Intelligent fire-fighting management system based on regional division and model and early warning method thereof

Abstract

The invention discloses an intelligent fire-fighting management system based on region division and a model and an early warning method thereof, and belongs to the technical field of intelligent fire fighting, and the method specifically comprises the steps that a system initialization unit is responsible for recording all equipment information in a monitoring region, subdividing the monitoring region into sub-regions based on equipment distribution, and associating the sub-regions with equipment; the region defining unit identifies the related sub-regions as to-be-confirmed regions when a fire occurs, and marks to-be-confirmed devices; the data acquisition unit collects real-time monitoring data of the equipment, and preliminarily analyzes and judges whether a fire is an electrical fault or not; and if yes, further collecting historical power consumption data. The accident analysis unit normalizes the historical data, constructs a time sequence model, and analyzes the current intensity change; if the real-time current value cluster deviates, a fire disaster caused by abnormal behaviors of the equipment is deduced; otherwise, the aging is caused by the electrical insulating material; according to the invention, the electrical fault is identified through the power consumption data.

Description

Technical Field

[0001] The present invention relates to the field of intelligent fire protection technology, and in particular to an intelligent fire protection management system and an early warning method thereof based on regional division and models. Background Art

[0002] With technological advancements and accelerated urbanization, the number of electrical devices in modern buildings and industrial facilities has increased significantly, leading to a corresponding increase in the risk of electrical failures. Traditional fire safety monitoring methods primarily rely on devices such as smoke detectors and temperature sensors, which often have already undergone significant physical changes or damage by the time they detect a fire. Furthermore, traditional monitoring systems often lack the ability to provide early warning of fires caused by electrical failures, making it difficult to accurately locate and analyze the cause of electrical failures.

[0003] To improve the intelligence and responsiveness of fire safety monitoring, intelligent fire management systems based on regional division and models have emerged in recent years. These systems integrate a variety of monitoring devices, such as energy meters, circuit protection devices, electrical fault warning sensors, intelligent sockets, and electrical fault prevention and control equipment, to comprehensively monitor electrical equipment within the monitoring area. By analyzing and processing monitoring data in real time, the system can promptly detect anomalies and provide early warnings before a fire occurs, effectively reducing the risks and losses associated with fires.

[0004] However, existing intelligent fire safety monitoring systems still have certain limitations in data processing and accident cause analysis. When determining the cause of electrical failures, existing systems mostly rely on simple data comparisons and lack in-depth mining and analysis of historical data and time series, making it difficult to accurately distinguish the cause of fires. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent fire management system and early warning method based on regional division and model, and to solve the following technical problems: When determining the cause of electrical failures, existing systems mostly rely on simple data comparisons and lack in-depth mining and analysis of historical data and time series, making it difficult to accurately distinguish the cause of fires.

[0006] The purpose of the present invention can be achieved through the following technical solutions: An intelligent fire management system based on regional division and model, comprising: The system initialization unit is used to sort out and record the information of all monitoring devices in the monitoring area. The monitoring devices include electricity meters, circuit protection devices, electrical fault warning sensors, intelligent sockets, and electrical fault prevention and control equipment. Based on the distribution of monitoring devices, the entire monitoring area is subdivided into several sub-areas, and each monitoring device is associated with a specific sub-area. The area definition unit is used to identify and mark the sub-area related to the fire event as the area to be confirmed when a fire event occurs, and mark the monitoring equipment in the area to be confirmed as the equipment to be confirmed; The data acquisition unit is used to collect real-time monitoring data of the marked equipment to be confirmed within a specific time window before the fire incident, and perform preliminary processing on the real-time monitoring data. By analyzing the real-time monitoring data, it is determined whether the fire was caused by an electrical fault. If the judgment result is positive, all historical electricity consumption data of the area to be confirmed before the fire occurred will be collected; otherwise, the judgment will be terminated; The accident analysis unit is used to normalize the collected historical electricity consumption data and build a time series model, extract the current value change sequence within a specific time window t before the fire occurs, mark it as a pending current value change sequence, and divide the historical electricity consumption data into several current value change sequences with a duration of t. All current value change sequences are clustered to identify whether the category cluster to which the pending current value change sequence belongs is offset in the cluster. If the offset exceeds a preset threshold, it is inferred that the cause of the electrical fault is the abnormal behavior of the electrical equipment; if it does not exceed the preset threshold, the electrical fault is attributed to the aging problem of the electrical insulation material.

