A power generation vehicle risk identification active early warning method

Abstract

The present application relates to the technical field of power generation vehicle monitoring, and discloses a kind of based on power generation vehicle risk identification active early warning method, comprising the following steps: S1, establish power generation vehicle knowledge expert library;S2, establish data model;S3, training and modification are carried out to data model;S4, collection environment data and operating data;S5, judge risk;S6, risk early warning.The present application obtains the environment data inside and outside power generation vehicle, image condition, real-time data condition of engine and generator using Internet of Things technology, and the real-time condition and risk condition of power generation vehicle operation are comprehensively evaluated using power generation vehicle knowledge expert library and power generation vehicle comprehensive data risk identification data model, the automatic identification and active early warning of power generation vehicle real-time operation condition and risk are formed, to help power generation vehicle security personnel to carry out auxiliary reference and analysis to the overall condition of power generation vehicle.

Description

Technical Field

[0001] This invention relates to the field of generator truck monitoring technology, specifically a proactive early warning method based on generator truck risk identification. Background Technology

[0002] When various types of generator cars, such as diesel and flywheel generator cars, are in operation, it is necessary to monitor their real-time unit operating status, environmental conditions, and video images to understand the generator's operating status and provide risk warnings, so as to carry out targeted maintenance and reduce losses caused by generator car failures.

[0003] However, human observation and judgment cannot accurately analyze the generator operation status of the generator truck, nor can they obtain real-time information on the internal and external environment and real-time images of the generator truck during power generation. Therefore, it is impossible to make a comprehensive judgment on the entire power generation process of the generator truck, to accurately arrange the maintenance plan of the generator truck, and to have a complete grasp of the overall status of the generator trucks of the company or unit. In addition, there is an urgent need to obtain early warning of potential risks during the operation of the generator truck, so as to help it quickly handle and predict potential risks. Summary of the Invention

[0004] The purpose of this invention is to provide a proactive early warning method based on generator vehicle risk identification to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A proactive early warning method based on generator vehicle risk identification includes a front-end module and a back-end platform module. The front-end module includes environmental sensors and vehicle-mounted video installed inside the generator vehicle, and a communication module installed inside the generator set controller. The environmental sensors include temperature and humidity sensors, noise sensors, and smoke sensors. The back-end platform module includes a generator vehicle knowledge expert database, a generator vehicle comprehensive data risk identification data model, a multi-class standard dataset, an IoT gateway, an IoT platform, and a message push center. The early warning method includes the following steps: S1. Utilize the professional standards, data, specifications, and environmental impact information related to the generator car and its engine, generator, and environmental impact during operation, as well as the generator car operation and maintenance specifications, standards, and data information, combined with relevant standards, specifications, and other relevant professional data information provided by experts, to establish a generator car knowledge expert database. S2. Based on the professional standards and standardized data information of the generator vehicle knowledge expert database in step S1, and combined with the historical operation failure information of the generator vehicle, determine the data and its related processes, define the data, ensure the integrity of the data, define the operation process, select the data storage technology, and use the generator vehicle environmental data, images, engine, generator and other comprehensive elements to determine the data model algorithm, thereby establishing a comprehensive data risk identification data model for generator vehicles. The method for establishing the comprehensive data risk identification model for generator vehicles is as follows: Different weights k are assigned to the generator vehicle data items based on the different types of data collected. n The probability L(θ) of identifying the generator vehicle risk is calculated, where P(Y) N Given N distinct data items, the probability of an event occurring is: L(θ) = P(Y1)k1P(Y2)k2...p(Y n ) k n P I =1 / (1+E -(α+βX I ) ) When multiple data items affect the operation of the generator car, N parameters and weight values ​​can be passed in. When the value of L(θ) is greater than 1, it indicates that there is a probability that the generator car will encounter a risk. S3. Based on the data model for risk identification of generator vehicle comprehensive data in step S2, import the data model algorithm, and use the generator vehicle knowledge expert database and its historical data to extract the parameter data of the data model algorithm to form a multi-class standard dataset. The data model algorithm is as follows: Let for Loss estimates for different data types, for The probability of occurrence of risks for different data types is used to calculate the risk assessment value by combining the loss estimate and probability of the same data type. :

