equipment operation status, timely discover potential problems, and self-adjust according to historical data, further improving the ability of fault prediction and analysis, and adapting to the operating characteristics of different equipment.

[0061] It should be understood by those skilled in the art that the embodiments of the present invention can be provided as methods, systems or computer program products. Therefore, the intelligent refrigeration equipment predictive maintenance system of the present invention can be in the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention can be in the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, magnetic disk or optical disk. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0062] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A smart refrigeration equipment predictive maintenance system, characterized in that: It includes data acquisition module, data analysis module and fault prediction module; The data acquisition module obtains the historical fault type of the device, obtains the historical fault parameters corresponding to the historical fault type, and assigns a first number to the historical fault type; The data analysis module performs a first preprocessing on the historical fault parameters, analyzes the preprocessed historical fault parameters, selects key parameters according to the analysis results, performs a second numbering on the key parameters, calculates the average of the key parameters, and calculates the standard deviation of the key parameters, constructs a confidence interval of the key parameters, constructs a first mapping relationship between the confidence interval of the key parameters and the historical fault type, obtains an internal infrared image corresponding to the historical fault parameters, performs a third numbering on the internal infrared image, performs a second preprocessing on the internal infrared image, calculates the average gray value of the internal infrared image after the second preprocessing, selects a fault corresponding image according to the average gray value, and establishes a second mapping relationship between the fault corresponding image and the confidence interval of the key parameters; The fault prediction module obtains the average gray value of the current fault corresponding map, obtains the corresponding confidence interval of the key parameter according to the second mapping relationship, and obtains the corresponding historical fault type according to the first mapping relationship; The data analysis module obtains any historical fault type and its corresponding key parameters, calculates the average of each key parameter, calculates the standard deviation of the key parameter through the standard deviation formula, and constructs the confidence interval of the key parameter; The construction logic of the confidence interval of the key parameter includes: Obtaining the standard deviation and the average of the key parameter, performing a difference calculation between the average of the key parameter and the standard deviation to obtain a first difference value, setting the first difference value as the lower limit of the confidence interval of the key parameter, performing a sum calculation between the average of the key parameter and the standard deviation to obtain a first sum value, setting the first sum value as the upper limit of the confidence interval of the key parameter, and constructing a confidence interval according to the lower limit of the confidence interval and the upper limit of the confidence interval; The data analysis module constructs a first mapping relationship between the confidence interval of the key parameter and the historical fault type, obtains the internal infrared image corresponding to the historical fault parameter, and assigns a third number to the internal infrared image, wherein the third number corresponds to the first number, and the third number is represented as , the internal infrared image is subjected to a second preprocessing, the second preprocessing includes graying processing and segmentation processing, the internal infrared image is formed by an infrared imager by monitoring the motor winding temperature, bearing temperature, gear temperature, exhaust port temperature and intake port temperature, and the logic of the segmentation processing includes: Recognize the motor winding, bearing, gear, exhaust port and air intake port images through machine vision, set the contour lines of the motor winding, bearing, gear, exhaust port and air intake port, and segment the internal infrared image after grayscale processing according to the contour lines to obtain infrared sub-images; Obtaining the grayscale value of the second preprocessed internal infrared image, calculating the average grayscale value of the second preprocessed internal infrared image, selecting a fault corresponding map according to the average grayscale value, and establishing a second mapping relationship between the fault corresponding map and the confidence interval of the key parameter, wherein the selection logic of the fault corresponding map includes: Traverse the average grayscale values ​​of the internal infrared images, sort the average grayscale values ​​of the internal infrared images in descending order, select the average grayscale value of the internal infrared image with the largest value, and set the infrared sub-image corresponding to the average grayscale value of the internal infrared image with the largest value as the fault corresponding image; The recognition logic of the machine vision includes: Obtain images of standard motor windings, standard bearings, standard gears, standard exhaust ports, and standard air inlets, respectively extract graphic feature quantities of the images of standard motor windings, standard bearings, standard gears, standard exhaust ports, and standard air inlets, record them as first feature quantities, obtain an internal infrared image after grayscale processing, extract graphic feature quantities of the internal infrared image after grayscale processing, record them as second feature quantities, calculate the similarity between the first feature quantity and the second feature quantity by a cosine similarity formula, traverse the similarities, sort the similarities in descending order, select the similarity with the largest numerical value, and set the standard image corresponding to the similarity with the largest numerical value as the current image, the standard image includes images of standard motor windings, standard bearings, standard gears, standard exhaust ports, and standard air inlets, and the current image includes images of motor windings, bearings, gears, exhaust ports, and air inlets; The fault prediction module obtains the current infrared sub-image, performs grayscale processing on the current infrared sub-image, calculates the average grayscale value of the current infrared sub-image, performs a difference calculation between the average grayscale value of the current infrared sub-image and the average grayscale value of the corresponding fault corresponding image to obtain a second difference value, sets a difference threshold value for a third value, and compares the second difference with the third value; When the second difference is less than or equal to the difference threshold, obtaining the corresponding confidence interval of the key parameter according to the second mapping relationship, and obtaining the corresponding historical fault type according to the first mapping relationship; When the second difference is greater than the difference threshold, jump to the next current infrared sub-image, and stop the jump operation when the second difference is less than or equal to the difference threshold; The fault prediction module outputs and displays the historical fault type and the current infrared sub-image on the OLED display.

