Pump lamp tube fault accurate diagnosis system based on multi-parameter data analysis

Abstract

The invention discloses a pump lamp tube fault accurate diagnosis system based on multi-parameter data analysis, and relates to the technical field of industrial automation monitoring, and the system comprises a multi-parameter acquisition and unique fault judgment algorithm, a multi-parameter acquisition module is connected with a high-precision current transformer, an infrared temperature sensor, a miniature pressure sensor, a photoelectric sensor and the like, and the multi-parameter acquisition module is connected with the unique fault judgment algorithm; parameters such as current, voltage, lamp tube surface temperature, internal gas pressure and luminous intensity during operation of a pumping lamp tube are accurately acquired, denoising and filtering parameters and a normalization method based on lamp tube rated parameter values are set by the preprocessing unit, and a fault diagnosis model is constructed by the data analysis module by selecting multiple learning algorithms. The collected and preprocessed parameter vector is X, a weight vector W and a bias item b are obtained through model training, a comprehensive score S is calculated, a fault judgment threshold T is set through a large amount of sample data training, and if S is larger than or equal to T, it is judged that a lamp tube has a fault, and a fault type mark is output; if Slt; and T, judging normal operation of the lamp tube.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial automation monitoring, and particularly to a precise fault diagnosis system for pump tubes based on multi-parameter data analysis. Background Art

[0002] As a key component in numerous lighting and specific industrial and scientific research fields (such as laser pumping, fluorescence detection equipment, etc.), the stable operation of the pump tube is crucial for ensuring the performance, reliability, and continuity of the entire system. In traditional application scenarios, once a pump tube fails, it often leads to equipment shutdown, production interruption, causing high economic losses, and may even affect the accuracy of scientific research experiment data with extremely high stability requirements, resulting in inestimable scientific research losses.

[0003] Currently, for the monitoring and fault diagnosis of the operating state of pump tubes, the industry mainly relies on measuring the current or voltage parameters of the pump tubes. The working current of the tubes is detected by a current transformer. When the current value exceeds the preset normal range, it is judged that the tube may have a fault. However, this single-parameter monitoring method has significant limitations: on the one hand, the electrical characteristics of the tubes vary greatly under different working conditions (such as ambient temperature, power supply voltage fluctuations, and the moment of equipment startup, etc.). Simply relying on current or voltage thresholds for judgment is extremely prone to false alarms, misjudging normal working condition fluctuations as faults; on the other hand, some initial subtle faults inside the tubes, such as local slight aging of the filament and the initial stage of cathode sputtering, do not immediately show obvious abnormalities in current and voltage, resulting in the inability to detect potential problems in a timely manner and missing the best maintenance opportunity. Therefore, we propose a precise fault diagnosis system for pump tubes based on multi-parameter data analysis to solve the above problems. Summary of the Invention

[0004] The purpose of the present invention is to provide a precise fault diagnosis system for pump tubes based on multi-parameter data analysis to solve the technical problem of inaccurate single-parameter monitoring mentioned in the above background art.

[0005] To achieve the above invention purpose, the present invention adopts the following technical solutions:

[0006] The precise fault diagnosis system for pump tubes based on multi-parameter data analysis provided by the present invention includes:

[0007] S101, a multi-parameter acquisition module: used to acquire multiple parameters during the operation of the pump tube, and the multiple parameters at least include current parameters, voltage parameters, tube surface temperature parameters, internal gas pressure parameters of the tube, and tube luminous intensity parameters;

[0008] S102. A data preprocessing unit, connected to the multi-parameter acquisition module, for denoising, filtering, and normalizing the acquired multi-parameter data to obtain standardized operation data;

[0009] S103. A data analysis module, connected to the data preprocessing unit, with a built-in fault diagnosis model. The fault diagnosis model is constructed based on machine learning algorithms, for receiving the standardized operation data and comprehensively analyzing the operation state of the pump lamp according to the pre-trained model parameters to establish a fault judgment algorithm;

[0010] The acquired parameter vector is X; the weight vector obtained by model training is W, and the bias term is b. Calculate the comprehensive score S:

