Power plant safety production management system

Abstract

The invention discloses a power plant safety production management system, and belongs to the technical field of power plant management. The system comprises a data acquisition layer, a data transmission layer, a data processing layer, a business application layer and a user interaction layer which are connected in sequence, and all the layers cooperate to realize safety production whole-process management; the data acquisition layer acquires hidden danger data through multiple types of sensors, intelligent instruments and the like; the data transmission layer ensures safe transmission of data; the business application layer provides the functions of equipment full-life-cycle management, real-time safety monitoring, personnel management, emergency processing and custom report statistics; integration and deep application of power plant safety production data are achieved, management efficiency and safety are effectively improved, accident loss is reduced, and the method is suitable for safety production management of power plants of various scales.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power plant management, in particular to a power plant safety production management system. BACKGROUND

[0002] In the current power plant operation process, safety production is the key to ensure the stable operation of the power plant and avoid personnel casualties and property losses. With the continuous expansion of the scale of power plants and the increasing complexity of power generation technology, the traditional safety production management method gradually exposes many problems. The traditional management relies on manual recording and manual inspection, which is not only low in efficiency, but also prone to human errors, resulting in that safety hazards cannot be discovered and handled in time. At the same time, the data of each production link is independent of each other, and the effective integration and sharing of data cannot be realized, and the management personnel cannot comprehensively and real-timely master the safety production status of the power plant, and when a sudden safety event occurs, a scientific and effective emergency treatment scheme cannot be quickly formulated, which may cause the expansion of the accident and bring huge economic losses and bad social influence to the power plant. Therefore, developing a management system capable of realizing the integration of power plant safety production data, real-time monitoring, intelligent early warning and efficient emergency treatment has become a problem to be solved in the current power plant safety production management field. SUMMARY

[0003] In view of the above problems in the prior art, the present application provides a power plant safety production management system, which can effectively solve the problems existing in the traditional power plant safety production management method.

[0004] A power plant safety production management system, comprising a data acquisition layer, a data transmission layer, a data processing layer, a business application layer and a user interaction layer connected in sequence, which cooperatively realize the whole-process management of power plant safety production:

[0005] The data acquisition layer is configured with temperature sensors, pressure sensors, vibration sensors, intelligent instruments, video monitoring equipment and personnel positioning terminals, and is used for collecting equipment operation parameters, environmental parameters, production area image information and staff position information;

[0006] The data transmission layer adopts a 5G and optical fiber hybrid transmission network, and is provided with a data encryption module based on an AES-256 encryption algorithm, which is used for encrypting and classifying the collected data and transmitting it to the data processing layer, wherein the 5G network transmits video monitoring data and personnel positioning data, and the optical fiber network transmits equipment historical operation data;

[0007] The data processing layer performs data cleaning, data fusion, data storage and data mining analysis: the data cleaning adopts rule screening combined with an isolation forest algorithm to remove outliers; the data fusion is based on D-S evidence theory, and the multi-source same type data is fused according to the following formula:

[0008]

[0009] Where m(A) represents the fused evidence, m1, m 2, …,m n For evidence from each acquisition device, A represents the possible values ​​of the data. i The data for the i-th piece of evidence may take the following values; data storage employs a distributed database to store non-time-series data and a time-series database to store time-series data; data mining analysis uses the LSTM model, according to the formula P = σ(W o ·tanh(C t )+b o Predict the probability of equipment failure, where P is the failure probability and W is the failure probability. o Let b be the output layer weight matrix. o C is the output layer bias vector. t σ represents the LSTM cell state, σ is the Sigmoid function, and tanh is the hyperbolic tangent function;

[0010] The business application layer includes an equipment management module, a safety monitoring module, a personnel management module, an emergency response module, and a report statistics module: the equipment management module generates equipment maintenance plans based on fault prediction results; the safety monitoring module monitors abnormal parameters in real time and triggers early warnings; the personnel management module generates training plans and restricts work areas according to job requirements; the emergency response module assesses the level of emergencies and generates handling solutions; and the report statistics module supports the generation of custom-formatted tables and charts.

[0011] The user interaction layer provides two access methods: web and mobile, and supports user-defined interface layout and function configuration.

[0012] Furthermore, the sensors and devices of the data acquisition layer meet the following parameters:

[0013] The temperature sensor is a PT100 platinum resistance thermometer, with a measurement range of -200℃ to 850℃ and an accuracy of ±0.1℃.

