RU2796623C1 - Administration system for risks and elimination of consequences of emergencies - Google Patents

Abstract

FIELD: ensuring natural and technogenic safety of subject territories.

SUBSTANCE: introducing automation tools for predicting the consequences of natural and man-made emergencies into the administration system for risks of occurrence or elimination of consequences of an emergency, assessing natural and man-made risks and issuing recommendations to reduce their level, taking into account distribution of population over the territory and modelled parameters of damage factors, support for making administrative decisions in case of a risk of occurrence or elimination of consequences of an emergency.

EFFECT: expansion of available means.

9 cl, 2 dwg

Description

The field of technology to which the invention belongs .

This invention relates to the field of ensuring natural and technogenic safety of the territories of a subject of the Russian Federation, which includes a subsystem for collecting and processing applications; decision support subsystem; subsystem of complex informing and notification; information storage and processing subsystem; integration geoinformation subsystem.

The following abbreviations are used in this description:

APK - hardware and software complex

IGI - integration geoinformation subsystem

IAP - information and analytical subsystem

CIS - corporate information systems

NNE - violation of normal operating conditions;

R & D - research work

NTI - scientific and technical information

NKI - national classification of inventions

OPTB - ensuring natural and technogenic safety

PAK - software and hardware complex

PAE - operation administration subsystem

PVKD - subsystem of interaction and coordination of actions

PI - Patent Research

PIB - information security subsystem

PID - data integration subsystem

PKD - production and control documentation;

PCIO - a subsystem of integrated information and warning

PKMP - subsystem of complex monitoring and forecasting

PPPR - decision support subsystem

PSOO - subsystem for collecting and processing requests

PTD - production and technological documentation;

PTK - software and hardware complex

PHO - information storage and processing subsystem

PUK - content management subsystem

PC - personal electronic computer

TCH - telecommunication subsystem

Emergency - emergency

EDSPR - expert decision support system

In this description, the following terms are used:

Server (English server) is an electronic device that performs service functions at the request of the client, providing him with access to certain resources. For the purposes of this description, a server is considered that has a permanent connection to an internetwork that can transmit data to the server from client devices. The server can process this data and send the result of processing back to the client device.

The internetwork, as well as all connections between all modules and blocks, includes various topologies, configurations and layouts of interconnection components, made with the ability to interconnect corporate, global and local computer networks, and includes, without limitation, traditional wired, wireless, satellite, optical and equivalent network technologies.

A module is the same as a compute module is a server module, which is a microprocessor specially adapted for signal processing.

Database or database storage module - this is a server module that stores data corresponding to this data module, which can be implemented as a hard disk drive, or as a flash memory (flash memory), which refers to electrically reprogrammable memory semiconductors. However, some modules can be combined in separate implementations. For example, different databases can be stored in the same memory module.

The level of technology.

At present, there are systems for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation. So from the prior art there is a system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation, which includes a subsystem for collecting and processing requests; decision support subsystem; subsystem of complex informing and notification; information storage and processing subsystem; integration geoinformation subsystem (patent for invention No. 2726957, publication date 07/20/2020).

This system is the closest in technical essence and achieved technical result and is chosen as a prototype of the proposed invention as a system.

The disadvantage of this prototype is also the lack of automation of predicting the consequences of natural and man-made emergencies; assessment of natural and technogenic risks and issuance of recommendations to reduce their level, taking into account the distribution of the population over the territory and the modeled parameters of damaging factors; support for management decision-making in the event of a threat of occurrence or elimination of the consequences of an emergency.

Disclosure of the invention.

The present invention mainly aims to propose a system for ensuring the natural and technogenic safety of the territories of the subject of the Russian Federation of the wire, which makes it possible to at least mitigate at least one of the above disadvantages, namely the possibility of automating the prediction of the consequences of natural and technogenic emergencies; assessment of natural and technogenic risks and issuance of recommendations to reduce their level, taking into account the distribution of the population over the territory and the modeled parameters of damaging factors; support for managerial decision-making in the event of a threat of occurrence or elimination of the consequences of an emergency, which is the task.

To achieve this goal, the system also includes

• a subsystem for interaction and coordination of actions to eliminate the CES;

• a subsystem for integrated monitoring and forecasting of the CES;

• data integration subsystem providing automated data exchange between subsystems;

• content management subsystem to provide an ordered register of information objects of the KiChS;

• information-analytical subsystem for displaying, exchanging information on CES and generating reporting forms based on information stored in the system database;

• telecommunications subsystem that provides data transfer and interaction between subsystems,

• operation administration subsystem for performing the tasks of monitoring the technical condition and functioning of subsystems;

• an information security subsystem to protect information and means of its processing of the system,

• moreover, the decision support subsystem in the event of a threat and elimination of the consequences of emergencies includes a component for assessing the probability of occurrence of forest fires, based on the principle of artificial intelligence.

Thanks to these advantageous characteristics, it becomes possible to automate the forecasting of the consequences of natural and man-made emergencies; assessment of natural and technogenic risks and issuance of recommendations to reduce their level, taking into account the distribution of the population over the territory and the modeled parameters of damaging factors; support for management decision-making in the event of a threat of occurrence or elimination of the consequences of an emergency. At the same time, the characteristics have the possibility of automating the forecasting of the probability of occurrence of forest fires based on the principle of artificial intelligence using a neural network.

Risk analysis is performed automatically:

• identification of hazard sources and determination of their characteristics;

• identification of possible scenarios for the initiation of sources of danger;

• modeling the consequences of the implementation of all possible scenarios for the initiation of sources of danger;

• determination of the probability of each of the considered scenarios.

There is an advantageous version of this system, in which the subsystem of integrated monitoring and forecasting of the CES includes:

• intelligent video surveillance module for smoke and fires;

• module for monitoring housing and communal services systems;

• fire situation monitoring module;

• module for monitoring the state of the environment to control the state of current threats of a natural nature;

• Forecasting component as part of the following modules:

• module of fires and explosions at potentially dangerous objects;

• module of accidents with the release of toxic substances at chemically hazardous facilities;

• module of floods caused by high water;

• module of forest and steppe fires.

Thanks to these advantageous characteristics, it becomes possible to comprehensively monitor and predict the fire situation, the state of the environment, fires and explosions at potentially hazardous facilities, accidents with the release of toxic substances at chemically hazardous facilities, floods caused by floods, forest and steppe fires.

There is another version of this system, in which the subsystem for decision support in the event of a threat and elimination of the consequences of an emergency includes:

• component of assessment and management of natural and technogenic risks;

• component of contingency plans;

• component of the calculation of forces and means for rescue and recovery operations;

• component of management documents preparation

Thanks to these advantageous characteristics, it becomes possible to automate the assessment and management of natural and man-made risks, situational plans, calculation of forces and means, and preparation of management documents.

There is also a version of this system, in which the subsystem of interaction and coordination of actions to eliminate the CES includes:

• information exchange component;

• execution control component.

Thanks to these advantageous characteristics, it becomes possible to automate the exchange of information and control of execution.

There is another version of this system, in which the data integration subsystem, which provides automated data exchange between subsystems, includes:

• component of interaction protocols for standardization of formats, rules and regulations of interaction between subsystems;

• data collection component;

• data transfer component;

• component of classifiers and directories.

Thanks to these advantageous characteristics, it becomes possible to automate interaction protocols, collect and transmit data, as well as create classifiers and directories.

There is also a version of this system, in which the content management subsystem for providing an ordered register of information objects of the KiChS includes:

• information object management component (IOC);

• Internet portal content management component (CSI).

Thanks to these advantageous characteristics, it becomes possible to automate the management of information objects, content management of the Internet portal.

There is another version of this system, in which the geoinformation subsystem for converting vector formats of geospatial data into raster formats and ensuring prompt display of information based on electronic maps of territories includes:

• component of transformation of cartographic data;

• a component for displaying the location of forces and means of the unified state system for the prevention and liquidation of emergency situations (RSChS);

• component for displaying cartographic information;

• component for editing passports of territories.

Thanks to these advantageous characteristics, it becomes possible to automate the transformation of cartographic data, the display of forces and means, as well as the display of cartographic information in general, as well as editing passports of territories.

