Rapid fault detection device and method for general air conditioning system

Abstract

The invention provides a rapid fault detection device and method for a general air conditioning system. The rapid fault detection device comprises: a housing; the data acquisition module is used for being in signal connection with an external data detection system arranged outside the air-conditioning system so as to acquire operation data of various operation parameters of the air-conditioning system in real time and pre-process the operation data; wherein the multiple operation parameters comprise vibration, wind speed, temperature, power and gas; the data processing module is in signal connection with the data acquisition module and is used for analyzing and processing the preprocessed operation data and obtaining a fault diagnosis result through a fault diagnosis model; and the display screen is arranged on the surface of the shell and is used for displaying the operation data and the fault diagnosis result. The problems that an existing air conditioner system is low in fault diagnosis efficiency and poor in diagnosis simplicity and convenience are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of air conditioning detection, and in particular to a general air conditioning system rapid fault detection device and method. Background Art

[0002] Traditional air conditioner fault diagnosis methods primarily include manual inspection and instrument testing. Manual inspection is the most basic diagnostic method, where maintenance personnel use their experience to determine the fault by observing the air conditioner's appearance, listening to sounds, and touching components. Instrument-based testing uses specialized tools for more accurate fault diagnosis. Typically, pressure gauges, thermometers, and ammeters are used to measure parameters such as pressure, temperature, and current in the air conditioner system to assist in fault diagnosis.

[0003] Traditional empirical diagnostic methods have the following problems: they rely on the rich experience and professional knowledge of maintenance personnel, which is difficult for ordinary users to master, resulting in increased maintenance costs; they are inefficient, and maintenance personnel need to check components one by one, which is time-consuming and labor-intensive, making it difficult to quickly locate faults and extending maintenance time; they have poor diagnostic accuracy, and relying solely on experience and simple tools can easily lead to misjudgment, resulting in incorrect maintenance directions and increased costs; they have poor applicability, and traditional experience may not be applicable to new air-conditioning systems or complex faults, making them difficult to diagnose accurately. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed to provide a general air-conditioning system rapid fault detection device and method that overcomes the above problems or at least partially solves the above problems, which can solve the problems of low efficiency and poor diagnostic simplicity of existing air-conditioning system fault diagnosis.

[0005] Specifically, the present invention provides a rapid fault detection device for a general air conditioning system, comprising:

[0006] case;

[0007] A data acquisition module, configured to connect to an external data detection system independently provided outside the air conditioning system to collect and pre-process operating data of various operating parameters of the air conditioning system in real time; wherein the various operating parameters include vibration, wind speed, temperature, power, and gas;

[0008] A data processing module, signal-connected to the data acquisition module, configured to analyze and process the pre-processed operating data and obtain a fault diagnosis result through a fault diagnosis model;

[0009] A display screen is provided on the surface of the housing and is used to display the operating data and the fault diagnosis results.

[0010] Optionally, the external data detection system includes:

[0011] A vibration sensor is provided on the housing of the air conditioner outdoor unit to obtain vibration data;

[0012] A wind speed sensor is installed at the air inlet and / or outlet of the outdoor unit to obtain wind speed data;

[0013] a temperature sensor, configured to be disposed at least one of an air inlet and / or air outlet of an outdoor air conditioner unit, an air inlet and / or air outlet of an indoor air conditioner unit, a condenser pipe, an evaporator inlet, an outdoor space where the outdoor air conditioner unit is located, and an indoor space where the indoor air conditioner unit is located, to obtain temperature data; the temperature data including at least one of an inlet temperature and / or an outlet temperature of the outdoor air conditioner unit, an inlet temperature and / or an outlet temperature of the indoor air conditioner unit, a surface temperature of an evaporator inlet pipe, a surface temperature of a condenser metal pipe, an indoor ambient temperature, and an outdoor ambient temperature;

[0014] A gas sensor is provided at the interface of the refrigerant pipeline to obtain gas data;

[0015] The power metering socket is used to be installed at the power input end of the air conditioning system to obtain power data.

[0016] Optionally, the temperature sensor includes:

[0017] Infrared temperature sensors are installed at the air inlet and outlet of the outdoor air conditioner and the air inlet and outlet of the indoor air conditioner to obtain the air inlet temperature and air outlet temperature of the outdoor air conditioner and the air inlet temperature and air outlet temperature of the indoor air conditioner; and / or

[0018] A patch temperature sensor is attached to the surface of the evaporator inlet pipe to obtain the surface temperature of the evaporator inlet pipe; and / or

[0019] A magnetic temperature sensor is attached to the surface of the condenser metal pipe to obtain the temperature of the condenser metal pipe surface; and / or

[0020] The wall-mounted temperature sensor is installed on the wall near the air inlet of the air conditioner indoor unit and / or on the side of the air conditioner outdoor unit mounting bracket to obtain the indoor ambient temperature and / or outdoor ambient temperature.

[0021] Optionally, the data acquisition module includes:

[0022] a multiplexer, connected to the external data detection system signal, for receiving the operation data;

[0023] a filter, signal-connected to the multiplexer, for pre-processing the operating data to remove noise and amplify the signal;

[0024] An analog-to-digital converter is connected to the filter signal signal, and is used to convert the pre-processed operating data from an analog signal to a digital signal, and transmit the digital signal to the data processing module through the output interface of the data acquisition module.

[0025] Optionally, the fault diagnosis model is a convolutional neural network model;

[0026] The data processing module includes:

[0027] A main control chip, wherein the main control chip has a built-in control program for controlling the multiplexer to switch different sensor channels, sequentially collecting digital signals of the operating data corresponding to each sensor channel, and performing data processing;

[0028] The coprocessor is connected to the main control chip signal to accelerate the execution speed of the algorithm;

[0029] Wireless communication module, used to transmit data to cloud servers in real time;

[0030] The data processing module is also configured to optimize the convolutional neural network model using the Adam optimization algorithm.

