KR20200107626A - Real-time accident prediction system using hazard assessment and method thereof - Google Patents

Abstract

The present invention relates to a real-time accident prediction system using hazard assessment and a method thereof. The real-time accident prediction system comprises: a communication unit collecting a measured signal value from one or more IoT sensors located inside or outside a facility; an analysis unit which changes one or more of individual frequency values set in accordance with possibility of an accident occurrence and ending values in accordance with severity of an ending in accordance with the accident occurrence with regard to an individual accident related in a signal within an abnormal range in an accident list stored in advance based on an event tree linked to the IoT sensor when the measured signal value is within the abnormal range, and calculating a hazard index using the changed individual frequency values and ending values; and a control unit calculating a priority with regard to the accident list in real time based on the hazard index and providing a list of predicted accidents rearranged in an order of high priority.

Description

Real-time accident prediction system using risk assessment and its method {REAL-TIME ACCIDENT PREDICTION SYSTEM USING HAZARD ASSESSMENT AND METHOD THEREOF}

A real-time accident prediction system and method using risk assessment are provided.

Nuclear power plants and nuclear cycle facilities require flawless safety management because the scale of damage and high risk in the event of an accident is high. Accordingly, a highly reliable safety management method for nuclear facilities is being studied.

In particular, rapid safety management through simulated management of emergency response in case of an accident, or limited and specialized systems in specific fields such as fire, gas, and electricity are being developed.

However, since these methods provide safety response measures in response to fixed scenarios, they are insufficient to be applied to an actual environment with many various variables, and there is a limit to application to various facilities.

For example, a nuclear power plant is a continuous and complex process that converts thermal energy into electrical energy through nuclear reactors, steam generators, turbines, and generators. While all processes are always operating, Nuclear Fuel cycle facility) is composed of several independent processes compared to nuclear power plants, so that the main processes of the nuclear cycle facility operate independently.

Therefore, risk and safety management for nuclear cycle facilities differs from hazardous materials and hazards depending on the process and facility, and the impact of an abnormal accident is limited to the process. It should be carried out in units of facilities.

Therefore, there is a need for a system that analyzes the risk of each facility in consideration of individual facility characteristics, analyzes a list of possible accidents, and predicts possible accidents in the facility through real-time IoT communication and responds in advance.

As a related prior document, Korean Patent No. 1,675,733 discloses a "nuclear accident integrated response system".

Korean Patent 1,675,733

An embodiment of the present invention is to provide a real-time accident prediction system for preventing accidents and minimizing adverse effects due to accidents through dynamic comprehensive safety analysis and real-time accident prediction using IoT.

One embodiment of the present invention is to implement a constant intensive monitoring system by preemptively providing a risk index and response priority according to the probability of occurrence of a predicted accident and the degree of harm caused by the accident.

One embodiment of the present invention is to provide a quick and accurate preemptive response plan and accident mitigation response plan by providing visualized information such as countermeasures, information on expected accidents, unusual elements, and abnormalities based on the expected accident grade. will be.

In addition to the above problems, the embodiments according to the present invention may be used to achieve other tasks not specifically mentioned.

The real-time accident prediction system according to an embodiment of the present invention includes a communication unit that collects a signal value measured from one or more IoT sensors located inside or outside a facility, and if the measured signal value is in an abnormal range, an event associated with the IoT sensor At least one of an individual frequency value set according to the likelihood of an accident and an ending value set according to the severity of the ending due to the occurrence of the accident for individual accidents related to the signal in the above range in the accident list stored in advance based on the event tree. An analysis unit that calculates a hazard index using the changed individual frequency value and ending value, and a predicted accident list that calculates the priority for the accident list in real time based on the hazard index, and rearranges the order of highest priority. It includes a control unit that provides in real time.

The accident prediction method of the real-time accident prediction system according to an embodiment of the present invention includes the steps of collecting measurement values from one or more IoT sensors located inside or outside a facility, and when the measurement values are included in the abnormal range, the IoT sensor One of the individual frequency values set according to the probability of the occurrence of the accident and the ending values set according to the severity of the ending due to the occurrence of the accident for an individual accident related to the signal in the above range in the accident list stored in advance based on the number of events associated with Changing the above, updating the changed data in the event tree and calculating the hazard index using individual frequency values and ending values, calculating the priority of the accident list in real time based on the hazard index, and prioritization It includes the step of providing in real time a list of predicted accidents rearranged in order of higher ranking.

According to an embodiment of the present invention, by making the most of information collected from IoT sensors and combining it with a dynamic comprehensive safety analysis, it is possible to perform an accident prediction that is most suitable for a site in which various variables are embedded.

According to an embodiment of the present invention, information can be provided so that preemptive countermeasures to prevent accidents can be taken in response to detected abnormal signs, and the spread of damage can inevitably be minimized when an accident occurs. It can provide fast and accurate initial response measures.

According to an embodiment of the present invention, by performing an overall safety check and safety analysis based on a detailed accident analysis, and by setting and providing priorities in real time, the facility manager is selective for accidents with a high probability of occurrence and a large impact of a harmful outcome. By focusing on it, safety management can be performed economically and efficiently.

According to one embodiment of the present invention, by performing constant intensive monitoring in vulnerable buildings such as nuclear facilities and nuclear-related facilities, implementing an immediate response system, and minimizing the impact of the end of harm through active accident prevention measures. It can increase the external safety of nuclear facilities or hazardous facilities and increase the acceptance of the public.

