KR102475148B1 - ESS Fire Prevention System and the method thereof - Google Patents

Abstract

The present invention relates to an ESS fire prevention system, and more particularly, to a cloud-based ESS integrated safety management system using a cloud system, an artificial intelligence system, or a BLE 5.0-based micro-wireless sensor, a BLE 5.0-based micro-wireless sensor module system, and It is about an ESS fire prevention system that can provide an AI-based Safety ESS Watchdog System.

Description

ESS fire prevention system and method thereof {ESS Fire Prevention System and the method thereof}

The present invention relates to an ESS fire prevention system, and more particularly, to a cloud-based ESS integrated safety management system using a cloud system, an artificial intelligence system, or a BLE 5.0-based micro-wireless sensor, a BLE 5.0-based micro-wireless sensor module system, and It is about an ESS fire prevention system that can provide an AI-based Safety ESS Watchdog System.

Energy storage refers to the storage of energy using a device or physical medium. The device used for this is called an accumulator, and the entire system in a wider range is called an "Energy Storage System (ESS)".

Batteries used in general households or small batteries used in electronic products can convert and store electrical energy into other forms of energy, but these small-scale power storage devices are not called ESS, and are generally single systems that store hundreds of kWh or more of power. is called "ESS". ESS can be installed and operated in power generation, transmission and distribution, and consumers in the power system, and provides various functions such as frequency regulation, renewable generator output stabilization, peak shaving, load leveling, and emergency power. used as a function. ESS can contribute to improving energy use efficiency, enhancing the utilization of new and renewable energy, and stabilizing the power supply system by storing electric energy when it is used less and supplying it when needed.

Currently, fires frequently occur in the energy storage system (ESS) due to causes such as pressure, overheating, vibration, shock, short circuit, overcharge, and discharge. The situation is.

The recent large-scale ESS fires are hindering the growth of the new energy industry, and a small wireless sensor-based edge safety ESS system is needed.

Therefore, it is necessary to apply edge computing technology to prevent large-scale ESS fires at the edge by predicting, detecting, and responding quickly to prevent them from expanding into larger accidents.

In particular, it is possible to block ESS fires and explosions at the source by collecting major fire-causing factors (off gas, humidity, temperature, smoke, vibration) in ESS through a micro-wireless sensor module and detecting pre-fire signs such as battery thermal runaway through deep learning analysis. There is a need to commercialize technologies/products that can

Korean Registered Patent No. 1039133 Korean Registered Patent No. 2045489 Korea Patent No. 2045871

The present invention has been made to solve the above problems, an edge computing-based safety ESS watchdog system, a sensor data analysis algorithm for predicting and diagnosing a deep learning-based ESS fire, and a cloud-based system for monitoring and managing the operation status of multiple ESSs at a distance. In relation to the ESS (Energy Storage System) fire explosion accidents that have recently frequently occurred using the ESS integrated monitoring system, the fire occurrence factors occurring inside the ESS are monitored in real time using ultra-small wireless sensor technology, and in particular, deep learning-based sensor data The purpose is to provide a system that can predict/detect/emergency control of fire and explosion accidents in advance through correlation analysis.

In particular, a magnet-attached sensor module incorporating BLE 5.0-based subminiature sensor technology that can detect off-gas early and detect temperature, humidity, vibration, and smoke in about 10 minutes of golden time from thermal runaway to fire. The purpose is to provide an ESS integrated safety management integrated solution that can prevent ESS fire accidents in advance through ESS operation emergency control through development and real-time ESS internal environment analysis/prediction.

To this end, a BLE 5.0-based low-power, ultra-small wireless sensor module capable of monitoring/monitoring the inside of the ESS in real time and detecting the cause of a fire in advance is manufactured using non-combustible materials, and an edge ( Edge) plans to design/develop a Safety ESS Watchdog System using computing technology, and aims to provide a cloud-based ESS integrated safety management system for real-time integrated monitoring/safety management of multiple ESSs at a distance.

In order to solve the above problems, the present invention provides a battery pack, a sensor node, and an ESS composed of a power conversion system that converts electrical characteristics to receive and store or discharge power from the battery pack; An Ai edge controller that verifies, monitors, manages, and controls the ESS; and the Ai edge controller includes a VEE function unit for validating the data of the sensor node, an automatic operation scheduler function of the ESS, and an alarm notification function for realization. It consists of an alarm notification and control unit, and a high-speed edge Rules Engine Service unit for monitoring and controlling the real-time operation status of the ESS.