[0031] Optionally, the rapid fault detection device further includes:

[0032] A plurality of input interfaces are provided on the surface of the housing, and the plurality of input interfaces are respectively connected to the signals of the vibration sensor, the wind speed sensor, the temperature sensor, and the gas sensor;

[0033] A battery module, used to supply power to the data acquisition module and the data processing module;

[0034] The data acquisition module, the data processing module and the battery module are sequentially arranged at intervals from top to bottom.

[0035] On the other hand, the present invention further provides a method for rapid fault detection of a universal air-conditioning system, the method being based on any one of the rapid fault detection devices described above and comprising:

[0036] S100, installing the external data detection system on the running air conditioning system to be tested, and turning on the rapid fault detection device;

[0037] S200, the data acquisition module collects the operating data of the various operating parameters of the air-conditioning system to be tested within a preset time period through the external data detection system, and performs preprocessing;

[0038] S300, the data processing module analyzes and processes the pre-processed operating data, and obtains a fault diagnosis result through the fault diagnosis model;

[0039] S400: Outputting the fault diagnosis result to the display screen.

[0040] Optionally, S200 includes the following steps:

[0041] The external data detection system collects temperature data, wind speed data, power data, vibration data, and gas data of the air conditioning system in real time within a preset time period to form multiple one-dimensional time series data;

[0042] Merging the plurality of one-dimensional time series data into a first multi-dimensional time series data matrix;

[0043] Wavelet soft threshold filtering is performed on each column of the first multidimensional time series data matrix to remove noise and retain the main features of the signal, thereby obtaining a second multidimensional time series data matrix.

[0044] Optionally, the fault diagnosis model is a convolutional neural network model;

[0045] S300 includes the following steps:

[0046] S310, Gram angle field conversion:

[0047] Normalizing each column of data in the second multidimensional time series data matrix to the interval [-1, 1] to obtain a third multidimensional time series data matrix;

[0048] Mapping each column of data in the third multi-dimensional time series data matrix to a polar coordinate system to obtain a polar angle matrix;

[0049] Calculating the Gram angle sum field matrix and the Gram angle difference field matrix according to the polar angle matrix to obtain multi-dimensional GAF image data;

[0050] S320, Convolutional Neural Network Feature Extraction:

[0051] Input the multi-dimensional GAF image data into a convolutional neural network model;

[0052] Perform convolution operation through the convolution layer to extract local features and obtain feature maps;

[0053] Reducing the dimension of the feature map through a pooling layer;

[0054] Flatten the feature map into a one-dimensional vector and input it into the fully connected layer for classification and recognition;

[0055] According to the output of the fully connected layer, the fault type of the air-conditioning system to be tested is determined through probability distribution.

[0056] Optionally, the rapid fault detection method further includes:

[0057] S500, training and evaluating the convolutional neural network model:

[0058] Selecting a cross entropy loss function as the loss function of the convolutional neural network model to measure the difference between the output of the fully connected layer and the true label;

[0059] The Adam optimization algorithm is used to optimize the parameters of the fault diagnosis model;

[0060] Dividing the second multi-dimensional time series data matrix X' into a training set and a test set; training the convolutional neural network model using the training set data, calculating a loss function value through forward propagation, and then updating the model parameters through backpropagation; repeating the training process for multiple iterations until the loss function value of the convolutional neural network model converges;

[0061] Use the test set data to evaluate the trained convolutional neural network model and calculate the accuracy of the convolutional neural network model;

[0062] Based on the accuracy, the structure and parameters of the convolutional neural network model are optimized.

[0063] The universal air conditioning system rapid fault detection device and method of the present invention, on the one hand, features a data acquisition module and a data processing module, enabling automated data collection and analysis. This significantly reduces reliance on the personal experience and expertise of maintenance personnel, allowing even ordinary users to perform preliminary fault diagnosis. This significantly lowers the maintenance threshold and cost, reducing the cost of a single repair by 80%. Users can also track the health status of the air conditioning system through a history recording function. On the other hand, the rapid fault detection device comprehensively processes multiple operating parameters in real time and quickly generates diagnostic results, eliminating the inefficient process of manual item-by-item troubleshooting. This significantly improves fault location speed and overall diagnostic efficiency, significantly reducing maintenance time. Furthermore, compared to traditional methods that rely on single experience or simple tools, the present invention significantly improves diagnostic accuracy and reliability, effectively reducing the occurrence of misdiagnosis. Furthermore, because the fault diagnosis model is updateable and learnable, the rapid fault detection device offers greater adaptability and scalability to new air conditioning systems and complex fault modes, addressing the limited applicability of traditional empirical methods. Furthermore, the display directly presents data and diagnostic results, making operation and result reading more intuitive and convenient.

[0064] Furthermore, the present invention's external measurement point design allows all sensors to be installed externally, eliminating the need to puncture the air conditioning system's piping or dismantle its casing. This allows the rapid fault detection device to be compatible with a wide range of models and brands of air conditioning equipment, eliminating the high costs associated with retrofits and further reducing implementation costs. Furthermore, the present invention is particularly suitable for older air conditioning systems, reducing the number of older units scrapped.

[0065] Based on the following detailed description of specific embodiments of the present invention in conjunction with the accompanying drawings, those skilled in the art will become more aware of the above and other objects, advantages and features of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Hereinafter, some specific embodiments of the present invention will be described in detail in an exemplary and non-limiting manner with reference to the accompanying drawings. The same reference numerals in the accompanying drawings indicate the same or similar components or parts. It should be understood by those skilled in the art that these drawings are not necessarily drawn to scale. In the accompanying drawings:

[0067] Figure 1 is a schematic structural diagram of a rapid fault detection device for a general air conditioning system according to one embodiment of the present invention;

[0068] Figure 2 1 is a schematic internal structure diagram of a rapid fault detection device for a general air-conditioning system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0069] Refer to the following Figures 1 to 2 To describe the rapid fault detection device and detection method for a general air-conditioning system according to an embodiment of the present invention. In the description of this embodiment, it should be understood that the terms "first" and "second" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features, that is, include one or more of the features. In the description of the present invention, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. When a feature "includes or contains" one or some of the features it covers, unless otherwise specifically described, this indicates that other features are not excluded and may further include other features.