[0007] As a further solution of the present invention: the monitoring parameters of the monitoring equipment include the fundamental wave of the current signal, the amplitude of the current signal, the harmonic components, and the voltage.

[0008] As a further solution of the present invention: the monitoring device calculates and generates power parameters based on the real-time voltage value and the real-time current value, and the power parameters include apparent power, active power, reactive power, and power factor.

[0009] As a further solution of the present invention: the process of the preliminary processing of the data acquisition unit is: Validate the monitoring parameters collected from various monitoring devices, supplement missing values ​​or outliers, convert date and time fields from strings to a unified format, detect and remove duplicate monitoring parameter records, remove outliers and invalid characters in monitoring parameters, and normalize the values ​​of different types of monitoring parameters.

[0010] As a further solution of the present invention: the process of normalization is: Calculate the mean value μ and standard deviation σ of any type of monitoring parameter, and the normalization formula is: ; Where x' represents the normalized monitoring parameter, and x represents the original monitoring parameter.

[0011] As a further solution of the present invention: the process of the data acquisition unit determining whether it is an electrical fault is as follows: Obtain the values ​​of all monitoring parameters and power parameters within a specific time window before the fire event occurs. If the real-time current value amplitude is 0, the fire event is considered not to be an electrical fault; if the real-time current value amplitude exceeds the preset standard range, the fire event is considered to be an electrical fault; if the real-time current value amplitude is within the preset standard range, obtain any real-time monitoring parameter or power parameter. When the value of any real-time monitoring parameter or power parameter exceeds the corresponding preset standard range, the fire event is considered to be an electrical fault.

[0012] As a further solution of the present invention: the process of dividing the current value change sequence by the accident analysis unit is as follows: Obtain the initial time series of historical electricity consumption data, collect the length of the initial time series and mark it as L, obtain the sliding step of the time series and mark it as S, the length of the current value change sequence is t, and the number of splits of the historical electricity consumption data is n=L / St.

[0013] As a further solution of the present invention: a control radius and a minimum similarity number min are set. For a pending current value change sequence, the number m of similar sequences within the control radius is detected. If m is less than the minimum similarity number min, the cause of the electrical fault is directly determined to be abnormal behavior of the electrical equipment. If m is greater than or equal to the minimum similarity number m, the pending current value change sequence is placed in a category cluster, and all current value change sequences are placed in the category cluster in turn and marked as abnormal. Repeat the above process and extract the cluster centers of the category clusters after several clusterings. Compare the number of cluster centers where the current change value sequence is located and the difference in the cluster center values. If the cluster center shifts significantly, it means that the cause of the electrical fault is the abnormal behavior of the electrical equipment.

[0014] As a further solution of the present invention: the formula for determining whether the cluster center is offset is: , ; Among them, S i is a set of data points, |S i ∣ is the number of data points in the class cluster, Δc i Indicates the change value of the i-th cluster center between the previous and next two iterations; c i,t and c i,t+ 1 represents the value of the i-th cluster center at time t and time t+1 respectively. When Δc i When it is greater than the preset threshold, it means that the cluster center has shifted significantly.