There is also a version of this system, in which the telecommunications subsystem that provides data transfer and interaction between subsystems includes:

• computer facilities;

• means of communication technology;

• means of organizational technology;

• hardware and software of the information storage and processing subsystem.

Thanks to these advantageous characteristics, it becomes possible to implement the telecommunications subsystem in the form of computing, communication and organizational equipment.

Finally, there is another version of this system, in which the information security subsystem for protecting information and means of its processing of the system includes:

• access control component;

• component of registration and accounting of user actions in the system;

• a component for ensuring secure interconnection;

• component of cryptographic protection of information;

• component of anti-virus protection;

• user identification component.

Thanks to these advantageous characteristics, it becomes possible to automate access control, register and record user actions in the system, ensure secure internetworking, cryptographic information protection, anti-virus protection, and user identification.

The set of essential features of the proposed invention is unknown from the prior art for methods of similar purpose, which allows us to conclude that the criterion "novelty" for the invention in relation to the method. In addition, this solution is not obvious to a person skilled in the art.

Brief description of the drawings.

Other features and advantages of the present invention will clearly appear from the description below, by way of illustration and without being restrictive, with reference to the accompanying drawings, in which:

- figure 1 depicts the first part of the functional diagram of the system for ensuring natural and technogenic safety of the territories of the subject of the Russian Federation according to the invention,

- figure 2 depicts the second part of the functional diagram of the system for ensuring natural and technogenic safety of the territories of the subject of the Russian Federation, according to the invention.

Figures 1, 2 show the functional structure of the system, including such elements as:

• Subsystem for collection and processing of appeals (PSOO);

• Subsystem of complex monitoring and forecasting (PKMP);

• Decision Support Subsystem (DSS);

• Subsystem of interaction and coordination of actions (PVKD);

• Subsystem of complex informing and warning (PCIO);

• Subsystem of information storage and processing (PHO);

• Data integration subsystem (PID);

• Content management subsystem (CMS);

• Integration geoinformation subsystem (IGS);

• Information-analytical subsystem (IAP);

• Telecommunication subsystem (TCH);

• Operation Administration Subsystem (PAE);

Subsystems for the collection and processing of requests (PSOO) provide for the reception, registration and classification of messages about crisis and emergency situations through the Internet portal, the subsystem for complex monitoring and forecasting, external information systems, and the user interface of the system.

PSOO provides the following functions:

• receiving messages through the Internet portal through integration with the community interaction portal;

• reception of messages coming from the subsystem of complex monitoring and forecasting;

• ensuring two-way interaction (through the Data Integration Subsystem) in terms of receiving and transmitting basic information about incidents with the existing “System for Calling Emergency Operational Services to a single number “112”;

• formation of an information card of the event and provision of opportunities for filling it out to the operator in the process of additional polling of the subscriber through integration with the Internet portal of interaction with the population;

• determination of the structure of RSChS day-to-day management bodies informed about the event, depending on the type of incident, with the possibility of adjusting this list by the operator;

• ensuring the exchange of textual and graphic information of the services involved in the response through a registered information card related to a specific CES;

• ensuring interaction with the decision support subsystem (through the data integration subsystem) in order to provide information about registered requests;

• ensuring interaction with the integration geoinformation subsystem in order to determine the geolocation of the scene;

• provision of analytical and statistical reports on received reports of CES.

The Integrated Monitoring and Forecasting Subsystem (PKMP) is designed to provide monitoring of data obtained from interacting automated (automatic) monitoring systems on the territory of a constituent entity of the Russian Federation (on the example of the Samara Region) in order to ensure forecasting, monitoring of threats of a natural, technogenic and biological and social nature on territory of the Samara region, including using modern “end-to-end technologies” (“Internet of things”, “big data”, “artificial intelligence”).

PCMP includes:

• a monitoring component as part of the following modules in case of providing organizational and technical capabilities for interfacing with external systems:

• intelligent video surveillance module;

• module for monitoring housing and communal services systems;

• fire situation monitoring module;

• module for monitoring the state of the environment;

• Forecasting component as part of the following modules:

• module of fires and explosions at potentially dangerous objects;

• module of accidents with the release of toxic substances at chemically hazardous facilities;

• module of floods caused by high water;

• module of forest and steppe fires.

The monitoring component provides interaction with information systems that control the operation of sensors installed on stationary and mobile monitoring objects located on the territory of municipalities of a constituent entity of the Russian Federation (in this project, on the example of the territory of the Samara region) by loading the accumulated data of continuous monitoring of parameters, the values of which are obtained in synchronous or asynchronous mode from external/adjacent systems.

The forecasting component provides the ability to quickly predict the consequences of emergencies and support decision-making and includes software modules for forecasting the situation and supporting decision-making to protect the population and territories in the event of:

• fires and explosions at potentially hazardous facilities;

• accidents with the release of toxic substances at chemically hazardous facilities;

• floods caused by high water;

• forest and steppe fires.

The data obtained from the components and modules of the PKMP are available to other subsystems of the system for ensuring natural and technogenic safety of the territories of the constituent entity of the Russian Federation.

It is possible to download historical data.

Information messages indicating a high probability of an emergency occurring are sent to the subsystem for collecting and processing requests.

The subsystems for collecting and processing requests, decision support and the integration geographic subsystem receive data from the components and modules of the PKMP to register the received information about the incident in an automated mode. The forecasting module receives data for building a forecast after processing emergency messages.

The intelligent video surveillance module provides integration with existing intelligent video surveillance systems of a constituent entity of the Russian Federation or a municipality and provides the ability to generate calls in the System in an automated mode when receiving information from intelligent video surveillance systems about the following types of incidents:

• threat of forest/steppe fires (smoke);

• occurrence of a forest/steppe fire (seed of ignition).

The module provides the ability to view video from a camera that has recorded the threat of occurrence or the occurrence of CES.

The housing and communal services monitoring module is designed to monitor and control the state of the existing infrastructure by integrating, if organizational and technical capabilities are provided, with the existing information monitoring systems of a subject of the Russian Federation or a municipality.

The module for monitoring housing and communal services systems allows you to generate requests to the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation when receiving from external integrated systems information about incidents at the following facilities:

• gas mains and pipelines laid on the territory of the municipality;

• gas supply to stationary monitoring objects and places of residence of residents of the municipality;

• water supply and sewerage facilities of the municipality;

• electrification of stationary monitoring objects and places of residence of the population of the municipal district.

The fire situation monitoring module provides information interaction in case of providing organizational and technical capabilities with existing fire situation information systems operating on the territory of a constituent entity of the Russian Federation or a municipality and allows you to automatically generate an event card when receiving information about it from external integrated systems.

The module for monitoring the state of the environment provides control over the state of current threats of a natural nature to the municipality by obtaining, if organizational and technical capabilities are provided, data from existing information systems for monitoring the environment installed on the territory of a constituent entity of the Russian Federation or a municipal district on the following events:

• excess content of harmful substances in the air;

• exceeding the safe water level in reservoirs (flood situation) on the territory of the municipality.

The module allows you to calculate the areas of action of damaging fire factors by simulating the following scenarios for the combustion of flammable liquids:

• combustion inside the tank;

• burning inside the embankment;

• burning at free spill;

• combustion in a given area;

• flare burning;

• fire ball;

• fire-flash.

The module also allows you to calculate the impact zones of damaging factors in the following types of explosions:

• explosions of gas-vapor-air mixtures;

• explosions of solid explosives;

• explosions of dusty air mixtures;

• explosions of tanks with superheated liquid;

• explosions of containers with compressed gases;

• explosions of containers with liquefied combustible gas;

• explosions of containers with liquefied non-combustible gas.

When calculating the damage to people in a fire and explosion, the module allows you to take into account the number of people living or working in residential and industrial buildings located in the fire and explosion zone, as well as the density of people in open areas and adjacent territories.

When calculating losses, the module takes into account crowded places and the number of people in them.

The user is given the opportunity to:

• changes in the number of people in buildings;

• adjusting the value of the density of people in open areas;

• Drawing places of mass congestion of people.

