Multi-source intelligent analysis system and method for safety management of personnel in power plant area

Abstract

The invention discloses a multi-source intelligent analysis system and method for power plant area personnel safety management. According to the system, multiple technologies including big data analysis, a deep learning algorithm, UWB high-precision positioning, Beidou positioning and navigation, a pressure sensor, a smoke alarm and an attitude sensor are integrated into an intelligent helmet worn by a worker, and a complete safety management solution is formed. The system can collect position information, behavior states and environmental data of operators in real time, perform multi-source data fusion processing, and intelligently analyze the standardization of operation behaviors and the safety of the surrounding environment. Through multi-dimensional risk assessment, early warning of safety risks of operators is realized, early warning prompts are sent to managers, and intelligent management and control support is provided for personnel safety in a power plant. The system is mainly composed of a data acquisition module, a data fusion processing module, a behavior analysis module and an early warning and feedback module, and can significantly improve the safety management level of a power plant area and the protection effect of operators.

Description

Technical Field

[0001] The present invention relates to the field of safety management and intelligent monitoring, and in particular to a multi-source intelligent analysis system and method for safety management of personnel in a power plant area. Background Art

[0002] With the development of industrial automation and intelligence, power plants, as important places for energy production, have become particularly critical for safety management. There are many potential safety hazards in the power plant area, such as high-temperature and high-pressure equipment, electrical facilities, chemicals and smoke, which pose severe challenges to the work behavior and safety of personnel. Traditional safety management methods mostly rely on manual inspections and video monitoring. These methods have problems such as blind spots in monitoring, slow response speed, and human misjudgment, making it difficult to detect and deal with safety hazards in a timely manner. Therefore, it is necessary to introduce more intelligent technical means to achieve real-time monitoring of personnel and analysis of work behavior, thereby improving the efficiency and accuracy of safety management.

[0003] At present, some advanced positioning technologies such as Ultra Wide Band (UWB) and Beidou navigation system can provide high-precision personnel location data; deep learning and big data technologies can intelligently analyze complex operating behaviors; weight sensors and smoke alarms can detect abnormal conditions in the environment. However, how to organically combine these technologies to achieve fusion processing of multi-source data and improve the accuracy and real-time performance of personnel safety management in power plant areas is still an urgent problem to be solved.

[0004] Chinese patent document CN 115240365A discloses a dynamic situational awareness solution that integrates multi-dimensional information such as UWB positioning and video surveillance. It mainly performs dynamic perception through UWB positioning, deep learning and behavior state recognition, and issues early warning based on safe distance. Although the present invention also involves UWB positioning and multi-dimensional information fusion, it expands more sensor data types, uses Beidou, pressure sensors, smoke alarms and attitude sensors, and combines algorithms of different technologies to form a special safety management system for power plant areas; the solution of Chinese patent document CN 115240365A is based on UWB positioning and video data, and the present invention also introduces rich data collection methods such as pressure sensors, smoke alarms and attitude sensors; Chinese patent document CN 115240365A focuses on dynamic monitoring of power workers, while the present invention targets the overall safety management needs of power plant areas and has a wider range of application scenarios and functions. Based on multi-dimensional data fusion, the present invention combines more environmental and personnel status information, is more sensitive to potential abnormal behavior recognition, and risk warnings are more accurate and immediate, which is suitable for high-risk environment management in power plants.

[0005] Chinese patent document CN 117274841A discloses the use of drones and deep learning networks to monitor the safety of non-stop operations in high-voltage distribution network areas in real time by detecting the clothing and behavior of operators. The present invention is aimed at the overall personnel safety management in the power plant area, and does not rely on a single high-voltage environment, but integrates multi-source sensors and positioning means, significantly enhancing the extensiveness and multi-dimensionality of environmental monitoring; Chinese patent document CN 117274841A mainly collects image data and locates through drones, while the present invention adopts a fixed multi-source sensor solution, including various sensors in helmet equipment; Chinese patent CN 117274841A focuses on non-stop operations, while the present invention is suitable for overall personnel dynamic monitoring and risk management in the power plant area; the present invention is not limited to a single dimension of the operating environment, but comprehensively perceives and warns of various dangerous conditions in the power plant, has stronger environmental adaptability, and can respond to more types of sudden safety incidents.