[0015] The present invention also includes an intelligent fire safety monitoring and early warning method based on a regional alarm model, comprising the following steps: Organize and record information on all monitoring devices within the monitoring area, including energy meters, circuit protection devices, electrical fault warning sensors, intelligent sockets, and electrical fault prevention and control equipment. Based on the distribution of monitoring devices, the entire monitoring area is subdivided into several sub-areas, with each monitoring device associated with a specific sub-area. When a fire incident occurs, the sub-area related to the fire incident is identified and marked as a pending confirmation area, and the monitoring equipment in the pending confirmation area is marked as a pending confirmation equipment; For the marked equipment to be confirmed, the real-time monitoring data within a specific time window before the fire incident is collected and preliminarily processed. By analyzing the real-time monitoring data, it is determined whether the fire was caused by an electrical fault. If the judgment result is positive, all historical electricity consumption data of the area to be confirmed before the fire occurred is collected; otherwise, the judgment is terminated. The collected historical electricity consumption data is normalized and a time series model is constructed to extract the current value change sequence within a specific time window t before the fire occurs, marked as the pending current value change sequence, and the historical electricity consumption data is divided into several current value change sequences with a duration of t. All current value change sequences are clustered to identify whether the category cluster to which the pending current value change sequence belongs is offset in the cluster. If the offset exceeds a preset threshold, it is inferred that the cause of the electrical fault is the abnormal behavior of the electrical equipment; if it does not exceed the preset threshold, the electrical fault is attributed to the aging problem of the electrical insulation material.

[0016] Beneficial effects of the present invention: The present invention automatically collects and analyzes electricity monitoring data, subdivides the monitoring area into multiple sub-areas, and quickly locks the involved sub-areas and corresponding monitoring equipment when a fire occurs. Through the regular processing and time series analysis of historical and real-time electricity consumption data, as well as the use of cluster analysis, it can effectively identify abnormal behavior of electrical equipment and line insulation aging problems, and can quickly and accurately determine the cause of the fire, especially distinguish between electrical faults and non-electrical faults, reducing the guesswork and manual judgment in traditional investigation methods and improving investigation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention will be further described below with reference to the accompanying drawings.

[0018] Figure 1 This is a module diagram of an intelligent fire management system based on area division and model in the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0020] See also Figure 1 As shown, the present invention is an intelligent fire management system based on regional division and model, including: System initialization unit: This unit is responsible for organizing and recording information from all monitoring devices within the monitoring area. This includes, but is not limited to, energy meters, circuit protection devices (such as circuit breakers and leakage protectors), electrical fault warning sensors, intelligent sockets, and electrical fault prevention and control equipment. Based on the distribution of these devices, the entire monitoring area is divided into several sub-areas, each corresponding to a specific set of monitoring devices. This not only helps to more accurately locate problems but also improves the efficiency of subsequent data processing.

[0021] Area Definition Unit: Once a fire incident occurs, this unit immediately activates its work, first identifying all sub-areas associated with the fire incident and marking them as "pending areas." Simultaneously, all monitoring devices within these pending areas are also marked as "pending devices." This step is crucial for narrowing the scope of the investigation and effectively focusing resources on suspected fault points for in-depth analysis.

[0022] Data acquisition unit: For the previously identified devices, this unit collects real-time monitoring data from a specific time window before the fire and performs preliminary processing on this raw data. By analyzing this data, the system can determine whether the fire was caused by an electrical fault. If the initial determination is positive, further data collection is performed from the start of power use to the present within the area to be confirmed. Otherwise, further data collection ceases.

[0023] Accident Analysis Unit: Once the fire is determined to be caused by an electrical fault, this unit conducts a more detailed analysis of the collected historical electricity usage data. Specifically, it first normalizes the data to ensure that information from different sources can be compared uniformly. Next, it uses a time series model to construct a pattern of current fluctuations, focusing specifically on current fluctuations during the period preceding the fire (assuming T hours). The current curve for this period is called the "undetermined current value change series."

[0024] In addition, the entire data set is divided into multiple small segments of the same length T, each of which represents an independent time series sample. Then, a clustering algorithm is used to classify all the generated time series in order to find patterns that may contain abnormal behavior or trends. The last step is to compare the degree of difference between the series of changes in the current value to be determined and other members of the category to which it belongs. If a significant deviation is found between the two, and this deviation exceeds a pre-set threshold, it can be inferred that the cause of the electrical fault may be due to abnormal operation or damage of some electrical equipment; if there is no obvious deviation, it may mean that the problem is caused by problems such as aging of insulation materials due to long-term use.

[0025] In another preferred embodiment of the present invention, the monitoring parameters of the monitoring device include current signal fundamental wave, current signal amplitude, harmonic components, and voltage.

