Multi-sensor fire-fighting early warning data processing method and system

Abstract

The invention discloses a multi-sensor fire-fighting early warning data processing method and system, relates to the field of multi-sensor fire-fighting early warning data processing, is used for accurately identifying a fire event, and comprises the following steps: receiving real-time environment data of detection equipment in a preset warehouse, the real-time environment data comprising carbon monoxide concentration, smoke concentration and environment temperature; the preset warehouse comprises different functional areas; identifying a functional area to which the detection equipment belongs according to the detection equipment identifier; based on the functional area to which the detection equipment belongs and the real-time environment data, identifying a fire event of the functional area to which the detection equipment belongs; and triggering an alarm of a corresponding level according to the identification result of the fire event.

Description

Technical Field

[0001] This invention relates to the field of multi-sensor fire early warning data processing, and in particular to a multi-sensor fire early warning data processing method and system. Background Technology

[0002] In complex application scenarios where both routine operational interference (such as exhaust gas from internal combustion engine equipment) and special fire hazards (such as thermal runaway of lithium batteries) coexist, existing multi-sensor fire early warning data processing methods are caught in a dilemma: in order to reduce false alarms caused by routine operational interference, the system adopts a rigid judgment logic that requires confirmation from multiple different types of detection information (such as smoke and temperature). However, this makes it insensitive to early warning signals of certain special fires (such as thermal runaway of lithium batteries) that initially only exhibit a single characteristic (such as a sharp increase in gas concentration), thus missing the best opportunity for early response.

[0003] Therefore, the current technical challenge is to design a data processing method that can effectively filter out conventional, non-hazardous interference signals while accurately identifying and responding to early hazard signals with non-traditional fire characteristics, thereby achieving a balance between low false alarm rate and high sensitivity to address diverse fire risks. Summary of the Invention

[0004] This invention provides a multi-sensor fire early warning data processing method for accurately identifying fire events.

[0005] Firstly, to address the aforementioned technical problems, this invention provides a multi-sensor fire early warning data processing method, comprising: receiving real-time environmental data from detection devices within a preset warehouse, the real-time environmental data including carbon monoxide concentration, smoke concentration, and ambient temperature; the preset warehouse comprising different functional areas; identifying the functional area to which the detection device belongs based on the device's identifier; and identifying a fire event in the functional area to which the detection device belongs based on the functional area and the real-time environmental data; wherein, when the functional area is a first-class functional area, if only the carbon monoxide concentration is greater than or equal to a carbon monoxide concentration threshold and the smoke concentration is less than a smoke concentration threshold... If the ambient temperature is below the threshold, it is identified as a non-fire event; the first type of functional area is an area with routine operational interference; when the functional area is the second type of functional area, if the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, it is identified as a fire warning event; the second type of functional area is the lithium battery storage area; when the functional area is the third type of functional area, if the smoke concentration is greater than or equal to the smoke concentration threshold and the ambient temperature is greater than or equal to the temperature threshold, it is identified as a fire warning event; the third type of functional area is the area other than the first and second types of functional areas; based on the fire event identification result, the corresponding level of alarm is triggered.

[0006] Optionally, based on the functional area to which the detection device belongs and real-time environmental data, fire events in the functional area to which the detection device belongs are identified, including: when the identified functional area is a second-class functional area and an operation mode activation signal is received, adjusting the signal processing logic of the carbon monoxide detector in the second-class functional area; the adjustment includes increasing the carbon monoxide concentration threshold and activating carbon monoxide concentration change rate monitoring; in operation mode, comparing the rate of change of carbon monoxide concentration with preset gradual increase thresholds and abrupt change thresholds; if the rate of change of carbon monoxide concentration is lower than the gradual increase threshold, the smoke concentration is lower than the smoke concentration threshold, and the ambient temperature is lower than the temperature threshold, it is judged as a non-fire event; if the rate of change of carbon monoxide concentration is greater than or equal to the abrupt change threshold, it is judged as a fire warning event; after the operation mode ends, restoring the signal processing logic of the carbon monoxide detector to the default warning state.

[0007] Optionally, after the operation mode ends, the signal processing logic of the carbon monoxide detector is restored to the default warning state, including: monitoring real-time environmental data in the second type of functional area; determining whether the monitored real-time environmental data is within the normal environmental baseline range for a continuous period of time, and determining whether there is any operating equipment in the second type of functional area; if the monitored real-time environmental data is within the normal environmental baseline range and no operating equipment is in the second type of functional area, then the signal processing logic of the carbon monoxide detector is restored to the default warning state.