[0006] The level of risk is determined by the risk assessment value; S4. Using Internet of Things (IoT) technology, environmental sensors and vehicle-mounted video are installed inside the generator vehicle, and a communication module is installed inside the generator set controller. The environmental sensors, vehicle-mounted video, and communication module are connected through communication lines, and communication is achieved using IoT gateways. S5. Use the Internet of Things platform to collect and analyze the data obtained in step S4. Integrate and refine the collected environmental, image, and unit data according to the parameter configuration of the data model algorithm, input it into the data model, compare the data model with the generator vehicle knowledge expert database, calculate the risk identification result, and generate a risk identification report. S6. Based on the risk identification report in step S5, when there is a potential risk to the generator vehicle, the system automatically sends the report to the message push center. The message push center accurately locates the relevant operation and maintenance personnel, management personnel, and on-site support personnel of the generator vehicle. Through technologies such as SMS and mobile terminal push, the system pushes the potential risks of the generator vehicle to the relevant personnel, providing data and reference support for the relevant personnel to handle the risks, thereby realizing proactive early warning of generator vehicle risks.

[0007] As a further aspect of the present invention: In step S1: information management is carried out on the data information related to the generator vehicle knowledge expert database, and the data content of the generator vehicle knowledge expert database is classified, organized, improved and processed to provide data support for the active identification model; the data information that has been entered into the database and classified, organized, improved and processed is cleaned by binning, clustering and regression methods. Data cleaning includes checking data consistency and handling invalid and missing values, thereby improving the data quality of the expert database.

[0008] As a further aspect of the present invention: In step S3, the sample extraction of the data model algorithm requires data training, and the data training method is as follows: S11. A supervised learning training mode is adopted. By continuously inputting parameter data, the data model is trained to predict the results. The prediction results are compared with the actual results of the training data, and the prediction model is continuously adjusted until the model's prediction results reach an expected accuracy. S12. The training mode adopts an unsupervised learning approach. By importing the parameter dataset, the training data is used to predict the results, and finally the prediction results are analyzed. S13. Train the data model using both supervised and unsupervised methods, employ cross-validation, observe the loss curve and test result curve to analyze the reasons, evaluate the data model, and adjust and optimize the parameters based on the evaluation results. S14. Through the adjusted data model and parameter optimization, continuous machine learning is carried out until it meets the actual standards.

[0009] As a further aspect of the present invention: In step S4: Based on the MODBUS protocol, configure commands for acquiring temperature, humidity, noise, and smoke detection data, as well as commands for acquiring unit data; by configuring video parameters, connect the video images to the IoT gateway; and send data commands through the gateway to acquire real-time data from environmental sensors and unit operation.

[0010] As a further aspect of the present invention: In step S4: Based on the MQTT protocol, connection, subscription, and publishing parameters and services are configured; the IoT gateway connects to the IoT platform; the IoT gateway collects data through the issued commands and publishes it to the IoT platform; the IoT platform receives the data published by the IoT gateway through the subscribed parameter information and parses the data according to different data types based on the communication protocol.

[0011] As a further aspect of the present invention: In step S5: using the risk identification results, the risk identification content is converted into a risk identification report using information technology, thereby realizing risk identification during the operation of the generator car.

[0012] As a further aspect of the present invention: the generator vehicle knowledge expert database in step S1 consists of a remote data entry and maintenance module, a query module, a classification and management module, a security and confidentiality module, and a system recovery module, wherein, The remote data entry and maintenance module allows experts to independently enter and maintain data, enabling timely updates. The query module is used for dynamic querying of generator vehicle data and outputs the queried information and related structures in report form. The classification and management module is used to categorize different information types. The security and confidentiality module is used to set different query permissions for different operators to prevent users with insufficient permissions from arbitrarily accessing and modifying data. The system recovery module includes the creation, backup, and recovery of the knowledge expert database. By regularly backing up data information, it prevents data loss in the event of system failure, thereby ensuring the normal operation of the system.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention utilizes IoT technology to acquire environmental data, images, and real-time data on the engine and generator inside and outside the generator vehicle. It then uses a generator vehicle knowledge expert database and a comprehensive data risk identification model to comprehensively assess the real-time operation and risks of the generator vehicle. This results in automatic identification and proactive early warning of the generator vehicle's real-time operating status and risks. This assists generator vehicle maintenance personnel in providing auxiliary reference and analysis of the overall status of the generator vehicle, facilitating corresponding measures. For maintenance personnel, it provides auxiliary reference by offering data on the generator vehicle's total power generation, power generation duration, and engine data, enabling better maintenance plans and targeted maintenance. For management personnel, it provides data support for understanding the overall status of the generator vehicle, facilitating better work arrangements. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating a proactive early warning method for risk identification based on generator vehicles. Detailed Implementation