1 is an exemplary diagram showing a communication network including a real-time accident prediction system according to an embodiment of the present invention.
2 is a block diagram showing a real-time accident prediction system according to an embodiment of the present invention.
3 is a block diagram showing an ISA analysis unit according to an embodiment of the present invention.
4 is an exemplary view for explaining a method of calculating a risk index through categories according to an embodiment of the present invention.
5 is a flowchart illustrating an accident prediction method of a real-time accident prediction system according to an embodiment of the present invention.
6 is an exemplary view showing a tree of events including a No.01-1 risk criterion event according to an embodiment of the present invention.
7 is an exemplary view showing a tree of events including a No. 01-2 risk criterion event according to an embodiment of the present invention.
8 is an exemplary view showing an accident situation with respect to a cause of fire according to an embodiment of the present invention.
9 is an exemplary view showing an accident procedure for a cause of a short circuit and an electric fire according to an embodiment of the present invention.
10 is an exemplary view showing an accident related to radioactive leakage according to an embodiment of the present invention.
FIG. 11 is an exemplary diagram for explaining priorities varying according to individual frequency values in a real-time predicted accident list according to an embodiment of the present invention.
12 is an exemplary diagram for explaining a priority that is changed by an increase in an ending value of an ending category in a real-time predicted accident list according to an embodiment of the present invention.
13 is an exemplary diagram for explaining a configuration for visualizing and providing environment or state information according to accident prediction information according to an embodiment of the present invention.

Embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art may easily implement the present invention. The present invention may be implemented in various different forms, and is not limited to the embodiments described herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and the same reference numerals are used for the same or similar components throughout the specification. Also, in the case of well-known technologies, detailed descriptions thereof will be omitted.

Throughout the specification, when a part "includes" a certain component, it means that other components may be further included rather than excluding other components unless otherwise stated.

In this specification,'hazard (hazard)' refers to a physical or chemical state that has the potential to cause damage to human, material, or environment, and'risk reference event' refers to the subject of analysis through risk identification. It refers to an event that is re-selected from the hazard events that are presumed to be likely to occur among the hazard events that have been identified and arranged for hazardous substances and potential hazards existing in the facility and its surrounding area.

In the present specification, the'event tree' represents a kind of event judgment, and in detail, it represents a logical tree used to find the possibility of a major failure or event occurring for a large-scale system.

Therefore, the event tree contains scenarios of potential accidents that may occur.

1 is an exemplary diagram showing a communication network including a real-time accident prediction system according to an embodiment of the present invention.

As shown in FIG. 1, the real-time accident prediction system 200 is connected to one or more IoT sensors 100, a related terminal 300, and a database 400 through a network to transmit and receive data.

Here, the network may include all types of communication networks that transmit data, such as a wired communication network, a short-range or long-distance wireless communication network, and a network in which they are mixed.

First, the IoT sensor 100 represents a sensor or device having a function (Internet Of Things, IoT) that transmits and receives data in real time through a network, and may include electricity, gas, temperature, fire detection, and the like. In addition, the IoT sensor 100 may include a sensor capable of detecting radiation, movement, vibration, flooding, flow rate, pressure, precipitation, fine dust measurement, wind speed, air volume, etc., but is not limited thereto.

In addition, the IoT sensor 100 may transmit the measured value to the real-time accident prediction system 200 in real time, or output as on/off when reaching a reference range according to the measured value and transmit it to the real-time accident prediction system 200. .

This configuration is set differently according to the type and characteristics of the IoT sensor 100, and can be easily changed and designed by an administrator later.

In addition, each IoT sensor 100 may measure the location information currently located in real time to check the actual location at the designated installation point, and transmit the measured value and the location information to the real-time accident prediction system 200 together. .

Next, the real-time accident prediction system 200 constructs data by calculating the accident list analyzed through comprehensive safety analysis (ISA), the frequency and ending of individual accidents, and measuring values reflecting the situation at the site through IoT sensors in real time. By constructing big data by using, dynamic comprehensive stability analysis (Dynamic-ISA) is performed.

Accordingly, the real-time accident prediction system 200 may select a real-time predicted accident priority using a dynamic comprehensive safety analysis and the Internet of Things, and propose a priority to be responded to from the entire list of predicted accidents.

In addition, the real-time accident prediction system 200 may transmit a notification message including a preemptive response plan and a follow-up plan to the related terminal 300 to enable a complex response to the expected accident.

Meanwhile, the real-time accident prediction system 200 is applied to a nuclear facility, a nuclear cycle facility, and a spent nuclear fuel management facility, but is not limited thereto, and can be applied to various facilities requiring accident prediction management.

In addition, the real-time accident prediction system 200 may be implemented in the form of a platform, which is an operating system based on a specific processor model that is a basis of a computer system and one computer system.

The relevant terminal 300 represents a terminal of a person in charge of a facility, a person in charge of safety management, or a worker registered in advance. The related terminal 300 includes, for example, a personal computer, a handheld computer, a personal digital assistant (PDA), a mobile phone, a smart device, a tablet, and the like.

In addition, the related terminal 300 may include not only a person in charge of related facilities, but also a server of another institution such as a fire department, a military unit, or a hospital.

Next, the database 400 stores all data calculated by real-time collection and analysis by the real-time accident prediction system 200.

In addition, the database 400 may include reference data for dynamic overall safety analysis, response plan data, contact network data of related parties, and spatial information data within the facility, and can be updated in real time.

Although the real-time accident prediction system 200 and the database 400 are separately illustrated in FIG. 1, the database 400 may be included in the real-time accident prediction system 200 and implemented by an administrator later.

Hereinafter, a real-time accident prediction system 200 using dynamic comprehensive safety analysis will be described in detail with reference to FIGS. 2 to 4.

