Instrument remote intelligent maintenance system based on real-time data

Abstract

The invention discloses an instrument remote intelligent maintenance system based on real-time data, which belongs to the technical field of instrument intelligent maintenance and is composed of an instrument, a two-way communication layer, an application server and a user terminal. The instrument acquires operation and health data through the data acquisition module, and the operation and health data are preprocessed by the edge calculation module and then uploaded to the cloud. The bidirectional communication layer adopts various communication technologies to guarantee real-time and stable transmission of data. And the application server constructs a predictive maintenance model by using big data analysis and a machine learning algorithm, and realizes anomaly detection, fault prediction and intelligent maintenance decision in combination with database data. The user terminal provides a visual interface, and supports remote operation and maintenance strategy customization. The system solves the problems of insufficient real-time performance, remote limitation and low intelligent degree of traditional instrument maintenance, realizes closed-loop management of real-time abnormal detection, fault prediction, remote calibration and intelligent maintenance decision of the instrument, and improves the operation efficiency and data reliability of the instrument.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent instrument maintenance, and particularly relates to an instrument remote intelligent maintenance system based on real-time data. Background Art

[0002] In the field of industrial automation, the stable operation of analytical instruments plays a key role in the monitoring, control, and optimization of the production process. However, there are many drawbacks in the current maintenance technology of analytical instruments.

[0003] Traditional instrument maintenance mainly relies on regular on-site manual inspections. During manual inspections, the status and data changes of the instruments cannot be immediately feedback. For example, in some chemical production scenarios, there may be abnormal fluctuations in parameters before the instrument fails, but due to the limitation of the inspection cycle, it cannot be discovered in time, resulting in a lag in fault warning. The equipment may continue to operate in an abnormal state, which not only affects production efficiency but may also cause more serious faults, resulting in huge economic losses.

[0004] In terms of remote operation, the traditional instrument maintenance method requires on-site operation, which greatly limits the maintenance work by manpower, time, and geographical location. For example, for instruments in remote areas or dangerous environments, it is time-consuming and laborious for maintenance personnel to go to the site, and there may even be safety risks. Some existing remote monitoring systems only achieve data upload, lack real-time decision-making capabilities, and cannot take corresponding measures in a timely manner according to data changes, resulting in low maintenance efficiency.

[0005] The degree of intelligence of the existing maintenance technology is also relatively low. Maintenance operations mostly rely on manual experience, and it is difficult to achieve predictive maintenance based on data analysis. Taking the maintenance of instruments in the power system as an example, due to the inability to effectively analyze real-time data and historical data, potential faults cannot be predicted in advance, and only repairs can be carried out after the faults occur, affecting the stability of power supply. Moreover, the existing technology cannot automatically optimize the maintenance cycle and parameter settings, and most of the few intelligent maintenance solutions are based on offline data analysis, and cannot dynamically adjust the maintenance strategy according to the real-time operation status of the instrument, making it difficult to adapt to the complex and changeable industrial environment, resulting in poor maintenance effects.

[0006] In summary, the existing maintenance technology of analytical instruments has obvious deficiencies in terms of real-time performance, remote operation, and intelligence, and there is an urgent need for an innovative maintenance system to solve these problems in order to improve the operation efficiency and data reliability of the instruments. Summary of the Invention

[0007] The main object of the present invention is to provide an instrument remote intelligent maintenance system based on real-time data, which can effectively solve the problems mentioned in the background art.