The sensor node applies a BLE 5.0-based low-power wireless communication method, so there is no need to supply constant power and build a separate communication line, and it is attached to a magnet.

The sensor node may include a temperature sensor for detecting heat generation of a battery of the ESS; a humidity sensor for detecting condensation of the ESS; a smoke sensor that detects the occurrence of a fire in the ESS in advance; and a vibration sensor for pre-sensing physical shock and vibration of the ESS.

The Ai edge controller analyzes, predicts, and alarms the data of the sensor node based on artificial intelligence.

The present invention generates sensor data including off gas, temperature, smoke, humidity, and vibration information using a BLE 5.0-based subminiature wireless sensor module in an ESS; In order to collect and detect major fire-causing elements in the ESS with function definition, cloud-based integrated DB implementation, and Device-Object structure for real-time integrated management of a plurality of remote ESSs, the system, status, statistics, and Implementation of emergency control function based on settings, integrated monitoring and individual monitoring steps; Pre-fire signs, including battery thermal runaway in the ESS, are learned through deep learning analysis, and the LSTM algorithm for correcting learning errors of the RNN technique is applied to analyze, predict, and in case of emergency according to the analysis and prediction of the sensor data alarm step; includes.

The real-time storage and management step through the cloud-based ESS integrated safety management system further includes a data validation step of the sensor node through the VEE function unit.

The real-time storage and management step through the cloud-based ESS integrated safety management system includes the step of implementing the automatic operation scheduler function and alarm notification function of the ESS through an alarm notification and control unit.

The real-time storage and management step through the cloud-based ESS integrated safety management system includes the real-time operation state monitoring and control step of the ESS through the high-speed edge Rules Engine Service unit.

The present invention made as described above can provide an edge safety ESS system based on a miniature wireless sensor for preventing fire accidents in an industrial ESS.

In addition, according to the present invention, it is possible to secure an emergency preventive action time (about 10 minutes) according to off-gas detection in the early stage of thermal runaway.

In addition, the present invention can secure emergency response time and prevent fire by predicting fire occurrence and detecting fire in advance through deep learning-based correlation analysis on ESS internal environmental data.

In addition, the present invention can provide a high-speed control function of the ESS operation state at the edge in an emergency through the application of edge computing technology.

In addition, the present invention can prevent fire and minimize damage through prior detection and prediction of fire spread before thermal runaway / before ignition / after ignition, and emergency measures.

1 is a diagram showing a conceptual diagram of a cloud-based ESS integrated safety management system according to the present invention.
2 is a diagram showing the overall configuration of an ESS fire prevention system according to an embodiment of the present invention.
3 is a diagram specifically showing an Ai edge controller of an ESS fire prevention system according to an embodiment of the present invention.
4 is a diagram showing BLE 5.0 based subminiature wireless sensor modules according to an embodiment of the present invention.
Figure 5 is a view showing the attachment position of the monitoring sensor of the battery rack / room according to another embodiment of the present invention.
6 is a diagram showing an Ai edge controller SW block according to an embodiment of the present invention.
7 is a diagram showing alarm/notification/alert steps for each step according to the present invention.
8 is a diagram showing conditions in which an alarm is generated when a sensor threshold condition for each alarm is met according to the present invention.
9 is a diagram showing manager notification, system stop, site alert functions, etc. according to the present invention.

In order to fully understand the present invention, preferred embodiments of the present invention will be described with reference to the accompanying drawings. Embodiments of the present invention may be modified in various forms, and the scope of the present invention should not be construed as being limited to the examples described in detail below. This embodiment is provided to more completely explain the present invention to those skilled in the art. Therefore, the shapes of elements in the drawings may be exaggerated to emphasize a clearer explanation. It should be noted that in each drawing, the same members are sometimes indicated by the same reference numerals. In addition, detailed descriptions of well-known functions and configurations that may unnecessarily obscure the subject matter of the present invention are omitted.

As shown in FIGS. 2 and 3 , the ESS 10 according to the present invention is composed of an Ai edge controller 20 and a third party interface 30 .