[0070] Unless otherwise expressly defined or limited, terms such as "disposed," "installed," "connected," "connected," "fixed," and "coupled" should be broadly interpreted. For example, they may refer to fixed or detachable connections, or integration; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two elements or interaction between two elements, unless otherwise expressly defined. A person of ordinary skill in the art should be able to understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0071] In addition, in the description of this embodiment, the first feature being "above" or "below" the second feature may include the first and second features being in direct contact, or may include the first and second features not being in direct contact but being in contact via another feature between them. That is, in the description of this embodiment, the first feature being "above," "above," and "above" the second feature includes the first feature being directly above or diagonally above the second feature, or simply indicates that the first feature is higher in level than the second feature. The first feature being "below," "below," or "below" the second feature may mean that the first feature is directly below or diagonally below the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0072] In the description of the present embodiment, reference to the terms "one embodiment," "some embodiments," "illustrative embodiments," "example," "specific example," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the exemplary expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any appropriate manner in any one or more embodiments or examples.

[0073] Figure 1 FIG. 1 is a schematic structural diagram of a rapid fault detection device for a general air conditioning system according to an embodiment of the present invention. Figure 1 As shown, and reference Figure 2 An embodiment of the present invention provides a rapid fault detection device for a general air-conditioning system, which includes a housing 1, a data acquisition module 2, a data processing module 3 and a display screen 5.

[0074] The data acquisition module 2 is connected to an external data detection system located outside the air conditioning system to collect and preprocess real-time operating data on various system operating parameters, including vibration, wind speed, temperature, power, and gas. The data processing module 3 is connected to the data acquisition module 2 and is responsible for analyzing and processing the preprocessed operating data and deriving fault diagnosis results using a fault diagnosis model. A display screen 5 is provided on the surface of the housing 1 and is used to display the operating data and fault diagnosis results.

[0075] Specifically, a mounting cavity is formed within the housing 1, within which the data acquisition module 2 and the data processing module 3 are disposed. The external data detection system is not part of the air conditioning system. When fault detection is required, it is installed at a corresponding location in the air conditioning system to collect data on the corresponding parameters.

[0076] When the rapid fault detection device of this embodiment is working, the data acquisition module 2 first collects the operating data of multiple key operating parameters such as vibration, wind speed, temperature, power, gas (refrigerant status) in real time through the external data detection system external to the air-conditioning system, and performs preliminary preprocessing (such as filtering, etc.) on these raw data. The preprocessed data is transmitted to the data processing module 3, which uses the built-in fault diagnosis model to perform in-depth analysis and calculation processing on the received multi-dimensional operating data, thereby obtaining a specific fault diagnosis result. The diagnostic results, together with the relevant operating parameter data, are finally intuitively presented to the user or maintenance personnel through the display screen 5 on the surface of the shell 1.

[0077] In this embodiment, the rapid fault detection device, comprising a data acquisition module 2 and a data processing module 3, enables automated data collection and analysis, significantly reducing reliance on the personal experience and expertise of maintenance personnel. This allows even ordinary users to perform preliminary fault diagnosis, significantly lowering the maintenance threshold and cost, reducing the cost of a single repair by 80%. Users can also trace the health status of the air conditioning system through a history recording function. Furthermore, the rapid fault detection device comprehensively processes multiple operating parameters in real time and rapidly generates diagnostic results, eliminating the inefficient process of manual item-by-item troubleshooting. This significantly improves fault location speed and overall diagnostic efficiency, significantly reducing repair time. Furthermore, compared to traditional methods that rely on single experience or simple tools, this embodiment significantly improves diagnostic accuracy and reliability, effectively reducing the occurrence of misdiagnosis. Furthermore, because the fault diagnosis model is updatable and learnable, the rapid fault detection device possesses greater adaptability and scalability to new air conditioning systems and complex fault modes, addressing the limited applicability of traditional empirical methods. Furthermore, the display screen 5 directly presents data and diagnostic results, making operation and result reading more intuitive and convenient.

[0078] Furthermore, the rapid fault detection device can diagnose the following fault types: evaporator obstruction, compressor overload, condenser obstruction, fan failure, refrigerant shortage or leakage, heat exchanger obstruction, capacitor aging or circuit overload, and bearing wear or refrigerant liquid shock.

[0079] In some optional embodiments of the present invention, the external data detection system includes a vibration sensor, a wind speed sensor, a temperature sensor, a gas sensor, and a power metering socket. The vibration sensor is used to be installed on the housing of the air conditioner outdoor unit to obtain vibration data of the air conditioner outdoor unit housing. The wind speed sensor is used to be installed at the air inlet and / or air outlet of the outdoor unit to obtain wind speed data. The temperature sensor is used to be installed at at least one of the air inlet and / or air outlet of the air conditioner outdoor unit, the air inlet and / or air outlet of the air conditioner indoor unit, the condenser pipe, the evaporator inlet, the outdoor space where the air conditioner outdoor unit is located, and the indoor space where the air conditioner indoor unit is located to obtain temperature data; the temperature data includes at least one of the inlet and / or outlet temperature of the air conditioner outdoor unit, the inlet and / or outlet temperature of the air conditioner indoor unit, the temperature of the evaporator inlet pipe surface, the temperature of the condenser metal pipe surface, the indoor ambient temperature, and the outdoor ambient temperature. The gas sensor is used to be installed at the refrigerant pipe interface (such as the four-way valve or expansion valve connection) to obtain gas data. The power metering socket is used to be installed at the power input end of the air conditioning system to obtain power data.

[0080] The rapid fault detection device may include an external data detection system, or may not include an external data detection system.

[0081] Specifically, there are no specific requirements for the vibration sensor, wind speed sensor, temperature sensor, gas sensor, and power metering socket, and those skilled in the art can select them as needed.