[0008] To achieve the above object, the technical solution adopted by the present invention is as follows: A remote intelligent maintenance system for instruments based on real-time data, comprising an instrument, a two-way communication layer, an application server, and a user terminal, characterized in that: The instrument is also a real-time data acquisition layer, which collects the operation data and health status data of the instrument in real time through a data acquisition module and supports multiple data interfaces (such as 4-20mA, RS485, MODBUS). Then, the edge computing module preprocesses the preliminary data and performs preliminary anomaly detection, and uploads the preprocessed data to the cloud server; The two-way communication layer is connected to the cloud platform using low-power wide-area Internet of Things (LPWAN, such as NB-IoT / LoRa) or Wi-Fi or 5G Internet of Things communication technology, etc., supports breakpoint resumption and data compression, and ensures real-time transmission of high-frequency data. The data compression technology and transmission redundancy mechanism are used to ensure network transmission stability, prevent data loss or tampering, and at the same time support two-way communication, facilitating remote operation and instruction issuance; The application server is the cloud intelligent analysis layer, mainly including a database, a predictive maintenance model, and a maintenance decision-making module; among them: The database mainly stores historical and real-time data and supports fast query; at the same time, it collects fault information and corresponding measures for daily maintenance, forms a complete system, and the subsequent fault system can quickly screen common corresponding measures.

[0009] The predictive maintenance model is based on big data analysis and artificial intelligence algorithms, and uses technologies such as knowledge decision trees, instrument operation mechanism models, and machine learning to clean, store, and analyze real-time data, realize anomaly detection, trend analysis, and fault prediction, and put forward corresponding maintenance suggestions and plans according to historical data and operating status.

[0010] The maintenance decision-making module pushes further analysis and diagnosis information of specific events to the user, can interact with the user remotely to complete further in-depth analysis and diagnosis of the event, so as to help the user find problems, locate the reasons, and repair them faster. It can also receive and issue instructions from the cloud or remotely, generate corresponding maintenance strategies at the instrument end and the user end, and automatically or remotely trigger maintenance operations such as calibration, cleaning, and parameter adjustment.

[0011] The results of remote maintenance operations and the feedback data of the instrument are transmitted back to the cloud as the input for the next round of data training, realizing the continuous adaptive optimization of maintenance strategies; the user operation records and maintenance effect data will form a closed-loop feedback; The user terminal provides a visualization interface based on the Web and mobile terminals, displaying device status, alarm information, and maintenance suggestions. It supports the AR remote assistance function, and the user can obtain the real scene of the instrument through the camera and receive the fault location guidance superimposed by AI. The user can also monitor the remote maintenance process through the interaction interface, and can manually adjust the maintenance strategy or directly issue maintenance instructions.

[0012] Preferably, the data acquisition module of the instrument uses a chip integrating multiple data acquisition functions, connects to the sensors of the instrument through the SPI interface, and acquires operation data and health status data according to a preset sampling frequency, which is set to 100Hz based on the dynamic characteristics of the instrument.

[0013] Preferably, the edge computing module of the instrument uses a microcontroller with multi-core processing capabilities, preprocesses the collected preliminary data using the Kalman filtering algorithm, and performs preliminary anomaly detection by comparing preset thresholds and data change trends. The thresholds are determined based on the statistical analysis of the historical operation data of the instrument.

[0014] Preferably, the two-way communication layer uses a 5G communication module, which is built with adaptive modulation and coding technology, adjusts the modulation method and coding rate in real time according to the signal strength and channel quality, and transmits data through the TCP / IP protocol to ensure the real-time transmission of high-frequency data.

[0015] Preferably, the data compression of the two-way communication layer uses the LZ77 algorithm, and the transmission redundancy mechanism uses Hamming code verification. The data is compressed and Hamming coded at the sending end, and the corresponding decompression and decoding operations are performed at the receiving end to ensure network transmission stability and prevent data loss or tampering.

[0016] Preferably, the database of the application server uses a non-relational database based on a distributed file system, stores data in slices through the hash algorithm, and uses the B+ tree index structure to achieve fast query of historical and real-time data, improving data storage and query efficiency.

[0017] Preferably, the predictive maintenance model is based on the LSTM neural network in deep learning, takes the real-time data of the instrument in time series as input, optimizes the network parameters through training, realizes anomaly detection, trend analysis and fault prediction, and generates maintenance suggestions and plans according to the prediction results combined with preset rules.

[0018] Preferably, the maintenance decision module communicates with the instrument side and the user side through the message queue middleware Kafka. After receiving the analysis results from the predictive maintenance model, it generates maintenance instructions according to the pre-set strategy and sends them to the Kafka queue. The instrument side and the user side obtain the instructions from the queue and execute the corresponding operations.