In addition, the Ai edge controller 20 includes a communication unit 21, a validation, estimation, editing (VEE) function unit 22, an aggregator 23, a DB 24, an alarm notification and control unit 25, an artificial It consists of an intelligence unit 26, a high-speed edge Rules Engine Service unit 27, and the like.

Cloud-based ESS integrated safety management system

Referring to (a) cloud-based ESS integrated safety management system conceptual diagram in FIG. 1, the present invention designs and complements the system architecture by reflecting it according to the requirements analysis of the ESS integrated safety management system.

It designs various ESS type setting functions and data maps for integrated management of ESS, and provides a device object (Device-Object) structure for integrated management of multiple ESSs at a distance.

In addition, it implements ESS Type support communication interface, detailed function definition for real-time integrated management of ESS, and cloud-based integrated DB.

Looking at (b) Safety ESS integrated safety management system data flow chart in FIG. 1, when ESS information including ESS ID, name, type, device ID, name, type, etc. is generated through the user WEB and transmitted to the EMS server, the EMS server Save each information and set Device Modbus Map.

For example, the REST (REpresentational State Transfer) API (Application Programming Interface) method used in the present invention transmits ESS ID, Device ID (1), Object (N), and isControl commands, and MQTT (Message Queuing Telemetry Transport) method includes (ACQ)Control/ESS ID/Device ID as Topic, ESS ID, Device ID, and Object (N pieces) as Data.

As shown in FIG. 3, the BLE 5.0 communication gateway main board is designed to be miniaturized to embed the Safety Watchdog System, and communication gateway firmware is also added.

Design basic operations such as Register, Update, Read, Write, Execute, and Discover for device registration, management and service implementation, and information reporting functions.

In addition, it has multiple communication interfaces to support heterogeneous equipment and external platform interworking, communication device failure tolerant interworking function, embedded logging function, and data stream transaction real-time processing function.

The present invention further includes a PKI algorithm security chipset application board for implementing edge device security authentication and server-linked communication security functions.

In addition, the present invention applies an artificial intelligence module for deep learning analysis of fire factor sensor data.

For example, the present invention can provide the AI edge controller SW block 50 to the AI edge controller 20, classify fire occurrence factors, receive sensor data, and perform deep learning analysis in depth.

BLE 5.0-based ultra-small wireless sensor module

As shown in FIG. 4, the present invention includes a BLE 5.0 based subminiature wireless sensor module (five types).

As an embodiment, the present invention includes a temperature, humidity, smoke, vibration, and off-gas sensor unit for pre-detection of fire trigger factors.

The wireless sensors are easy to install and include a magnet (Neodymium) attachable sensor case using a non-combustible material to enable communication and sensing functions to be performed for a certain period of time (more than 5 minutes) in the event of a fire.

The wireless sensors include a BLE 5.0 communication module applying a low-power wireless communication method to secure a long-term battery replacement cycle.

As shown in FIG. 5, the installation place of the miniature wireless sensor module according to the present invention is composed of a battery room monitoring sensor 41, a battery rack monitoring sensor 42, and the like.

The battery room monitoring sensor 41 is attached to the ceiling to monitor the battery room and detects temperature and humidity inside the battery room.

The battery rack monitoring sensor 42 is attached to the upper side or side of the battery rack to monitor the battery rack, and is attached to increase according to the number of parallel arrangements of the battery rack.

As shown in FIG. 6, the Ai edge controller SW block 50 is composed of a SUPPORT SERVICES unit 51, a CORE SERVICES unit 52, and a DEVICE SERVICES unit 53.

Again, the SUPPORT SERVICES unit 51 includes an ML module, an off-gas supervisory module, a vibration supervisory module, a temperature supervisory module, and a Device Auth. It is divided into & Setting module, Database module, Rule Engine module, VEE module, Logging module, Alert & Notification module, Control module, Aggregator module, 3rd Party Interface module, etc.

In addition, the CORE SERVICES unit 52 is divided into a CORE DATA module, a COMMAND module, a META DATA module, and a REGISTRY&CONFIG module.

The DEVICE SERVICES unit 53 is divided into a REST module, a MODBUS module, an MQTT module, and an ADDITIONAL DEVICES module.