After the calculation is completed, the module issues a report in the following format:

• compatible with Microsoft Word, LibreOffice, Apache OpenOffice indicating the radii of damage to people, the radii of the ignition zones of objects, the radii of destruction of buildings and their elements and materials, as well as the number of dead and injured people in buildings and in open areas. A list of all buildings that fell into the fire and explosion zone should also be displayed, indicating the number of affected people;

• objects of the layer of the integration geographic subsystem indicating the radii of damage to people, the radii of the ignition zones of objects, the radii of destruction of buildings and their elements and materials.

The module of accidents with the release of toxic substances at chemically hazardous facilities allows you to simulate the spread of a cloud of toxic substances, taking into account wind parameters and calculate the consequences of an accident.

When calculating population losses in case of accidents at chemically hazardous facilities, the module takes into account the number of people living or working in residential and industrial buildings located in the contaminated zone, as well as the density of people in open areas. The user is given the opportunity to change the number of people in buildings and adjust the value of the density of people in open areas.

The user is also given the opportunity to apply crowded places. When calculating losses, the module takes into account crowded places and the number of people in them.

After the calculation is completed, the module generates a report in the following formats:

• compatible with Microsoft Word, LibreOffice, Apache OpenOffice indicating the maximum values of the zones of destruction of people, as well as the number of dead and injured people in buildings and in open areas;

• objects of the layer of the integration geographic subsystem with indication of the maximum values of the zones of destruction of people, as well as the number of dead and injured people in buildings and in open areas.

The module allows you to display a list of all buildings that have fallen into the infection zone, indicating the number of dead and injured in each building.

The Flood Caused by Flood Module provides for the calculation using the initial data obtained both from the integrated monitoring and forecasting subsystem, and through manual input of the initial data by the System user:

• coordinates of the river section;

• level of water rise in the river.

The program module generates the results of the calculation of the consequences of floods:

• total area of flooding broken down by depth of flooding;

• the total number of people in the flood zone;

• list of urban-type settlements in the flood zone;

• list of rural settlements in the flood zone;

• forces and means to eliminate the consequences of the flood.

• graphical display on the map.

By means of the module, a graphical representation on the map of the flood zone is formed with the ability to determine the depth of flooding at each point.

The program module implements the function of calculating the consequences of flooding both for the entire zone of the river, and for a separately selected area.

After the calculation is completed, the module generates a report in the following formats:

• compatible with Microsoft Word, LibreOffice, Apache OpenOffice indicating flood zones and the number of affected people.

• objects of the layer of the integration geographic subsystem with indication of flood zones and flood depth.

The module of forest and steppe fires provides calculations using the initial data obtained both from the Integrated Monitoring and Forecasting Subsystem, and through manual input of the initial data by the System user:

• forest fire characteristics:

• type of fire;

• forest fire class;

• the duration of the fire;

• weather fire hazard class;

• weather conditions at the time of the forest fire:

• wind direction;

• wind speed.

When carrying out the calculation, the user of the software module has the ability to plot a forest fire on a map.

The program module generates the results of calculating the consequences of a forest fire:

• total area of the forest fire zone;

• forces and means to eliminate the consequences of a forest fire.

By means of the module, a graphical representation on the map of forest fire zones is formed.

After the calculation is completed, the module generates a report in the following formats:

• compatible with Microsoft Word, LibreOffice, Apache OpenOffice with indication of fire zones;

• objects of the layer of the integration geographic subsystem with indication of fire zones.

The decision support subsystem (DSS) is designed to provide information support for making managerial decisions in the event of a threat and liquidation of the consequences of emergencies, based on situational plans that represent a formalized description of a set of actions to prevent or eliminate emergencies.

The main goals of the subsystem functioning are the formation of situational action plans in emergency situations, the formation of recommendations for making managerial decisions, the analysis and adjustment of plans based on the results of the analysis in order to optimize and enter them into the Base of Situational Plans (BSP), as well as the formation of management documents. When operating the subsystem, data from the passports of the territories, data on the threat of occurrence or the occurrence of emergencies, data from the subsystem of complex monitoring and forecasting, data on incidents received from the subsystem for collecting and processing requests are taken into account. The decision support subsystem includes the following functional components:

• component of assessment and management of natural and technogenic risks;

• component of contingency plans;

• component of calculation of forces and means;

• component of preparation of management documents.

For advance planning of measures to protect the population, taking into account the large-scale nature of the consequences of technogenic and natural emergencies, the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation provides for a component for assessing and managing natural and technogenic risks, which implements the ability to build risk maps of the territories of a constituent entity of the Russian Federation in the event of a threat of occurrence Emergency.

The component of assessment and management of natural and technogenic risks allows solving problems of information and analytical support for the activities of municipalities when organizing emergency response activities.

The natural-technogenic risk assessment and management component provides automation of solving risk assessment tasks and developing risk management decisions in the event of a threat or occurrence of at least:

Technogenic emergencies caused by:

• accidents at fire hazardous facilities;

• accidents at explosive facilities;

• accidents at chemically hazardous facilities;

Natural emergencies caused by:

• floods;

• forest fires.

The component allows you to perform risk analysis both on the availability of information about the actual sources of danger, and on the basis of statistical data.

The component provides forecasting of emergencies (probability of damage) on the territory of a municipal district by analyzing current hazards. To predict the occurrence of emergencies, information about the identification of emergencies is used, which comes from integrated systems for analyzing social networks and open sources of information.

The component of situational plans provides a formalized presentation of data on operational planning and measures to eliminate emergencies, attracting forces and means to eliminate emergencies, providing them with the necessary resources in terms of:

• planning the composition of events;

• planning the grouping of forces and assets;

• planning for attracting resources;

• optimization of existing contingency plans for liquidation of the consequences of emergencies.

The force and equipment calculation component is designed to calculate the number of forces and assets required to perform urgent rescue and recovery operations, including:

• determination of the required number of beds, incl. specialized, indicating specific institutions for the hospitalization of victims;

• determination of the need for the forces and means of the RSChS, engineering equipment necessary to eliminate the consequences of an emergency, indicating the names of organizations from among those not included in the emergency zone;

• determination of the need for equipment for evacuation;

• determination of the required number of sets of food and clothing supplies;

• Updating the results of calculations of forces and means when information about the situation changes.

The force and means calculation component provides operational information to the user in order to make managerial decisions both in the format of information contained in screen forms and in the form of objects of the integration geoinformation subsystem layer.

The management documents preparation component is intended for issuing formalized documents.

The formalized documents include:

• reporting and information output documents to higher management bodies;

• output documents for subordinate and interacting organizations, including statistical reports;

• administrative output documents.

The output documents generated for transfer to higher management bodies, cooperating and subordinate organizations in case of a threat, occurrence, localization and liquidation of the consequences of emergency situations (ES) are:

• message about the threat (forecast) of emergencies;

• report on the fact and the main parameters of the emergency;

• reporting on measures to protect the population and territories, conducting emergency rescue and other urgent work;

• report on the forces and means involved in the liquidation of emergencies;

• Report of the operational duty officer of the EDDS to the TsUKS of the Main Directorate of the Ministry of Emergency Situations of Russia of a constituent entity of the Russian Federation on the situation in the emergency zone;

• daily report of the operational duty officer of the EDDS to the TsUKS of the Main Directorate of the Ministry of Emergency Situations of Russia of the constituent entity of the Russian Federation on the situation on the territory of the constituent entity of the Russian Federation;

• working map of the situation in the emergency zone;

• working map of the situation on the territory of the municipality of the Subject of the Russian Federation;

• consolidated report with statistics of the CES and response to emergencies.

The decision support subsystem performs the functions of generating proposals on the composition of the grouping of forces and means in the event of a threat of an emergency of a man-made and natural nature based on the results of calculations carried out in the components of the subsystem and the results of calculations carried out in the monitoring and forecasting subsystem.

The subsystem of interaction and coordination of actions carries out prompt communication of information to state authorities and management bodies of the RSChS on the territory of municipalities of a constituent entity of the Russian Federation (in this project, on the example of the Samara region) in accordance with the approved response plans with a task statement and control over execution.

The subsystem of interaction and coordination of actions includes the following functional components:

• information exchange component (CI);

• Execution control component (QC).

The information exchange component allows delivering messages to users with text, hypertext links and/or files (of a limited size) with a guarantee of delivery and integrity by ensuring the interaction of state authorities and RSChS management bodies using the toolkit of situational plans.

