Operation and maintenance management and operation state monitoring method of online quality inspection system

Abstract

The invention belongs to the technical field of online quality inspection, and relates to an operation and maintenance management and running state monitoring method of an online quality inspection system, which comprises the following steps: deploying sensors and monitoring equipment at key nodes of the quality inspection system, and collecting equipment running state data, quality inspection data and key parameters of specific algorithm running; and carrying out calibration, verification and encrypted transmission on the acquired data, and setting a synchronization period to synchronize the data to a central server. By constructing an accurate equipment state evaluation and fault diagnosis model, the accuracy of fault early warning and positioning is improved, and powerful support is provided for rapid fault repair; maintenance arrangement is carried out in combination with a production plan, the influence of maintenance on production is reduced to the greatest extent, a system performance optimization module improves the load balancing capacity and the alarm response efficiency of the system, the stability and reliability of the online quality inspection system are enhanced, high-efficiency proceeding of quality inspection work is guaranteed, the production efficiency and the product quality of enterprises are improved, and the economic benefits of enterprises are improved. And the market competitiveness of enterprises is enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of online quality inspection, and particularly relates to a method for operation and maintenance management and operation status monitoring of an online quality inspection system. Background Art

[0002] Online quality inspection systems are widely used in various manufacturing industries and are key technical means to ensure product quality and improve production efficiency. With the continuous expansion of production scale and the increasing complexity of production processes, the scale and complexity of online quality inspection systems are also rapidly increasing, which poses higher requirements for their operation and maintenance management and operation status monitoring.

[0003] In terms of operation and maintenance management of existing online quality inspection systems, fault discovery mainly relies on manual regular inspections. For example, as mentioned in patent CN118200355A regarding the monitoring of artificial intelligence quality inspection equipment, the traditional method requires manual inspection of the operation status of each device one by one, which is inefficient and prone to omissions. Due to the inability to monitor in real time, fault discovery often lags, resulting in production interruptions, an increase in defective products, an increase in production costs, and an impact on the economic benefits of enterprises.

[0004] In terms of operation status monitoring, the existing technology does not comprehensively and deeply monitor the data of quality inspection equipment. Most only collect the basic operation parameters of the equipment, making it difficult to effectively evaluate the accuracy, stability, and reliability of quality inspection data. It is difficult to synchronize and integrate data between different quality inspection devices, and a unified and comprehensive view of the system operation status cannot be formed, resulting in difficulty for operation and maintenance personnel to grasp the overall system situation and increasing the complexity and difficulty of operation and maintenance management.

[0005] At the same time, the existing system lacks an effective monitoring and warning mechanism for key factors affecting system performance such as system load changes and alarm response delays. Similar to the problems existing in the operation and maintenance management platform monitoring and warning system in patent CN118538010A, if an online quality inspection system cannot timely discover and handle these problems, it will lead to unstable system operation and a decrease in detection accuracy, unable to meet the requirements of production for the efficient and reliable operation of the quality inspection system. Summary of the Invention

[0006] The present invention provides a method for operation and maintenance management and operation status monitoring of an online quality inspection system to solve the problems raised in the background art.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for operation and maintenance management and operation status monitoring of an online quality inspection system includes the following steps:

[0009] Deploy sensors and monitoring devices at key nodes of the quality inspection system to collect data on the operating status of devices, quality inspection data, and key parameters for the operation of specific algorithms. Calibrate, verify, and encrypt the transmitted collected data, and set a synchronization period to synchronize the data to the central server;

[0010] Construct a device status evaluation model using the long short-term memory network of deep learning, and construct a data quality monitoring model based on statistical principles to monitor the operating status of the device and the quality inspection data, and judge whether the device status is abnormal and whether the quality inspection data is accurate;

[0011] Adopt a fault diagnosis model based on support vector machine combined with a fault feature library for fault diagnosis, combine a decision tree algorithm for fault warning, and introduce an interface of the production plan management system to formulate a personalized maintenance plan;

[0012] Monitor the system load in real time. When the system load is too high, adopt a task allocation strategy based on priority and load balancing; establish a hierarchical alarm system, optimize the alarm content, and monitor the alarm response time.

[0013] As a further solution of the present invention: the device operating status data includes the temperature, voltage, and operating duration of the device; the quality inspection data includes the detection accuracy and defect judgment result; the key parameters for the operation of the specific algorithm include the model training accuracy, loss function value, and gradient change in the deep learning algorithm.

[0014] As a further solution of the present invention: calibrate, verify, and encrypt the transmitted collected data. Specifically, calibrate the sensors regularly, adopt a redundant sensor design, and use a data verification algorithm to perform consistency verification on the data collected by multiple sensors. Use the AES encryption algorithm to encrypt the data, and use the CRC check code to ensure data integrity.