2 is a block diagram illustrating a real-time accident prediction system according to an embodiment of the present invention, and FIG. 3 is an exemplary view for explaining a semi-quantitative safety analysis method according to an embodiment of the present invention. And Figure 4 is a block diagram showing the ISA analysis unit according to an embodiment of the present invention.

As shown in FIG. 2, the real-time accident prediction system 200 includes a communication unit 210, an analysis unit 220, a control unit 230, a notification unit 240, and an ISA analysis unit 250.

First, the communication unit 210 collects signal values measured from one or more IoT sensors 100 located inside or outside the facility. In this case, the communication unit 210 may collect a signal value as a value measured in real time or an on/off value output from the IoT sensor 100.

In addition, the communication unit 210 may collect corresponding data by accessing a separate external institution server, site, blog, or the like. For example, the communication unit 210 may collect weather information and climate change data of the Meteorological Agency.

If the measured signal value is in the abnormal range, the analysis unit 220 determines the frequency of individual accidents related to the abnormality in the accident list stored in advance based on the event tree linked to the IoT sensor 100 (accident occurrence). Possibility), or the ending value (accident severity) of an individual incident related to the anomaly.

In other words, the IoT sensor 100 is associated with one or more event trees, and is linked with individual frequency values and ending values of accident lists included in the event tree.

In addition, the analysis unit 220 calculates a hazard index using individual frequency values and ending categories based on the changed data.

Referring to FIG. 3, a frequency calculation value (Likelihood, individual frequency value) and a consequence value are divided into categories each having a value of 1 to 4, and are stored in a 4*4 matrix structure.

That is, a hazard index having a value of 1 to 16 is calculated through the multiplication of the frequency calculation value (Likelihood, individual frequency value) and the ending (Consequence).

The area A of FIG. 3 is calculated as Not Acceptable to request prompt action and confirmation, and the hatched area is calculated as Acceptable, but it may be set as requiring attention to the expected accident.

In this way, each hazard index can be grouped and a corresponding step can be set for each group.

In addition, although the basic 4*4 matrix structure is shown in FIG. 3, it can be easily changed by the administrator in the future according to the usage environment such as 3*3 or 4*5 according to individual frequencies and categories of endings.

This configuration can be economically and practically applied to small-scale facilities that cannot use probabilistic safety analysis (PSA) as a semi-quantitative safety analysis method, and facilities with low potential risk.

Next, the controller 230 calculates a priority for the accident list in real time based on the hazard index, and provides a list of predicted accidents rearranged in the order of high priority in real time.

The control unit 230 may collect the list of accidents when the risk index is more than a certain amount and calculate the priority, and automatically update the list of expected accidents when the risk index changes.

Next, the notification unit 240 sets one response step among new entry, standby, monitoring, accident, immediate response, or additional response to the expected accident based on the hazard index and priority. The configuration of these response steps can be changed and designed to be suitable for the facility later by the manager.

In addition, the notification unit 240 transmits a notification message corresponding to the corresponding step to the related terminal 300.

Next, the ISA analysis unit 250 analyzes and evaluates the risk through Integrated Safety Analysis (ISA) for the facility, and the list of possible accidents, the frequency of individual accidents in the accident list (probability), and conclusion Calculate and store categories.

As shown in FIG. 4, the ISA analysis unit 250 includes a risk analysis unit 251, a reference event selection unit 252, an event tree development unit 253, and a key point definition unit 254.

The risk analysis unit 251 identifies hazardous substances and potential hazards that exist in the facility to be analyzed, for example, process design information, zone definition of processes and facilities, identification of hazardous materials of processes and facilities, process and Hazards can be identified by checking the risk factors of the facility.

In other words, the risk analysis unit 251 identifies the hazardous substances handled by the facility and potential hazards including radiological, chemical, physical, natural disasters, and external threats, The frequency of occurrence of individual events (individual frequency values) and ending categories can be calculated.

Here, the frequency of occurrence (individual frequency value) is calculated by evaluating and analyzing the impact and the frequency of occurrence for each accident situation through accident data and experiences accumulated from similar facilities and previous points in the corresponding facility.

Specifically, the frequency of occurrence category (individual frequency value) is the accident occurrence record at the facility, the function failure record of the IROFS (Items Relied on for Safety, safety related items) at the facility, and the applicable accident data for similar systems, It can be based on objective qualitative criteria that control the damage rate and availability of the system, or other methods that are objectively verified.

In addition, the frequency of occurrence can be calculated by considering the duration of the accident and the common cause of failure depending on the situation.

In addition, the ending category (ending value) is a composition that analyzes the severity of the ending that hazardous substances affect workers, the public, and the environment due to hazards, and is classified based on the estimated radiological, chemical, or environmental effects. Can be.

Next, the reference event selection unit 252 includes one or more risk-based events (accident list), including a design-based accident selected based on the results of the risk analysis unit 251 and a risk event that is likely to occur in relation to the facility. ) Is selected.

In this case, the reference event selection unit 252 may include the reference events that are estimated to need consideration as all the considerations.

The event tree development unit 253 analyzes the risk-based events and generates an event tree that can know the possibility of the risk-based events and the potential accidents.

Next, the core point definition unit 254 selects an IoT sensor that determines whether an event tree occurs, and provides an installation point by designating it.

In addition, when the selected IoT sensor transmits a signal at the installation point, the core point definition unit 254 may link the event tree to the IoT sensor. Alternatively, you can check the installation point of the IoT sensor by confirming the administrator who installed the IoT sensor, and link the event tree to the IoT sensor.

In this case, although it is shown that the ISA analysis unit 250 is included in FIG. 2, the ISA analysis unit 250 may exist as a separate device or server, and may be implemented in a form of sharing a linked database.