[0006] Chinese patent document CN 110119656A mainly uses video surveillance and face recognition technology to detect violations of operating personnel. The system identifies violations on site through video acquisition units and face detection modules, and sends warnings. The present invention, by fusing multiple sensor data, including UWB, Beidou, pressure sensors and smoke alarms, uses different deep learning algorithms to conduct a comprehensive analysis of behavior, location and environmental factors, and provides real-time warnings; Chinese patent document CN110119656A relies on video surveillance and face detection, while the present invention integrates multiple data sources, not limited to image data; the violation monitoring of Chinese patent document CN 110119656A is limited to people appearing in video images, while the intelligent analysis system of the present invention covers multi-dimensional behavior recognition, including posture, position offset and environmental anomalies.

[0007] Chinese patent document CN 118506215A discloses an image detection method based on the YOLOv5s network and the ECA-Net attention mechanism module, which mainly detects whether the operator is wearing a helmet. The patent uses image recognition technology to achieve real-time monitoring of helmet wearing. The present invention combines multiple data sources such as UWB, Beidou positioning, pressure sensors, smoke alarms and attitude sensors, and adopts deep learning algorithms and data fusion technology. It can not only monitor the wearing of helmets by operators, but also conduct a comprehensive risk assessment of personnel positions, working behaviors and environmental data; Chinese patent document CN 118506215A focuses on image recognition technology based on deep learning, mainly detecting the wearing of helmets, while the present invention adopts a multi-sensor fusion method to realize multi-dimensional data fusion analysis of personnel positioning, behavior and environment; Chinese patent document CN 118506215A is suitable for single helmet detection, while the present invention covers the dynamic management of overall personnel behavior and environmental data; the present invention can perceive the global risk status of the power plant area in real time. Compared with a single detection solution, the comprehensiveness provided by multi-source data improves the accuracy and reliability of safety monitoring, and effectively makes up for the limitations of single image detection in complex environments. Summary of the invention

[0008] The technical problem to be solved by the present invention is that the present invention provides a multi-source intelligent analysis system and method for personnel safety management in power plant areas. The system is finally integrated into the staff helmet to form a smart helmet by combining big data analysis, deep learning algorithms, UWB high-precision positioning, Beidou positioning and navigation devices, pressure sensors, smoke alarms and attitude sensors, so as to achieve real-time monitoring of operating personnel and early warning of safety risks. The system can not only locate the position of personnel in real time, but also detect whether their operating behavior is standardized, and combine environmental data (such as pressure changes, smoke concentration, etc.) to perform multi-dimensional risk assessment, thereby realizing intelligent safety management and control of personnel in power plants.

[0009] To achieve the above objectives, the present invention proposes a multi-source intelligent analysis system and method for personnel safety management in power plant areas, which includes: a data acquisition module, a data fusion processing module, a behavior analysis module, and an early warning and feedback module.

[0010] The data acquisition module is composed of data provided by UWB and Beidou positioning and navigation devices, pressure sensors, smoke alarms and attitude sensors; further, the UWB and Beidou positioning provide high-precision personnel location data; further, the navigation device guides workers to work areas; further, the pressure sensor provides data on whether the workers are wearing helmets; further, the smoke alarm provides data on smoke concentration in the air; the attitude sensor provides data on workers' attitude and movement amplitude.

[0011] The data fusion processing module is designed to deeply integrate and analyze the diversified information obtained by the data acquisition module, including but not limited to precise location coordinates, real-time pressure values, detailed smoke concentration monitoring data, and detailed personnel posture recognition information. Through the comprehensive operation of advanced algorithms, it can efficiently generate a comprehensive and accurate safety risk assessment report, providing solid data support for subsequent safety warnings and emergency responses; further, the advanced algorithm is to fuse various sensor data and calculate the risk assessment value, so that managers can make decisions in a timely manner. The specific implementation formula is as follows:

[0012]

[0013] Where R(t) represents the original risk assessment value at time t, is the final risk assessment value, F(R(t)) is a nonlinear transformation of the original risk assessment value R(t), i represents different data sources (position, pressure, smoke concentration, posture, etc.), n represents the total number of data sources, X i (t) is the data vector for each data source at time t, ω i (t) is the weight of each data source at time t, k is a scaling factor used to adjust the amplitude of the function output, b is a bias term that helps the model capture the basic risk level and output a certain risk assessment even when all input data are very low, and m is the range threshold of the output value, ensuring that the risk assessment value is within a reasonable range.