[0026] In a preferred case of this embodiment, the monitoring device calculates and generates power parameters based on the real-time voltage value and the real-time current value. The power parameters include apparent power, active power, reactive power, and power factor.

[0027] In another preferred embodiment of the present invention, the preliminary processing process of the data acquisition unit is: Validate the monitoring parameters collected from various monitoring devices, supplement missing values ​​or outliers, convert date and time fields from strings to a unified format, detect and remove duplicate monitoring parameter records, remove outliers and invalid characters in monitoring parameters, and normalize the values ​​of different types of monitoring parameters.

[0028] In a preferred embodiment of the present invention, the normalization process is as follows: Calculate the mean value μ and standard deviation σ of any type of monitoring parameter, and the normalization formula is: ; Where x' represents the normalized monitoring parameter, and x represents the original monitoring parameter.

[0029] In another preferred embodiment of the present invention, the process of the data acquisition unit determining whether an electrical fault occurs is as follows: 1. Data Collection and Preprocessing After a fire occurs, the system first obtains the values ​​of all monitoring and power parameters from monitoring devices within a specific time window before the fire. These parameters include, but are not limited to, key indicators such as current, voltage, power factor, and temperature. Using this real-time monitoring data, the system constructs a comprehensive snapshot of the electrical status, providing a foundation for subsequent analysis.

[0030] 2. Preliminary current value check Zero Current: If the real-time current amplitude within this time window is zero, it indicates that no current is flowing, meaning that the fire event is unlikely to be caused by an electrical fault. In this case, the possibility of an electrical fault can be ruled out and other possible causes of the fire can be considered, such as combustible accumulation and external fire sources.

[0031] Current value exceeds the preset standard range: If the real-time current value exceeds the upper or lower limit of the pre-set standard range, it indicates that there is abnormally high current flow, which may be caused by a short circuit, overload, or other electrical problem. In this case, the system will preliminarily determine that the fire incident is caused by an electrical fault.

[0032] Current value is within the preset standard range: When the real-time current value amplitude falls within the preset standard range, the current value alone cannot directly determine whether there is an electrical fault. In this case, further investigation of other monitoring parameters and power parameters is necessary.

[0033] 3. In-depth parameter analysis Even if the current value is normal, if any real-time monitoring parameter (such as temperature, humidity) or power parameter (such as active power, reactive power) exceeds its corresponding preset standard range, this may indicate a potential electrical safety hazard or an ongoing electrical fault. For example, excessively high temperature may mean that a component is overheating; abnormally high active power may be caused by reduced equipment efficiency or some form of loss. Therefore, in this case, the system will also determine that the fire incident was caused by an electrical fault.

[0034] In another preferred embodiment of the present invention, the process of dividing the current value change sequence by the accident analysis unit is as follows: Obtain the initial time series of historical electricity consumption data, collect the length of the initial time series and mark it as L, obtain the sliding step of the time series and mark it as S, the length of the current value change sequence is t, and the number of splits of the historical electricity consumption data is n=L / St.

[0035] In another preferred embodiment of the present invention, a control radius and a minimum similarity number min are set. For a pending current value change sequence, the number m of similar sequences within the control radius is detected. If m is less than the minimum similarity number min, the cause of the electrical fault is directly determined to be abnormal behavior of the electrical equipment. If m is greater than or equal to the minimum similarity number m, the pending current value change sequence is placed in a category cluster, and all current value change sequences are placed in the category cluster in turn and marked as abnormal. Repeat the above process and extract the cluster centers of the category clusters after several clusterings. Compare the number of cluster centers where the current change value sequence is located and the difference in the cluster center values. If the cluster center shifts significantly, it means that the cause of the electrical fault is the abnormal behavior of the electrical equipment.

[0036] In another preferred embodiment of the present invention, the formula for determining whether the cluster center is offset is: , ; Among them, S i is a set of data points, |S i ∣ is the number of data points in the class cluster, Δc i Indicates the change value of the i-th cluster center between the previous and next two iterations; c i,t and c i,t+ 1 represents the value of the i-th cluster center at time t and time t+1 respectively. When Δc i When it is greater than the preset threshold, it means that the cluster center has shifted significantly.