[0014] The pressure sensor is of the diffused silicon type, with a measurement range of 0–10 MPa and an accuracy of ±0.2% FS;

[0015] The vibration sensor is a piezoelectric acceleration type, with a measurement range of 0–500g and a frequency range of 0.1Hz–10kHz;

[0016] The intelligent instrument is an electromagnetic flowmeter with a measurement range of 0–1000 m³. 3 / h, accuracy ±0.5% FS;

[0017] The video surveillance equipment is a 4K resolution high-definition camera with a frame rate of 25fps and supports night vision;

[0018] The personnel positioning terminal is of the UWB type, with a positioning accuracy of ±10cm and a communication distance of 0~100m.

[0019] Furthermore, the hardware configuration of the data transmission layer includes:

[0020] 5G base stations support the NR band, with a maximum downlink speed of 10Gbps and a maximum uplink speed of 2Gbps;

[0021] The fiber optic transmission equipment has a transmission rate of 10Gbps and a transmission distance of 0-100km.

[0022] The data encryption module uses a national cryptographic algorithm encryption machine and is compatible with the AES-256 encryption standard.

[0023] Furthermore, the hardware and software configuration of the data processing layer includes:

[0024] The hardware consists of a server configured with two Intel Xeon Gold 6348 processors, 128GB of DDR4 memory, and ten 1TB SSDs.

[0025] The software uses a MySQL distributed database and an InfluxDB time series database, and is built on the TensorFlow framework to create an LSTM model runtime environment that runs on the Linux operating system.

[0026] Furthermore, the emergency response module of the business application layer has the following triggering logic: when the security monitoring module detects that the parameters exceed the preset threshold or the video identifies the violation, it automatically calls the equipment operation data and personnel location data of the data processing layer, assesses the event level according to "the scope of the event's impact × the degree of harm", and generates a processing plan by matching preset emergency rules and historical cases, and pushes it to the web and mobile terminals of managers and rescue personnel.

[0027] Furthermore, the data cleaning process of the data processing layer is as follows:

[0028] The first step is to filter according to preset rules: delete obviously abnormal data with temperature >850℃ or <-200℃ and pressure >10MPa or <0MPa;

[0029] The second step is to input the Isolation Forest algorithm model, set the outlier detection threshold to 0.8, remove the outlier data identified by the algorithm, and output the cleaned dataset.

[0030] Furthermore, the maintenance plan generation logic of the equipment management module in the business application layer is as follows: when the equipment failure probability P predicted by the LSTM model is ≥ 0.6, an "emergency maintenance plan within 72 hours" is automatically generated; when 0.3 ≤ P < 0.6, a "routine maintenance plan within 7 days" is generated; when P < 0.3, a quarterly maintenance plan is generated according to the equipment operation cycle, and all plans are synchronously pushed to the maintenance personnel's terminal and the execution progress is recorded.

[0031] The beneficial effects of this invention are as follows:

[0032] 1. This invention, by setting up a data acquisition layer, a data transmission layer, a data processing layer, a business application layer, and a user interaction layer, achieves comprehensive collection, secure transmission, efficient processing, and in-depth application of power plant safety production data. It breaks the situation where data in each link of traditional power plant safety production management is independent, enabling managers to have a comprehensive and real-time grasp of the power plant's safety production status and providing strong data support for power plant safety production management decisions.

[0033] 2. During the data processing, a data cleaning method combining rules and machine learning, along with a data fusion method based on DS evidence theory, was adopted, which effectively improved the accuracy and reliability of the data. At the same time, artificial intelligence algorithms such as LSTM were used to mine and analyze the data, enabling accurate prediction of equipment failures. This allows for the early detection of potential equipment failures, facilitating timely maintenance by maintenance personnel, reducing equipment downtime, and improving the power plant's production efficiency.

[0034] 3. The system is equipped with comprehensive safety monitoring and emergency response modules, which can monitor the power plant's safety production status in real time. When safety hazards or sudden safety incidents are detected, it can issue early warning signals in a timely manner and generate scientific and effective emergency response plans, which can help to quickly control the development of the situation, reduce the losses caused by safety accidents, and ensure the safe production of the power plant. Attached Figure Description

[0035] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a diagram of the overall system architecture of the present invention;

[0037] Figure 2 This is a schematic diagram of the data acquisition layer of the present invention;

[0038] Figure 3This is a diagram showing the association between the business application layer modules of the present invention. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0040] A power plant safety production management system includes a data acquisition layer, a data transmission layer, a data processing layer, a business application layer, and a user interaction layer. These layers are connected sequentially and work together to achieve efficient management of the entire process of power plant safety production.

