CN-122023085-A - Geographic information management system and management method for emergency treatment

CN122023085ACN 122023085 ACN122023085 ACN 122023085ACN-122023085-A

Abstract

The invention relates to the technical field of geographic information management, in particular to a geographic information management system for emergency treatment, which comprises a device end and a cloud platform; the invention processes and analyzes the data by collecting the real-time data, thereby monitoring the fluctuation condition of the index in real time, judging whether the current situation is a critical event by combining with the built-in threshold alarm algorithm rule, and automatically sending out emergency alarm to the relevant units in the region when the critical event such as leakage, fire and the like occurs, thereby shortening the response time of the relevant units, reducing the life and property loss of personnel and daily updating the information of the emergency management department in a certain range of the management area, thereby ensuring that the emergency alarm information can be sent accurately.

Inventors

LI PEI
XU ZIQIANG
REN XIAORONG
WANG JIAXIN
SONG TAO
ZHANG LIANG
HUANG TIANHU
REN YANBING
ZHANG JINSUO
ZHAO XIAOCHUN
WANG ZHONGHAI
XU WENLONG

Assignees

中国石油天然气股份有限公司

Dates

Publication Date: 20260512
Application Date: 20241112

Claims (9)

1. The system is characterized by comprising an equipment end and a cloud platform; the equipment end is provided with a display screen and is used for displaying an operation interface to a user, acquiring data through a sensor arranged on the site and transmitting the acquired data to the cloud platform in a 5G/4G, WIFI, VPN communication mode; the cloud platform stores and analyzes the data acquired by the equipment end and comprises a data processing and analyzing module, a data storage management module and an emergency alarm module; The data processing and analyzing module processes and processes the collected information and feeds back the result to the emergency alarm module, and the data processing and analyzing comprises the following steps: (1) Data cleaning ‌ is carried out, original data is processed, and noise, errors and redundant information in the original data are removed; (2) Data normalization processing, which converts data of different scales and units into the same standard form; (3) ‌ discretization and serialization processing of data Discretizing converts continuous data into discrete data for classification and grouping analysis; Serializing to convert the discrete data into continuous data for regression and predictive analysis; (4) ‌ data analysis, which is to count and analyze the cleaned data, extract valuable information and conclusions, and the data analysis method comprises descriptive statistics, inference statistics, data mining and visual analysis; (5) Feature engineering, including feature selection, feature construction, feature transformation, and feature encoding; (6) Model construction, wherein the model comprises linear regression, logistic regression, decision tree, random forest, support vector machine and neural network, and the linear regression model is expressed as Wherein Is a dependent variable, Is an independent variable matrix, Is a regression coefficient vector, Is an error term, and the formula of the logistic regression model is , Representing the probability of outputting y=1 given an input x, w being a feature weight vector, b being a bias term, e being the base of the natural logarithm, The formula of the decision tree model is that the data set D= is set up ),( ),...( ) Wherein Is an input feature vector, Is the output category, and the random forest model formula is Wherein y represents a target variable, x represents an independent variable, Representing the output structure of each tree, M representing the total number of trees in the forest, and the model formula of the support vector machine is X is an input feature vector, Is a normal vector of the hyperplane, b is a bias term, Representing variable, the formula of the neural network model is Wherein , Representing the input of the neuron(s), , Representing the weight corresponding to each input, b being a bias term, Representing an activation function, y representing the output of the neuron, training using a training data set and validating and adjusting model parameters using a validation data set when constructing the model; The data storage management module stores the processed and analyzed data into a database, and comprises a storage unit and a management unit, wherein the storage unit stores the data in a structured mode, and the management unit maintains and manages the data in the database; The emergency alarm module comprises an emergency management department information management unit and an alarm unit, wherein the emergency management department information management unit can manually update the emergency management department information of the area where the emergency management department information is located, an early warning threshold value of each index is set in the alarm unit, the fluctuation condition of the index is monitored in real time according to the data processing analysis result, whether the emergency event is a critical event is judged by combining with a built-in threshold alarm algorithm rule, and emergency alarm is sent to relevant units of the area where the emergency event occurs; The threshold value alarm algorithm is as follows, the pre-warning threshold values of the pressure in the pipeline, the liquid temperature in the pipeline, the flow rate in the pipeline, the air temperature, the carbon monoxide gas concentration, the carbon dioxide gas concentration and the methane gas concentration are set, and the pre-warning threshold values are respectively used The representation is made of a combination of a first and a second color, For real-time values, if The result is recorded as 1, if The result is recorded as 0, then the comparison is sequentially carried out, the structure is accumulated, when the result is 0, the normal state is judged, when the result is 1~d, the unstable state is judged, the result is sent to a system manager, the system manager invites the technical staff to judge and overhaul, when the result is > d, the emergency event is judged, and the emergency alarm is sent to the relevant units of the region where the emergency event is located.
2. The geographical information management system for emergency treatment according to claim 1, wherein the display screen is a liquid crystal spliced screen, a ‌ DLP rear projection large screen or a ‌ small-space LED display screen, and the data collected by the sensor comprises pressure in a pipeline, liquid temperature in the pipeline, flow in the pipeline, air temperature, carbon monoxide gas concentration, carbon dioxide gas concentration and methane gas concentration.
3. The system of claim 1, wherein the data cleansing comprises processing missing values, duplicate values, and outliers, wherein the method of processing missing values comprises deleting missing values, filling missing values, interpolating or extrapolating missing values, wherein the method of processing duplicate values comprises deleting duplicate lines, retaining first or last lines, or using custom methods to determine which duplicates to retain or delete, and wherein the method of processing outliers comprises removing, pruning extreme values, and replacing extreme values with specified values that are closer to other data points.