[0082] The vibration sensor is preferably a microelectromechanical system (MEMS) vibration sensor, which can capture high-frequency harmonics and low-frequency impact signals from compressor operation. Using a magnetic base that tightly adheres to the air conditioner's outdoor unit casing, it converts the metal-conducted vibration signal into a digital spectrum. The MEMS vibration sensor, combined with a historical database and spectrum templates, can automatically distinguish mechanical faults such as bearing wear and loose anchor bolts. This sensor can detect specific frequency vibration harmonics generated by subtle wear in the compressor's internal bearings months in advance, providing early warning of faults, addressing the shortcomings of traditional detection methods and ensuring the healthy operation of the air conditioner's mechanical system.

[0083] The preferred wind speed sensor is a vane-type wind speed sensor, installed at the edge of the air conditioner's outdoor unit's protective screen. Using the rotating stainless steel impeller and the Hall effect, it converts speed into an electrical signal to monitor the difference in wind speed at the inlet and outlet. If the wind speed difference exceeds the specified value, the system will detect dust accumulation in the heat exchanger or fan aging and provide a cleaning recommendation. The vane-type wind speed sensor is easy to install, secured by a metal bracket and oriented perpendicular to the airflow, ensuring precise measurement. Its anodized surface provides resistance to heat and humidity, ensuring high reliability.

[0084] Preferably, the temperature sensor includes at least one of an infrared temperature sensor, a patch temperature sensor, a magnetic temperature sensor, and a wall-mounted temperature sensor. Further preferably, the temperature sensor includes an infrared temperature sensor, a patch temperature sensor, a magnetic temperature sensor, and a wall-mounted temperature sensor.

[0085] Specifically, the temperature sensor includes multiple platinum resistance sensors, which offer high accuracy (±0.5°C) and excellent stability. The magnetic temperature sensor's probe attaches to the surface of the condenser's metal pipe, making it easy to install and remove. The magnetic temperature sensor is used to obtain the surface temperature of the condenser's metal pipe. With a temperature resistance exceeding 120°C, it can stably capture temperature changes and trigger overload protection when the temperature is abnormal. The patch temperature sensor can be directly attached to the surface of the evaporator's inlet pipe (copper pipe) using quick-drying adhesive, eliminating the need for drilling. This non-destructive adhesive monitoring of refrigerant status and combining power data to accurately locate leaks simplifies traditional leak detection. The wall-mounted temperature sensor can be fixed to the wall near the indoor unit's air inlet using screws or double-sided tape, 10 to 20 cm from the inlet, to obtain the indoor ambient temperature. The wall-mounted temperature sensor can also be placed on the side of the air conditioner's outdoor unit mounting bracket to avoid direct sunlight or interference from hot air, to obtain the outdoor ambient temperature.

[0086] The temperature sensor also includes an infrared temperature sensor, which can be installed at the air inlet and / or air outlet of the air conditioner outdoor unit, as well as the air inlet and / or air outlet of the air conditioner indoor unit. Preferably, the infrared temperature sensor is installed at the air inlet and air outlet of the air conditioner outdoor unit and the air conditioner indoor unit to obtain the air inlet temperature and air outlet temperature of the air conditioner outdoor unit and the air inlet temperature and air outlet temperature of the air conditioner indoor unit, and then obtain the air inlet and outlet temperature difference of the outdoor unit and the indoor unit.

[0087] Gas sensors are used to detect refrigerant leakage concentrations or the rate of change in refrigerant concentration. These gas sensors are preferably semiconductor-type, with probes that selectively respond to molecules of refrigerants such as R22 and R410A. During testing, the probe scans along the pipeline, displaying concentration data in real time. This can accurately locate minute leaks at valve interfaces or welds, reducing refrigerant replenishment by 40% and minimizing ozone depletion.

[0088] When using the power metering socket, plugging the air conditioner into the power metering socket allows real-time monitoring of voltage, current, power factor, and active power, indirectly reflecting the operating status of the air conditioning system. The power metering socket can synchronize real-time power data to the rapid fault detection device host via Bluetooth or Wi-Fi wireless transmission.

[0089] In some optional embodiments of the present invention, the rapid fault detection device further includes a switch button 7 to facilitate the control of opening or closing the device.

[0090] In some optional embodiments of the present invention, the rapid fault detection device further includes a battery module 4, which is used to supply power to the data acquisition module 2 and the data processing module 3. During the entire fault detection process, the rapid fault detection device is powered by the built-in battery module 4, ensuring its independent operation.

[0091] In some optional embodiments of the present invention, the data acquisition module 2, the data processing module 3, and the battery module 4 are sequentially spaced from top to bottom. This embodiment effectively avoids heat conduction between modules through the above-mentioned layered isolation, improves heat dissipation efficiency, and ensures the accuracy of data acquisition and processing.

[0092] In some optional embodiments of the present invention, the rapid fault detection device further includes multiple input interfaces, which are arranged on the surface of the shell 1 and are respectively connected to the vibration sensor, wind speed sensor, temperature sensor, and gas sensor signals.

[0093] Specifically, the input interfaces are connected to the data acquisition module 2. The number of input interfaces corresponds to the number of external sensors.

[0094] Preferably, the plurality of input interfaces are placed at the upper portion of the side surface of the housing 1 and are sequentially spaced from front to back.

[0095] The multiple input interfaces are respectively a first input interface 61, a second input interface 62, a third input interface 63, a fourth input interface 64 and a fifth input interface 65. The first input interface 61 is connected to the wind speed sensor, the second input interface 62 is connected to the gas sensor, the third input interface 63 is connected to the infrared temperature sensor, the fourth input interface 64 is connected to the vibration sensor, and the fifth input interface 65 is connected to the magnetic temperature sensor, the patch temperature sensor and the wall-mounted temperature sensor.

[0096] This embodiment achieves fast and stable direct connection with various sensors (vibration, wind speed, temperature, gas, etc.) by setting up multiple dedicated input interfaces on the surface of the shell 1, significantly improving the operating convenience of the detection device and the efficiency and reliability of multi-parameter synchronous acquisition, while eliminating the error risks that may be introduced by additional switching links.