[0008] Optionally, if the work mode times out or no work mode end instruction is received, the method further includes: sending a verification reminder to a designated person, the verification reminder including the real-time environmental parameters of the second type of functional area, the duration of the work mode, and the last recorded type of work activity; while sending the verification reminder, continuously monitoring the changing trend of real-time environmental data within the second type of functional area; and sending the changing trend analysis results as auxiliary information to the designated person receiving the data.

[0009] Optionally, continuously monitor the changing trend of real-time environmental data within the second type of functional area, including: monitoring real-time environmental data within the second type of functional area; calculating the average rate of change and amplitude of change of the monitored real-time environmental data within a preset time window; identifying and filtering out short-term, high-frequency instantaneous fluctuations based on the average rate of change and amplitude of change; identifying persistent deviations in real-time environmental data; persistent deviations are defined as the average rate of change of real-time environmental data consistently exceeding or falling below a preset normal fluctuation threshold over multiple consecutive monitoring periods; identifying coordinated changes in real-time environmental data; coordinated changes are defined as at least two parameters in the real-time environmental data exhibiting synchronous, abnormal changing trends within the same time period; and obtaining the changing trend of real-time environmental data within the second type of functional area based on the real-time environmental data after filtering out instantaneous fluctuations, identifying persistent deviations, and identifying coordinated changes.

[0010] Optionally, identifying coordinated changes in real-time environmental data includes: determining whether at least two parameters in the real-time environmental data exhibit synchronous, abnormal change trends within the same time period; if at least two parameters exhibit synchronous, abnormal change trends within the same time period, determining whether the coordinated change matches a specific coordinated pattern of multiple preset independent, non-hazardous interference sources; if the coordinated change matches a specific coordinated pattern of multiple preset independent, non-hazardous interference sources, determining it as a non-fire event; if the coordinated change does not match a specific coordinated pattern of multiple preset independent, non-hazardous interference sources, determining it as a fire warning event.

[0011] Optionally, the method further includes: if the cooperative change does not conform to any specific cooperative pattern, then mark the cooperative change as an unknown cooperative pattern; record the characteristic information of the unknown cooperative pattern and submit the unknown cooperative pattern to the manual review queue; receive the analysis results of the manual reviewers on the unknown cooperative pattern; if the analysis results indicate that the unknown cooperative pattern is a new type of non-hazardous interference pattern, then add the characteristic information of the new type of non-hazardous interference pattern to the preset interference pattern library, and set the judgment logic of the new type of non-hazardous interference pattern to determine whether the cooperative change of the unknown cooperative pattern conforms to the preset interference pattern; if the analysis results indicate a fire event, then trigger a fire warning.

[0012] Optionally, the feature information of the novel non-hazardous interference pattern is added to a preset interference pattern library, and the judgment logic for the novel non-hazardous interference pattern is set, including: standardizing the feature parameters of the novel non-hazardous interference pattern; the standardization process includes formatting the parameter range, rate of change, and cooperative relationship; comparing the standardized features of the novel pattern with the features of existing interference patterns in the preset interference pattern library to obtain feature similarity; if the feature similarity is greater than or equal to the similarity threshold, the manual reviewer is prompted to optimize or merge the existing patterns; if the feature similarity is less than the similarity threshold, the feature information of the novel pattern is added as an independent entry to the preset interference pattern library; based on the feature parameters of the novel pattern, a set of judgment rules based on thresholds and logical relationships is generated, and the judgment rules set an initial judgment range and an initial confidence weight; a verification period is started, during which real-time environmental data similar to the novel pattern are continuously monitored, and the number of times the novel pattern judgment is triggered and whether false alarms are caused are recorded; based on the false alarm rate during the verification period, the confidence weight and the threshold of the judgment rules are adjusted.

[0013] Optionally, based on the false alarm rate during the validation period, the confidence weights and thresholds of the judgment rules are adjusted, including: classifying false alarm events according to classification criteria during the validation period; the classification criteria include environmental parameter characteristics, work activity type, and duration at the time of the false alarm; identifying the dominant factors causing the false alarms based on the classification results of the false alarm events; adjusting the confidence weights and thresholds of the judgment rules in stages for the dominant factors, including: adjusting the parameter thresholds related to the specific work activity for false alarms caused by it; adjusting the confidence weights for false alarms caused by instantaneous environmental fluctuations; after the adjustment, starting a new validation period and comparing the false alarm rate in the new validation period with the false alarm rate before the adjustment; if the decrease in the false alarm rate in the new validation period is less than the magnitude threshold or a new type of false alarm appears, then the parameter configuration before the adjustment is reviewed and the adjustment strategy is re-evaluated.