[0014] The behavior analysis module uses advanced deep learning algorithms to deeply analyze the posture and motion amplitude data collected in real time by the data acquisition module. Through precise pattern recognition and feature extraction technology, it intelligently determines whether the behavior of the operator complies with safety regulations, and effectively identifies and warns of potential abnormal operating behaviors or safety risks. Furthermore, the deep learning algorithm is specifically implemented as follows:

[0015]

[0016] Where R(t) represents the original risk assessment value at time t, X pose (t) is the data related to the posture, X motion (t) is the data related to the range of motion, f CNN (X pose (t)) represents the evaluation value obtained by processing the posture data using a convolutional neural network (CNN), and f RNN (X motion(t)) represents the evaluation value obtained by processing the motion amplitude data using a recurrent neural network (RNN). is the final risk assessment value, f(R(t)) is a nonlinear transformation of the original risk assessment value R(t), i represents different data sources (position, pressure, smoke concentration, posture, etc.), n represents the total number of data sources, X i (t) is the data vector for each data source at time i, ω i (t) is the weight of each data source at time t, k is a scaling factor used to adjust the amplitude of the function output, b is a bias term that helps the model capture the basic risk level and output a certain risk assessment even when all input data are very low, and m is the range threshold of the output value, ensuring that the risk assessment value is within a reasonable range.

[0017] The early warning and feedback module, when detecting an abnormal situation, the system will issue an alarm through sound, light, text message, etc., to prompt relevant personnel to pay attention; further, the sound is an alarm sound played in the helmet; further, the light is a red light installed in the helmet; further, the text message is a warning message sent to the helmet owner and administrator through the back-end server; further, the abnormal situation is specifically as follows:

[0018] When indoors, UWB positioning data is used as the main data. When the operator's positioning information exceeds the range threshold specified within the factory, the operator's position is judged to be abnormal;

[0019] When working outdoors, the Beidou positioning data is mainly used. When the positioning information of the operator exceeds the range threshold specified within the factory area and the Beidou planned path, the operator's position is judged to be abnormal;

[0020] When the pressure sensor value in the helmet worn by the operator exceeds the range threshold specified within the factory, it is determined that the operator is not wearing a helmet, which is considered abnormal;

[0021] When the smoke sensor in the helmet worn by the operator detects that the smoke concentration exceeds the range threshold specified within the factory, it is determined that the operator is smoking during the operation, which is considered abnormal;

[0022] When an operator is performing precise and meticulous operations, if the operator's posture and movement range exceed the range threshold specified within the factory, the operator is deemed to have not performed the work in a standardized manner and is considered abnormal.

[0023] At the same time, the system will record abnormal information for easy subsequent analysis and tracing.

[0024] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0025] (1) The present invention uses information fusion of multiple data sources (including UWB positioning, Beidou positioning, pressure sensors, smoke alarms and attitude sensors, etc.) and conducts comprehensive analysis of multivariate data through deep learning algorithms to achieve real-time accurate positioning and behavior analysis of personnel. This multi-source data fusion method can effectively make up for the lack of information caused by a single data source and improve the accuracy and reliability of security monitoring.

[0026] (2) By using deep learning algorithms and advanced pattern recognition technology, the present invention can intelligently determine whether the operator's behavior complies with safety regulations, timely identify high-risk behaviors such as abnormal operations, falls or loss of balance, and issue warnings. Compared with traditional rule-based monitoring systems, the present invention can autonomously learn and adapt to new risk patterns based on dynamic changes in data, thereby achieving more intelligent risk management.

[0027] (3) The present invention integrates multiple technologies into one system to form a complete safety management solution. This comprehensive approach can provide multi-level protection for personnel safety management in the power plant area, including precise positioning, behavior analysis, smoke detection, etc., effectively reducing the probability of safety accidents.

[0028] (4) The system of the present invention has a rapid response capability, can process various collected data in real time, and quickly make risk judgments and early warnings. Compared with traditional manual monitoring or post-analysis, the present invention provides a more efficient real-time safety management method, which helps to take timely response measures and reduce accident losses.

[0029] (5) The present invention is flexible in design and can adapt to the safety management requirements of power plants of different sizes and their different regions. The modular design of the system enables it to configure different sensors and data processing algorithms according to actual needs, and is suitable for various complex operating environments.