[0019] Preferably, the Web-based visualization interface of the user terminal is developed using the React framework, uses the Echarts chart library to display device status, alarm information and maintenance suggestions, and obtains data from the application server through the RESTful API interface and updates it in real time.

[0020] Preferably, the user terminal calls the notification management interface of the mobile phone system on the mobile side to display alarm information in the form of pop-up windows and vibrations, and maintains a long connection with the application server through the WebSocket protocol to receive the latest device status and maintenance suggestion data in real time.

[0021] Compared with the prior art, the present invention has the following beneficial effects: On the instrument side of the present invention, data is collected and preprocessed in real time through the data collection module and the edge computing module, and is uploaded securely in real time with the help of communication technologies such as 5G, breaking the problem of lag in data acquisition in traditional manual inspections, and being able to detect and maintain abnormalities in a timely manner; the application server uses big data analysis and machine learning algorithms to build a predictive maintenance model, combines database data to accurately analyze and predict faults, abandons the dependence on manual experience, improves the accuracy of maintenance, and reduces maintenance costs; the two-way communication layer supports two-way communication, cooperates with the visualization interface and remote operation function of the user terminal, breaks through geographical restrictions, improves operation convenience, enables users to quickly respond to emergencies, reduces equipment downtime, improves operation efficiency, and the remote maintenance results and feedback data are transmitted back to the cloud to optimize the maintenance strategy, continuously improving the system performance and maintenance effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the composition structure and process of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] In order to make the technical means, creative features, achieved purposes and effects of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0024] As Figure 1 shown in the composition structure and process schematic of an instrument remote intelligent maintenance system based on real-time data, the following will be described in conjunction with a detailed embodiment. The specific implementation process is as follows: I. Instrument side Suppose there are various types of instruments in the system, such as temperature instruments, pressure instruments and flow instruments. Taking the temperature instrument as an example, the data collection module of the instrument side uses a high-precision thermistor sensor, whose resistance value changes precisely with temperature, so as to achieve precise measurement of temperature. The data collection module converts the analog signal collected by the sensor into a digital signal through the built-in analog-to-digital conversion circuit. This module supports multiple data interfaces such as 4-20mA and RS485, and the temperature data can be output through the RS485 interface.

[0025] The edge computing module uses a chip that integrates an ARM core and a digital signal processing (DSP) unit. After receiving the temperature data, it preprocesses the data using the median filtering algorithm to remove outliers caused by interference. At the same time, it performs preliminary anomaly detection by comparing with the pre-set normal temperature range threshold. For example, if the normal operating temperature range of a certain device is 20°C - 50°C, when the collected temperature data continuously falls below 20°C or exceeds 50°C, the edge computing module marks the data as abnormal and uploads the preprocessed data and the anomaly mark to the cloud server through the two-way communication layer.

[0026] II. Two-way Communication Layer The two-way communication layer selects a 5G communication module to connect the instrument end and the cloud platform. This module integrates 5G radio frequency front-end, baseband chip, and multi-antenna transceiver technology to ensure that data can be transmitted at high speed and low latency.

[0027] During the data transmission process, the two-way communication layer uses the Huffman coding algorithm for data compression, which can effectively reduce the amount of data transmitted. At the same time, it uses the cyclic redundancy check (CRC) algorithm for transmission redundancy protection. The CRC check code is calculated for the data at the sending end and appended to the data. The receiving end verifies the integrity of the data by recalculating the CRC check code. If the receiving end finds that the data verification fails, it will send a retransmission request to the sending end to ensure reliable data transmission.

[0028] III. Application Server (I) Database The database of the application server adopts a distributed storage architecture and selects CockroachDB, a relational database that supports distribution. The database table structure design includes instrument operation data tables, fault information tables, maintenance record tables, etc. The instrument operation data table stores real-time data and historical data of various instruments, such as temperature, pressure, flow values, and corresponding collection times, instrument numbers, etc.; the fault information table records the instrument status, fault phenomena, fault causes, etc. each time a fault occurs; the maintenance record table stores detailed information of maintenance operations, including maintenance time, maintenance personnel, maintenance content, etc. By establishing indexes and foreign key constraints, the efficiency and accuracy of data query and management are ensured.