4. The geographical information management system for emergency treatment of claim 1, wherein the data normalization method comprises a min-max normalization, a Z-score normalization and a decimal scaling normalization, the data discretization method comprises an equal width discretization, an equal frequency discretization and a cluster discretization, and the data serialization method comprises an interpolation method, a regression method and a smoothing method.
5. The system of claim 1, wherein the descriptive statistics are used for basic statistical analysis of the data, including counting, summing, average and median, the inferred statistics are used for inferred analysis of the sample data, including hypothesis testing and confidence interval estimation, the data mining uses data mining algorithms and techniques to mine hidden association rules, trends and pattern information, and the visual analysis uses graphical tools or programming languages to visually display the analysis results to help users understand and interpret the data more intuitively.
6. The geographical information management system for emergency treatment according to claim 1, wherein the feature selection is performed by using a filtering method, a packing method or an embedding method, the feature construction can create a ratio, a difference value, a product or a composite feature, the feature conversion method comprises polynomial regression, logarithmic transformation, power transformation and principal component analysis, and the feature encoding method comprises single-heat encoding, tag encoding and ordinal encoding.
7. The geographical information management system for emergency treatment of claim 1, wherein the storage unit uses a disk to store data during the data storage process, and records the modification operation of the data through the log file, and simultaneously adopts the data compression technology to save the storage space, and improves the data storage efficiency through the partition and slicing technology.
8. The geographical information management system for emergency treatment according to claim 1, wherein the content of the management unit comprises data backup and recovery, data security, data integrity and data audit, wherein the data backup and recovery periodically backs up data in the database to a secure storage medium, the data security protects confidentiality, integrity and availability of the data through access control, encryption and audit means, the data integrity ensures consistency and accuracy of the data through constraint conditions, triggers and storage process means, and the data audit records and analyzes database operation logs, monitors the use condition of the database, and discovers and prevents potential security threats.
9. The geographical information management method for emergency treatment, characterized by comprising the geographical information management system for emergency treatment according to any one of claims 1 to 8, comprising the steps of: the method comprises the steps that firstly, a sensor collects data and transmits the collected data to a cloud platform in a 5G/4G, WIFI, VPN communication mode; Step two, data processing analysis (1) Data cleaning ‌ is carried out, original data is processed, and noise, errors and redundant information in the original data are removed; (2) Data normalization processing, which converts data of different scales and units into the same standard form; (3) ‌ discretization and serialization processing of data Discretizing converts continuous data into discrete data for classification and grouping analysis; Serializing to convert the discrete data into continuous data for regression and predictive analysis; (4) ‌ data analysis, which is to count and analyze the cleaned data, extract valuable information and conclusions, and the data analysis method comprises descriptive statistics, inference statistics, data mining and visual analysis; (5) Feature engineering, including feature selection, feature construction, feature transformation, and feature encoding; (6) Model construction, wherein the model comprises linear regression, logistic regression, decision tree, random forest, support vector machine and neural network, and the linear regression model is expressed as Wherein Is a dependent variable, Is an independent variable matrix, Is a regression coefficient vector, Is an error term, and the formula of the logistic regression model is , Representing the probability of outputting y=1 given an input x, w being a feature weight vector, b being a bias term, e being the base of the natural logarithm, The formula of the decision tree model is that the data set D= is set up ),( ),...( ) Wherein Is an input feature vector, Is the output category, and the random forest model formula is Wherein y represents a target variable, x represents an independent variable, Representing the output structure of each tree, M representing the total number of trees in the forest, and the model formula of the support vector machine is X is an input feature vector, Is a normal vector of the hyperplane, b is a bias term, Representing variable, the formula of the neural network model is Wherein , Representing the input of the neuron(s), , Representing the weight corresponding to each input, b being a bias term, Representing an activation function, y representing the output of the neuron, training using a training data set and validating and adjusting model parameters using a validation data set when constructing the model; Step three, the data after processing and analysis are stored in a database; Step four, monitoring fluctuation conditions of indexes in real time according to data processing analysis results, and judging whether the emergency event is a critical event by combining with a built-in threshold alarm algorithm rule, and sending emergency alarm to relevant units in the region when the critical event occurs; The threshold value alarm algorithm is as follows, the pre-warning threshold values of the pressure in the pipeline, the liquid temperature in the pipeline, the flow rate in the pipeline, the air temperature, the carbon monoxide gas concentration, the carbon dioxide gas concentration and the methane gas concentration are set, and the pre-warning threshold values are respectively used The representation is made of a combination of a first and a second color, For real-time values, if The result is recorded as 1, if The result is recorded as 0, then the comparison is sequentially carried out, the structure is accumulated, when the result is 0, the normal state is judged, when the result is 1~d, the unstable state is judged, the result is sent to a system manager, the system manager invites the technical staff to judge and overhaul, when the result is > d, the emergency event is judged, and the emergency alarm is sent to the relevant units of the region where the emergency event is located.