[0011]

[0012] Through model data training, set the fault discrimination threshold T;

[0013] If S≥T, it is determined that the lamp has a fault, and the fault type label is output;

[0014] If S<T, it is determined that the lamp is in a normal operation state;

[0015] S104. An early warning module, connected to the data analysis module. When the data analysis module determines that the lamp has a fault, the early warning module is used to send a fault warning signal, and the fault warning signal includes fault type information;

[0016] S105. A storage module, respectively connected to the multi-parameter acquisition module and the data analysis module, for storing the acquired original parameter data and the diagnosis result data processed by the data analysis module for subsequent query and analysis.

[0017] Preferably, the current parameter acquisition in the multi-parameter acquisition module is realized by a high-precision current transformer, the surface temperature parameter of the lamp is acquired by an infrared temperature sensor attached to the surface of the lamp, the internal gas pressure parameter of the lamp is acquired by a micro pressure sensor installed at the sealed end of the lamp, and the luminous intensity parameter of the lamp is acquired by a photoelectric sensor facing the luminous direction of the lamp.

[0018] Preferably, the machine learning algorithms include but are not limited to decision tree algorithms, neural network algorithms, and support vector machine algorithms, which are suitable for fault diagnosis of pump lamps under different working conditions.

[0019] Preferably, the fault warning signal sent by the early warning module also includes fault severity information, and the fault severity is classified according to the prediction of the fault development trend by the data analysis module, divided into minor faults, moderate faults, and severe faults, corresponding to different warning methods.

[0020] Preferably, the storage module also has a data backup function, regularly backing up the stored data to an external storage device with a backup period of 168 hours to prevent data loss. Moreover, the data storage format of the storage module follows the general industrial data storage standard to facilitate data interaction with other devices.

[0021] Preferably, the system further includes a human-machine interface connected to the data analysis module and the storage module, which is used for operators to intuitively view the real-time operation parameters, diagnostic results, and historical data trend charts of the pump lamp, and can manually adjust some parameters of the fault diagnosis model through the human-machine interface.

[0022] Preferably, when the data preprocessing unit normalizes the parameter data, it adopts a normalization method based on the rated parameter values of the lamp tubes, that is, dividing each collected time parameter value by the corresponding rated parameter value of the lamp tube to make all data within the same comparable range for the subsequent processing of the data analysis model. And during the normalization process, if a parameter value exceeds or is lower than the rated parameter value, an abnormal alarm signal is immediately triggered and sent to the early warning module.

[0023] Preferably, after the data analysis module determines whether there is a fault in the lamp tube and the type of the fault, it further generates a fault cause analysis report. The fault cause analysis report is obtained based on the comprehensive analysis of the change trends of multiple parameters and the comparison with historical fault data, stored through the storage module, and can be viewed on the human-machine interface.

[0024] Preferably, the system has the ability of self-learning and self-adaptation, and regularly retrains the fault diagnosis model as new operation data of the pump lamp tubes are continuously collected.

[0025] Preferably, the multi-parameter acquisition module also has a sensor status self-checking function, which performs a self-check on each sensor every 24 hours. The self-check content includes whether the sensor connection is normal, whether the measured data is within a reasonable range, etc. If an abnormality is found, a sensor fault early warning is immediately sent to the early warning module, and at the same time, a backup sensor is started for data acquisition to ensure the continuity of the system operation.

[0026] Compared with the prior art, the above one or more technical solutions have the following beneficial effects:

[0027] This accurate fault diagnosis system for pump lamp tubes based on multi-parameter data analysis provides support for the reliable operation of pump lamp tubes by virtue of multi-parameter acquisition and a unique fault judgment algorithm.

[0028] In the system setup phase, the multi-parameter acquisition module accurately acquires parameters such as current, voltage, surface temperature of the lamp tube, internal gas pressure, and luminous intensity during the operation of the pump lamp tube through high-precision current transformers, infrared temperature sensors, micro pressure sensors, and photoelectric sensors. Each sensor is stably connected to the data preprocessing unit to ensure smooth data transmission. The preprocessing unit sets denoising, filtering parameters, and a normalization method based on the rated parameter values of the lamp tube to ensure accurate and comparable data.