Meanwhile, the real-time accident prediction system 200 may be a server, a terminal, or a combination thereof.

The terminal is collectively referred to as a device having computational processing capability by having a memory and a processor, respectively. For example, there are a personal computer, a handheld computer, a personal digital assistant (PDA), a mobile phone, a smart device, a tablet, and the like.

The server is a memory in which a plurality of modules are stored, and a processor that is connected to the memory and reacts to the plurality of modules and processes service information provided to the terminal or action information that controls the service information, communication It may include means, and a UI (user interface) display means.

Memory is a device that stores information, and is non-volatile such as high-speed random access memory, magnetic disk storage, flash memory device, and other non-volatile solid-state memory devices. It may include various types of memories such as volatile memory.

The communication means transmits and receives service information or action information to and from the terminal in real time.

The UI display means outputs system service information or action information in real time. The UI display means may be an independent device that directly or indirectly outputs or displays the UI, or may be a part of the device.

Hereinafter, a process of providing information by predicting an accident in real time using risk assessment and transmitting a notification message corresponding to each stage of the predicted accident will be described in detail with reference to FIGS. 5 to 13.

Hereinafter, a real-time accident prediction method using risk assessment will be described in detail with reference to FIG. 5.

5 is a flowchart illustrating an accident prediction method of a real-time accident prediction system according to an embodiment of the present invention.

As shown in FIG. 5, the real-time accident prediction system 200 identifies and analyzes the hazards related to the facility, and calculates the predicted events that may occur due to the hazards, the frequency of occurrence and the ending category for each event (S510). ).

The real-time accident prediction system 200 may identify and analyze hazardous substances handled by facilities and potential hazards including radiological, chemical, physical, natural disasters, and external threats.

In addition, the hazards related to the facility include, but are not limited to, all the hazards of the facility itself and various structures, systems, and components that exist inside and outside the facility.

In addition, the real-time accident prediction system 200 calculates a predicted event that can occur, an occurrence frequency and an ending category for each event using the analyzed hazard.

Next, the real-time accident prediction system 200 selects and analyzes a reference event (accident list) from among the possible expected events, and generates an event tree (S520). As such, the real-time accident prediction system 200 selects one or more risk-based events by analyzing risk factors, and analyzes the risk-based events. In addition, it is possible to create a tree of events in which the probability of occurrence of risk-based events and the history of potential accidents can be identified.

Next, the real-time accident prediction system 200 connects an IoT sensor that determines whether an event tree has occurred (S530).

The real-time accident prediction system 200 selects an IoT sensor that determines whether or not an event tree occurs, designates an installation point, and provides it to the related terminal 300.

Accordingly, the real-time accident prediction system 200 may link the event tree to the IoT sensor when the selected IoT sensor transmits a signal at the installation point.

The selection of risk-based events through risk analysis of the real-time accident prediction system 200, generation of event trees, and connection between IoT sensors and event trees will be described in detail with reference to FIGS. 6 and 7 below.

6 is an exemplary view showing a tree of events including a No. 01-1 risk criterion event according to an embodiment of the present invention, and FIG. 7 is an exemplary view showing a No. 01-2 risk criterion according to an embodiment of the present invention. It is an exemplary diagram showing the number of events including the event.

First, FIG. 6 includes a risk criterion event related to a fire due to overheating of a hot wire for preventing freezing when the external temperature decreases.

In order to prevent freezing, in case the outside temperature in winter falls below freezing, measures are taken to wind some pipes with heat wires and wrap them with insulation to prevent freezing of exposed pipes. At this time, when it detects that the external temperature falls below the set temperature, electricity is automatically supplied to the heater to prevent the water inside the pipe from freezing, and heat is generated by a constant output to the heating wire connected thereto.

The real-time accident prediction system 200 may analyze such risk-based events to check the probability of occurrence and the accident status one by one.

As shown in Fig. 6, when analyzing the accidents on fire due to overheating of the hot wire for freeze protection, it can be divided into external temperature measurement sensor abnormality, electric distribution board abnormality, heater output abnormality, heating wire temperature abnormality, overheating of the heating wire, and fire occurrence. I can.

In response to these accidents, the real-time accident prediction system 200 identifies the core points, which are the installation points of IoT sensors, as an external temperature measurement point, an abnormality measurement point for an electric distribution board, a heater output (current) measurement point, and a heat wire temperature measurement point in the external pipe insulation. , It can be selected as a measurement point for thermal imaging cameras outside the building, and as a fire detector.

Accordingly, IoT sensors are installed at each of the key points, and the real-time accident prediction system 200 can store and manage in connection with the corresponding event tree.

On the other hand, FIG. 7 includes a risk reference event indicating a fire due to a short circuit in the heater for preventing freezing in an environment where the external temperature is normal.

As shown in Fig.7, when analyzing the accident situation of the fire due to a short circuit in the heater for freeze protection, it can be divided into an external temperature measurement sensor abnormality, an electric distribution board abnormality, a heater output abnormality, a heating wire temperature abnormality, overheating of the heating wire, and fire occurrence. I can.

In response to these accidents, the real-time accident prediction system 200 is the installation point of the IoT sensor, which is an external temperature measurement point, an electric distribution panel abnormality measurement point, a heater output (current) measurement point, a heat wire temperature measurement point in an external pipe insulation material, and a building exterior. In addition to the thermal imaging camera measurement point and the fire detector, a thermal imaging camera measurement point inside the building may be further included.