The execution control component is intended for automated control of the execution of instructions and allows you to assign tasks, notify performers about assigned tasks, track execution progress, notify the author about execution progress, and maintain an archive of completed tasks.

The execution control component allows you to form the order of actions for system users for the situation under consideration. The order of actions is a list of tasks with reference to the time of execution. The order of actions can be formed both by the appearance of an event in the system (for example, receiving monitoring data with a value exceeding the threshold, generating an event card by the operator or receiving it from an external system, receiving values exceeding the specified values as the results of analytical calculations, changing the status of one of the tasks, and etc.). With the help of the execution control component, work with situational plans will be implemented.

It is possible to attach documents to each of the actions. The task as one of the properties has a status that can take one of the following values: not started, in progress, completed. The task status can be assigned by condition or by the user. The task is assigned a run time.

All user actions in the component are logged with reference to time. Tasks that are not started or completed within the set time are marked as overdue, and the task performer and his manager are informed about this. There is an opportunity to distribute the roles of users with the possibility of appointing managers who have access to consolidated information about all tasks and statuses in the group.

The subsystem of interaction and coordination of actions allows you to transfer at least 50 documents per minute to external systems.

The integrated information subsystem provides issuance of recommendations on the launch of information and notification tools, obtaining data on the status of the notification and the performance of the notification devices, visualization of the implementation of the notification through integration interaction with external information systems of information and notification, if technical and organizational capabilities are provided.

The complex informing subsystem provides the following functionality:

• sending recommendations on the start of the informing and notification process to the operator of the external information system;

• receiving data from an external information system on the current status of the informing and notification process;

• visualization of notification execution;

• informing the operators of the system for ensuring natural and technogenic safety of the territories of a subject of the Russian Federation according to a predetermined list of the need to inform;

• informing the population by posting information about the incident on the Internet portal;

• ensuring the storage of information about the results of monitoring the performance of an external warning and information system.

The integrated informing subsystem provides the operator with tools for working with information about public informing and warning systems, mainly at social facilities (schools, kindergartens, etc.), in shopping and entertainment centers, crowded places, at enterprises and organizations with municipal public address system.

The integrated informing subsystem receives from the database of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation information about the locations of the terminal equipment of warning systems and its performance. In the event that the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation receives information about a crisis situation (from an external system or entered manually by the user), an event card is formed, the user has the opportunity to assess the consequences of a possible emergency as a result of which the territory is zoned according to damaging factors, for example, in the area of chemical contamination. The integrated information subsystem, if data is available, allows you to generate a list of warning and information systems that fall into the zone of damaging factors and form a task for the user responsible for notification and information to use the data of the warning and information system to transmit the generated warning message to the population. If the notification and informing systems used have the technical ability to transmit information about the end of the notification process on the end devices, this information can be visualized for the operator responsible for notification and informing. If there is information in the database of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation about the coverage areas of end warning devices and information about the distribution of the population on the territory, the subsystem will generate a report indicating the systems involved, end devices, their response time, coverage area, and the predicted number of alerted population .

The information storage and processing subsystem includes a DBMS, databases and software for organizing the execution environment for the executable modules of the system for ensuring natural and technogenic safety of the territories of a subject of the Russian Federation, performing calculations, storing and transmitting data.

The data in the system is stored in the form of a set of databases containing structured information necessary for the operation of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation.

The subsystem uses both a centralized and a decentralized way of organizing data storage. The sources of stored data can be both elements of the system for ensuring the natural and technogenic safety of the territories of a constituent entity of the Russian Federation, and external systems that transmit data through the data integration subsystem.

The structure of the databases allows the storage of information objects of all formats that are used by the subsystems of the system for ensuring the natural and technogenic safety of the territories of the constituent entity of the Russian Federation.

The data storage subsystem allows increasing both the number of storage nodes and the memory capacity of individual nodes.

The data integration subsystem is an integral part of the system for ensuring the natural and technogenic safety of the territories of a constituent entity of the Russian Federation, which provides automated and automatic exchange of heterogeneous data between existing systems.

The data integration subsystem includes the following functional components:

• component of interaction protocols;

• data collection component;

• data transfer component;

• component of classifiers and directories.

The interaction protocols component provides standardization of formats, rules and regulations for interaction between all participants in the information exchange within the framework of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation and allows you to manage interaction protocols with specific automated systems for automated data exchange. The interaction protocols include both formats and methods of message transmission, as well as data on authorization and addressing. The interaction protocols component converts the received data about objects into the standard format of the system for ensuring natural and technogenic safety of the territories of the constituent entity of the Russian Federation and provides information exchange with other subsystems in a standardized format.

The format of data from external systems and the logic of their conversion are configured using the parameters of the configuration files.

The component of interaction protocols determines the rules for access to data for all participants in information interaction within the framework of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation.

The data collection component, based on information from the interaction protocols component, accesses external information systems at the specified address using pre-configured communication channels (including secure ones) and authentication data.

The data collection component allows you to interact with external systems and telemetry sensors in synchronous or asynchronous modes, by polling external monitoring systems or receiving data sent by external systems and devices (depending on the technical capabilities of external systems). For integration with telemetry end devices, the LwM2M/MQTT/TCP protocols are implemented in the part necessary for integration with specific end devices.

The component of classifiers and directories provides for obtaining a list of directories, loading selected directories, unloading selected directories, transferring this data to other systems.

The directories and classifiers component provides the user with the following directories:

• units of measurements;

• automotive technology;

• types of ASF;

• types of precipitation;

• types of governing bodies;

• types of crisis situations;

• classification of emergencies;

• types of emergency levels.

The data integration subsystem implements receiving from other systems and visualization:

data of passports of potentially dangerous and socially significant objects;

data on the situation on the territory of the municipality (municipalities) and the subject of the Russian Federation (as a whole);

data on decisions and measures taken to prevent, mitigate the consequences and eliminate emergencies on the territory of the municipality (municipalities) and the constituent entity of the Russian Federation (in general).

As part of the creation of the data integration subsystem, the following are integrated:

1. External information systems:

• AIMS RSChS-2030;

• APK "Safe City" (ESOR);

• AS EDDS;

• AS of the management bodies of the RSChS;

• Multi-level navigation and information system for monitoring vehicles of the Ministry of Emergency Situations of Russia.

2. External information systems belonging to the following classes:

• fire alarm;

• monitoring of forest/steppe fires;

• geoinformation systems;

• hydrological monitoring;

• satellite monitoring of forest fires;

• Internet portal for interaction with the population;

• monitoring the content of hazardous substances in the air;

• video surveillance (including intelligent video surveillance);

• monitoring of potentially dangerous objects;

• monitoring of moving objects;

• monitoring of the housing and communal services infrastructure;

• systems of coordination and interaction of forces and means of the RSChS;

• receiving and processing messages about the CES;

• informing and alerting the population of the municipal district, as well as public authorities and management bodies of the RSChS on the territory of the constituent entity of the Russian Federation;

• analysis of social networks;

• analysis of open sources of information posted on the Internet.

The data integration subsystem allows processing and storing in the database at least 200 events received from external systems per minute, no more than 1 kilobyte each, at least 1500 telemetry data packets received from external systems per minute, no more than 1 kilobyte each.

The content management subsystem includes:

• information object management component (IOC);

• Internet portal content management component (CSI).

The information object management component maintains an ordered register of all information objects of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation, supports loading information model profile extensions (that is, adding new classes of objects, attributes, etc.), has interfaces for exchanging an information model with third party systems in XML format.

The component provides the following functions:

• displaying a list of types of information objects;

• request and display of attributive information about a given type of information objects, including information about their total number in the database;

• query and output of lists of information objects of a given type and given attribute information, with the possibility of limiting their number (for example, from 100 to 200);

• creation, editing, deletion of objects and types of objects;

• generation of notifications when changes are made to objects;

• creation of hierarchies of object types with inheritance of property definitions from upper to lower levels;

• applicability of the same property to objects of different types;

• storage of information about specific information objects, with the possibility of assigning multiple values to each property of each specific object, assigning each object to several types at the same time.

Each information object of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation is characterized by a set of attributes of various types of data.