[0015] In this embodiment of the invention, a proactive early warning method based on generator vehicle risk identification includes a front-end module and a back-end platform module. The front-end module includes environmental sensors and vehicle-mounted video installed inside the generator vehicle, and a communication module installed inside the generator set controller. The environmental sensors include temperature and humidity sensors, noise sensors, and smoke sensors. The back-end platform module includes a generator vehicle knowledge expert database, a generator vehicle comprehensive data risk identification data model, a multi-class standard dataset, an IoT gateway, an IoT platform, and a message push center.

[0016] The early warning method includes the following steps: S1. Utilize the professional standards, data, specifications, and environmental impact information related to the generator car and its engine, generator, and environmental impact during operation, as well as the generator car operation and maintenance specifications, standards, and data information, combined with relevant standards, specifications, and other relevant professional data information provided by experts, to establish a generator car knowledge expert database. S2. Based on the professional standards and standardized data information of the generator vehicle knowledge expert database in step S1, and combined with the historical operation failure information of the generator vehicle, determine the data and its related processes, define the data, ensure the integrity of the data, define the operation process, select the data storage technology, and use the generator vehicle environmental data, images, engine, generator and other comprehensive elements to determine the data model algorithm, thereby establishing a comprehensive data risk identification data model for generator vehicles. The method for establishing the comprehensive data risk identification model for generator vehicles is as follows: Different weights k are assigned to the generator vehicle data items based on the different types of data collected. n The probability L(θ) of identifying the generator vehicle risk is calculated, where P(Y) N Given N distinct data items, the probability of an event occurring is: L(θ) = P(Y1)k1P(Y2)k2...p(Y n ) k n P I =1 / (1+E -(α+βX I ) ) When multiple data items affect the operation of the generator car, N parameters and weight values ​​can be passed in. When the value of L(θ) is greater than 1, it indicates that there is a probability that the generator car will encounter a risk. S3. Based on the data model for risk identification of generator vehicle comprehensive data in step S2, import the data model algorithm, and use the generator vehicle knowledge expert database and its historical data to extract the parameter data of the data model algorithm to form a multi-class standard dataset. The data model algorithm is as follows: Let for Loss estimates for different data types, for The probability of occurrence of risks for different data types is used to calculate the risk assessment value by combining the loss estimate and probability of the same data type. :

[0017] The level of risk is determined by the risk assessment value; S4. Using Internet of Things (IoT) technology, environmental sensors and vehicle-mounted video are installed inside the generator vehicle, and a communication module is installed inside the generator set controller. The environmental sensors, vehicle-mounted video, and communication module are connected through communication lines, and communication is achieved using IoT gateways. S5. Use the Internet of Things platform to collect and analyze the data obtained in step S4. Integrate and refine the collected environmental, image, and unit data according to the parameter configuration of the data model algorithm, input it into the data model, compare the data model with the generator vehicle knowledge expert database, calculate the risk identification result, and generate a risk identification report. S6. Based on the risk identification report in step S5, when there is a potential risk to the generator vehicle, the system automatically sends the report to the message push center. The message push center accurately locates the relevant operation and maintenance personnel, management personnel, and on-site support personnel of the generator vehicle. Through technologies such as SMS and mobile terminal push, the system pushes the potential risks of the generator vehicle to the relevant personnel, providing data and reference support for the relevant personnel to handle the risks, thereby realizing proactive early warning of generator vehicle risks.

[0018] Preferably, in step S1: the data information related to the generator vehicle knowledge expert database is managed through information technology, and the data content of the generator vehicle knowledge expert database is classified, organized, improved and processed to provide data support for the active identification model; the data information that has been entered into the database and classified, organized, improved and processed is cleaned by binning, clustering and regression methods. Data cleaning includes checking data consistency and handling invalid and missing values, thereby improving the data quality of the expert database.