On the other hand, since Figure 7 shows the same event at the same location caused by a different cause from Figure 6, looking at each event tree, the event for the distribution panel current anomaly, heater output anomaly, heating wire temperature anomaly, thermal imaging camera thermal rise observation It can be seen that the change values of the probability of occurrence of trees are the same.

In other words, if an abnormality in the distribution panel current is measured, the probability of fire is changed equally to 2, and if an abnormality is measured at the heater output, the probability of occurrence is changed equally to 3, and there is an abnormality in the temperature of the heating wire and heat rise due to the thermal imaging camera. When is observed, the probability of fire is changed equally to 4.

In this way, since the accident list is generated based on the event tree, two or more accident lists can be linked from the same incident tree to share the changed individual frequency values.

In addition, the real-time accident prediction system 200 may determine that the fire is caused by different causes by using the difference in signals at different measurement points in the situations of FIGS. 6 and 7 having the same event tree.

Next, the real-time accident prediction system 200 stores the number of events calculated as a result of performing the comprehensive safety analysis for the facility, a list of possible accidents, an individual frequency of the accident list, and an ending category (S540).

In this way, a series of processes of analyzing the risk of the facility and the number of incidents may be analyzed and updated in real time, or may be updated according to a preset scheduling.

On the other hand, the real-time accident prediction system 200 may collect the result of comprehensive safety analysis for a facility to build big data, and learn big data built using artificial intelligence (AI).

Accordingly, the real-time accident prediction system 200 may analyze hazard factors and derive risk-based events through artificial intelligence (AI), and more specifically and accurately analyze the event occurrence.

Next, the real-time accident prediction system 200 collects measurement values from one or more IoT sensors located at a predetermined location (S550).

In addition, when the measured value is included in the abnormal range, the real-time accident prediction system 200 changes individual frequency values or ending values of accident lists based on the number of events (S560).

The real-time accident prediction system 200 may change an ending value, which is a value related to the severity of the ending predicted according to the occurrence of the accident, in addition to the individual frequency of the accident, in the accident list stored in advance based on the event tree linked to the IoT sensor.

At this time, the real-time accident prediction system 200 presets individual frequency values according to the sequence of IoT sensors described in the linked event tree, and when detecting the corresponding IoT sensor signal, the individual frequency value of the accident list is set as a preset individual frequency value. You can change it.

Alternatively, the real-time accident prediction system 200 continuously calculates the actual individual frequency from the signal value measured from the IoT sensor through artificial intelligence (AI) using big data including the climate, the characteristics of the facility, and the record at the previous time point. Calculate as.

Alternatively, the real-time accident prediction system 200 may change an individual frequency value by using a probability value calculated in real time in a probabilistic safety analysis (PSA) as a signal value.

In addition, the real-time accident prediction system 200 may change the individual frequency value of the accident list to a frequency value corresponding to the calculated actual individual frequency.

These individual frequency values can be easily designed and changed by the administrator in the future.

Meanwhile, the real-time accident prediction system 200 may change the ending category (ending value) when the individual frequency value is changed according to the number of events and the predicted damage becomes severe as the accident progresses. The change of the ending category can also be changed by using a preset value or by calculating the damage scale and damage degree through artificial intelligence using big data, in the same way as the method of changing the individual frequency values described above.

In addition, the real-time accident prediction system 200 may estimate a probability of occurrence of an overlapping accident within a short distance having a distance less than or equal to a threshold value from the predicted accident location. At this time, if the possibility of an overlapping accident is confirmed, the ending category for the predicted accident can be changed.

The real-time accident prediction system 200 may increase the ending value of the ending category in response to a higher probability of occurrence of an overlapping accident.

Next, the real-time accident prediction system 200 calculates a hazard index using an individual frequency value and an ending category (S570).

The real-time accident prediction system 200 updates the changed data in the event tree and multiplies the individual frequency value and the ending value to calculate a hazard index. In this case, the real-time accident prediction system 200 may calculate a hazard index, which is a value obtained by multiplying the individual frequency value and the ending value, through a category consisting of a matrix structure in which the individual frequency value and the ending value are constant.

In addition, the real-time accident prediction system 200 calculates the priority for the accident list having the detected hazard index in the facility (S580).

The real-time accident prediction system 200 may select a predicted accident that requires confirmation first for a comprehensive determination as a priority.

In this case, when selecting a priority, the real-time accident prediction system 200 may preferentially select a case having a high risk index as a priority. At this time, if the risk index is the same, the probability category and the ending category are compared, and those with a lower ending category can be selected as a subordinate priority.

For example, when it occurs among two predicted accidents with the same risk index, the predicted accident with a larger risk can be selected as a priority.

When these risk indices are the same, the composition applying the priority can be applied in the same way even when an accident occurs.

Next, the real-time accident prediction system 200 provides a list of predicted accidents rearranged in the order of high priority and performs a step-by-step alarm (S590).

The real-time accident prediction system 200 calculates the priority of the accident list in real time based on the hazard index and provides a list of predicted accidents rearranged in the order of high priority in real time.

At this time, in providing a list of predicted accidents, when the relevant terminal 300 selects each predicted accident list, the accident history including the name, location, time and content of the predicted accident, contact information for the accident response team, and corresponding to the expected accident Countermeasures, urgency of processing, and surrounding CCTV images can be provided.

Meanwhile, the real-time accident prediction system 200 may set a response step for a predicted accident based on a hazard index and a priority, and perform a step-by-step alarm.

Here, the response step includes a new entry, standby, monitoring, possible accident, accident, immediate response or additional response, and one response step may be set according to each expected accident. In addition, the real-time accident prediction system 200 may extract a notification message configuration corresponding to the response step and transmit it to the related terminal 300.