[0041] Data Acquisition Layer: This layer collects safety-related data from all production stages of the power plant, including equipment operation data, environmental parameter data, personnel operation data, and safety hazard data. It is equipped with various types of acquisition devices, such as sensors, smart meters, video surveillance equipment, and personnel positioning terminals. Sensors include temperature sensors, pressure sensors, vibration sensors, current sensors, and voltage sensors, used to collect operating parameters such as temperature, pressure, vibration, current, and voltage from various power plant equipment. Smart meters collect parameters such as flow rate and liquid level during the power plant's production process. Video surveillance equipment collects real-time image information from various production areas of the power plant to promptly detect personnel violations and abnormal equipment conditions. Personnel positioning terminals are worn by power plant workers to obtain their real-time location information, ensuring that personnel operate within safe areas.

[0042] Data Transmission Layer: This layer transmits various types of data collected by the data acquisition layer to the data processing layer. It employs a hybrid transmission network based on 5G and fiber optics. The 5G network is used to transmit data with high real-time requirements, such as video surveillance data and personnel positioning data; the fiber optic network is used to transmit larger volumes of data with relatively lower real-time requirements, such as historical equipment operation data. Simultaneously, the data transmission layer also includes a data encryption module, using the AES-256 encryption algorithm to encrypt the transmitted data, ensuring its security and confidentiality during transmission.

[0043] Data processing layer: Used to process and analyze data transmitted from the data transmission layer, including data cleaning, data fusion, data storage, and data mining analysis.

[0044] Data cleaning: A combination of rule-based and machine learning methods is used to remove noise, redundant data, and outliers from the data. First, the data is initially screened according to preset rules, such as deleting data that is clearly outside the reasonable range. Then, the Isolation Forest algorithm is used to detect outliers in the initially screened data to further remove outliers and obtain clean and accurate data.

[0045] Data fusion: Data fusion is performed using the DS evidence theory to fuse data of the same type from different acquisition devices, thereby improving the accuracy and reliability of the data. Let the evidence from the same type of data from different acquisition devices be m1, m2, m3, m4, m5, m6, m7, m8, m9, m1, m1, m1, m2, m1, m2, m3, m4, m5, m6, m7, m8, m9, m1, m1, m1, m2, m1, m2, m1, m2, m1, m2, m3 ... 2, …,m n The fused evidence m is then calculated according to the following formula:

[0046]

[0047] Where A represents the possible values ​​of the data. i Let A be the possible values ​​of the data for the i-th piece of evidence. i This represents the intersection of the possible values ​​of the data in all the evidence.

[0048] Data storage: A combination of distributed and time-series databases is used for data storage. The distributed database is used to store non-time-series data, such as personnel information, basic equipment information, and safety regulations; the time-series database is used to store time-series data, such as equipment operating parameters and environmental parameters, to facilitate time-series analysis and querying of the data.

[0049] Data mining and analysis: Utilizing big data analytics and artificial intelligence algorithms to mine and analyze processed data, including equipment failure prediction, safety risk assessment, and personnel behavior analysis. In equipment failure prediction, an LSTM (Long Short-Term Memory) model is used, employing historical equipment operating parameters as input to construct a failure prediction model that predicts the probability of equipment failure within a future period. Let the historical equipment operating parameter sequence be X = [x1, x2, ..., x...]. T ], where x t Let represent the device operating parameter vector at time t. The LSTM model is used to process this sequence to obtain the probability P of the device failing at time T+k in the future. The formula is as follows: P = σ(W o ·tanh(C t )+b o )

[0050] Among them, W o and b o These are the weight matrix and bias vector of the output layer, respectively. tLet represent the cell state of the LSTM cell at time t, σ be the Sigmoid activation function, and tanh be the hyperbolic tangent activation function.

[0051] Business application layer: This is the core application part of the system. Based on the analysis results of the data processing layer, it provides various business application functions for power plant safety production management, including equipment management module, safety monitoring module, personnel management module, emergency response module, and report statistics module.

[0052] Equipment Management Module: This module manages all types of equipment in the power plant throughout their entire lifecycle, including equipment ledger management, equipment maintenance plan development, equipment maintenance record management, and equipment fault handling. Based on the equipment failure probability predicted by the data processing layer, it automatically generates personalized maintenance plans for each piece of equipment and promptly reminds maintenance personnel to perform maintenance. When equipment fails, it automatically generates a fault report and provides suggested fault handling solutions.