The SUPPRT SERVICES part is a support service, 1) ML (Machine Learning) part - ESS fire prediction and detection algorithm through sensor data correlation analysis, 2) Off gas supervisory part - Off gas detection function: thermal runaway / Preliminary detection of ignition / wire sheath heating, 3) Vibration supervisor - Vibration detection function: Physical shock / earthquake detection, 4) Temperature supervisor - Temperature detection function: Heat, electric leakage environment advance Detection / Thermal runaway detection, 5) Humidity supervisor - Humidity detection function: Leakage / ignition condition detection in advance, 6) Smoke supervisor - Smoke (Co2, tvoc) detection function: Thermal runaway / Combustibility before ignition Substance ejection detection, 7) Device authorization and setting part - Device authorization and setting registration and modification function, 8) Database - Facility, device registration information alarm setting information, 9) Rule engine part - SUPPRT SERVICES function command and management, - Control, Alert & Notification by monitoring sensor data, - Aggregator, Supervisory, Logging management, 10) VEE - Sensor data validation (Validation, Estimation, Editing), 11) Logging ) part - log registration: collection data, alarm detection, communication, error data collection, 12) alert & notification part - user notification: alarm and SMS transmission according to alarm, 13) control part - MC (Magnetic Contactor) control according to alarm level, 14) Aggregator part - Data collection and conversion function, 15) Third Party Interface part - Third party application support function have.

The CORE SERVICES part 52 is a device that configures and sets the META DATA generation system through core data storage and analysis, 1) CORE DATA - equipment and environmental sensor collection data, 2) COMMAND - ESS Safety Watchdog System command, 3) META DATA - Data analysis and management purpose produced after CORE DATA analysis, 4) REGISTRY & CONFIG - System configuration and setting information registration function.

The DEVICE SERVICES part (53) is a device that collects and converts data, transmits it to the EMS Server, and manages devices. 1) REST - Data collection and EMS Server data transmission (HTTP Protocol), 2) MODBUS - Facility data collection (PCS, BMS) , PowerMeter), 3) MQTT - Transmission to EMS Server, 4) ADDITIONAL DEVICES - Addition of devices (facility and environment sensor).

Specifically, the Ai edge controller 20 has (1) Safety ESS Watchdog System automatic operation scheduler function, (2) Event and alarm logging and management function, (3) Configure system setpoints (optimized set point) setting function, (4 ) Real time and historic data visualization function, (5) Monitor, Forecast, and Dispatch; It includes generation, demand, and storage functions, (6) Real-Time Stream Data analysis function, etc.

AI-based Safety ESS Watchdog System

The artificial intelligence unit 26 can provide a safety ESS watchdog function through learning using an ESS fire prediction and detection algorithm through deep learning-based sensor data correlation analysis.

For example, it is possible to develop an HMI to check the operation status of the Safety Watchdog System and provide a setting function by applying the LSTM algorithm for correcting learning errors of the RNN technique using data measured in real-time 1 second to 1 minute cycles.

Cloud and BLE 5.0 based ESS fire prevention system

As an embodiment, the ESS fire prevention system according to the present invention includes a Representational State Transfer (REST) method, an Application Programming Interface (API) method, and a Message Queuing Telemetry Transport (MQTT) method.

The MQTT implements a two-way pub/sub messaging service for efficient communication with a plurality of BLE 5.0-based subminiature wireless sensor modules, and there is no restriction on the size of messaging, and interfaces between sensors or ESS devices using different protocols between heterogeneous ESS facilities ESS fire prevention system based on a protocol converter that converts the It consists of applications and web services that can be controlled/monitored.

That is, by checking errors that may occur when collecting data in a cloud environment that has the possibility of causing serious errors in various types of BLE 5.0-based micro wireless sensor modules, and by normalizing data, the stability and operational efficiency of IoT-based communication devices are secured. The system can continue to operate.

The error checking syntax inserted to check the error of the plurality of BLE 5.0 based micro wireless sensor modules is replaced with the error checking syntax according to the judgment result, and according to the result of the substitution, exception handling for errors occurring within the function ) and logging error information according to exception handling, there is an effect of automatically detecting and logging errors.

In addition, the plurality of BLE 5.0-based subminiature wireless sensor modules include a communication interface for supporting heterogeneous communication protocols, a VEE function unit 22 for validating sensor data, and a high-speed edge Rules Engine Service unit for monitoring and controlling real-time ESS operation status ( 27) is additionally connected so that, for example, the VEE function unit 22 can check, evaluate, and correct errors or missing data after processing VEE (Validation, Estimation and Edit) through the VEE module. do.