At this time, the real-time accident prediction system 200 includes notification settings such as a notification signal, lighting of a specific color, and blinking in the notification message in the case of a high-level response stage such as an accident possibility, accident, or immediate response according to the corresponding response stage. I can make it.

In addition, the real-time accident prediction system 200 sets up a communication line directly with external organizations such as fire departments, military units, police officers, government offices, hospitals, as well as internal accident response teams, officials, and managers, and sends a notification message to enable emergency response. Can be transmitted.

Meanwhile, in the notification message, one or more of the IoT sensor measurement value, the CCTV image in the facility, and the corresponding step-by-step alarm may be visualized in real time on a 3D model generated based on spatial information of the facility and provided.

Hereinafter, a process of changing the probability of occurrence (individual frequency) of the real-time accident prediction system 200 according to a specific accident situation will be described using FIGS. 8 to 10.

FIG. 8 is an exemplary view showing an accident procedure for a cause of a fire according to an embodiment of the present invention, and FIG. 9 is an exemplary view showing an accident procedure for a cause of a short circuit and an electric fire according to an exemplary embodiment of the present invention. . 10 is an exemplary view showing an accident related to radioactive leakage according to an embodiment of the present invention.

8 shows a scenario corresponding to the same reference event as in FIGS. 6 and 7. 8 shows the accident situation in which a fire occurred in a freeze prevention facility.

First, No. 01-1 of FIG. 8 relates to the accident situation in case of a fire due to overheating of the hot wire for freeze protection, and the heater was operated at a temperature below zero, but overheating in the hot wire due to output abnormality due to a heater failure or aging of the hot wire. Occurred, and as a result, the thermal insulation material surrounding the heating wire rose above the ignition point, resulting in a fire.

No. 01-1 of FIG. 8 changes the likelihood of occurrence (individual frequency value) by applying a value measured from an IoT sensor for each event tree to an individual frequency value, as shown in FIG. 6.

On the other hand, N0.01-2 of FIG. 8 relates to a fire caused by overheating of the hot wire for freeze protection, but shows an accident situation for a fire caused by a different cause than No. 01-1.

In No. 01-2 of FIG. 8, the heater should not operate at the temperature of the image, but unnecessary output occurred due to a heater failure or a short circuit, and a fire occurred after overheating due to a short circuit or overcurrent in the heater itself.

At this time, event trees 2 and 3 of No. 01-1 in FIG. 8 and event trees 2 and 3 of No. 01-2 share the same event tree, so that the results can be shared. In addition, by confirming that the external temperature is an image from the IoT sensor, it is possible to estimate whether the cause of the accident is a heater leakage.

In other words, when the real-time accident prediction system 200 calculates the likelihood of occurrence for No.01-1, even if the likelihood of occurrence is not separately calculated for No.01-2, the previously calculated change of the likelihood of occurrence is shared and applied. can do. On the other hand, No. 02-1 of FIG. 9 shows the accidents of a short circuit and electric fire due to window breakage and flooding due to strong rain and wind, and No. 02-2 shows a short circuit and electric fire due to flooding due to pipe breakage. .

In No. 02-1 of FIG. 9, a flooded area occurred in the interior due to the influence of strong rain and wind in the external environment, and a current leakage occurred in the flooded area, resulting in an abnormal output of the distribution board current, resulting in a fire due to a short circuit.

On the other hand, No. 02-2 of Fig. 9 detects whether there is an abnormality in the flow rate flowing in the pipe in a situation where flooding has occurred, and if there is no change in the flow rate, the expected result is derived that this is not caused by a pipe break even if a flooding accident occurs Show that you can.

In addition, even if a fire occurs somewhere, it can be confirmed that flooding due to pipe breakage, leakage due to flooding, and fire around the pipe did not occur.

As shown in FIG. 9, the real-time accident prediction system 200 is not limited to the measured value of the IoT sensor inside and outside the facility, and may collect further environmental information by accessing a server, blog, or site of another institution through a network.

At this time, since event trees 2, 3, and 4 of No. 02-1 in Fig. 9 and event trees 2, 3, and 4 of No. 02-2 share the same event tree, the results can be shared.

Next, No. 03-1 of FIG. 10 shows the accidents related to the occurrence of exposure to workers and provision of mitigation countermeasures due to the leakage of radioactive substances in the area where the radioactivity meter is not installed. Indicate the history of the accident.

No. 03-1 and No. 03-2 of FIG. 10 diagnose the possibility of leakage of radioactive materials due to physical, chemical, and natural disasters in the area to be measured, respectively, and estimate the rate of increase by confirming the increase or decrease of the surrounding radioactivity level. , A radioactive material with a rate of increase of more than half of the standard value was leaked.

No. 03-1 and No. 03-2 in Fig. 10 show the accident situation with the same result as a radioactive material, although the area in which the generated area is different.

In the same manner as described above, event lists 2 and 3 in FIG. 10 can be shared because the same accident status and change value of the probability of occurrence are applied.

And since the final probability values of No. 03-1 and No. 03-2 of FIG. 10 are the same, and have the same result, the provided countermeasures may be the same.

For example, the real-time accident prediction system 200 estimates a predicted contaminated area by collecting real-time measurement values by inputting a radioactivity meter formed in a fixed or mobile type. In addition, the real-time accident prediction system 200 may transmit the pollution status or the location of the worker to the control room, and transmit data including the expected evacuation route to the terminal of the worker.

In addition, the real-time accident prediction system 200 may blink the LED lamp of the predicted evacuation route in white or the LED lamp of the inaccessible area in red. Accordingly, the degree of contamination of the work area can be visually lit so that workers can quickly and easily recognize it.