Description

Geographic information management system and management method for emergency treatment Technical Field The invention relates to the technical field of geographic information management, in particular to a geographic information management system and a geographic information management method for emergency treatment. Background Oil pipeline leakage is a common event in the links of oil and gas exploitation, transportation and storage, if not handled in time, serious pollution is caused to the environment, life and property safety of people is threatened, and fire disaster also causes serious threat to the life and property safety of personnel, so that the establishment of a complete emergency treatment scheme is vital, and when critical events such as leakage, fire disaster and the like occur, the emergency treatment scheme can rapidly respond and handle, thereby accident injury and property loss can be reduced to the maximum extent by ‌, and negative influence is reduced by ‌. The geographic information management system ‌ enables a user to obtain information and insight required by decision making through analysis and simulation of data, so that the user is helped to make an intelligent decision, but when a crisis event occurs, the conventional geographic information management system can only send alarm information to the user of the system, and then send emergency alarm to a related unit (department) of the area where the emergency event is located through manpower, so that the emergency event cannot be responded quickly at the first time of occurrence of the crisis event, the optimal opportunity of emergency disposal is delayed, and great harm is brought to life and property safety of personnel. Therefore, a system and a management method for sending out emergency alarm to the relevant units in the region can be designed aiming at the problem that the emergency alarm can not be automatically sent out to the relevant units in the region. Disclosure of Invention In order to solve the problem that emergency alarm can not be automatically sent to relevant units in the area. The technical scheme of the invention is that the geographic information management system for emergency treatment comprises an equipment end and a cloud platform; the equipment end is provided with a display screen and is used for displaying an operation interface to a user, acquiring data through a sensor arranged on the site and transmitting the acquired data to the cloud platform in a 5G/4G, WIFI, VPN communication mode; the cloud platform stores and analyzes the data acquired by the equipment end and comprises a data processing and analyzing module, a data storage management module and an emergency alarm module; The data processing and analyzing module processes and processes the collected information and feeds back the result to the emergency alarm module, and the data processing and analyzing comprises the following steps: (1) Data cleaning ‌ is carried out, original data is processed, and noise, errors and redundant information in the original data are removed; (2) Data normalization processing, which converts data of different scales and units into the same standard form; (3) ‌ discretization and serialization processing of data Discretizing converts continuous data into discrete data for classification and grouping analysis; Serializing to convert the discrete data into continuous data for regression and predictive analysis; (4) ‌ data analysis, which is to count and analyze the cleaned data, extract valuable information and conclusions, and the data analysis method comprises descriptive statistics, inference statistics, data mining and visual analysis; (5) Feature engineering, including feature selection, feature construction, feature transformation, and feature encoding; (6) Model construction, wherein the model comprises linear regression, logistic regression, decision tree, random forest, support vector machine and neural network, and the linear regression model is expressed as WhereinIs a dependent variable,Is an independent variable matrix,Is a regression coefficient vector,Is an error term, and the formula of the logistic regression model is,Representing the probability of outputting y=1 given an input x, w being a feature weight vector, b being a bias term, e being the base of the natural logarithm,The formula of the decision tree model is that the data set D= is set up),(),...() WhereinIs an input feature vector,Is the output category, and the random forest model formula isWherein y represents a target variable, x represents an independent variable,Representing the output structure of each tree, M representing the total number of trees in the forest, and the model formula of the support vector machine isX is an input feature vector,Is a normal vector of the hyperplane, b is a bias term,Representing variable, the formula of the neural network model isWherein,Representing the input of the neuron(s),,Representing the weig