[0029] The data analysis module selects a variety of learning algorithms to construct a fault diagnosis model. Its fault judgment algorithm is as follows: The parameter vector after acquisition and preprocessing is X, the weight vector W and bias term b are obtained through model training, and the comprehensive score S is calculated, that is, the weighted sum of each parameter based on the weights. A fault discrimination threshold T is set through training with a large number of sample data. If S≥T, it is determined that the lamp tube has a fault and the fault type label is output; if S<T, it is determined that the lamp tube is operating normally.

[0030] When a lamp tube fault is judged, the warning module issues different warnings according to the fault type and severity. At the same time, a fault cause analysis report is generated and stored in the storage module. The multi-parameter collaborative acquisition and analysis, combined with a scientific fault judgment algorithm, enable the system to efficiently and accurately monitor and diagnose pump lamp tube faults. Brief Description of the Drawings

[0031] The schematic diagrams in the specification that form a part of the present invention are used to provide a further understanding of the present invention. The illustrative embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention.

[0032] Figure 1 is a schematic diagram of the overall system architecture proposed according to an embodiment of the present invention;

[0033] Figure 2 is a schematic diagram of the core process of the system proposed according to an embodiment of the present invention;

[0034] Figure 3 is a detailed structural schematic diagram of the multi-parameter acquisition module proposed according to an embodiment of the present invention;

[0035] Figure 4 is a schematic diagram of the data preprocessing process and normalization logic proposed according to an embodiment of the present invention;

[0036] Figure 5 is a schematic diagram of the data analysis module and the fault diagnosis model proposed according to an embodiment of the present invention;

[0037] Figure 6 is a schematic diagram of the warning, storage, and human-machine interaction module proposed according to an embodiment of the present invention. Detailed Embodiments

[0038] To enable those skilled in the art to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the scope of protection of this application.

[0039] Please refer to Figures 1-6 , the present invention provides a precise pump tube fault diagnosis system based on multi-parameter data analysis;

[0040] Install the multi-parameter acquisition module according to the design requirements, accurately connect the high-precision current transformer to the current loop of the pump tube to ensure accurate acquisition of current parameters; closely attach the infrared temperature sensor to the surface of the tube to achieve reliable acquisition of the surface temperature parameters of the tube; properly install the micro pressure sensor at the sealed end of the tube to measure the internal gas pressure parameters of the tube; and correctly orient the photoelectric sensor towards the light-emitting direction of the tube to ensure accurate acquisition of the light intensity parameters of the tube. At the same time, ensure that each sensor establishes a stable data transmission connection with the subsequent data preprocessing unit.

[0041] Initialize the data preprocessing unit, set the algorithm parameters for denoising and filtering, such as using appropriate filter cut-off frequencies, denoising thresholds, etc., so that it can effectively remove the noise interference in the acquired data and ensure the accuracy of the data. At the same time, according to the rated parameter values of the pump tube used, configure the normalization processing module, determine the normalization method based on the rated parameter values of the tube, that is, prepare the rated values corresponding to each parameter, so as to divide the acquired real-time parameter values by the rated parameter values later to achieve data normalization and make all data within the same comparable range.

[0042] Select and construct the fault diagnosis model in the data analysis module, and select a suitable one or a combination from various machine learning algorithms such as decision tree algorithm, neural network algorithm, support vector machine algorithm, etc. according to the actual application scenario and the working characteristics of the pump tube. Use the known normal operation data and various fault data of the pump tube for model training, determine the weight vector W and bias term b obtained from the model training, and through a large number of sample data training, set a reasonable fault discrimination threshold T to enable the model to accurately judge the operating state of the tube.

[0043] Connect the warning module, configure the communication link between the warning module and the data analysis module, ensure that when the data analysis module determines that there is a fault in the lamp tube, the warning module can receive the fault information in a timely manner, and according to the pre-set rules, prepare the corresponding fault warning signal sending methods for different fault types and severities, such as different forms of warnings like light flashing and sound prompts.