[0030] (6) The intelligent system of the present invention can significantly reduce the manual intervention required for traditional security monitoring and improve management efficiency. Through automated monitoring and early warning, managers can devote more energy to emergency response and decision-making, effectively improving the overall security management level. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. The drawings are only used to illustrate the preferred embodiments and are not considered to be limitations of this patent.

[0032] Figure 1 A schematic diagram of the process architecture of a multi-source intelligent analysis system and method for personnel safety management in a power plant area according to an embodiment of the present invention;

[0033] Figure 2 A schematic diagram of the appearance of a helmet in a multi-source intelligent analysis system and method for personnel safety management in a power plant area according to an embodiment of the present invention; DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] like Figure 1 As shown, the logic flow diagram of the multi-source intelligent analysis system and method for power plant area personnel safety management includes a data acquisition module, a data fusion processing module, a behavior analysis module, and an early warning and feedback module; the logic includes:

[0036] S1, consists of data provided by UWB and Beidou positioning and navigation devices, pressure sensors, smoke alarms and attitude sensors; further, the UWB and Beidou positioning provide high-precision personnel location data; further, the navigation device guides the staff to the work area; further, the pressure sensor provides data on whether the staff is wearing a helmet; further, the smoke alarm provides data on the smoke concentration in the air; the attitude sensor provides data on the staff's attitude and movement amplitude.

[0037] S2, in order to deeply integrate and analyze the diversified information obtained by the data acquisition module, including but not limited to precise location coordinates, real-time pressure values, detailed smoke concentration monitoring data and detailed personnel posture recognition information, through the comprehensive operation of advanced algorithms, a comprehensive and accurate safety risk assessment report can be efficiently generated to provide solid data support for subsequent safety warnings and emergency responses; further, the advanced algorithm is to fuse various sensor data and calculate the risk assessment value, so that managers can make decisions in a timely manner. The specific implementation formula is as follows:

[0038]

[0039] Where R(t) represents the original risk assessment value at time t, is the final risk assessment value, f(R(t)) is a nonlinear transformation of the original risk assessment value R(t), i represents different data sources (position, pressure, smoke concentration, posture, etc.), n represents the total number of data sources, X i (t) is the data vector for each data source at time t, ω i(t) is the weight of each data source at time t, k is a scaling factor used to adjust the amplitude of the function output, b is a bias term that helps the model capture the basic risk level and output a certain risk assessment even when all input data are very low, and m is the range threshold of the output value, ensuring that the risk assessment value is within a reasonable range.

[0040] S3 uses advanced deep learning algorithms to deeply analyze the posture and motion amplitude data collected in real time by the data acquisition module. Through precise pattern recognition and feature extraction technology, it intelligently determines whether the behavior of the operator complies with safety regulations, effectively identifies and warns of potential abnormal operating behaviors or safety risks. Furthermore, the deep learning algorithm is specifically implemented as follows:

[0041]

[0042] Where R(t) represents the original risk assessment value at time t, X pose (t) is the data related to the posture, X motion (t) is the data related to the range of motion, f CNN (X pose (t)) represents the evaluation value obtained by processing the posture data using a convolutional neural network (CNN), and f RNN (X motion (t)) represents the evaluation value obtained by processing the motion amplitude data using a recurrent neural network (RNN). is the final risk assessment value, f(R(t)) is a nonlinear transformation of the original risk assessment value R(t), i represents different data sources (position, pressure, smoke concentration, posture, etc.), n represents the total number of data sources, X i (t) is the data vector for each data source at time t, ω i (t) is the weight of each data source at time t, k is a scaling factor used to adjust the amplitude of the function output, b is a bias term that helps the model capture the basic risk level and output a certain risk assessment even when all input data are very low, and m is the range threshold of the output value, ensuring that the risk assessment value is within a reasonable range.