The component provides an administrative user interface that allows you to create and modify information object types.

The component provides a unified access interface both to information objects contained in a single data storage and to external information objects stored in third-party information systems, automatically accessing the information interaction subsystem if necessary.

The content management component of the Internet portal provides the following functions:

• display information on the map using the tools of the Integration Geoinformation Subsystem of the System for Ensuring Natural and Technogenic Safety of the Territories of a Subject of the Russian Federation;

• filter events displayed as map-based markers and lists;

• configure the parameters of subscription to system events;

• customize the list of statistical data displayed by the list on system events;

• configure the list (of the available ones) and display parameters of information layers of cartographic information;

• customization of types, displayed data and list of their characteristics.

The integration geoinformation subsystem is designed to convert geospatial data formats and ensure prompt display of information based on electronic maps. The subsystem provides a mechanism for regular updating of electronic maps of the subsystem to ensure the relevance of cartographic information and supports the mif/mid data exchange formats.

The subsystem user interface provides the following functionality:

• attributive search on the map of objects of classified types;

• indication and clarification of the location of objects associated with the incident, both with the help of visual graphic means, and with the help of direct input of coordinates;

• displaying information about incidents in the form of appropriate markers with the ability to provide prompt access to information about the incident;

• displaying the results of the execution of algorithms of the subsystem of complex monitoring and forecasting and the subsystem of decision support.

The integration geoinformation subsystem includes the following functional components:

• component of transformation of cartographic data;

• display component of forces and means;

• component for displaying cartographic information;

• component for editing passports of territories.

The cartographic data conversion component provides the ability to convert a vector cartographic base into a raster one for its display in the Web interface of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation.

The force and means display component is designed to receive and display data on the location of the RSChS forces and means using GLONASS technology.

During the operation of the component, information about the location of the forces and means of the RSChS is provided only to authorized users in accordance with their access rights. User access rights can be configured separately for each type of departmental affiliation of forces and assets.

The component for displaying cartographic information supports the display of freely distributed cartographic substrates (YandekMaps, OpenStreetMap, GoogleMaps). Maps are displayed in the WGS-84 coordinate system.

The component allows you to get semantic information about geometric objects displayed on top of the cartographic base in custom layers.

The component provides the ability to create custom layers to display one or more types of system information objects. Each layer has the following settings:

• name and description of the layer;

• layer type (object layer, region coloring layer);

• initial data for the layer (types of information objects with possible filtering);

• enable pop-up hints for layer objects;

• enabling layer tiles caching;

• combining layers into an integral layer;

• clustering of point objects of the layer;

• limiting the display scale of the layer;

• layer styles, separately for line (line thickness, color and transparency), polygons (hatch line thickness, hatch size and style, fill color and transparency), point (symbol or point type, size, color and transparency) and text ( font color and size, text offset relative to the symbol, text pattern).

The cartographic information display component includes a list of layers:

• map of the municipal district (in the form of a cartographic substrate);

• locations of the administrative bodies of the territorial subsystem of the RSChS;

• areas of responsibility of EDDS and DDS;

• location of healthcare facilities;

• operational situation in emergency situations;

• locations of permanent readiness forces and their areas of responsibility;

• locations of helipads;

• display of population density in the zone of occurrence of emergencies;

• locations of potentially dangerous objects and their territories;

• location of warehouses for emergency reserves;

• location of hydraulic structures;

• mobile facilities equipped with GLONASS terminals (with the possibility of separating: passenger vehicles, vehicles for transporting schoolchildren, TS 03, housing and communal services TS, OIV TS, TS of forces and means of the RSChS);

• public transport routes;

• locations of the system's CCTV cameras (including location information);

• locations of meteorological and environmental monitoring sensors;

• locations of gauging stations;

• locations of facilities equipped with fire alarm systems;

• forecasting layers: the threat of flooding and flooding, fire, AHOV;

• spatial layer of relief level lines for the area of interest;

• spatial layer of mobile communication coverage;

• spatial polygonal layer of the territories of the borders of settlements;

• Spatial layer of development of settlements;

• Spatial layer of forest quarters of the area of interest;

• spatial layer of roads of the area of interest and streets of settlements;

• spatial linear layer of stream axes;

• spatial polygonal layer of rivers;

• socially significant objects, including those with round-the-clock stay of people;

• objects with mass stay of people;

• display of information about traffic congestion (if necessary data are available).

The component for editing territory passports provides the user of the system for ensuring natural and technogenic safety of territories of a constituent entity of the Russian Federation with the functions of editing the structure and data of territory passports.

The integration geoinformation subsystem allows displaying up to 20 cached layers in the user's browser in 15 seconds (with a fully generated cache). Refreshing data in non-cached layers (containing no more than 20 thousand objects) occurs no later than 15 seconds (if there are no other updates).

The information and analytical subsystem provides information coverage of the operational situation in the region, provides an opportunity for interaction between officials of regional and municipal authorities, state organizations and the population on issues of ensuring public safety, law and order and environmental safety through the Internet portal.

The subsystem of information and analytical support includes an Internet portal designed to display and exchange information about CES on the territory of pilot municipalities (location, status, involved forces and means) to participants in information exchange, including territorial executive authorities, subordinate institutions, EDDS, DDS, municipal services and population.

The information-analytical subsystem provides the user with the functionality to generate reporting forms based on information stored in the database of the system for ensuring natural and technogenic safety of the territories of the constituent entity of the Russian Federation. In reporting forms, textual, tabular, graphical representation of data can be used. Information-analytical allows you to generate summary data on the information stored in the database.

The telecommunications subsystem includes:

• computer facilities;

• means of communication technology;

• means of organizational technology;

• hardware and software of the information storage and processing subsystem;

• general system software.

Computer facilities provide the implementation of complex technologies for processing and storing information and be the basis for the integration of technical means to ensure the management of information resources.

Communication technology ensures the implementation of data transmission technologies and involves functioning in conjunction with computer technology.

Organizational technology provides the implementation of technologies for storing, presenting and using information, as well as the performance of various auxiliary operations within the framework of certain technologies for information support of management activities.

General system software is compatible with special software developed within the framework of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation.

The operation process administration subsystem is designed to perform the tasks of monitoring the technical condition and functioning of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation, as well as setting, changing and monitoring the main parameters of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation during operation.

The administration subsystem includes the following components:

• a component of centralized administration of software, network and server equipment, which is part of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation;

• a component of serviceability control, collection and storage of data on the parameters of functioning of the main elements of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation;

• a component of accounting and diagnosing malfunctions in the operation of network, software and hardware systems for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation.

The administration subsystem records all registered failures in the operation of both software and hardware, indicating the location, type and cause of failures, and also has the ability to store them for a long time and generate reports for an arbitrary period of time.

The functions of the component of centralized administration of software, network and server equipment, which is part of the system for ensuring natural and technogenic safety of the territories of the subject of the Russian Federation, are performed by the system-wide software, as well as part of the functions of the health monitoring component.

The component for monitoring the serviceability, collecting and storing data on the parameters of the functioning of the main elements of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation provides a centralized collection and display on information radiators of data on the parameters of the functioning of subsystems and modules of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation, including:

• number of incoming HTTP requests from external systems;

• the number of errors that occur when processing incoming HTTP requests to the system;

• the number of created documents in the document repository;

• number of completed forecasts for the calculation of emergency risks;

The component for recording and diagnosing malfunctions in the operation of network, software and hardware of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation performs the following control checks:

• availability of devices, servers, subsystems, components, databases, services

• sufficient reserve of free system resources: disk space, RAM, processor cores.

The functions of the administration subsystem can be performed by means of freely distributed information systems. As a component of accounting and diagnosing malfunctions in the operation of network, software and hardware of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation, freely distributed special Zabbix software is used. The functionality of this software allows you to automatically diagnose in real time all significant parameters of the state of equipment and software. It also provides automatic notification of administrators and other responsible persons about all events.

The information security subsystem (ISS) is designed to protect information and means of its processing. PIB is a set of organizational measures, technical, software and hardware means of information protection and means of monitoring the effectiveness of information protection. The composition and technical implementation of the PIB is determined based on the analysis of the developed threat model and the violator model in accordance with the legislation of the Russian Federation in the field of state secret protection and protection of personal data and other confidential information.