[0037] The present invention also includes an intelligent fire safety monitoring and early warning method based on a regional alarm model, comprising the following steps: Organize and record information on all monitoring devices within the monitoring area, including energy meters, circuit protection devices, electrical fault warning sensors, intelligent sockets, and electrical fault prevention and control equipment. Based on the distribution of monitoring devices, the entire monitoring area is subdivided into several sub-areas, with each monitoring device associated with a specific sub-area. When a fire incident occurs, the sub-area related to the fire incident is identified and marked as a pending confirmation area, and the monitoring equipment in the pending confirmation area is marked as a pending confirmation equipment; For the marked equipment to be confirmed, the real-time monitoring data within a specific time window before the fire incident is collected and preliminarily processed. By analyzing the real-time monitoring data, it is determined whether the fire was caused by an electrical fault. If the judgment result is positive, all historical electricity consumption data of the area to be confirmed before the fire occurred is collected; otherwise, the judgment is terminated. The collected historical electricity consumption data is normalized and a time series model is constructed to extract the current value change sequence within a specific time window t before the fire occurs, marked as the pending current value change sequence, and the historical electricity consumption data is divided into several current value change sequences with a duration of t. All current value change sequences are clustered to identify whether the category cluster to which the pending current value change sequence belongs is offset in the cluster. If the offset exceeds a preset threshold, it is inferred that the cause of the electrical fault is the abnormal behavior of the electrical equipment; if it does not exceed the preset threshold, the electrical fault is attributed to the aging problem of the electrical insulation material.

[0038] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An intelligent fire management system based on regional division and model, characterized in that: include: The system initialization unit is used to sort out and record the information of all monitoring devices in the monitoring area. The monitoring devices include electricity meters, circuit protection devices, electrical fault warning sensors, intelligent sockets, and electrical fault prevention and control equipment. Based on the distribution of monitoring devices, the entire monitoring area is subdivided into several sub-areas, and each monitoring device is associated with a specific sub-area. The area definition unit is used to identify and mark the sub-area related to the fire event as the area to be confirmed when a fire event occurs, and mark the monitoring equipment in the area to be confirmed as the equipment to be confirmed; The data acquisition unit is used to collect real-time monitoring data of the marked equipment to be confirmed within a specific time window before the fire incident, and perform preliminary processing on the real-time monitoring data. By analyzing the real-time monitoring data, it is determined whether the fire was caused by an electrical fault. If the judgment result is positive, all historical electricity consumption data of the area to be confirmed before the fire occurred will be collected; otherwise, the judgment will be terminated; The accident analysis unit is used to normalize the collected historical electricity consumption data and build a time series model. It extracts the current value change sequence within a specific time window t before the fire occurs, marks it as a pending current value change sequence, and divides the historical electricity consumption data into several current value change sequences with a duration of t. All current value change sequences are clustered to identify whether the category cluster to which the pending current value change sequence belongs is offset within the cluster. If the offset exceeds a preset threshold, it is inferred that the cause of the electrical fault is abnormal behavior of the electrical equipment. If the preset threshold is not exceeded, the electrical fault is attributed to aging of the electrical insulation material.

2. The intelligent fire management system based on regional division and model according to claim 1 is characterized in that: The monitoring parameters of the monitoring device include current signal fundamental wave, current signal amplitude, harmonic components, and voltage.

3. The intelligent fire management system based on regional division and model according to claim 2 is characterized in that: The monitoring device calculates and generates power parameters according to the real-time voltage value and the real-time current value. The power parameters include apparent power, active power, reactive power, and power factor.

4. The intelligent fire management system based on regional division and model according to claim 2 is characterized in that: The process of preliminary processing of the data acquisition unit is as follows: Validate the monitoring parameters collected from various monitoring devices, supplement missing values ​​or outliers, convert date and time fields from strings to a unified format, detect and remove duplicate monitoring parameter records, remove outliers and invalid characters in monitoring parameters, and normalize the values ​​of different types of monitoring parameters.