[0029] (II) Predictive Maintenance Model The predictive maintenance model utilizes the Long Short-Term Memory (LSTM) algorithm in deep learning. Taking a temperature instrument as an example, the model takes the historical temperature data of the instrument, the operating duration of the equipment, and the status data of associated equipment (such as equipment load affecting temperature changes) as inputs. During the training phase, the model automatically extracts features and patterns from a large amount of historical data and adjusts its own parameters to optimize the prediction performance. After training is completed, the model performs real-time analysis on real-time data. If the model detects an abnormal upward trend in the temperature data and determines based on historical data that this trend may cause an instrument failure, the model will generate maintenance suggestions according to historical data and operating status, such as arranging maintenance personnel to check the cooling system in advance or calibrating the temperature sensor.

[0030] (III) Maintenance Decision Module After receiving the maintenance suggestions from the predictive maintenance model, the maintenance decision module makes a comprehensive decision in combination with the actual situation. For example, when detecting a fault risk in a certain temperature instrument, the maintenance decision module first evaluates the urgency of the fault. If it is judged as an urgent fault, such as a sharp rise in temperature that may damage the equipment, the maintenance decision module immediately sends a stop work instruction to the instrument end through the two-way communication layer to protect the equipment, and at the same time pushes an alarm message and detailed maintenance suggestions, including speculation on the cause of the fault, repair steps, etc., to the user terminal of the maintenance personnel.

[0031] After receiving the information, the maintenance personnel can view the real-time status and historical data of the instrument through the visualization interface of the user terminal to further understand the fault situation. If remote operation is required, maintenance instructions such as calibration and restart can be issued at the user terminal, and the instructions are transmitted to the instrument end through the two-way communication layer for execution. The maintenance decision module records the maintenance operation results and instrument feedback data into the database as inputs for the next round of data training to continuously optimize the maintenance strategy.

[0032] IV. User Terminal Maintenance personnel and management personnel use a user terminal based on the Web and mobile devices. Taking the mobile device as an example, the system is logged in through a specially developed APP. The APP interface displays the equipment status with intuitive visualization charts, such as using a line chart to show the temperature change trend and using a bar chart to compare the number of faults of various instruments in different time periods.

[0033] When an abnormality occurs in a certain instrument, the APP notifies the maintenance personnel through vibration, sound, and pop-up windows. The maintenance personnel click on the pop-up window to enter the detailed fault page to view the fault alarm information, maintenance suggestions, as well as the real-time data and historical curve of the instrument. For example, for an abnormal temperature instrument, the maintenance personnel can remotely view the temperature change curve for a period of time through the APP to assist in judging the cause of the fault.

[0034] In addition, the user terminal supports historical data query and maintenance operation record viewing. Maintenance personnel can query the historical operation data of a certain instrument at any time, analyze the operation trend, and provide a reference for subsequent maintenance work. By viewing the maintenance operation records, they can understand the previous maintenance operations on the instrument and their effects, so as to formulate a more reasonable maintenance plan.