[0053] Safety monitoring module: Used for real-time monitoring of the power plant's safety production status, including equipment operation status monitoring, production area environmental monitoring, and personnel operation behavior monitoring. By acquiring the analysis results from the data processing layer in real time, when it is found that equipment operating parameters exceed the normal range, production area environmental parameters are abnormal, or personnel are engaging in unauthorized operations, an early warning signal is immediately issued and displayed in the user interaction layer.

[0054] Personnel Management Module: Used for managing power plant staff, including personnel information management, personnel training management, personnel attendance management, and personnel work permission management. It develops personalized training plans for personnel based on job requirements and skill levels, and records personnel training progress; it obtains personnel location information through personnel positioning terminals, restricts personnel work areas, and issues warning signals when personnel enter unauthorized work areas.

[0055] Emergency Response Module: This module provides rapid and effective emergency response solutions in the event of a sudden safety incident at the power plant. When the system detects a sudden safety incident, it automatically retrieves relevant data from the data processing layer to assess the incident's severity and impact. Based on preset emergency response rules and historical cases, it generates an emergency response plan and sends the plan to relevant management and emergency response personnel to ensure the smooth operation of the emergency response.

[0056] The report statistics module is used to statistically analyze various data in the power plant's safety production management process and generate corresponding reports. Report types include equipment operation reports, safety accident reports, and personnel training reports. Users can customize the report format and content according to their needs. Reports can be displayed in tables, charts, and other formats, facilitating data analysis and decision-making for management personnel.

[0057] The user interaction layer enables interaction between users and the system, including system login, data query, function operation, and viewing of warning information. This layer provides both web and mobile access methods, allowing users to log in to the system anytime, anywhere via computers, mobile phones, and other devices to obtain information related to power plant safety production management. Furthermore, the user interaction layer features a personalized user interface, allowing users to customize the layout and functions according to their usage habits, improving system usability.

[0058] System Workflow

[0059] Data acquisition phase: Various acquisition devices in the data acquisition layer collect safety-related data from various production stages of the power plant in real time according to the preset acquisition frequency, such as equipment operating parameters, environmental parameters, personnel location information, and video surveillance data.

[0060] Data transmission phase: After being encrypted by the encryption module of the data transmission layer, the collected data is transmitted to the data processing layer via a hybrid 5G and fiber optic network. Data with high real-time requirements is prioritized for transmission via the 5G network, while larger volumes of non-real-time data are transmitted via the fiber optic network.

[0061] Data processing stage: The data processing layer first cleans the transmitted data to remove noise, redundant data and outliers; then, it uses DS evidence theory to fuse multi-source data of the same type to improve the accuracy and reliability of the data; the fused data is stored in a distributed database and a time-series database respectively; finally, artificial intelligence algorithms such as LSTM are used to mine and analyze the data to obtain equipment failure prediction results, safety risk assessment results and personnel behavior analysis results, etc.

[0062] Business Application Phase: The business application layer implements various business functions based on the analysis results of the data processing layer. For example, the equipment management module formulates maintenance plans based on equipment failure prediction results; the safety monitoring module issues early warning signals based on safety risk assessment results and real-time data monitoring; the emergency response module generates emergency response plans when a sudden safety incident is detected; and the report statistics module generates corresponding statistical reports based on various types of data.

[0063] User interaction phase: Users log in to the system via web or mobile devices, perform data queries and function operations in the user interaction layer, view system-generated warning information and reports, and personalize the system according to their own needs.

[0064] In summary, the power plant safety production management system provided by this invention can effectively solve the problems existing in traditional power plant safety production management methods, and has high practicality and promotion value.