[0044] Install the storage module, establish the connection between the storage module and the multi-parameter acquisition module and the data analysis module, set the data storage format, and make it follow the general industrial data storage standard to facilitate subsequent data interaction with other devices. At the same time, turn on the data backup function of the storage module, set the backup period to 168 hours, and regularly back up the stored data to an external storage device to ensure the security of the data and prevent data loss.

[0045] The system is equipped with a human-machine interaction interface, which is connected to the data analysis module and the storage module, and the corresponding software functions are developed and deployed to enable the operator to intuitively view the real-time operating parameters, diagnostic results, historical data trend charts, etc. of the pump lamp tube. At the same time, it provides an interface interaction function that allows the operator to manually adjust some parameters of the fault diagnosis model through the human-machine interaction interface. For example, under special working conditions, the operator can fine-tune the model parameters according to experience to meet the actual needs.

[0046] The multi-parameter acquisition module continuously acquires multiple parameters during the operation of the pump lamp tube, including current parameters, voltage parameters, lamp tube surface temperature parameters, lamp tube internal gas pressure parameters, and lamp tube luminous intensity parameters. The acquisition frequency can be set according to actual needs, such as acquiring data once per second, to ensure that the change in the operating state of the lamp tube can be reflected in a timely manner. Each sensor transmits the acquired data to the data preprocessing unit in real time.

[0047] After receiving the acquired data, the data preprocessing unit first performs denoising on the data, using a suitable denoising algorithm such as wavelet denoising to remove the high-frequency noise interference in the data; then performs filtering, using low-pass, high-pass or band-pass filters to filter out the interference signals in specific frequency bands to make the data smoother and more stable; then performs normalization processing according to the pre-set normalization method based on the rated parameter values of the lamp tube, that is, dividing each acquired real-time parameter value by the corresponding rated parameter value of the lamp tube. During the normalization process, if a parameter value exceeds or is lower than the rated parameter value, an abnormal alarm signal is immediately triggered and sent to the warning module. After receiving the signal, the warning module issues an abnormal alarm prompt in a timely manner, such as popping up an alarm window on the human-machine interaction interface to remind the operator to pay attention to the abnormal situation of the lamp tube parameters.

[0048] The preprocessed standardized operation data is transmitted to the data analysis module. The fault diagnosis model receives this data and comprehensively analyzes the operating status of the pumping lamp tube based on the pre-trained model parameters;

[0049] The parameter vector after its acquisition and preprocessing is X = [X1, X2, X3... X n , where X1 to X n correspond to different operating parameters (such as the normalized values of current, voltage, temperature, etc. mentioned above);

[0050] A set of weight vectors obtained from model training is W = [W1, W2, W3... W n , and the bias term is b. First, calculate the comprehensive score S:

[0051]

[0052] S is a weighted sum of each parameter based on its weight. The weight reflects the relative importance of different parameters in judging faults, and the bias term can be used to adjust the overall discrimination benchmark.

[0053] Set the fault discrimination threshold T: By training and learning with a large amount of data of pumping lamp tubes in known normal and faulty states, determine a suitable threshold T.

[0054] If the S values calculated from the normal lamp tube data in the training set are generally concentrated in a certain range, while the S values corresponding to the faulty lamp tube data deviate significantly, a suitable T can be selected according to these statistical characteristics to distinguish the two.

[0055] Judge whether there is a fault: If S ≥ T, it is determined that the lamp tube has a fault, and output the fault type label. The fault type label can be determined by subsequent more refined classification rules to further analyze which parameters have a large degree of abnormality to correspond to different fault types;

[0056] If S < T, it is determined that the lamp tube is in a normal operating state. When actually constructing the fault diagnosis model: Machine learning algorithms (such as neural networks, support vector machines, etc.) will automatically find the optimal weight vector W and bias term b for model training.

[0057] The training data contains multi-parameter data of normal and faulty lamp tubes under different working conditions. The model aims to minimize the error between the prediction result and the true label (i.e., the actual fault or not and type), and continuously adjusts W and b. As the amount of data increases and the model complexity improves, the model can capture more complex parameter relationships, thus achieving more accurate fault diagnosis.