The information security subsystem includes the following functional components:

• access control component;

• component of registration and accounting of user actions in the system;

• a component for ensuring secure interconnection;

• component of cryptographic protection of information;

• component of anti-virus protection;

• user identification component.

The PIB also includes all the necessary tools for administering information security issues and carrying out preventive work.

User authorization log is maintained.

Access control component

Performs identification and authentication of access subjects when entering the system by identifier (login) and password of a semi-permanent action with restrictions on the minimum length and complexity.

Identification of terminals, network nodes, communication channels, external devices by logical names is carried out if technically possible.

Subjects' access to protected resources is controlled in accordance with the access matrix.

Component of registration and accounting of user actions in the system

Registration of entry (exit) of subjects of access to the system (from the system) is carried out. Logging out of the system or stopping is not carried out at the time of a hardware shutdown.

Secure Interworking Component

Due to the connection of the system for ensuring natural and technogenic safety of the territories of the constituent entity of the Russian Federation to public communication networks, this functional component is implemented by using firewalls. Firewalls are purchased, installed and administered under a separate contract or existing in the subject can be used.

Component of cryptographic information protection

CIPF used in the system of ensuring natural and technogenic safety of the territories of the subject of the Russian Federation have all the relevant certificates. CIPF are purchased, installed and administered under a separate agreement or existing in the subject can be used.

Component of anti-virus protection

As part of the system for ensuring the natural and technogenic safety of the territories of a constituent entity of the Russian Federation, certified anti-virus protection tools are used on workstations and servers in order to protect confidential information, personal data, and software and hardware from the effects of malicious software. Anti-virus protection means are purchased, installed and administered under a separate contract or existing in the subject can be used.

User identification component

The user identification component provides a unified accounting of users and management of their rights.

The component is scalable to support newly created subsystems within the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation. User authorization (determination of the user's rights to information resources and a set of actions with them) is determined by each functional subsystem separately, with reference to unique user identifiers and taking into account their roles.

The administrator has the ability to maintain a database containing for each user its current state, roles, as well as other information necessary for the functioning of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation.

After identification, work with various functional subsystems does not require re-identification (except for the case of user inactivity for more than a configurable period of time - session timeout).

The component provides a programming interface for interaction that allows you to:

• identify the user by name and password, with the issuance of a unique session key (USK);

• extend the validity of the USC;

• request the validity of the USC;

• request the identifier of the user and his roles according to USC;

• forcibly terminate the validity of the USC (logout of the user from the system).

The identification component provides a subsystem administration user interface.

Usernames and passwords (or their encrypted representation) are transmitted to the system via an encrypted HTTPS connection.

For software tools used to protect information in the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation, control is provided for the absence of unauthorized access. All software and hardware that implements the functionality of information security is certified in the certification system of the FSTEC of Russia.

As the hardware and system-wide software of the information security subsystem, the software and hardware existing in the subject can be used or new ones deployed by the Customer can be organized.

In the course of developing a system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation, the following tools are provided:

• identification and authentication of access subjects and access objects;

• access control of access subjects to access objects;

• logging of identification and authentication events in the system;

• view and analyze information about attempts to identify and authenticate in the system.

Tools for identification and authentication of access subjects and access objects provide assignment of a unique attribute (identifier) to access subjects and objects, comparison of the identifier presented by the access subject (object) with the list of assigned identifiers, as well as verification of ownership of the identifier presented by the access subject (object) (authentication) .

Access control tools for access subjects to access objects provide management of the rights and privileges of access subjects, access control of access subjects to access objects based on a set of access control rules established in the information system, and also provide control over compliance with these rules.

Logging tools collect, record, store and protect information about identification and authentication events in the system, and viewing tools provide the ability to view and analyze information about such events and respond to them.

Implementation of the invention.

The system for ensuring natural and technogenic safety of the territories of the constituent entity of the Russian Federation is as follows. We give the most comprehensive example of the implementation of the invention. Bearing in mind that this example does not limit the application of the invention.

From various sources of external data, through wired communication lines over a unified network, that is, wireless communication, data enters the various modules and subsystems described above, where they are processed. We will show how the use of artificial intelligence, big data and the Internet of things is carried out

Determination of the probability of occurrence of forest fires, their area and geographical coordinates is carried out depending on weather conditions, the availability of infrastructure facilities.

As input data, a data vector of numerical and categorical features is taken: date, time, coordinates of a thermopoint (confirmed thermopoint - fire seat), area of the fire seat, weather conditions (temperature, precipitation, wind, pressure, thunderstorm activity, solar activity), weather forecast conditions for three days, fire hazard class, distance to buildings, distance to the railway section, distance to the highway section.

The output data is the area of a probable thermopoint, its coordinates. In the case of a binary classification, zero is returned if there is no thermopoint and one if there is one, as well as the probability.

Since artificial intelligence models accept only numerical data as input, it is necessary to preprocess categorical data (for example, a description of clouds, weather conditions, etc.).

The following methods can be used:

The One Hot Encoding method consists in the formation of a binary code of a fixed length, containing only one 1 - a direct unitary code, or only one 0 - an inverse (inverse) unitary code. The length of the code is determined by the number of encoded objects, that is, each object corresponds to a separate bit of the code, and the value of the code is determined by the position 1 or 0 in the categorical attribute. By default, One Hot Encoding converts the data to a sparse matrix.

Processing the dimension of input data

Many machine learning algorithms perform better when the features are relatively the same scale and close to normal.

The MinMaxScaler method is to reduce the data dimension. For each value in the MinMaxScaler object, subtracts the minimum value in the object and then divides by the range. The range is the difference between the original high and the original low. MinMaxScaler retains the form of the original distribution. This does not significantly change the information embedded in the original data. The default range for the function returned by MinMaxScaler is 0 to 1.

,

,

– новое значение;Where

– new value;

– текущее значение;

- present value;

– минимальное значение в диапазоне;

– minimum value in the range;

– максимальное значение в диапазоне.

– the maximum value in the range.

The StandardScaler method standardizes a function by subtracting the mean and then scaling to the variance of units. The unit of deviation means dividing all values by the standard deviation.

The StandardScaler averages the distribution to 0, i.e. about 68% of the values will be between -1 and 1.

,

,

where z is the standardized value;

– текущая выборка;

is the current sample;

– среднее текущей выборки;

is the average of the current sample;

– стандартное отклонение.

- standard deviation.

Next, work is performed with an unbalanced data set.

In the collected dataset, there are about 420 records of thermal points for 8200 observed records, which is 5% of the total amount of data. Such a dataset can be considered unbalanced.

Since the number of thermal points is small, we do not have the ability to remove the extra ones, in which case it is necessary to create synthetic data.

SMOTE - Synthesized Minorities Resampling Method, whereby we create elements in close proximity to those already existing in a smaller set.

For the operation of the system, deep machine learning algorithms can be used:

1. Fully connected neural network (multilayer perceptron)

Multilayer perceptron

A perceptron is an elementary neuron, which is a linear adder, each of the input signals of which is multiplied by a certain weight factor, and the output total signal is non-zero if the sum exceeds a certain threshold value.

A multilayer perceptron is a classic example of a unidirectional (feed-forward) neural network, i.e., a neural network in which there are no feedback effects of output neuron signals on the input of the network, and no connections between neurons of the same layer.

2. Convolutional Neural Networks

Convolutional neural networks (Convolutional Neural Networks) consist of several types of layers: a convolutional layer, a subsampling layer, a fully connected network layer.

The convolutional layer is a set of feature maps, each map has a synaptic nucleus. The subsampling layer performs a reduction in the size of the input map, characteristics (usually 2 times). This can be done in many ways, for example, you can use the method of selecting the maximum element (max-pooling). These two types of layers (convolutional, subsampling), interleaved with each other, form the input feature vector for the last layer type, the multilayer perceptron. The network can be trained using gradient methods.

3. Recurrent neural networks

Recurrent neural networks are networks that contain feedback and allow you to save information in the learning process.

4. LSTM layer (Long short-term memory, long short-term memory) is a type of recurrent neural network architecture capable of learning long-term dependencies. Figure B.6 shows the LSTM diagram of a neural connection.