[0065] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A power plant safety production management system, characterized in that, It includes a data acquisition layer, a data transmission layer, a data processing layer, a business application layer, and a user interaction layer, which are connected in sequence. These layers work together to achieve full-process management of safe production in power plants. The data acquisition layer is equipped with temperature sensors, pressure sensors, vibration sensors, smart meters, video monitoring equipment, and personnel positioning terminals, which are used to collect equipment operating parameters, environmental parameters, production area image information, and staff location information. The data transmission layer adopts a hybrid 5G and fiber optic transmission network and is equipped with a data encryption module based on the AES-256 encryption algorithm. This module is used to encrypt the collected data and then classify and transmit it to the data processing layer. The 5G network transmits video surveillance data and personnel positioning data, while the fiber optic network transmits historical operating data of the equipment. The data processing layer performs data cleaning, data fusion, data storage, and data mining analysis: data cleaning uses rule-based filtering combined with the isolated forest algorithm to remove outliers; data fusion is based on the DS evidence theory and fuses multi-source data of the same type according to the following formula: Where m(A) represents the fused evidence, m1, m 2, …,m n For evidence from each acquisition device, A represents the possible values ​​of the data. i The data for the i-th piece of evidence may take the following values; data storage employs a distributed database to store non-time-series data and a time-series database to store time-series data; data mining analysis uses the LSTM model, according to the formula P = σ(W o ·tanh(C t )+b o Predict the probability of equipment failure, where P is the failure probability and W is the failure probability. o Let b be the output layer weight matrix. o C is the output layer bias vector. t σ represents the LSTM cell state, σ is the Sigmoid function, and tanh is the hyperbolic tangent function; The business application layer includes an equipment management module, a security monitoring module, a personnel management module, an emergency response module, and a report statistics module: the equipment management module generates equipment maintenance plans based on fault prediction results; the security monitoring module monitors abnormal parameters in real time and triggers early warnings; The personnel management module generates training plans and restricts work areas based on job requirements; the emergency response module assesses the severity of emergencies and generates response plans; the report statistics module supports the generation of custom-formatted tables and charts. The user interaction layer provides two access methods: web and mobile, and supports user-defined interface layout and function configuration.

2. The power plant safety production management system according to claim 1, characterized in that: The sensors and devices in the data acquisition layer meet the following parameters: The temperature sensor is a PT100 platinum resistance thermometer, with a measurement range of -200℃ to 850℃ and an accuracy of ±0.1℃. The pressure sensor is of the diffused silicon type, with a measurement range of 0–10 MPa and an accuracy of ±0.2% FS; The vibration sensor is a piezoelectric acceleration type, with a measurement range of 0–500g and a frequency range of 0.1Hz–10kHz; The intelligent instrument is an electromagnetic flowmeter with a measurement range of 0–1000 m³. 3 / h, accuracy ±0.5%FS; The video surveillance equipment is a 4K resolution high-definition camera with a frame rate of 25fps and supports night vision; The personnel positioning terminal is of the UWB type, with a positioning accuracy of ±10cm and a communication distance of 0~100m.

3. The power plant safety production management system according to claim 1, characterized in that: The hardware configuration of the data transmission layer includes: 5G base stations support the NR band, with a maximum downlink speed of 10Gbps and a maximum uplink speed of 2Gbps; The fiber optic transmission equipment has a transmission rate of 10Gbps and a transmission distance of 0-100km. The data encryption module uses a national cryptographic algorithm encryption machine and is compatible with the AES-256 encryption standard.

4. The power plant safety production management system according to claim 1, characterized in that: The hardware and software configuration of the data processing layer includes: The hardware consists of a server configured with two Intel Xeon Gold 6348 processors, 128GB of DDR4 memory, and ten 1TB SSDs. The software uses a MySQL distributed database and an InfluxDB time-series database, and is built on the TensorFlow framework to create an LSTM model runtime environment that runs on the Linux operating system.

5. A power plant safety production management system according to claim 1, characterized in that: The emergency response module of the business application layer is triggered by the following logic: when the security monitoring module detects that the parameters exceed the preset threshold or the video identifies the violation, it automatically calls the equipment operation data and personnel location data of the data processing layer, assesses the event level according to "the scope of the event's impact × the degree of harm", and generates a processing plan by matching the preset emergency rules and historical cases, and pushes it to the web and mobile terminals of the managers and rescue personnel.

6. The power plant safety production management system according to claim 1, characterized in that: The data cleaning process of the data processing layer is as follows: The first step is to filter according to preset rules: delete obviously abnormal data with temperature >850℃ or <-200℃ and pressure >10MPa or <0MPa; The second step is to input the Isolation Forest algorithm model, set the outlier detection threshold to 0.8, remove the outlier data identified by the algorithm, and output the cleaned dataset.

7. A power plant safety production management system according to claim 1, characterized in that: The device management module of the business application layer has the following maintenance plan generation logic: when the device failure probability P predicted by the LSTM model is ≥ 0.6, an "emergency maintenance plan within 72 hours" is automatically generated; when 0.3 ≤ P < 0.6, a "routine maintenance plan within 7 days" is generated; when P < 0.3, a quarterly maintenance plan is generated according to the device operation cycle, and all plans are pushed to the maintenance personnel's terminal simultaneously and the execution progress is recorded.

Images