For example, it is possible to control the color of the LED lighting installed in the area according to the expected level of contamination in the order of blue, green, yellow, orange, and red, or the confirmed area is lit and the expected contamination area blinks.

Hereinafter, a configuration for setting a priority by calculating a hazard index of the real-time accident prediction system 200 will be described using FIGS. 11 and 12.

FIG. 11 is an exemplary diagram for explaining the priority varying according to individual frequency values in a real-time predicted accident list according to an embodiment of the present invention, and FIG. 12 is an ending in a real-time predicted accident list according to an embodiment of the present invention. This is an exemplary diagram for explaining the priority that changes due to an increase in the value of a category.

First, referring to FIG. 11, with a list of predicted accidents in building A, priority consisting of priority (Rank), probability category (L), ending category (C), risk index (RI=L*C), and remarks. You can check the list.

The real-time accident prediction system 200 calculates a hazard index by multiplying the occurrence probability category and the ending category, and sets priorities for predicted events that require confirmation first through a comprehensive judgment.

In Fig. 11A is the first time point, B is the second time point, and C is the third time point.As the value of the probability category for fire due to overheating of the hot wire for preventing pipe freezing increases in chronological order (incident probability increases ) It can be seen that the calculated risk index increases.

For example, at the second time point (B), if only an abnormality in the current value is detected due to an abnormality in the temperature sensor and the distribution board signal, at the third time point (C), overheating of the temperature sensor is detected at a specific point of the heating wire. It can be seen that the index increases.

At this time, the real-time accident prediction system 200, as in the third time point (C), if it is seen that there is a possibility of an accident, the accident history including the expected accident name, location, time and content, and corresponding to the expected accident Including basic countermeasures, additional countermeasures, urgency and priority of processing, and surrounding CCTVs, a notification message can be automatically transmitted to the terminal 300 concerned with the incident response team.

On the other hand, FIG. 12 shows that the calculated risk index increases as the value of the probability category increases at the first time point (A) and the second time point (B), and then discovers the possibility of a near-distance accident at the third time point (C). This indicates a situation in which the value of the ending category is elevated (increased severity of the accident).

For example, at the second point in time (B), only an abnormality in the current value is detected due to an abnormality in the temperature sensor and the distribution panel signal, and thus the index has risen. At the third point in time (C), the temperature sensor It can be seen that overheating of is detected and the value of the ending category is increased.

In other words, the value of the probability category did not increase because overheating of the temperature sensor at the direct location was not detected in the list of expected accidents, but the overheating of the temperature sensor was detected at a point located at a close distance, and the risk will rise sharply in the event of an accident. Is estimated.

As such, the real-time accident prediction system 200 estimates the possibility of expanding to a larger accident due to events sharing the same tree of events according to the accident situation, and when the severity of the accident is deepened according to the estimated probability, the ending category (Ending value) can be changed. In addition, the real-time accident prediction system 200 may recalculate and provide a hazard index based on the changed ending category.

On the other hand, in response to a near overlapping accident, a notification message requesting an immediate response to the predicted accident may be transmitted to the terminal 300 concerned with the accident response team.

13 is an exemplary diagram for explaining a configuration for visualizing and providing environment or state information according to accident prediction information according to an embodiment of the present invention.

13(a) is an exemplary diagram visualizing the environment of a facility, and (b) is an exemplary diagram visualizing status information.

As shown in FIG. 13, the real-time accident prediction system 200 interlocks various IoT information on a 3D model generated based on spatial information of an actual facility to enable security and environmental control in a 3D virtual space. ) Can be provided.

In addition, in providing information in a 3D virtual space, the real-time accident prediction system 200 provides a 3D model interface in which a real-life 3D model can freely zoom in, zoom out, and rotate, and is linked with not only IoT sensor measurement values, but also CCTV images in the facility. Can be provided.

As such, the real-time accident prediction system 200 may provide information on the sequence according to the event tree through a 3D virtual space. On the other hand, this real-time accident prediction system 200 continuously collects data before or when an accident occurs to build big data, and iteratively learns and applies using artificial intelligence (AI) to improve accuracy of accident prediction and response measures. Reliability can be improved.

As described above, the dynamic comprehensive safety analysis and the real-time accident prediction system using the Internet of Things accurately analyzes and predicts real-time facility conditions, thereby providing a preemptive and sophisticated countermeasure capable of complex responses to abnormality detection.

A program for executing a method according to an embodiment of the present invention may be recorded on a computer-readable recording medium.

The computer-readable medium may include program instructions, data files, data structures, and the like alone or in combination. The medium may be specially designed and configured, or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magnetic-optical media such as floppy disks, and ROM, RAM, flash memory, and the like. Hardware devices specifically configured to store and execute the same program instructions are included. Here, the medium may be a transmission medium such as an optical or metal wire, a waveguide, etc. including a carrier wave that transmits a signal specifying a program command, a data structure, and the like. Examples of the program instructions include not only machine language codes such as those produced by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

Although the preferred embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the present invention defined in the following claims are also present. It belongs to the scope of rights of

100: IoT sensor 200: real-time accident prediction system
210: communication unit 220: analysis unit
230: control unit 240: notification unit
250: ISA analysis unit 251: risk analysis unit
252: Reference event selection department 253: Incident tree development department
254: Core Designated Definition Department 300: Related Terminal
400: database

Claims (17)