[0097] In some optional embodiments of the present invention, the data acquisition module 2 includes a multiplexer, a filter, and an analog-to-digital converter. The multiplexer is connected to an external data detection system for receiving operating data; the filter is connected to the multiplexer for preprocessing the operating data to remove noise and amplify the signal; and the analog-to-digital converter is connected to the filter for converting the preprocessed operating data from an analog signal to a digital signal, and transmitting the digital signal to the data processing module 3 via the output interface of the data acquisition module 2.

[0098] Specifically, the multiplexer efficiently integrates data streams from multiple sensors into a single data stream, which not only simplifies the hardware interface but also significantly reduces system complexity and cost.

[0099] This embodiment integrates a multiplexer, a filter, and an analog-to-digital converter within the data acquisition module 2 to achieve efficient selection and acquisition, effective noise suppression, signal conditioning, and high-precision digital conversion of multi-channel sensor signals, thereby significantly improving the quality, reliability, and processing efficiency of the original operating data, and laying a solid foundation for accurate fault diagnosis by the subsequent data processing module 3.

[0100] In some optional embodiments of the present invention, the fault diagnosis model is a convolutional neural network (CNN) model. The data processing module 3 includes a main control chip and a coprocessor. The main control chip has a built-in control program for controlling the multiplexer to switch between different sensor channels, sequentially collecting the digital signals (ADC data) of the operating data corresponding to each sensor channel, and performing data processing. The coprocessor is signal-connected to the main control chip and is used to accelerate the execution speed of the algorithm. Specifically, the coprocessor significantly accelerates the execution speed of the algorithm, working in conjunction with the main processor to accelerate computing tasks.

[0101] Preferably, the model of the main control chip is STM32H743VIT6.

[0102] This embodiment can achieve intelligent, high-precision and efficient fault diagnosis by utilizing the main control chip and the coprocessor.

[0103] In some optional embodiments of the present invention, the data processing module 3 further includes a wireless communication module for transmitting data to a cloud server in real time.

[0104] Preferably, the wireless communication module can be an ESP32-C3 Wi-Fi module.

[0105] This embodiment achieves real-time cloud-based synchronization of detection data and fault diagnosis results by adding a wireless communication module. On the one hand, it breaks the physical distance limitation and supports remote monitoring and efficient management; on the other hand, it provides a data basis for cloud-based big data analysis and iterative optimization of diagnostic models.

[0106] In some optional embodiments of the present invention, the data processing module 3 is further configured to: optimize the convolutional neural network model using the Adam optimization algorithm.

[0107] This embodiment uses the Adam optimization algorithm to automatically tune the hyperparameters of the convolutional neural network model. On the one hand, it significantly improves the model convergence speed and diagnostic accuracy, reducing training costs; on the other hand, it enhances the model's generalization ability and robustness to complex fault modes, avoiding falling into local optimal solutions, thereby comprehensively improving the accuracy and reliability of fault diagnosis.

[0108] An embodiment of the present invention further provides a method for rapid fault detection of a universal air-conditioning system. The method is based on any of the rapid fault detection devices described above and comprises the following steps:

[0109] S100, installing an external data detection system on the operating air conditioning system to be tested and turning on the rapid fault detection device;

[0110] S200, the data acquisition module 2 collects operating data of various operating parameters of the air-conditioning system to be tested within a preset time period through an external data detection system and performs preprocessing;

[0111] S300, the data processing module 3 analyzes and processes the pre-processed operating data and obtains a fault diagnosis result through a fault diagnosis model;

[0112] S400: Output the fault diagnosis result to display screen 5.

[0113] This embodiment leverages the hardware advantages of the aforementioned device to achieve rapid diagnosis of air conditioner faults through a process of automatic data collection, intelligent analysis, and result output. This simplifies the complex diagnostic process into a single click, significantly reducing both user experience and manual effort. Furthermore, leveraging the device's multidimensional data processing capabilities and optimization model, it ensures both rapid and accurate diagnostic results, significantly improving maintenance efficiency and accuracy.

[0114] In some optional embodiments of the invention, S200 includes the following steps:

[0115] S201, the data acquisition module 2 collects the temperature data, wind speed data, power data, vibration data, and gas data of the air conditioning system in real time within a preset time period through an external data detection system to form multiple one-dimensional time series data. Among them, the temperature data collected by the temperature sensor is [T1, T2, ..., Tn ], the wind speed data collected by the wind speed sensor is [W1,W2,…,W n ], the power data collected by the power sensor is [P1, P2, ..., P n ], the vibration data collected by the vibration sensor is [V1, V2,…, V n ], the gas concentration data collected by the gas sensor is [G1,G2,…,G n ].

[0116] S202: The data acquisition module 2 combines the multiple one-dimensional time series data into a first multi-dimensional time series data matrix X.

[0117]

[0118] S203 , the filter performs wavelet soft threshold filtering on each column of the first multi-dimensional time series data matrix to remove noise and retain the main features of the signal, thereby obtaining a second multi-dimensional time series data matrix X′.

[0119] Specifically, taking temperature data as an example, the wavelet soft threshold filtering process is as follows:

[0120] (1) Select an appropriate wavelet basis function and decomposition layer number 1.

[0121] (2) For the temperature time series data [T1, T2, ..., T n ]Perform wavelet decomposition to obtain wavelet coefficients cA (approximation coefficient) and cD (detail coefficient).

[0122] (3) Perform soft threshold processing on the wavelet coefficient cD, the formula is: cD′=singn(cD)·max(|cD|-λ,0), where λ is the threshold.

[0123] (4) Using the soft threshold processed wavelet coefficients cA and cD' to perform inverse wavelet transform, the denoised temperature time series data [T'1, T'2, ..., T' n ].

[0124] (5) Similarly, the time series data of wind speed, power, vibration, gas and other parameters are processed by wavelet soft threshold filtering to obtain the denoised multi-dimensional time series data matrix X'.

[0125] In some optional embodiments of the invention, the fault diagnosis model is a convolutional neural network model.

[0126] S300 includes the following steps: S310, Gram angular field transformation; S320, convolutional neural network feature extraction.

[0127] S310 specifically includes the following steps:

[0128] S311, normalization: normalize each column of data in the second multidimensional time series data matrix X' to the interval [-1, 1] to obtain a third multidimensional time series data matrix X".