[0035] The above has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A remote intelligent maintenance system for instruments based on real-time data, comprising an instrument, a two-way communication layer, an application server and a user terminal, characterized in that: The instrument is also a real-time data collection layer, which collects the operation data and health status data of the instrument in real time through the data collection module and supports multiple data interfaces; The edge computing module then performs preprocessing and preliminary anomaly detection on the preliminary data, and uploads the preprocessed data to the cloud server; The two-way communication layer uses low-power wide-area Internet of Things or Wi-Fi or 5G Internet of Things communication technology to connect to the cloud platform, supports breakpoint resumption and data compression, uses data compression technology and transmission redundancy mechanism to ensure network transmission stability, and supports two-way communication; The application server is the cloud-based intelligent analysis layer, which mainly includes a database, a predictive maintenance model and a maintenance decision module; wherein: the database mainly stores historical and real-time data and supports fast query; at the same time, it collects fault information and countermeasures for daily maintenance; the predictive maintenance model is based on big data analysis and artificial intelligence algorithms, and uses knowledge decision trees, instrument operation mechanism models, and machine learning technologies to clean, store and analyze real-time data, realize anomaly detection, trend analysis and fault prediction, and put forward corresponding maintenance suggestions and plans based on historical data and operating status; The maintenance decision module pushes further analysis and diagnosis information of specific events to users, can remotely interact with users to complete further in-depth analysis and diagnosis of events, and can also receive and issue instructions from the cloud or remotely, generate corresponding maintenance strategies on the instrument side and user side, and automatically or remotely trigger calibration, cleaning, and parameter adjustment maintenance operations; The user terminal provides a Web- and mobile-based visualization interface to display equipment status, alarm information and maintenance suggestions; it also supports AR remote assistance function, where users can obtain the real view of the instrument through the camera and receive AI-superimposed fault location guidance; users can also monitor the remote maintenance process through the interactive interface, and can manually adjust the maintenance strategy or directly issue maintenance instructions.

2. The remote intelligent maintenance system for instruments based on real-time data according to claim 1 is characterized by: The data acquisition module of the instrument adopts a chip integrating multiple data acquisition functions, which is connected to the sensor of the instrument through the SPI interface, and collects operation data and health status data according to a preset sampling frequency. The sampling frequency is set to 100 Hz according to the dynamic characteristics of the instrument.

3. The remote intelligent maintenance system for instruments based on real-time data according to claim 2 is characterized by: The edge computing module of the instrument uses a microcontroller with multi-core processing capabilities, and utilizes the Kalman filtering algorithm to pre-process the collected preliminary data, and performs preliminary anomaly detection by comparing preset thresholds and data change trends. The thresholds are determined based on statistical analysis of the instrument's historical operating data.

4. The remote intelligent maintenance system for instruments based on real-time data according to claim 3 is characterized by: The two-way communication layer adopts a 5G communication module, which has built-in adaptive modulation and coding technology. It adjusts the modulation mode and coding rate in real time according to signal strength and channel quality, and transmits data through the TCP / IP protocol.

5. The remote intelligent maintenance system for instruments based on real-time data according to claim 4 is characterized by: The data compression of the bidirectional communication layer adopts the LZ77 algorithm, and the transmission redundancy mechanism adopts the Hamming code check. The data is compressed and Hamming code encoded at the sending end, and the receiving end performs corresponding decompression and decoding operations.

6. The remote intelligent maintenance system for instruments based on real-time data according to claim 5 is characterized by: The database of the application server adopts a non-relational database based on a distributed file system, stores data in slices through a hash algorithm, and uses a B+ tree index structure to achieve fast query of historical and real-time data.

7. The remote intelligent maintenance system for instruments based on real-time data according to claim 6 is characterized by: The predictive maintenance model is based on the LSTM neural network in deep learning. It takes the real-time data of the instrument in time series as input, optimizes the network parameters through training, realizes anomaly detection, trend analysis and fault prediction, and generates maintenance suggestions and plans based on the prediction results combined with preset rules.

8. The remote intelligent maintenance system for instruments based on real-time data according to claim 7 is characterized by: The maintenance decision module communicates with the instrument end and the user end through the message queue middleware Kafka. After receiving the analysis results from the predictive maintenance model, it generates maintenance instructions according to the pre-set strategy and sends them to the Kafka queue. The instrument end and the user end obtain the instructions from the queue and perform corresponding operations.

9. The remote intelligent maintenance system for instruments based on real-time data according to claim 8 is characterized by: The web-based visualization interface of the user terminal is developed using the React framework, uses the Echarts chart library to display device status, alarm information and maintenance suggestions, and obtains data from the application server through the RESTful API interface and updates it in real time.

10. The remote intelligent maintenance system for instruments based on real-time data according to claim 9 is characterized in that: The user terminal displays the alarm information in the form of pop-up windows and vibrations by calling the notification management interface of the mobile phone system on the mobile terminal, maintains a long connection with the application server through the WebSocket protocol, and receives the latest equipment status and maintenance suggestion data in real time.

Images