[0019] Preferably, in step S3: the sample extraction for the data model algorithm requires data training, and the data training method is as follows: S11. A supervised learning training mode is adopted. By continuously inputting parameter data, the data model is trained to predict the results. The prediction results are compared with the actual results of the training data, and the prediction model is continuously adjusted until the model's prediction results reach an expected accuracy. S12. The training mode adopts an unsupervised learning approach. By importing the parameter dataset, the training data is used to predict the results, and finally the prediction results are analyzed. S13. Train the data model using both supervised and unsupervised methods, employ cross-validation, observe the loss curve and test result curve to analyze the reasons, evaluate the data model, and adjust and optimize the parameters based on the evaluation results. S14. Through the adjusted data model and parameter optimization, continuous machine learning is carried out until it meets the actual standards.

[0020] Preferably, in step S4: based on the MODBUS protocol, configure commands to acquire temperature, humidity, noise, and smoke detection data, as well as commands to acquire unit data; by configuring video parameters, connect the video images to the IoT gateway; and send data commands through the gateway to acquire real-time data from environmental sensors and unit operation.

[0021] Preferably, in step S4: based on the MQTT protocol, configure parameters and services such as connection, subscription, and publication; the IoT gateway connects to the IoT platform; the IoT gateway collects data and publishes it to the IoT platform through the issued commands; the IoT platform receives the data published by the IoT gateway through the subscribed parameter information and parses the data according to different data types based on the communication protocol.

[0022] Preferably, in step S5: using the risk identification results, the risk identification content is converted into a risk identification report using information technology, thereby realizing risk identification during the operation of the generator car.

[0023] Preferably, the generator vehicle knowledge expert database in step S1 consists of a remote data entry and maintenance module, a query module, a classification and management module, a security and confidentiality module, and a system recovery module. The remote data entry and maintenance module allows experts to independently enter and maintain data, enabling timely updates; the query module is used for dynamic querying of generator vehicle data and outputs the queried information and related structures in report form; the classification and management module is used to classify different information types; the security and confidentiality module is used to set different query permissions for different operators to prevent users with insufficient permissions from arbitrarily accessing and modifying data; the system recovery module includes the creation, backup, and recovery of the knowledge expert base, and by regularly backing up data information, it avoids data loss in the event of system failure, thereby ensuring the normal operation of the system.

[0024] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A proactive early warning method for risk identification based on generator vehicles, comprising a front-end module and a back-end platform module, characterized in that, The front-end module includes environmental sensors and vehicle-mounted video installed inside the generator vehicle, and a communication module installed inside the generator set controller. The environmental sensors include temperature and humidity sensors, noise sensors, and smoke sensors. The back-end platform module includes a generator vehicle knowledge expert database, a generator vehicle comprehensive data risk identification data model, a multi-class standard dataset, an IoT gateway, an IoT platform, and a message push center. The early warning method includes the following steps: S1. Utilize the professional standards, data, specifications, and environmental impact information related to the generator car and its engine, generator, and environmental impact during operation, as well as the generator car operation and maintenance specifications, standards, and data information, combined with relevant standards, specifications, and other relevant professional data information provided by experts, to establish a generator car knowledge expert database. By implementing information-based management of data related to the generator vehicle knowledge expert database, and classifying, organizing, improving, and processing the data content of the database, data support is provided for the active identification model. The data that has been entered into the database and classified, organized, improved, and processed is then cleaned using binning, clustering, and regression methods. Data cleaning includes checking data consistency and handling invalid and missing values, thereby improving the quality of the expert database data. S2. Based on the professional standards and standardized data information of the generator vehicle knowledge expert database in step S1, and combined with the historical operation failure information of the generator vehicle, determine the data and its related processes, define the data, ensure the integrity of the data, define the operation process, select the data storage technology, and use the comprehensive elements of generator vehicle environmental data, images, engine, and generator to determine the data model algorithm, thereby establishing a comprehensive data risk identification data model for generator vehicles. The method for establishing the comprehensive data risk identification model for generator vehicles is as follows: Different weights k are assigned to the generator vehicle data items based on the different types of data collected. n The probability L(θ) of identifying the generator vehicle risk is calculated, where P(Y) N Given N distinct data items, the probability of an event occurring is: L(θ) = P(Y1)k1P(Y2)k2...p(Y n ) k n P I =1 / (1+E -(α+βX I ) ) When multiple data items affect the operation of the generator car, N parameters and weight values ​​can be passed in. When the value of L(θ) is greater than 1, it indicates that there is a probability that the generator car will encounter a risk. S3. Based on the data model for risk identification of generator vehicle comprehensive data in step S2, import the data model algorithm, and use the generator vehicle knowledge expert database and its historical data to extract the parameter data of the data model algorithm to form a multi-class standard dataset. The data model algorithm is as follows: Let for Loss estimates for different data types, for The probability of occurrence of risks for different data types is used to calculate the risk assessment value by combining the loss estimate and probability of the same data type. : The level of risk is determined by the risk assessment value; S4. Using Internet of Things (IoT) technology, environmental sensors and vehicle-mounted video are installed inside the generator vehicle, and a communication module is installed inside the generator set controller. The environmental sensors, vehicle-mounted video, and communication module are connected through communication lines, and communication is achieved using IoT gateways. S5. Use the Internet of Things platform to collect and analyze the data obtained in step S4. Integrate and refine the collected environmental, image, and unit data according to the parameter configuration of the data model algorithm, input it into the data model, compare the data model with the generator vehicle knowledge expert database, calculate the risk identification result, and generate a risk identification report. S6. Based on the risk identification report in step S5, when there is a potential risk to the generator vehicle, the system automatically sends the report to the message push center. The message push center accurately locates the relevant operation and maintenance personnel, management personnel, and on-site support personnel of the generator vehicle, and pushes the potential risks of the generator vehicle to the relevant personnel through SMS and mobile terminal push technology, providing data and reference support for the relevant personnel to handle risks, thereby realizing proactive early warning of generator vehicle risks.