A communication unit that collects signal values measured from one or more IoT sensors located inside or outside the facility,
If the measured signal value is in the abnormal range, based on the event tree associated with the IoT sensor, for an individual accident related to the signal in the abnormal range in a pre-stored accident list, set according to the probability of occurrence of an accident. An analysis unit that changes one or more of the set ending values according to the individual frequency value and the severity of the ending due to the occurrence of the accident, and calculates a risk index using the changed individual frequency value, and
A control unit that calculates a priority for an accident list based on the risk index, and provides a list of predicted accidents rearranged in the order of higher priority in real time,
Real-time accident prediction system comprising a. In claim 1,
The ISA analysis unit further develops the event tree through Integrated Safety Analysis (ISA) for the facility, and calculates the list of possible accidents and the frequency of occurrence and ending category for individual accidents in the accident list. Real-time accident prediction system including.
In paragraph 2,
The ISA analysis unit,
A risk analysis unit that identifies hazards to the above facilities and calculates possible expected events, frequency of occurrence and ending category for individual events,
A reference event selection unit for configuring the accident list by selecting one or more risk-based events from among the predicted events based on the result of the risk analysis unit,
An event tree development department that analyzes the risk-based events to generate the event tree, and
A key point definition unit that selects an IoT sensor that determines whether the event tree has occurred, specifies and provides an installation point, and connects the event tree to the IoT sensor,
Real-time accident prediction system comprising a.
In claim 1,
The analysis unit
Real-time accident prediction in which the individual frequency value is preset according to the sequence of the IoT sensors described in the linked event tree, and when the corresponding IoT sensor signal is detected, the individual frequency value of the accident list is changed with the preset individual frequency value. system.
In claim 1,
The analysis unit
The actual individual frequency is continuously calculated from the signal value measured from the IoT sensor through artificial intelligence (AI) using big data including the climate, the characteristics of the facility and the record at the previous time point, and the calculated actual A real-time accident prediction system that changes the individual frequency value of the accident list to a frequency value corresponding to the individual frequency.
In claim 1,
The analysis unit,
Calculating the hazard index, which is a value obtained by multiplying the individual frequency value and the ending value through a category consisting of a matrix structure in which the individual frequency value and the ending value are constant,
A real-time accident prediction system in which the hazard index calculated for each event and the individual frequency values and ending values as the basis are stored by time.
In claim 1,
The analysis unit,
When the individual frequency value is changed according to the IoT sensor signal value, and the severity of the accident is intensified according to the progress of the accident based on the linked event tree, the ending value is changed, and the changed ending value is used. A real-time accident prediction system that calculates a hazard index.
In claim 1,
The analysis unit,
A real-time accident prediction system for estimating a probability of occurrence of an overlapping accident in a short distance having a distance less than a threshold value from the predicted accident location, and changing the ending value for the predicted accident when the possibility of the overlapping accident occurs.
In claim 1,
A real-time accident prediction system further comprising a notification unit configured to set a response step for the predicted accident based on the hazard index and the priority, and transmit a notification corresponding to the response step.
In claim 9,
The notification includes a notification message,
The notification message includes IoT sensor measurement values, possible accident details, accident prevention measures, checklist, expected accident location, time or contents according to the expected accident situation, urgency to deal with, priority, CCTV video around the expected accident location, contact information , Real-time accident prediction system including one or more data in response to the response step from among the response organizations.
Collecting measurement values from one or more IoT sensors located inside or outside the facility at a predetermined location,
If the measured value is included in the abnormal range, for each accident related to the signal of the abnormal range in the accident list stored in advance based on the event tree linked to the IoT sensor, the individual frequency value and accident set according to the probability of occurrence of the accident Changing one or more of the set ending values according to the severity of the ending according to the occurrence,
Updating data changed according to the number of events and calculating a hazard index using the individual frequency value and the ending value,
Calculating a priority for the accident list in real time based on the hazard index, and
Accident prediction method of a real-time accident prediction system comprising the step of providing in real time a list of predicted accidents rearranged in the order of the highest priority.
In clause 11,
Identifying and analyzing hazards to the facility, and calculating possible expected events, frequency of occurrence and ending category for individual events,
Constructing the accident list by selecting one or more risk-based events among the predicted events based on the calculation result, and generating the event tree by analyzing the risk-based events,
Selecting an IoT sensor that determines whether the event tree has occurred, designating an installation point and providing it, and linking the event tree to the IoT sensor, and
The accident prediction method of a real-time accident prediction system further comprising the step of storing the event tree, the list of possible accidents, the occurrence frequency and the ending category.
In clause 11,
When two or more accidents share the same event tree, the accident prediction method of a real-time accident prediction system that shares the changed individual frequency values linked from the same event tree.
In clause 11,
If you change the individual frequency values in the accident list above,
A signal value measured from the IoT sensor using a preset individual frequency value according to the sequence of the IoT sensors described in the linked event tree, or using big data including climate, characteristics of the facility, and records at a previous time point Accident prediction method of a real-time accident prediction system that continuously calculates the actual individual frequency from and changes the individual frequency value of the accident list.
In clause 11,
The step of calculating the hazard index,
An accident prediction method of a real-time accident prediction system for calculating the hazard index, which is a value obtained by multiplying the individual frequency value and the ending value through a category consisting of a matrix structure in which the individual frequency value and the ending value are constant.
In clause 11,
If you change the ending value of the accident list,
When the individual frequency value is changed according to the number of events and the accident progresses, the ending value is changed, or
An accident prediction method of a real-time accident prediction system that changes the ending category when the individual frequency value is changed according to the IoT sensor signal value and the severity of the accident is intensified according to the progress of the accident based on the linked event tree .
In clause 11,
The accident prediction method of a real-time accident prediction system, further comprising the step of setting a response step for the predicted accident based on the hazard index and the priority, and transmitting a notification corresponding to the response step.

Images