[0058] When the data analysis module determines that there is a fault in the lamp tube, it immediately sends a fault message to the warning module. The warning module generates a fault warning signal containing fault type information and fault severity information based on the fault type and the prediction of the fault development trend. The fault severity is divided into minor faults, moderate faults, and severe faults, corresponding to different warning methods. For example, a minor fault may only display a prompt icon on the human-machine interface, a moderate fault uses sound plus icon prompts, and a severe fault triggers a strong sound and light alarm to ensure that the operator can be informed of the fault situation and its severity in a timely manner.

[0059] Meanwhile, after the data analysis module determines whether there is a fault in the lamp tube and the fault type, it further generates a fault cause analysis report. This report is obtained based on the comprehensive analysis of the change trends of multiple parameters and the comparison with historical fault data. For example, if it is found that the lamp tube current gradually increases and the luminous intensity gradually weakens, and considering that similar situations in historical data are mostly caused by lamp tube aging, it is pointed out in the fault cause analysis report that the lamp tube aging is likely. The fault cause analysis report is stored through the storage module and can be viewed on the human-machine interface, facilitating the operator to deeply understand the root cause of the fault and providing a basis for subsequent maintenance.

[0060] During the entire operation of the system, the storage module continuously stores the collected original parameter data and the diagnostic result data processed by the data analysis module, and stores them according to the set data storage format for subsequent query and analysis. For example, after the lamp tube fails, maintenance personnel can query the historical data in the storage module to understand the parameter changes of the lamp tube in a period of time before the fault, assisting in fault troubleshooting and maintenance.

[0061] The multi-parameter acquisition module has a sensor status self-check function, and automatically performs a self-check on each sensor every 24 hours. The self-check content includes whether the sensor connection is normal, which is judged by sending a detection signal and receiving feedback; whether the measured data is within a reasonable range, which is judged based on the measurement characteristics of each sensor and the parameter range during the normal operation of the lamp tube. If an abnormality is found, a sensor fault warning is immediately sent to the warning module, and at the same time, a backup sensor is activated for data acquisition to ensure the continuity of the system operation. The switching process of the backup sensor should be designed as a seamless switch to avoid data acquisition interruption.

[0062] The system has the ability of self-learning and self-adaptation. As new operation data of the pump lamp tube is continuously collected, the fault diagnosis model is retrained regularly, and the model is updated and trained once with the newly collected data. By continuously incorporating new data samples, the fault diagnosis model can adapt to the changes of the pump lamp tube under different working conditions and different usage stages, continuously improving the accuracy of fault diagnosis and ensuring that the accurate diagnosis ability of the entire system for the pump lamp tube fault is always in good condition.

[0063] The pump lamp fault precise diagnosis system based on multi-parameter data analysis can efficiently and accurately realize functions such as monitoring, diagnosis, early warning, and data storage and analysis of pump lamp faults, providing a strong guarantee for the reliable operation of pump lamps.

[0064] The above are only the preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes should be covered within the protection scope of the present invention.

Claims

1. A precise diagnosis system for pump lamp tube faults based on multi-parameter data analysis, characterized in that, Including: S101. Multi-parameter acquisition module: used to acquire multiple parameters during the operation of the pump lamp, and the multiple parameters at least include current parameter, voltage parameter, lamp surface temperature parameter, internal gas pressure parameter of the lamp, and lamp luminous intensity parameter; S102. Data preprocessing unit, connected to the multi-parameter acquisition module, used to perform denoising, filtering, and normalization processing on the acquired multiple parameter data to obtain standardized operation data; S103. Data analysis module, connected to the data preprocessing unit, with a built-in fault diagnosis model. The fault diagnosis model is constructed based on machine learning algorithms, used to receive the standardized operation data, and comprehensively analyze the operation state of the pump lamp according to the pre-trained model parameters, and establish a fault judgment algorithm; The acquired parameter vector is X; the weight vector obtained by model training is W, and the bias term is b. Calculate the comprehensive score S: Through model data training, set the fault discrimination threshold T; If S≥T, it is determined that the lamp has a fault, and a fault type mark is output; If S<T, it is determined that the lamp is in a normal operation state; S104. Warning module, connected to the data analysis module. When the data analysis module determines that the lamp has a fault, the warning module is used to send a fault warning signal, and the fault warning signal contains fault type information; S105. Storage module, respectively connected to the multi-parameter acquisition module and the data analysis module, used to store the acquired original parameter data and the diagnostic result data processed by the data analysis module for subsequent query and analysis.