The key component of an LSTM is the cell state, the horizontal line that runs across the top of the diagram.

The first step in an LSTM is to determine what information can be discarded from the cell state.

, (В.3)

, (AT 3)

– вектор вентиля забывания, вес запоминания старой информации;Where

is the forgetting gate vector, the weight of remembering old information;

– функция активации на основе сигмоиды;

is the activation function based on the sigmoid;

,

,

– матрицы параметров и вектора вентиля забывания;

,

,

are the parameter matrices and forget gate vectors;

– входной вектор;

is the input vector;

– выходной вектор.

is the output vector.

The second step is to decide what new information will be stored in the state of the cell. This stage consists of two parts. First, a sigmoidal layer called the "input layer gate" determines which values to update. Then the tanh layer builds a vector of new candidate values:

, (В.4)

, (AT 4)

– вектор входного вентиля, вес получения новой информации;Where

is the input gate vector, the weight of obtaining new information;

,

,

– матрицы параметров и вектора входного вентиля.

,

,

– matrices of parameters and input gate vector.

,

,

– вектор состояний;Where

is the state vector;

– функция активации на основе гиперболического тангенса;

is the activation function based on the hyperbolic tangent;

,

,

– матрицы параметров и вектора состояний.

,

,

are the parameter matrices and state vectors.

The third step is to decide what information we want to receive as an output. The output will be based on our cell state and some filters will be applied to it.

,

,

– вектор выходного вентиля, кандидат на выход;Where

is the output gate vector, a candidate for the output;

,

,

– матрицы параметров и вектора выходного вентиля.

,

,

are the matrix of parameters and the vector of the output gate.

,

,

– выходной вектор.Where

is the output vector.

Since thermopoints arise and are recorded in time, it can be assumed that when using neurons with memory, it is possible to trace some dependencies that affect the occurrence of fires.

If we take thermal points as an anomalous phenomenon, then it seems possible to use algorithms aimed at identifying these anomalies.

5. The dLSTM method (delayed long-short term memory, delayed long short-term memory) is a neural network training method. We first build a predictive model from the normal (not anomalous) training data and then perform anomaly detection based on the prediction error on the observed data. After building the models, we compare the forecast errors on normal and abnormal data. Obviously, the error will be higher on anomalous data.

6. Regularization

Regularization is a method of adding constraints to model parameters instead of reducing them.

Dropout regularization means that at each training step, each neuron turns off with some probability p. On such a thinned network, the error is backpropagated (while the excluded neurons do not affect the learning process), after which all the turned off neurons are returned to the neural network. Thus, at each step, one of the possible 2N network architectures is configured, where N is the total number of neurons. When using a neural network, the neurons are no longer turned off, but the output of each neuron is multiplied by (1 − p) - thanks to this, the output of the neuron will be the mathematical expectation of its response over all 2N architectures. Thus, a neural network trained using Dropout regularization can be considered as the result of averaging 2N networks.

7 Activation functions

When constructing the architectures of neural networks, it is supposed to solve the problems of regression, as well as classification, taking into account which the following activation functions are used:

In Sigmoid, in the range of X values from -2 to 2, the Y values change very quickly. This means that any small change in the value of X in this area entails a significant change in the value of Y. This behavior of the function indicates that Y tends to nestle towards one of the edges of the curve.

The sigmoid is not linear in nature and the combinations are not linear either. So you can connect layers to each other. The sigmoid is not binary, which makes the activation analog. It has a smooth gradient. This activation function looks like a suitable function for classification problems. It tends to bring the values to one of the sides of the curve (for example, to the top - at X = 2 and the bottom - at X = -2). This behavior allows you to find clear boundaries in the prediction.

The sigmoid has a fixed range of function values - [0,1]. This property of the sigmoid is very useful, as it does not lead to errors in the case of large activation values. As you approach the ends of the sigmoid, the Y values tend to react weakly to changes in X. This means that the gradient in such areas takes on small values, which in turn leads to problems with the fading gradient. In such a case, the gradient value is small or disappears (will not lead to significant changes due to the extremely small value). The neural network refuses to train further or does so extremely slowly (depending on the usage or until the gradient/calculation starts to suffer from floating point limitations).

ReLU function

ReLU returns x if x is positive and 0 otherwise.

ReLU is non-linear in nature and ReLU combination is non-linear as well. (Actually, such a function is a good approximator, since any function can be approximated by a ReLU combination.) This means that we can connect layers to each other. ReLU is less demanding on computing resources than hyperbolic tangent or sigmoid, as it performs simpler mathematical operations. Therefore, it makes sense to use ReLU when creating deep neural networks.

Softmax function

The Softmax function is used in machine learning for classification problems when the number of possible classes is more than two (a logistic function is used for two classes).

Softmax is used for the last layer of deep neural networks for classification tasks. To train the neural network, cross entropy is used as a loss function.

Assessment Metrics

Confusion matrix is a performance measurement for a machine learning classification problem where the output can be two or more classes (Figure B.10). This is a table with 4 different combinations of predicted and actual values.

It is extremely useful for calculating recall, accuracy, specificity.

TP - true positive solution: Interpretation: A positive number was predicted, which turned out to be true.

TN - true negative solution: Interpretation: A negative value was predicted, which turned out to be true.

FP – False Positive Solution (Type 1 Error): Interpretation: Positive value predicted, which is false (value is actually negative).

FN - False-Negative (Type 2 Error): Interpretation: Predicted a negative value, which is false (the value is actually positive).

Metrics allow you to fully evaluate the ability of the algorithm to recognize.

Precision means how many correct predictions were made from all classes. The accuracy should be as high as possible.

Recall (responsiveness) means how many correct predictions were made from all positive classes. This setting should also be as high as possible.

F-measure. It is difficult to compare two models with low accuracy and high responsiveness, or vice versa. Therefore, to make them comparable, we use the F-measure. F-measure helps to measure recall and precision at the same time. It uses the harmonic mean instead of the arithmetic mean, penalizing extreme values more.

The Accuracy and Specificity metrics are almost never used due to their partial bias.

Loss/error functions

The task of training neural networks is to minimize the error.

For classification tasks, Categorical cross-entropy is used - categorical cross-entropy. On the output layer - Softmax, then the values \u200b\u200bwill take values from 0 to 1

Data must be submitted in One Hot Encoding format

,

,

where p is the "true" distribution, observed only partially;

t is the input value;

C is the number of classes;

o - some observation.

For regression problems, use:

MSE - root mean square error

The essence of the method is essentially to minimize the sum of squared deviations of the actual values from the calculated ones (SSE - "Sum of Squared Errors"). If the resulting sum is divided by the number of observations, then the same MSE will be obtained. The objective function formula in this case is as follows:

,

,

– истинное значение;Where

- true value;

– расчетное значение;

– calculated value;

n is the number of observations.

MAE - mean absolute error - the average value of the modulus of the difference between prediction and real values /

The gradient is the same at all points, because there is a module. (decreases when approaching the minimum) The undoubted advantage of MAE is that the modules do not multiply the deviations considered as outliers. Therefore, this estimate is more robust than MSE and actually corresponds to the median.

Fully Connected Neural Network Architecture

Layer Options Input Number of neurons: 128, 256, 512
Input dimension: vector [1]
ReLU activation function, Sigmoid Layer Options First hidden layer Number of neurons: 64, 128, 256
ReLU activation function, Sigmoid Regularization batch normalization
Dropout: 0.18 0.2; 0.25; 0.3 Second hidden layer Number of neurons: 256, 512, 1024
Activation function: ReLU, Sigmoid Regularization batch normalization
Dropout: 0.18 0.2; 0.25; 0.3 third hidden layer Number of neurons: 32, 64, 128
Activation function: ReLU, Sigmoid Regularization Dropout: 0.18 0.2; 0.25; 0.3 Output fully connected layer Number of neurons: 3
Activation function: Sigmoid, Softmax Compilation Error function: Categorical cross-entropy, MAE, MSE
Optimizer: Adam
Metric: accuracy, precision, recall

An experiment is performed for each parameter.