[0015] As a further solution of the present invention: using the long short-term memory network to construct a device status evaluation model is to input the device operating parameters into the LSTM network for training according to the time series, establish a normal operating parameter prediction model, and judge the device status by comparing the real-time data with the predicted value; the data quality monitoring model constructed based on statistical principles is to calculate statistics such as the mean and standard deviation of the quality inspection data, set a fluctuation range, and judge whether the quality inspection data is abnormal in combination with the comparison results of the quality inspection data of multiple devices.

[0016] As a further solution of the present invention: the fault diagnosis model based on support vector machine combined with a fault feature library is to collect device fault data to construct a fault feature library, extract the data features of the faulty device and transform them into the feature vectors of the SVM model for fault classification prediction; the combination of the decision tree algorithm for fault warning is to trigger a warning based on the threshold of the device operating parameters.

[0017] As a further solution of the present invention: introducing the production plan management system interface to formulate personalized maintenance plans, comprehensively considering the equipment operation status, historical fault data, maintenance records and production plans, preferentially arranging equipment maintenance during production breaks or low-peak periods, formulating emergency plans for key equipment, and dynamically adjusting maintenance strategies according to the equipment operation conditions.

[0018] As a further solution of the present invention: real-time monitoring of system load, including monitoring CPU usage rate, memory occupancy rate, and network bandwidth; the task allocation strategy based on priority and load balancing divides priorities according to the urgency and resource requirements of quality inspection tasks, preferentially allocating low-priority tasks to devices with lower loads, and considering the device processing capacity and the current task queue length.

[0019] As a further solution of the present invention: the hierarchical alarm system sets different alarm levels according to the severity of anomalies and adopts different notification methods; the optimized alarm content includes providing a trend chart of equipment operation data before the fault occurs and details of quality inspection data anomalies; monitoring and optimizing the alarm response time to ensure that operation and maintenance personnel can respond and handle problems in a timely manner.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] 1. The operation and maintenance management and operation status monitoring method of this online quality inspection system realizes real-time and comprehensive monitoring of the online quality inspection system, can timely detect equipment failures and data quality problems, effectively reduce production interruptions and the generation of defective products, reduce production costs, improve the accuracy of fault warning and positioning by constructing a precise equipment status evaluation and fault diagnosis model, and provide strong support for quickly repairing faults; the formulation of personalized maintenance plans, combined with production plans for maintenance arrangements, minimizes the impact of maintenance on production. The system performance optimization module improves the load balancing ability and alarm response efficiency of the system, enhances the stability and reliability of the online quality inspection system, ensures the efficient progress of quality inspection work, improves the production efficiency and product quality of the enterprise, and enhances the market competitiveness of the enterprise. Specific embodiments

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0023] Embodiment

[0024] The present invention provides the following technical solutions:

[0025] The operation and maintenance management and operation status monitoring method of the online quality inspection system includes:

[0026] The data acquisition and synchronization module deploys various types of sensors and monitoring equipment at key nodes of the quality inspection system, such as the core processing unit, sensors, and data transmission line interfaces of the quality inspection equipment, to achieve comprehensive collection of equipment operating status and quality inspection data. The collected data not only covers basic operating data such as the temperature, voltage, and operating time of the equipment, but also includes quality inspection data such as detection accuracy and defect judgment results. For systems that use specific algorithms (such as deep learning and machine vision algorithms) for quality inspection, the key parameters of the algorithm operation are collected in depth, such as the model training accuracy, loss function value, and gradient changes in the deep learning algorithm.

[0027] To ensure the accuracy and reliability of data collection, a regular calibration mechanism is established to calibrate sensors regularly. At the same time, redundant sensor design is adopted, and data verification algorithms are used to verify the consistency of data collected by multiple sensors. During data transmission, encryption and verification technologies are used, such as using the AES encryption algorithm to encrypt data and using CRC checksums to ensure data integrity, to prevent data from being stolen, tampered with, or lost during transmission.

[0028] According to the real-time requirements and data volume of the data, a flexible data synchronization cycle is set. With the help of network communication technologies such as 5G and industrial Ethernet, data synchronization between quality inspection equipment and the central server is achieved, and breakpoint resumption technology is used to ensure the integrity of data transmission.