2. The pump lamp tube fault precise diagnosis system based on multi-parameter data analysis according to claim 1, characterized in that The acquisition of the current parameter in the multi-parameter acquisition module is realized through a high-precision current transformer. The acquisition of the lamp surface temperature parameter is realized by attaching an infrared temperature sensor to the lamp surface. The acquisition of the internal gas pressure parameter of the lamp is realized by a micro pressure sensor installed at the sealed end of the lamp. The acquisition of the lamp luminous intensity parameter is realized by a photoelectric sensor facing the luminous direction of the lamp.

3. The pump lamp tube fault precise diagnosis system based on multi-parameter data analysis according to claim 1, characterized in that, The machine learning algorithms include but are not limited to decision tree algorithm, neural network algorithm, and support vector machine algorithm, which are suitable for fault diagnosis of pump lamps under different working conditions.

4. The pump lamp tube fault precise diagnosis system based on multi-parameter data analysis according to claim 1, characterized in that The fault warning signal sent by the warning module also includes fault severity information. The fault severity is classified according to the prediction of the fault development trend by the data analysis module, and is divided into minor faults, moderate faults, and severe faults, corresponding to different warning methods.

5. The pump lamp tube fault precise diagnosis system based on multi-parameter data analysis according to claim 1, characterized in that, The storage module also has a data backup function, regularly backing up the stored data to an external storage device, with a backup period of 168 hours to prevent data loss. And the data storage format of the storage module follows the general industrial data storage standard to facilitate data interaction with other devices.

6. The pump lamp tube fault precise diagnosis system based on multi-parameter data analysis according to claim 1, characterized in that, The system also includes a human-machine interface, connected to the data analysis module and the storage module, used for operators to intuitively view the real-time operation parameters, diagnostic results, and historical data trend charts of the pump lamp, and can manually adjust some parameters of the fault diagnosis model through the human-machine interface.

7. The pump lamp fault precise diagnosis system based on multi-parameter data analysis according to claim 1, characterized in that When normalizing the parameter data, the data preprocessing unit adopts a normalization method based on the rated parameter values of the lamp tubes, that is, dividing each collected time parameter value by the corresponding rated parameter value of the lamp tube to make all data within the same comparable range, which is convenient for the subsequent data analysis model to process. And during the normalization process, if a parameter value exceeds or is lower than the rated parameter value, an abnormal alarm signal is immediately triggered and sent to the early warning module.

8. The accurate fault diagnosis system for pump tubes based on multi-parameter data analysis according to claim 1, characterized in that, After the data analysis module determines whether there is a fault in the lamp tube and the type of the fault, it further generates a fault cause analysis report. The fault cause analysis report is obtained based on the comprehensive analysis of the change trends of multiple parameters and the comparison with historical fault data, stored through the storage module, and can be viewed on the human-machine interaction interface.

9. The accurate fault diagnosis system for pump tubes based on multi-parameter data analysis according to claim 1, wherein The system has the ability of self-learning and self-adaptation. As new operation data of the pump lamp tubes are continuously collected, the fault diagnosis model is retrained regularly.

10. The pump lamp tube fault precise diagnosis system based on multi-parameter data analysis according to claim 1, characterized in that, The multi-parameter acquisition module also has a sensor status self-checking function. Each sensor is self-checked once every 24 hours. The self-checking content includes whether the sensor connection is normal, whether the measured data is within a reasonable range, etc. If an abnormality is found, a sensor fault early warning is immediately sent to the early warning module, and at the same time, a standby sensor is started to collect data to ensure the continuity of the system operation.

Images