In addition, the VEE function unit 22 verifies in real time according to a predefined rule equally applied to all data regardless of the type of data generated by the plurality of BLE 5.0 based subminiature wireless sensor modules. Based on these verified results, the VEE module classifies, stores, and manages normal data and data with errors.

BLE 5.0 and artificial intelligence based ESS fire prevention system

The artificial intelligence unit 26 may use an ESS fire prediction and detection algorithm through deep learning-based sensor data correlation analysis.

In other words, off gas sensor, temperature sensor’s thermal runaway phenomenon, electric leakage, heat data, arc generation, electric leakage, and ignition condition preliminary detection data of vibration sensor and humidity sensor, electric leakage, thermal runaway, electric wire sheath heat detection data of smoke sensor ESS fire prediction and detection algorithms are created through data correlation analysis.

Performs RNN (Recurrent Neural Network) model learning in a supervised learning method using data generated in the BLE 5.0-based ultra-small wireless sensor module (see FIG. 4) according to the present invention and measured in real-time 1 second to 1 minute cycles can do.

As an embodiment, the present invention can apply the LSTM algorithm for learning error correction of the RNN technique using data measured by the temperature, humidity, smoke, vibration, and off-gas sensor unit for pre-detection of fire trigger factors.

The wireless sensors include a BLE 5.0 communication module applying a low-power wireless communication method to secure a long battery replacement cycle. A process of classifying in detail (larger value of replacement cycle is larger) may be added.

It is possible to design a fire prediction inference engine through the correlation analysis of the wireless 5-type sensor data, and it is possible to analyze 5-type sensor data and implement a learning algorithm by using an auto-encoder technique.

In particular, after the VEE function unit 22 processes VEE (Validation, Estimation and Edit) through the VEE module, errors in the five types of sensor data are checked, evaluated, and corrected.

In addition, the Ai edge controller 20 verifies in real time according to a predefined rule that is equally applied to all data regardless of data type. Based on these verified results, the VEE module classifies, stores, and manages normal data and data with errors.

In addition to applying the LSTM algorithm for correcting learning errors of the above-described RNN technique, predicted data may be calculated using interpolation based on normal data of the data map.

Specifically, by performing model learning of the Safety ESS Watchdog System automatic operation scheduler according to the present invention, when an error occurs, the schedule is switched to the dual schedule mode so that the local scheduler is operated so that the local scheduler uses the same schedule as the schedule of the ESS fire prevention system. All schedule control signals and virtual control signals from a previous point in time of switching to the dual mode to the point of switching to the schedule duplication mode are transmitted to ESS facilities to ensure normal operation of the ESS facilities.

That is, learning of a recurrent neural network (RNN) model may be performed in a supervised learning method using the schedule control signal and the virtual control signal transmitted in real time.

In addition, model learning can be performed with Event Log Analysis Software, monitoring tools, searching engine, etc. collected through the aggregator 23 and the like.

In this case, a Recurrent Neural Network (RNN) model may be trained in a supervised learning method using log data measured in real time.

For example, a supervised learning method allows administrators to create multiple alarm transmission method models, test them, and perform them with optimized models.

To this end, a video camera for generating sensor data is added to the ESS.

는 배터리 룸 내부의 연기, 오프가스 이미지 훈련영상이며,

실제 배터리 룸 내부의 연기, 오프가스 이미지 영상으로 각각 +1 또는 -1의 값을 가진다. Table 1 shows steps for constructing a strong classifier by linearly combining weak classifiers using the Adaboost algorithm.

is a training video of smoke and off-gas images inside the battery room,

It is an image of smoke and off-gas inside the actual battery room, and has a value of +1 or -1, respectively.

가중치 오차를 최소화하는 약한 분류기

를 생성하여 각 분류기에 대한 센서 데이터 가중치를 갱신한다.

는 모든 데이터

에 대하여 수식을 정규화한다.and

A weak classifier that minimizes weight errors

to update the sensor data weights for each classifier.

is all data

Normalize the expression for

는 최종적으로 생성된 강한 분류기이다. 강한 분류기는 벡터

를 입력으로 받으며 훈련된 약한 분류기

의 합으로 센서 데이터 분류를 수행하게 된다.

is the finally generated strong classifier. A strong classifier is a vector

Weak classifier trained taking as input

The sensor data classification is performed by the sum of .