[0129] Taking temperature data as an example, the normalization formula is:

[0130]

[0131] Among them, T i ″ represents the normalized temperature data, max(T′) and max(T′) represent the minimum and maximum values ​​of the denoised temperature data, respectively. Similarly, the data of wind speed, power, vibration, gas and other parameters are normalized to obtain the normalized third multidimensional time series data matrix X".

[0132] S312, polar coordinate mapping: Map each column of data in the third multi-dimensional time series data matrix X'' to a polar coordinate system to obtain a polar angle matrix Φ.

[0133] Specifically, taking temperature data as an example, the polar coordinate mapping formula is:

[0134] φ i =arccos(T i ″)

[0135] Where Φi represents the polar angle corresponding to the temperature data. Similarly, the data of wind speed, power, vibration, gas and other parameters are mapped to polar coordinates to obtain the polar angle matrix Φ.

[0136] S313, generating GASF and GADF matrices: calculating the Gram angle sum field matrix and the Gram angle difference field matrix according to the polar angle matrix Φ, and obtaining multi-dimensional GAF image data.

[0137] Taking temperature data as an example, the calculation formulas for GASF and GADF are:

[0138] GASF ij =cos(φ i +φ j )

[0139] GADF ij = sin(φ i -φ j )

[0140] Among them, GASFij and GADF ijRepresent the elements in the GASF and GADF matrices respectively. Similarly, the GASF and GADF matrices are calculated for the data of wind speed, power, vibration, gas and other parameters to obtain multi-dimensional GAF image data.

[0141] S320 includes the following specific steps:

[0142] S321, data input: multi-dimensional GAF image data is input into the convolutional neural network model.

[0143] Specifically, assume that the input GAF image data is I, and its dimension is C×H×W, where C represents the number of channels (corresponding to different parameters such as temperature, wind speed, etc.), and H and W represent the height and width of the image, respectively.

[0144] S322, performing a convolution operation through a convolution layer to extract local features and obtain a feature map.

[0145] Specifically, taking the first convolutional layer as an example, the convolution operation formula is:

[0146]

[0147] in, Represents the value of the output feature map of the lth convolutional layer at position (i, j), C l-1 represents the number of channels in the l-1 layer, K represents the size of the convolution kernel, Represents the weight of the convolution kernel of the lth layer, I i+u-1,j+v-1 Represents the value of the input feature map at position (i+u-1,j+v-1), b(l) represents the bias term, and f represents a nonlinear activation function (such as the ReLU function).

[0148] S323: Reduce the dimension of the feature map through a pooling layer to reduce the amount of calculation.

[0149] Specifically, taking maximum pooling as an example, the pooling operation formula is:

[0150]

[0151] in, It represents the value of the output feature map of the l-th pooling layer at position (i, j), and S represents the size of the pooling window.

[0152] S324, after feature extraction by multiple convolutional layers and pooling layers, the feature map is flattened into a one-dimensional vector and input into the fully connected layer for classification and recognition.

[0153] Specifically, assuming that the flattened feature vector is F, the output calculation formula of the fully connected layer is:

[0154] y=σ(W·F+b)

[0155] Among them, y represents the output of the fully connected layer, σ represents the activation function (such as the softmax function), W and b represent the weight and bias terms of the fully connected layer, respectively.

[0156] S324: Determine the fault type of the air conditioning system under test using the probability distribution based on the output of the fully connected layer. For example, if the value of the kth element in the output y is the largest, it is determined that the air conditioning system has a kth type of fault.

[0157] This embodiment achieves three beneficial effects by integrating Gram Angular Field (GAF) and Convolutional Neural Network (CNN) technology: 1. Efficient feature encoding of multi-dimensional time series data: GAF is used to convert multi-parameter time series data such as temperature and wind speed into spatial correlation images (GASF / GADF), retaining the coupling relationship and time dependency between parameters, and providing structured input for CNN. 2. Automatic extraction capability of deep fault features: CNN automatically learns local fault modes (such as abnormal vibration and temperature distribution distortion) in GAF images through a convolution-pooling hierarchical structure, overcoming the limitations of traditional methods that rely on manual feature engineering. 3. Improved diagnostic accuracy for complex faults: Combining the spatial encoding capability of GAF with the visual perception advantages of CNN, the recognition accuracy of complex faults (such as vibration + refrigerant leakage) is significantly improved, and the false alarm rate is significantly reduced.

[0158] In some optional embodiments of the present invention, the rapid fault detection method further includes the following steps: S500, training and evaluating a convolutional neural network model. Specifically, S500 is performed after S300, and there is no order between S500 and S400.

[0159] S500 specifically includes the following steps:

[0160] S501, selecting a cross entropy loss function as a loss function of the convolutional neural network model to measure the difference between the output of the fully connected layer and the true label.

[0161] Specifically, assuming the model output is y and the true label is t, the cross entropy loss function formula is:

[0162]

[0163] Where C represents the total number of fault categories, t i Represents the one-hot encoding of the true label, y i Represents the probability distribution of the model output.

[0164] S502: Optimize the parameters of the fault diagnosis model using the Adam optimization algorithm, where Adam stands for Adaptive Moment Estimation.

[0165] Specifically, the update formula corresponding to the zebra optimization algorithm is:

[0166]

[0167] Among them, m t and v t are the first-order moment estimate and the second-order moment estimate of the current time step, m t-1 and v t-1 is the first-order moment estimate and the second-order moment estimate of the previous time step, β1 and β2 are the decay rates of the first-order moment estimate and the second-order moment estimate, respectively. is the gradient of the loss function at the current time step, and is the corrected first-order and second-order moment estimate, t is the current time step (number of iterations), which is used to correct the zero bias problem at the initial moment. α represents the learning rate, and ε is a minimum value used to prevent the denominator from being zero. θ t Updated model parameters, θ t-1 is the parameter of the previous time step.