[0043] S4, when an abnormal situation is detected, the system will issue an alarm through sound, light, text message, etc. to prompt relevant personnel to pay attention; further, the sound is an alarm sound played in the helmet; further, the light is a red light installed in the helmet; further, the text message is a warning message sent to the helmet owner and administrator through the backend server; further, the abnormal situation is specifically as follows:

[0044] When indoors, UWB positioning data is used as the main data. When the operator's positioning information exceeds the range threshold specified within the factory, the operator's position is judged to be abnormal;

[0045] When working outdoors, the Beidou positioning data is mainly used. When the positioning information of the operator exceeds the range threshold specified within the factory area and the Beidou planned path, the operator's position is judged to be abnormal;

[0046] When the pressure sensor value in the helmet worn by the operator exceeds the range threshold specified within the factory, it is determined that the operator is not wearing a helmet, which is considered abnormal;

[0047] When the smoke sensor in the helmet worn by the operator detects that the smoke concentration exceeds the range threshold specified within the factory, it is determined that the operator is smoking during the operation, which is considered abnormal;

[0048] When an operator is performing precise and meticulous operations, if the operator's posture and movement range exceed the range threshold specified within the factory, the operator is deemed to have not performed the work in a standardized manner and is considered abnormal.

[0049] At the same time, the system will record abnormal information for easy subsequent analysis and tracing.

[0050] Figure 2 The figure is a schematic diagram of the appearance of the helmet in the present invention, which is made of glass fiber reinforced plastic. The use of this material can significantly improve the impact resistance and heat resistance of the helmet, while having good electrical insulation and chemical corrosion resistance, ensuring reliable protection for operators under the high safety requirements of the power plant area.

[0051] The above is a detailed introduction to the multi-source intelligent analysis system and method for personnel safety management in power plant areas provided by the embodiments of the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A multi-source intelligent analysis system and method for personnel safety management in power plant areas, characterized in that: include: The data acquisition module is used to collect the position data, pressure data, smoke concentration data, and posture and motion amplitude data of personnel in the power plant area, wherein the position data is obtained through UWB and Beidou positioning and navigation devices, the pressure data is obtained through pressure sensors, the smoke concentration data is obtained through smoke alarms, and the posture and motion amplitude data is obtained through posture sensors; A data fusion processing module is used to deeply integrate and analyze the diversified information obtained by the data acquisition module, and generate a comprehensive and accurate security risk assessment report through comprehensive calculations using advanced algorithms; The behavior analysis module uses a deep learning algorithm to deeply analyze the posture and movement amplitude data collected in real time by the data acquisition module, and intelligently judges whether the behavior of the operator complies with safety regulations; The early warning and feedback module will issue an alarm through sound, light, text message, etc. when an abnormal situation is detected, and the abnormal information will be recorded.

2. The method of generating a comprehensive and accurate security risk assessment report through comprehensive calculation using advanced algorithms according to claim 1, characterized in that: The specific implementation formula of the advanced algorithm is: Where R(t) represents the original risk assessment value at time t, is the final risk assessment value, f(R(t)) is a nonlinear transformation of the original risk assessment value R(t), i represents different data sources, n represents the total number of data sources, X i (t) is the data vector for each data source at time t, ω i (t) is the weight of each data source at time t, k is a scaling factor, b is a bias term, and m is the range threshold of the output value.

3. The method of claim 1, wherein the method comprises: The specific implementation formula of the deep learning algorithm is: Among them, X pose (t) is the data related to the posture, X motion (t) is the data related to the range of motion, f CNN (X pose (t)) represents the evaluation value obtained by processing the posture data using the convolutional neural network, and f RNN (X motion (t)) represents the evaluation value obtained by processing the motion amplitude data using a recurrent neural network.

4. The system according to claim 1, characterized in that The early warning and feedback module issues an alarm and records abnormal information in the following situations: When indoors, the UWB positioning data exceeds the range threshold specified within the factory; When outdoors, the Beidou positioning data is inconsistent with the range threshold specified within the factory and the Beidou planned path; The pressure sensor value exceeds the range threshold specified within the factory; The smoke concentration detected by the smoke sensor exceeds the range threshold specified within the factory; The operator's posture and movement range exceeded the range threshold specified within the factory.

5. The system according to claim 1, comprising the steps of: Collect the position data, pressure data, smoke concentration data, posture and movement amplitude data of personnel in the power plant area; In-depth integration and analysis of the collected diversified information to generate a comprehensive and accurate security risk assessment report; Use deep learning algorithms to deeply analyze the posture and motion amplitude data collected in real time to determine whether the operator's behavior complies with safety regulations; When an abnormal situation is detected, an alarm will be issued through sound, light, text message, etc., and the abnormal information will be recorded.

Images