For example, the algorithm is used to express local texture features of smoke or gas, such as edges, dots, and corners, using pixel values of local binary pattern features used to detect smoke and off-gas images inside a battery room. It has the advantages of high processing speed with simple calculation, discrimination, and robustness to changes in lighting.

The local binary pattern feature of the sensor data is expressed as in Equation 1, and the difference between the pixel value at the current location and the neighboring pixel value is expressed as values of 0 and 1.

와

는 중심화소의 화소값 및 이웃화소의 화소값을 의미한다.In Equation 1, P and R mean the number of adjacent pixels and the radius of the circle of the smoke and offgas images, respectively,

Wow

denotes a pixel value of a central pixel and a pixel value of neighboring pixels.

For example, in the case of LBP(8, 1), the pixel values of 8 neighboring pixels are compared based on the center pixel value of the smoke or off-gas image. It can be obtained by bit encoding.

Thereafter, the pixel value image is analyzed, converted into text, and trained through deep learning to predict and detect fire occurrence in advance to secure emergency response time and prevent fire.

In addition, by the above-described configuration, the portion consisting only of the image can be converted into textual text, and an alarm can be specified by specifying the fire occurrence prediction step 1 and step 2.

For example, it is possible to predict the timing of fire occurrence by dividing it into fire prediction steps 1-1, 1-2 steps, etc. and analyzing it with a linear regression analysis model.

Alternatively, the Ai edge controller 20 may calculate fire prediction step 1-2 prediction data by interpolation based on normal data of a data map expressing pixel values of fire prediction step 1-1.

Therefore, even beginners can quickly respond to an emergency without judging by looking at the image.

Hereinafter, a fire prevention method for implementing the present invention will be described in detail.

The present invention includes (1) generating sensor data including off gas, temperature, smoke, and vibration information using a BLE 5.0-based ultra-small wireless sensor module in the ESS; (2) Function definition for real-time integrated management of a plurality of remote ESSs, cloud-based integrated DB implementation and Device-Object structure, to collect and detect major fire-causing elements in the ESS, system, status, Emergency control function implementation, integrated monitoring and individual monitoring steps based on statistics and settings; (3) The pre-fire signs including battery thermal runaway in the ESS are learned through deep learning analysis, and the LSTM algorithm for correcting learning errors of the RNN technique is applied to analyze, predict, and analyze and predict the sensor data. An alarm step in an emergency situation; includes.

For example, as shown in FIG. 7 , a notification through a mobile alarm 62 and an alarm 61 display stage in case of an emergency at each stage may be used.

Additionally, in the present invention, the real-time storage and management step through the cloud-based ESS integrated safety management system includes the data validation step of the sensor node 14 through the VEE (Validation, Estimation, Editing) function unit 22, alarm notification and It consists of implementing the automatic operation scheduler function of the ESS 10 through the control unit 25 and monitoring and controlling the real-time operation state of the ESS 10 through the high-speed edge Rules Engine Service unit 27.

As shown in FIG. 8 , in the present invention, an alarm is generated when a sensor threshold condition for each alarm is met, and is released when not applicable.

For example, it is divided into 4 categories including normal, and alarms are classified into 3 colors (separate display) of Occurrence/Danger/Caution.

The normal standard was based on the recommended environment for ESS operation including secondary batteries (lithium ion batteries).

As shown in FIG. 9, the present invention provides manager notification (Web alert notification, SMS transmission), system stop (PCS system stop), site alert (operation of warning light at the top of the ESS container), electrical system cutoff (blocking the PCS system of the emergency control panel) ), fire suppression (emergency control panel automatic fire extinguishing device operation) function.

Therefore, when a specific power plant is clicked on the integrated monitoring page, it moves to the corresponding page, and today's/cumulative power generation, transmission, charge, and discharge of the currently entered power plant can be summarized and displayed.

That is, the present, today's, cumulative power generation, today's charge/discharge, and SOC are displayed, and the flow can be confirmed on the line of the simplified system diagram according to the state.