5. The intelligent fire management system based on regional division and model according to claim 3 is characterized in that: The normalization process is: Calculate the mean value μ and standard deviation σ of any type of monitoring parameter, and the normalization formula is: ； Where x' represents the normalized monitoring parameter, and x represents the original monitoring parameter.

6. The intelligent fire management system based on regional division and model according to claim 3 is characterized in that: The process of the data acquisition unit determining whether it is an electrical fault is as follows: Obtain the values ​​of all monitoring parameters and power parameters within a specific time window before the fire event occurs. If the real-time current value amplitude is 0, the fire event is considered not to be an electrical fault; if the real-time current value amplitude exceeds the preset standard range, the fire event is considered to be an electrical fault; if the real-time current value amplitude is within the preset standard range, obtain any real-time monitoring parameter or power parameter. When the value of any real-time monitoring parameter or power parameter exceeds the corresponding preset standard range, the fire event is considered to be an electrical fault.

7. The intelligent fire management system based on regional division and model according to claim 1 is characterized in that: The process of dividing the current value change sequence by the accident analysis unit is as follows: Obtain the initial time series of historical electricity consumption data, collect the length of the initial time series and mark it as L, obtain the sliding step of the time series and mark it as S, the length of the current value change sequence is t, and the number of splits of the historical electricity consumption data is n=L / St.

8. The intelligent fire management system based on regional division and model according to claim 7 is characterized in that: Set the control radius and the minimum similarity number min. For the pending current value change sequence, detect the number m of similar sequences within its control radius. If m is less than the minimum similarity number min, directly determine that the cause of the electrical fault is the abnormal behavior of the electrical equipment. If m is greater than or equal to the minimum similarity number m, place the pending current value change sequence into a category cluster, and then place all current value change value sequences into the category cluster in turn and mark them as abnormal. Repeat the above process and extract the cluster centers of the category clusters after several clusterings. Compare the number of cluster centers where the current change value sequence is located and the difference in the cluster center values. If the cluster center shifts significantly, it means that the cause of the electrical fault is the abnormal behavior of the electrical equipment.

9. The intelligent fire management system based on regional division and model according to claim 8 is characterized in that: The formula for determining whether the cluster center is offset is: ， ； Among them, S i is a set of data points, |S i ∣ is the number of data points in the class cluster, Δc i Indicates the change value of the i-th cluster center between the previous and next two iterations; c i,t and c i,t+ 1 represents the value of the i-th cluster center at time t and time t+1 respectively. When Δc i When it is greater than the preset threshold, it means that the cluster center has shifted significantly.

10. An intelligent fire safety monitoring and early warning method based on a regional alarm model, characterized in that: The following steps are involved: Organize and record information on all monitoring devices within the monitoring area, including energy meters, circuit protection devices, electrical fault warning sensors, intelligent sockets, and electrical fault prevention and control equipment. Based on the distribution of monitoring devices, the entire monitoring area is subdivided into several sub-areas, with each monitoring device associated with a specific sub-area. When a fire incident occurs, the sub-area related to the fire incident is identified and marked as a pending confirmation area, and the monitoring equipment in the pending confirmation area is marked as a pending confirmation equipment; For the marked equipment to be confirmed, the real-time monitoring data within a specific time window before the fire incident is collected and preliminarily processed. By analyzing the real-time monitoring data, it is determined whether the fire was caused by an electrical fault. If the judgment result is positive, all historical electricity consumption data of the area to be confirmed before the fire occurred is collected; otherwise, the judgment is terminated. The collected historical electricity consumption data is normalized and a time series model is constructed. The current value change sequence within a specific time window t before the fire occurs is extracted and marked as a pending current value change sequence. The historical electricity consumption data is then divided into several current value change sequences of duration t. All current value change sequences are clustered to identify whether the category cluster containing the pending current value change sequence has shifted within the cluster. If the shift exceeds a preset threshold, it is inferred that the cause of the electrical fault is abnormal behavior of the electrical equipment. If the preset threshold is not exceeded, the electrical fault is attributed to aging of the electrical insulation material.

Images