[0168] S503, training process: dividing the second multidimensional time series data matrix X' into a training set and a test set; using the training set data to train the convolutional neural network model, calculating the loss function value through forward propagation, and then updating the model parameters through backpropagation; repeating the training process for multiple iterations until the loss function value of the convolutional neural network model converges.

[0169] S504, evaluation process: Use the test set data to evaluate the trained convolutional neural network model and calculate the accuracy of the convolutional neural network model.

[0170] Specifically, assuming that there are N samples in the test set and the number of samples correctly diagnosed by the model is M, the accuracy calculation formula of the model is:

[0171]

[0172] S505: Optimize the structure and parameters of the convolutional neural network model based on the accuracy.

[0173] This embodiment achieves continuous optimization and reliability assurance of fault diagnosis model performance by establishing a standardized, model training, and evaluation process. Since it fully covers the entire chain of loss function selection (cross entropy) → parameter optimization (Adam optimization algorithm) → data set partitioning → iterative training → accuracy evaluation → structural tuning, on the one hand, it significantly improves the model's ability to fit complex fault patterns and convergence efficiency, achieving a diagnostic accuracy of 94.65%. On the other hand, the objective evaluation mechanism (S504) and dynamic feedback optimization (S505) based on the test set accuracy fundamentally eliminate the risk of model overfitting, ensuring that it maintains high generalization on unknown fault data. Furthermore, the coordinated application of the cross entropy loss function and the intelligent optimization algorithm significantly accelerates the model's learning ability for class-imbalanced fault samples (such as rare faults), reducing the missed diagnosis rate to below 3%, thereby providing an intelligent diagnosis solution for air conditioning systems that combines high precision, strong robustness, and engineering practicality.

[0174] In some optional embodiments of the present invention, the rapid fault detection method further includes the following steps: S600, outputting the fault diagnosis result and operation data to a cloud server. There is no existing sequence between S600 and S400.

[0175] In summary, the device and method of the present invention have the following beneficial effects:

[0176] 1. Cost reduction: The external measurement point design eliminates the need for large-scale modifications to existing air-conditioning systems. It can be adapted to various models and brands of air-conditioning equipment, avoiding the high costs associated with modifications and further reducing implementation costs.

[0177] 2. Improved efficiency:

[0178] High-performance hardware and optimized algorithms: The device is equipped with a high-performance main control chip and coprocessor, combined with optimized algorithms, to achieve efficient data acquisition and processing. In the data acquisition process, the multiplexer significantly improves collection efficiency, allowing data to enter the processing stage quickly.

[0179] During the data processing process, from converting time series data into GAF images to using CNN and wavelet filtering for feature extraction, fusion and analysis, the entire process is efficient and fast, greatly shortening fault diagnosis time and improving maintenance efficiency.

[0180] Real-time monitoring and instant feedback: Utilizing a wireless communication module, the device monitors the air conditioner's operating status in real time and provides timely feedback to the user. Users can view the air conditioner's operating status in real time through their terminal device and quickly respond to fault alerts, effectively reducing downtime and ensuring the system's continued stable operation.

[0181] 3. Improved accuracy:

[0182] Innovative Model and Structure: This invention utilizes a lightweight space-time fusion model, combined with GAF and CNN architecture, to accurately capture dynamic changes in air conditioner operating status. GAF transformation effectively enhances the signal's periodicity, removes redundant information, and improves the model's noise immunity and adaptability, resulting in more accurate and reliable fault diagnosis results.

[0183] 4. Intelligent upgrade:

[0184] Adaptive Learning and Predictive Maintenance: The device features adaptive learning capabilities, optimizing diagnostic models based on continuously collected data and enabling predictive maintenance. Through in-depth analysis of air conditioner operating data, it proactively identifies potential faults and issues alerts, prompting users to perform preventive maintenance. This feature not only extends the life of the equipment but also reduces maintenance costs.

[0185] Remote monitoring and data sharing: The wireless communication module supports remote monitoring and data sharing. Users can view the air conditioner's operating status in real time, receive fault alerts, and perform remote management through mobile phones, computers, and other terminal devices. This improves the intelligent level of air conditioning system management and provides users with a more convenient and efficient user experience.

[0186] Improved reliability and adaptability:

[0187] Multiple fault identification and suggestions: The device can accurately identify a variety of common air conditioning faults, such as evaporator obstruction, insufficient refrigerant, fan failure, compressor overload, etc., and provide specific maintenance suggestions for each fault to help users quickly solve the problem.

[0188] Stable operation and anti-interference: The use of advanced filtering technology and anti-interference design ensures that the device can still operate stably in complex field environments, provide reliable diagnostic results, and adapt to different working scenarios and environmental conditions.

[0189] 5. Portability and low power consumption:

[0190] Portable design: The device adopts a portable design with a compact size, which is easy to carry and install, and can easily adapt to the deployment requirements of small spaces and different on-site environments.

[0191] Low-power operation: The optimized power management strategy ensures that the device can still operate stably in low-power mode, extending battery life, reducing usage costs, and further improving the practicality and economy of the device.

[0192] At this point, those skilled in the art will recognize that, although a number of exemplary embodiments of the present invention have been shown and described in detail herein, many other variations or modifications consistent with the principles of the present invention may be directly determined or derived from the disclosure of the present invention without departing from the spirit and scope of the present invention. Therefore, the scope of the present invention should be understood and deemed to cover all such other variations or modifications.

Claims

1. A rapid fault detection device for a general air conditioning system, characterized in that: include: case; A data acquisition module, configured to connect to an external data detection system independently provided outside the air conditioning system to collect and pre-process operating data of various operating parameters of the air conditioning system in real time; wherein the various operating parameters include vibration, wind speed, temperature, power, and gas; A data processing module, signal-connected to the data acquisition module, configured to analyze and process the pre-processed operating data and obtain a fault diagnosis result through a fault diagnosis model; A display screen is provided on the surface of the housing and is used to display the operating data and the fault diagnosis results.