In addition, summary information of the power plant entered on the integrated monitoring page is displayed, and the current weather information and average temperature and humidity of each sensor are displayed.

For example, the facility communication status of the entered power plant is expressed as communication/communication interruption, today's alarm statuses generated in the entered power plant are expressed, and today's power generation and charge/discharge amount are expressed as hourly charts.

At this time, when the user has a plurality of power plants, he or she may directly move to another power plant through a corresponding button.

As an embodiment, fire prevention and damage can be minimized through preliminary detection and prediction of fire spread before thermal runaway / before ignition / after ignition in the fire occurrence environment and fire occurrence detection mode for each sensor of the present invention, and emergency measures.

For example, the temperature sensor has a function to detect heat and electric leakage environment in advance / thermal runaway, the vibration sensor has a function to detect physical shock / earthquake, the humidity sensor has a function to detect electric leakage / fire condition in advance, and the off gas sensor has a function to detect It has a function to detect thermal runaway / ignition of combustible material ejection, and the smoke sensor has a function to detect thermal runaway / ignition / wire sheath heating in advance.

Terms used in this document are only used to describe a specific embodiment, and may not be intended to limit the scope of other embodiments.

Singular expressions may include plural expressions unless the context clearly dictates otherwise. Terms used herein, including technical or scientific terms, may have the same meaning as commonly understood by a person of ordinary skill in the technical field described in this document.

Among the terms used in this document, the terms defined in general dictionaries may be interpreted as having the same or similar meaning to the meaning in the context of the related art, and unless explicitly defined in this document, in an ideal or excessively formal meaning. not interpreted In some cases, even terms defined in this document cannot be interpreted to exclude the embodiments of this document.

10: ESS
12: battery pack
14: sensor node
16: power conversion system
20: Ai Edge Controller
22: with VEE function
25: alarm notification and control unit
27: High-speed Edge Rules Engine Service Department
40: monitoring sensor
41: battery room monitoring sensor
42: battery rack monitoring sensor
50: Ai Edge Controller SW Block
51: SUPPORT SERVICES part
52: CORE SERVICES part
53: DEVICE SERVICES part

Claims (8)