[0029] The operation status monitoring module uses the long short-term memory network (LSTM) technology in deep learning to build an equipment status assessment model. The various parameters of equipment operation are input into the LSTM network in time series for training, so that the model can learn the relationship and change pattern between the parameters under the normal operation of the equipment, thereby establishing a normal operation parameter prediction model. By comparing the real-time collected data with the predicted value, when the deviation exceeds the preset threshold, the equipment status is judged to be abnormal;

[0030] A data quality monitoring model is constructed based on statistical principles. The mean, standard deviation and other statistical quantities of the quality inspection data are calculated. A reasonable fluctuation range is set. Combined with the comparison results of quality inspection data from multiple devices, the quality inspection data is judged to be abnormal when the data fluctuation exceeds the range and the differences in the inspection results from multiple devices are significant. At the same time, time series analysis methods are used to monitor the changing trends of quality inspection data and timely discover potential data quality problems.

[0031] The operation and maintenance management module adopts a fault diagnosis model based on support vector machine (SVM) combined with a fault feature library, collects information such as the operation data, fault phenomena, and maintenance records of the equipment under different fault types, constructs a fault feature library, extracts the data features of the faulty equipment and converts them into feature vectors recognizable by the SVM model, classifies and predicts the fault types through the SVM model, realizes the accurate positioning of the fault cause and location, combines the decision tree algorithm for fault warning, triggers the warning based on the equipment operation parameter threshold, and provides sufficient processing time for the operation and maintenance personnel;

[0032] Introduce the interface of the production plan management system, obtain the production task information, comprehensively consider the equipment operation status, historical fault data, maintenance records, and production plan, formulate a personalized maintenance plan, preferably arrange equipment maintenance during production breaks or low-peak periods, formulate emergency plans for key equipment, and dynamically adjust the maintenance strategy according to the equipment operation situation, such as increasing the maintenance frequency for equipment with frequent faults.

[0033] The system performance optimization module monitors the system load in real time, including key indicators such as CPU usage rate, memory occupancy rate, and network bandwidth, sets the judgment criteria for excessive system load, and when the system load reaches this standard, adopts a task allocation strategy based on priority and load balancing, divides the priorities according to the urgency and resource requirements of the quality inspection tasks, preferably assigns low-priority tasks to devices with lower load, and at the same time considers the processing capacity of the devices and the current task queue length to achieve balanced task allocation and improve the overall performance of the system;

[0034] Establish a hierarchical alarm system, set different alarm levels according to the severity of the anomalies, such as serious faults, general faults, and prompt messages, and adopt different notification methods, such as sound and light alarms, text messages, emails, etc., optimize the alarm content, provide information such as the equipment operation data trend chart before the fault occurs and the details of the quality inspection data anomalies, monitor the alarm response time, analyze the influencing factors and optimize them to ensure that the operation and maintenance personnel can respond and handle problems in a timely manner.

[0035] Implementation steps of the monitoring method:

[0036] 1. Data collection: Deploy sensors and monitoring devices at the key nodes of the online quality inspection system to collect equipment operation status data (such as temperature, voltage, operation duration), quality inspection data (detection accuracy, defect judgment results), and key parameters for the operation of specific algorithms (such as training accuracy, loss function value, gradient change of deep learning algorithms).

[0037] 2. Data processing and synchronization: Calibrate the sensors regularly, adopt redundant sensor design and use data verification algorithms to ensure the accuracy of the collected data, ensure the security of data transmission through encryption technology, set the synchronization period according to the data characteristics, and synchronize the data to the central server by means of network communication technology.

[0038] 3. Operating Status Monitoring

[0039] Equipment Status Evaluation: Use the Long Short-Term Memory Network (LSTM) to train the equipment operation parameters according to the time series, construct a normal operation parameter prediction model, collect data in real time and compare it with the predicted value. If the deviation exceeds the threshold, it is determined that the equipment status is abnormal;

[0040] Data Quality Monitoring: Calculate statistics such as the mean and standard deviation of the quality inspection data, set a reasonable fluctuation range, and combine the comparison results of the quality inspection data of multiple devices. When the data fluctuates abnormally and the detection results of multiple devices are significantly different, it is determined that the quality inspection data is abnormal.

[0041] 4. Fault Warning and Diagnosis

[0042] Fault Warning: Combine the decision tree algorithm to trigger a fault warning based on the equipment operation parameter threshold, and timely notify the operation and maintenance personnel to pay attention to potential problems of the equipment;

[0043] Fault Diagnosis: Collect equipment fault data to build a fault feature library, extract the data features of the faulty equipment and convert them into feature vectors of the Support Vector Machine (SVM) model, and predict the fault type through the SVM model to locate the fault cause and location.

[0044] 5. Maintenance Plan Formulation: Introduce the interface of the production plan management system, comprehensively consider the equipment operation status, historical fault data, maintenance records and production plans, give priority to arranging equipment maintenance during production breaks or low peaks, formulate emergency plans for key equipment, and dynamically adjust the maintenance strategy according to the actual operation of the equipment.