2. The rapid fault detection device according to claim 1, characterized in that: The external data detection system includes: A vibration sensor is provided on the housing of the air conditioner outdoor unit to obtain vibration data; A wind speed sensor is installed at the air inlet and / or outlet of the outdoor unit to obtain wind speed data; a temperature sensor, configured to be disposed at least one of an air inlet and / or air outlet of an outdoor air conditioner unit, an air inlet and / or air outlet of an indoor air conditioner unit, a condenser pipe, an evaporator inlet, an outdoor space where the outdoor air conditioner unit is located, and an indoor space where the indoor air conditioner unit is located, to obtain temperature data; the temperature data including at least one of an inlet temperature and / or an outlet temperature of the outdoor air conditioner unit, an inlet temperature and / or an outlet temperature of the indoor air conditioner unit, a surface temperature of an evaporator inlet pipe, a surface temperature of a condenser metal pipe, an indoor ambient temperature, and an outdoor ambient temperature; A gas sensor is provided at the interface of the refrigerant pipeline to obtain gas data; The power metering socket is used to be installed at the power input end of the air conditioning system to obtain power data.

3. The rapid fault detection device according to claim 2, characterized in that: The temperature sensor comprises: Infrared temperature sensors are installed at the air inlet and outlet of the outdoor air conditioner and the air inlet and outlet of the indoor air conditioner to obtain the air inlet temperature and air outlet temperature of the outdoor air conditioner and the air inlet temperature and air outlet temperature of the indoor air conditioner; and / or A patch temperature sensor is attached to the surface of the evaporator inlet pipe to obtain the surface temperature of the evaporator inlet pipe; and / or A magnetic temperature sensor is attached to the surface of the condenser metal pipe to obtain the temperature of the condenser metal pipe surface; and / or The wall-mounted temperature sensor is installed on the wall near the air inlet of the air conditioner indoor unit and / or on the side of the air conditioner outdoor unit mounting bracket to obtain the indoor ambient temperature and / or outdoor ambient temperature.

4. The rapid fault detection device according to claim 1, characterized in that: The data acquisition module includes: a multiplexer, connected to the external data detection system signal, for receiving the operation data; a filter, signal-connected to the multiplexer, for pre-processing the operating data to remove noise and amplify the signal; An analog-to-digital converter is connected to the filter signal signal, and is used to convert the pre-processed operating data from an analog signal to a digital signal, and transmit the digital signal to the data processing module through the output interface of the data acquisition module.

5. The rapid fault detection device according to claim 4, characterized in that: The fault diagnosis model is a convolutional neural network model; The data processing module includes: A main control chip, wherein the main control chip has a built-in control program for controlling the multiplexer to switch different sensor channels, sequentially collecting digital signals of the operating data corresponding to each sensor channel, and performing data processing; The coprocessor is connected to the main control chip signal to accelerate the execution speed of the algorithm; Wireless communication module, used to transmit data to cloud servers in real time; The data processing module is also configured to optimize the convolutional neural network model using the Adam optimization algorithm.

6. The rapid fault detection device according to claim 2, characterized in that: The rapid fault detection device also includes: A plurality of input interfaces are provided on the surface of the housing, and the plurality of input interfaces are respectively connected to the signals of the vibration sensor, the wind speed sensor, the temperature sensor, and the gas sensor; A battery module, used to supply power to the data acquisition module and the data processing module; The data acquisition module, the data processing module and the battery module are sequentially arranged at intervals from top to bottom.

7. A rapid fault detection method for a general air conditioning system, characterized in that: The method is based on the rapid fault detection device according to any one of claims 1 to 6, and the method includes: S100, installing the external data detection system on the running air conditioning system to be tested, and turning on the rapid fault detection device; S200, the data acquisition module collects the operating data of the various operating parameters of the air-conditioning system to be tested within a preset time period through the external data detection system, and performs preprocessing; S300, the data processing module analyzes and processes the pre-processed operating data, and obtains a fault diagnosis result through the fault diagnosis model; S400: Outputting the fault diagnosis result to the display screen.

8. The rapid fault detection method according to claim 7, characterized in that: S200 includes the following steps: The external data detection system collects temperature data, wind speed data, power data, vibration data, and gas data of the air conditioning system in real time within a preset time period to form multiple one-dimensional time series data; Merging the plurality of one-dimensional time series data into a first multi-dimensional time series data matrix; Wavelet soft threshold filtering is performed on each column of the first multidimensional time series data matrix to remove noise and retain the main features of the signal, thereby obtaining a second multidimensional time series data matrix.

9. The rapid fault detection method according to claim 8, characterized in that: The fault diagnosis model is a convolutional neural network model; S300 includes the following steps: S310, Gram angle field conversion: Normalizing each column of data in the second multidimensional time series data matrix to the interval [-1, 1] to obtain a third multidimensional time series data matrix; Mapping each column of data in the third multi-dimensional time series data matrix to a polar coordinate system to obtain a polar angle matrix; Calculating the Gram angle sum field matrix and the Gram angle difference field matrix according to the polar angle matrix to obtain multi-dimensional GAF image data; S320, Convolutional Neural Network Feature Extraction: Input the multi-dimensional GAF image data into a convolutional neural network model; Perform convolution operation through the convolution layer to extract local features and obtain feature maps; Reducing the dimension of the feature map through a pooling layer; Flatten the feature map into a one-dimensional vector and input it into the fully connected layer for classification and recognition; According to the output of the fully connected layer, the fault type of the air-conditioning system to be tested is determined through probability distribution.

10. The rapid fault detection method according to claim 9, characterized in that: Also includes: S500, training and evaluating the convolutional neural network model: Selecting a cross entropy loss function as the loss function of the convolutional neural network model to measure the difference between the output of the fully connected layer and the true label; The Adam optimization algorithm is used to optimize the parameters of the fault diagnosis model; Dividing the second multi-dimensional time series data matrix X' into a training set and a test set; training the convolutional neural network model using the training set data, calculating a loss function value through forward propagation, and then updating the model parameters through backpropagation; repeating the training process for multiple iterations until the loss function value of the convolutional neural network model converges; Use the test set data to evaluate the trained convolutional neural network model and calculate the accuracy of the convolutional neural network model; Based on the accuracy, the structure and parameters of the convolutional neural network model are optimized.

Images