An ESS (10) composed of a battery pack (12), a sensor node (14), and a power conversion system (16) that converts electrical characteristics in order to receive power from the battery pack (12) and store or discharge power to the system;
An Ai edge controller 20 that verifies, monitors, manages, and controls the ESS 10;
The Ai edge controller 20 analyzes, predicts, and alarms the data of the sensor node 14 based on artificial intelligence,
Among the SW blocks 50 of the Ai edge controller 20, the SUPPORT SERVICES unit 51,
ML (Machine Learning) unit including ESS fire prediction and detection algorithm through sensor data correlation analysis as a support service;
An off-gas supervisory unit that pre-detects thermal runaway, ignition, and electric wire coating heating as an off-gas detection function of the ESS 10;
A vibration supervisor having a physical shock and earthquake sensing function of the ESS 10;
As a temperature sensing function of the ESS 10, a temperature supervisor capable of detecting heat generation, an electric leakage environment in advance, and a thermal runaway phenomenon;
A humidity supervisor capable of detecting electric leakage and ignition conditions in advance as a humidity sensing function of the ESS (10);
Smoke supervisory unit capable of detecting thermal runaway as a function of detecting smoke (Co2, tvoc) / ejection of combustible substances before ignition;
a device authority and setting unit having a function of registering and modifying device authority and setting of the ESS (10);
a database for storing facility and device registration information and alarm setting information of the ESS (10);
As the SUPPRT SERVICES function command and management function of the ESS 10, a control unit, an alert and notification unit, an aggregator unit, a supervisor unit, and a rule engine unit managing a logging unit.
VEE for validating (Validation, Estimation, Editing) the sensor data of the ESS (10);
a logging unit that collects collected data, alarm detection, communication, and error data through log registration of the ESS (10);
An alert and notification unit capable of transmitting an alarm and SMS according to an alert as a user notification of the ESS 10;
A control unit capable of controlling a magnetic contactor (MC) according to the alarm level of the ESS (10);
An aggregator unit having a data collection and conversion function of the ESS (10); and
An ESS fire prevention system consisting of a third party interface (Third Party Interface) unit including a third party application support function, which is an external app according to the ESS (10). According to claim 1,
The Ai edge controller 20 includes a VEE function unit 22 for validating data of the sensor node 14, an alarm notification and control unit 25 for implementing an automatic operation scheduler function of the ESS 10, and the ESS ( 10) ESS fire prevention system characterized in that it is composed of a high-speed edge Rules Engine Service unit 27 for real-time monitoring and control of operating conditions. According to claim 1,
The sensor node 14,
Off gas sensor and temperature sensor for pre-detecting thermal runaway phenomenon, electric leakage, heat in the ESS (10);
a vibration sensor and a humidity sensor for pre-sensing arc occurrence, electric leakage, and ignition conditions in the ESS (10);
a smoke sensor for preliminarily detecting electric leakage, thermal runaway, and heat from electric wire coating in the ESS (10); An ESS fire protection system including one of delete An ESS (10) composed of a battery pack (12), a sensor node (14), and a power conversion system (16) that converts electrical characteristics in order to receive power from the battery pack (12) and store or discharge power to the system; In the ESS fire prevention method using the ESS fire prevention system consisting of; Ai edge controller 20 that verifies, monitors, manages and controls the ESS 10,
Generating sensor data including off gas, temperature, smoke, and vibration information using a BLE 5.0-based ultra-small wireless sensor module in the ESS;
In order to collect and detect major fire-causing elements in the ESS with function definition, cloud-based integrated DB implementation, and Device-Object structure for real-time integrated management of a plurality of remote ESSs, the system, status, statistics, and Implementation of emergency control function based on settings, integrated monitoring and individual monitoring steps;
Pre-fire signs, including battery thermal runaway in the ESS, are learned through deep learning analysis, and the LSTM algorithm for correcting learning errors of the RNN technique is applied to analyze, predict, and in case of emergency according to the analysis and prediction of the sensor data alarm stage;
The real-time storage and management step through the cloud-based ESS integrated safety management system includes a data validation step of the sensor node 14 through the VEE function unit 22 included in the Ai edge controller 20,
Among the SW blocks 50 of the Ai edge controller 20, the SUPPORT SERVICES unit 51,
ML (Machine Learning) unit including ESS fire prediction and detection algorithm through sensor data correlation analysis as a support service;
An off-gas supervisory unit that pre-detects thermal runaway, ignition, and electric wire coating heating as an off-gas detection function of the ESS 10;
A vibration supervisor having a physical shock and earthquake sensing function of the ESS 10;
As a temperature sensing function of the ESS 10, a temperature supervisor capable of detecting heat generation, an electric leakage environment in advance, and a thermal runaway phenomenon;
A humidity supervisor capable of detecting electric leakage and ignition conditions in advance as a humidity sensing function of the ESS (10);
Smoke supervisory unit capable of detecting thermal runaway as a function of detecting smoke (Co2, tvoc) / ejection of combustible substances before ignition;
a device authority and setting unit having a function of registering and modifying device authority and setting of the ESS (10);
a database for storing facility and device registration information and alarm setting information of the ESS (10);
As a SUPPRT SERVICES function command and management function of the ESS (10), sensor data is monitored to control a control unit, an alert and notification unit, an aggregator unit, a supervisor unit, and a rule engine unit managing a logging unit.
VEE for validating (Validation, Estimation, Editing) sensor data of the ESS (10);
a logging unit that collects collected data, alarm detection, communication, and error data through log registration of the ESS (10);
An alert and notification unit capable of transmitting an alarm and SMS according to an alert as a user notification of the ESS 10;
A control unit capable of controlling a magnetic contactor (MC) according to the alarm level of the ESS (10);
An aggregator unit having a data collection and conversion function of the ESS (10); and
ESS fire prevention method comprising a; third party interface (Third Party Interface) unit including a third party application support function, which is an external app according to the ESS (10). delete According to claim 5,
In the real-time storage and management step through the cloud-based ESS integrated safety management system,
An alarm notification included in the Ai edge controller 20 and implementing an automatic operation scheduler function of the ESS 10 through the control unit 25; ESS fire prevention method including. According to claim 5,
In the real-time storage and management step through the cloud-based ESS integrated safety management system,
ESS fire prevention method including; real-time operation state monitoring and control step of the ESS 10 through the high-speed edge Rules Engine Service unit 27 included in the Ai edge controller 20

Images