[0045] 6. System Performance Optimization

[0046] Load Monitoring and Task Allocation: Real-time monitor load indicators such as the CPU usage rate, memory occupancy rate, and network bandwidth of the system. When the system load is too high, divide the priorities according to the urgency of the quality inspection tasks and resource requirements, and give priority to allocating low-priority tasks to devices with lower loads. At the same time, consider the device processing capacity and task queue length to achieve balanced task allocation;

[0047] Alarm System Optimization: Establish a hierarchical alarm system, set different levels according to the severity of the abnormality and adopt different notification methods, optimize the alarm content, provide information such as the trend chart of the equipment operation data before the fault occurs and the details of the quality inspection data abnormality, monitor the alarm response time, analyze and optimize the influencing factors to ensure that the operation and maintenance personnel can handle problems in a timely manner.

[0048] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for operation and maintenance management and operation status monitoring of an online quality inspection system, characterized in that: The following steps are involved: Deploy sensors and monitoring equipment at key nodes of the quality inspection system to collect equipment operating status data, quality inspection data, and key parameters of specific algorithm operation, calibrate, verify, and encrypt the collected data for transmission, and set a synchronization cycle to synchronize the data to the central server; Use the long short-term memory network of deep learning to build an equipment status assessment model, and build a data quality monitoring model based on statistical principles to monitor the equipment operation status and quality inspection data to determine whether the equipment status is abnormal and whether the quality inspection data is accurate; A fault diagnosis model based on support vector machine combined with fault feature library is used for fault diagnosis, a decision tree algorithm is used for fault warning, and a production plan management system interface is introduced to formulate personalized maintenance plans; Monitor system load in real time, and adopt a task allocation strategy based on priority and load balancing when the system load is too high; establish a hierarchical alarm system, optimize alarm content and monitor alarm response time.

2. The operation and maintenance management and operation status monitoring method of the online quality inspection system according to claim 1 is characterized by: The equipment operation status data includes the temperature, voltage, and operating time of the equipment; the quality inspection data includes detection accuracy and defect judgment results; the key parameters for the operation of the specific algorithm include the model training accuracy, loss function value, and gradient change in the deep learning algorithm.

3. The operation and maintenance management and operation status monitoring method of the online quality inspection system according to claim 1 is characterized by: The collected data is calibrated, verified and encrypted for transmission. Specifically, the sensors are calibrated regularly, redundant sensor design is adopted and data verification algorithm is used to verify the consistency of data collected by multiple sensors, AES encryption algorithm is used to encrypt data, and CRC check code is used to ensure data integrity.

4. The operation and maintenance management and operation status monitoring method of the online quality inspection system according to claim 1 is characterized by: The equipment status assessment model is constructed using the long short-term memory network, which is to input the equipment operating parameters into the LSTM network in time series for training, establish a normal operating parameter prediction model, and judge the equipment status by comparing the real-time data with the predicted value; the data quality monitoring model is constructed based on statistical principles, which is to calculate the mean, standard deviation and other statistical quantities of the quality inspection data, set the fluctuation range, and judge whether the quality inspection data is abnormal based on the comparison results of multiple equipment quality inspection data.

5. The operation and maintenance management and operation status monitoring method of the online quality inspection system according to claim 1 is characterized by: The fault diagnosis model based on support vector machine combined with fault feature library collects equipment fault data to build a fault feature library, extracts the data features of faulty equipment and converts them into feature vectors of SVM model for fault classification prediction; the fault warning combined with decision tree algorithm triggers warning according to the threshold of equipment operation parameters.

6. The operation and maintenance management and operation status monitoring method of the online quality inspection system according to claim 1 is characterized by: The production planning management system interface is introduced to formulate personalized maintenance plans, which comprehensively considers the equipment operating status, historical fault data, maintenance records and production plans, prioritizes equipment maintenance during production gaps or off-peak periods, formulates emergency plans for key equipment, and dynamically adjusts maintenance strategies according to equipment operating conditions.

7. The operation and maintenance management and operation status monitoring method of the online quality inspection system according to claim 1 is characterized by: Real-time monitoring of system load, including monitoring of CPU usage, memory occupancy, and network bandwidth; the task allocation strategy based on priority and load balancing is to prioritize quality inspection tasks according to their urgency and resource requirements, and prioritize low-priority tasks to devices with lower loads, taking into account device processing capabilities and current task queue length.

8. The method for operation and maintenance management and operation status monitoring of an online quality inspection system according to claim 1, characterized in that: The hierarchical alarm system sets different alarm levels according to the severity of the abnormality and adopts different notification methods; the optimized alarm content includes providing equipment operation data trend charts before the failure occurs and details of quality inspection data abnormalities; the alarm response time is monitored and optimized to ensure that operation and maintenance personnel respond to and handle problems in a timely manner.