2. The proactive early warning method based on generator vehicle risk identification according to claim 1, characterized in that, In step S3: the sample extraction for the data model algorithm requires data training, and the data training method is as follows: S11. A supervised learning training mode is adopted. By continuously inputting parameter data, the data model is trained to predict the results. The prediction results are compared with the actual results of the training data, and the prediction model is continuously adjusted until the model's prediction results reach an expected accuracy. S12. The training mode adopts an unsupervised learning approach. By importing the parameter dataset, the training data is used to predict the results, and finally the prediction results are analyzed. S13. Train the model using supervised and unsupervised methods, employ cross-validation, observe the loss curve, analyze the reasons for the test result curve, evaluate the data model, and adjust and optimize the parameters based on the evaluation results. S14. Through the adjusted data model and parameter optimization, continuous machine learning is carried out until it meets the actual standards.

3. The proactive early warning method based on generator vehicle risk identification according to claim 1, characterized in that, In step S4: Based on the MODBUS protocol, configure commands to acquire temperature, humidity, noise, and smoke detection data, as well as commands to acquire unit data. By configuring video parameters, connect the video images to the IoT gateway, and send data commands through the gateway to acquire real-time data from environmental sensors and unit operation.

4. The proactive early warning method based on generator vehicle risk identification according to claim 1, characterized in that, In step S4: Based on the MQTT protocol, the parameters and services for connection, subscription, and publication are configured. The IoT gateway connects to the IoT platform. The IoT gateway collects data through the issued commands and publishes it to the IoT platform. The IoT platform receives the data published by the IoT gateway through the subscribed parameter information and parses the data according to different data types based on the communication protocol.

5. The proactive early warning method based on generator vehicle risk identification according to claim 1, characterized in that, In step S5: using the risk identification results, the risk identification content is converted into a risk identification report using information technology, thereby realizing risk identification during the operation of the generator car.

6. The proactive early warning method based on generator vehicle risk identification according to claim 1, characterized in that, The generator vehicle knowledge expert database in step S1 consists of a remote data entry and maintenance module, a query module, a classification and management module, a security and confidentiality module, and a system recovery module. The remote data entry and maintenance module allows experts to independently enter and maintain data, enabling timely updates. The query module is used for dynamic querying of generator vehicle data and outputs the queried information and related structures in report form. The classification and management module is used to categorize different information types. The security and confidentiality module is used to set different query permissions for different operators to prevent users with insufficient permissions from arbitrarily accessing and modifying data. The system recovery module includes the creation, backup, and recovery of the knowledge expert database. By regularly backing up data information, it prevents data loss in the event of system failure, thereby ensuring the normal operation of the system.

Images