Convolutional Neural Network Architecture (one-dimensional convolution)

Layer Options input layer Number of feature cards: 6, 12,24,32
Convolution kernel size: 2,3,6
Input size: 0 x vector matrix [1]
padding padding: same
Activation function: ReLU First convolution layer Number of feature cards: 6, 12,24,32
Convolution kernel size: 2,3,6
Input size: 0 x vector matrix [1]
padding padding: same
Activation function: ReLU Second layer of convolution Number of feature cards: 6, 12,24,32
Convolution kernel size: 2,3,6
Input size: 0 x vector matrix [1]
padding padding: same
Activation function: ReLU Layer Options Regularization Flatten - flatten to vector [1] Fully Connected Layer Number of neurons: 10, 32.64
Input dimension: vector [1]
ReLU activation function, Sigmoid Regularization batch normalization
Dropout: 0.18 0.2; 0.25; 0.3 Output fully connected layer Number of neurons: 3
ReLU activation function, Sigmoid, softmax Compilation Error function: Categorical cross-entropy, MAE, MSE
Optimizer: Adam
Metric: accuracy, precision, recall

An experiment is performed for each parameter.

Recurrent neural network architecture with LSTM layers

Layer Options input layer Number of neurons: 5, 10, 20.50
Input size: vector [1]
Return sequence: true
(for docking with the next layer)
Activation function: ReLU First hidden layer Number of neurons: 10,15,20,40
Return sequence: true
(for docking with the next layer)
Activation function: ReLU Second hidden layer Number of neurons: 5, 10,15,20
Return sequence: true
(for docking with the next layer)
Activation function: third hidden layer Number of neurons: 5,10,15,20
Return sequence: false
(for docking with the next layer)
Activation function: Regularization Flatten - flatten to vector [1] Output fully connected layer Number of neurons: 3
Activation function: ReLU, Sigmoid, Softmax Compilation Error function: Categorical cross-entropy, MAE, MSE
Optimizer: Adam
Metric: accuracy, precision, recall

For each parameter, an experiment is performed, selected with the best accuracy

Industrial applicability.

The proposed system for ensuring the natural and technogenic safety of the territories of a constituent entity of the Russian Federation can be implemented by a specialist in practice and, when implemented, ensure the implementation of the declared purpose, which allows us to conclude that the invention meets the criterion of "industrial applicability".

Six servers united in failover clusters can be used as the main computing facilities. Two virtual machines can be deployed on each of the clusters: a database server, an application server, and a mapping server.

To solve the problem of predicting the probability of occurrence of forest fires, leased computing resources with the characteristics of 16vCPUx3GHz, 400 GB RAM, 800 GB SSD-Disk, 1 Tesla T4 16 GB can be used.

For routing between data storage devices, a high-speed data transmission network with a capacity of at least 10 Gbps can be used; for other needs, communication channels with a bandwidth of at least 100 Mbps are used.

The system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation provides automation of the process, collection and processing of information on the level of natural and technogenic risks in the territory of municipalities of a constituent entity of the Russian Federation.

The system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation ensures the implementation of the following procedures:

1. receiving, processing and forwarding requests;

2. collection and analytical processing of signals from associated threat monitoring systems, including video surveillance systems, fire safety systems and traffic safety systems, transport monitoring systems (GLONASS), environmental, hydrological and meteorological monitoring systems;

3. coordination, management and support of interdepartmental interaction when responding to incoming calls and signals with the display of the operational situation on an electronic map, according to the received forecasts for the development of the situation;

4. informing the population through integration with the Internet portal for interaction with the population and posting information messages on it;

5. forecasting the consequences of natural and man-made emergencies;

6. assessment of natural and technogenic risks and issuance of recommendations to reduce their level, taking into account the distribution of the population over the territory and the modeled parameters of damaging factors;

7. information interface with monitoring systems of objects equipped with sensors for monitoring functioning parameters through the exchange of information exchange cards and instructions, as well as data from interfaced information systems;

8. promptly informing the population about the threat of occurrence or occurrence of the CES, state authorities and management bodies of the RSChS on the territory of municipal districts, by issuing recommendations on the launch of existing warning systems;

9. support for the adoption of managerial decisions in the event of a threat of occurrence or elimination of the consequences of an emergency.

The system for ensuring the natural and technogenic safety of the territories of a constituent entity of the Russian Federation was developed using the Internet of Things, Big Data and Artificial Intelligence technologies to predict the likelihood of occurrence of forest fires and recognize fires in the territory under consideration by interfacing with operated forest fire detection systems.

The functionality of the system for ensuring natural and technogenic safety of the territories of a constituent entity of the Russian Federation for automating response to crisis situations that are not classified as emergencies makes it possible to increase the efficiency of organizing interaction between various emergency services and departments, management bodies of the RSChS, dispatching services of enterprises and regional executive authorities in the event of incidents , which is a significant advantage compared to the currently operated Integrated Life Safety System and System-112.

Claims (58)

1. Decision-making system in case of threat and liquidation of consequences of emergencies, including interconnected • a subsystem for collecting and processing appeals on crisis and emergency situations (CEC) and a subsystem for comprehensive information and notification of CES; • decision support subsystem in case of threat and liquidation of consequences of emergencies; • subsystem of information storage and processing; • an integration geoinformation subsystem for converting vector formats of geospatial data into raster formats and ensuring prompt display of information based on electronic maps of territories; characterized in that the system also includes: • a subsystem for interaction and coordination of actions to eliminate the CES; • a subsystem for integrated monitoring and forecasting of the CES; • data integration subsystem providing automated data exchange between subsystems; • content management subsystem to provide an ordered register of information objects of the KiChS; • information-analytical subsystem for displaying, exchanging information about CES and generating reporting forms based on information stored in the system database; • telecommunications subsystem that provides data transfer and interaction between subsystems, • operation administration subsystem for performing the tasks of monitoring the technical condition and functioning of subsystems; • an information security subsystem to protect information and means of its processing of the system, • moreover, the decision support subsystem in the event of a threat and elimination of the consequences of emergencies includes a component for assessing the probability of occurrence of forest fires, based on the principle of artificial intelligence. 2. The system according to claim 1, characterized in that the subsystem of integrated monitoring and forecasting of the KiChS includes: • intelligent video surveillance module for smoke and fires; • module for monitoring housing and communal services systems; • fire situation monitoring module; • module for monitoring the state of the environment to control the state of current threats of a natural nature; • Forecasting component as part of the following modules: • module of fires and explosions at potentially dangerous objects; • module of accidents with the release of toxic substances at chemically hazardous facilities; • module of floods caused by high water; •module of forest and steppe fires. 3. The system according to claim 1, characterized in that decision support subsystem in the event of a threat and elimination of the consequences of emergencies includes: • component of assessment and management of natural and technogenic risks; • component of contingency plans; • component of the calculation of forces and means for rescue and recovery operations; • component of preparation of management documents. 4. The system according to claim 1, characterized in that the subsystem of interaction and coordination of actions for the elimination of the CES includes: • information exchange component; • execution control component. 5. The system according to claim 1, characterized in that data integration subsystem that provides automated data exchange between subsystems, includes: • component of interaction protocols for standardization of formats, rules and regulations of interaction between subsystems; • data collection component; • data transfer component; • component of classifiers and directories. 6. The system according to claim 1, characterized in that the content management subsystem for providing an ordered register of information objects of the KiChS includes: • component of information object management; • Internet portal content management component. 7. The system according to claim 1, characterized in that the geoinformation subsystem for converting vector formats of geospatial data into raster formats and ensuring prompt display of information based on electronic maps of territories includes: • component of transformation of cartographic data; • a component for displaying the location of forces and means of the unified state system for the prevention and liquidation of emergency situations (RSChS); • component for displaying cartographic information; • component for editing passports of territories. 8. The system according to claim 1, characterized in that the telecommunications subsystem that provides data transfer and interaction between subsystems includes: • computer facilities; • means of communication technology; • means of organizational technology; • hardware and software of the information storage and processing subsystem. 9. The system according to claim 1, characterized in that the information security subsystem for protecting information and means of its processing of the system includes: • access control component; • component of registration and accounting of user actions in the system; • a component for ensuring secure interconnection; • component of cryptographic protection of information; • component of anti-virus protection; • user identification component.

Images