2. The intelligent refrigeration equipment predictive maintenance system according to claim 1, characterized in that: The data acquisition module obtains the device model, retrieves the fault database, inputs the device model into the fault database, matches the historical fault type corresponding to the device model, the historical fault type includes bearing wear, gear wear, winding wear, overload, poor air intake and poor exhaust, and assigns a first number to the historical fault type, the first number is represented as , is a natural number; obtaining historical fault parameters corresponding to the historical fault type, wherein the historical fault parameters include bearing vibration frequency, gear vibration frequency, motor voltage, motor current, noise intensity, suction pressure, discharge pressure and winding vibration frequency.

3. The intelligent refrigeration equipment predictive maintenance system according to claim 1, characterized in that: The data analysis module performs a first preprocessing on the historical fault parameters, wherein the first preprocessing includes a process of removing duplicate values ​​and a normalization process, wherein the normalization process causes the values ​​of the historical fault parameters to be distributed between [0 and 1]; The pre-processed historical fault parameters are analyzed, key parameters are selected according to the analysis results, and the key parameters are second numbered, the second number corresponds to the first number, and the second number is expressed as: , is a natural number, is a natural number, among which, It is indicated as a subscript in the first number, and the second number corresponds to the first number; The selection logic of the key parameters includes: The historical fault type is set as the dependent variable, and the historical fault parameter is set as the independent variable. Multiple regression analysis is performed to obtain the coefficients corresponding to the historical fault parameters. The first value is set as the correlation threshold, and the coefficients corresponding to the historical fault parameters are sorted in descending order. The coefficients corresponding to the historical fault parameters after descending order are compared with the first value, and historical fault parameters greater than or equal to the first value are selected, and the historical fault parameters greater than or equal to the first value are set as key parameters.

4. The intelligent refrigeration equipment predictive maintenance system according to claim 2, characterized in that: When the historical fault type is bearing wear, the corresponding key parameters include bearing vibration frequency and noise intensity; When the historical fault type is gear wear, the corresponding key parameters include bearing vibration frequency, noise intensity, gear vibration frequency and noise intensity; When the historical fault type is winding wear, the corresponding key parameters include winding vibration frequency and motor current; When the historical fault type is overload, the corresponding key parameters are motor current and motor voltage; When the historical fault type is poor air intake, the corresponding key parameter is suction pressure; When the historical fault type is poor exhaust, the corresponding key parameter is the exhaust pressure.

Images