[0014] Secondly, the present invention provides a multi-sensor fire early warning data processing system, the system comprising: The area information preset module is used to receive real-time environmental data from the detection devices in the preset warehouse. The real-time environmental data includes carbon monoxide concentration, smoke concentration, and ambient temperature. The preset warehouse includes different functional areas. The real-time data receiving module is used to identify the functional area to which the detection device belongs based on the detection device identifier; The functional area identification module is used to identify fire events in the functional area to which the detection device belongs, based on the functional area to which the detection device belongs and real-time environmental data. Among them, when the functional area is a Class I functional area, if only the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, the smoke concentration is less than the smoke concentration threshold, and the ambient temperature is less than the temperature threshold, it is identified as a non-fire event; the Class I functional area is an area where there is interference from routine operations. When the functional area is a second type of functional area, if the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, it is identified as a fire warning event; the second type of functional area is the lithium battery storage area. When the functional area is a third type of functional area, if the smoke concentration is greater than or equal to the smoke concentration threshold and the ambient temperature is greater than or equal to the temperature threshold, it is identified as a fire warning event; the third type of functional area is the area other than the first type of functional area and the second type of functional area. The alarm triggering module is used to trigger alarms of the corresponding level based on the identification results of fire events.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This application provides a multi-sensor fire early warning data processing method. It receives real-time environmental data from pre-set detection devices within a warehouse and identifies the functional area to which each device belongs based on its identification. Based on this, the system can intelligently identify fire events according to the functional area and real-time environmental data. This method sets differentiated fire identification logic for different functional areas (such as a first-class functional area with routine operational interference, a second-class functional area for lithium battery storage, and other third-class functional areas). For example, in the first-class functional area, a non-fire event is determined only when the carbon monoxide concentration reaches a threshold, but smoke and temperature do not, effectively avoiding false alarms caused by routine operations. In the second-class functional area, a fire early warning is triggered when the carbon monoxide concentration reaches the threshold, demonstrating sensitivity to the early characteristics of lithium battery fires. In the third-class functional area, a judgment is made based on a combination of smoke concentration and ambient temperature. Finally, an alarm of the corresponding level is triggered based on the fire event identification result. Through this regionalized and differentiated processing approach, this application effectively solves the problem in existing technologies of finding a balance between reducing false alarm rates and improving early warning sensitivity, significantly improving the accuracy and reliability of the fire early warning system, avoiding frequent false alarms, and ensuring timely early warnings for various types of fires. Attached Figure Description

[0016] Figure 1 This is a schematic flowchart of a multi-sensor fire early warning data processing method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of another multi-sensor fire early warning data processing method provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a multi-sensor fire early warning data processing system provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0018] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0019] The implementation environment of this application is typically an intelligent fire early warning platform integrating various detection devices, data processing units, and alarm systems. Detection devices transmit real-time environmental data to the data processing unit via wired or wireless means. The data processing unit analyzes the data according to preset algorithms and logic, and the alarm system ultimately triggers the corresponding alarm based on the analysis results.

[0020] The following specific embodiments will provide a detailed description and explanation of a multi-sensor fire early warning data processing method provided in this application.

[0021] Reference Figure 1 This invention provides a multi-sensor fire early warning data processing method, comprising the following steps: S1 receives real-time environmental data from the detection devices within the preset warehouse.

[0022] The real-time environmental data includes carbon monoxide concentration, smoke concentration, and ambient temperature.

[0023] The pre-designated warehouse comprises different functional areas. Each area may have different fire risk characteristics and environmental disturbance sources. Detection equipment refers to various sensors deployed within the pre-designated warehouse for real-time monitoring of environmental parameters, such as carbon monoxide detectors, smoke detectors, and temperature sensors. Each detection device has a unique identifier used by the system to identify its corresponding functional area.

[0024] As one possible implementation, the system can employ a wired network connection, directly connecting each detection device to a central data acquisition server via Ethernet or an industrial bus (such as Modbus or CAN bus). The detection devices periodically (e.g., every 5 or 10 seconds) package and transmit the collected carbon monoxide concentration, smoke concentration, and ambient temperature data to the system via the network. Correspondingly, the system receives real-time environmental data from detection devices within a pre-defined storage area.

[0025] S2. Identify the functional area to which the detection equipment belongs based on the equipment identification.

[0026] One possible implementation is to assign a unique detector identifier to each device during system initialization, and then bind this identifier to the device's physical location in a pre-defined warehouse and its corresponding functional area, storing this information in a regional information database. When the system receives real-time environmental data containing the detector identifier, it can quickly identify the functional area to which the device belongs by querying this database.

[0027] For example, the detection device identifier "CO-001" is pre-defined as belonging to the "lithium battery storage area." When data is received from "CO-001," the system identifies it as originating from the second functional area. Another approach is to utilize a Geographic Information System (GIS) or indoor positioning technology. The boundaries of each functional area are pre-defined on a warehouse map, and the precise coordinates of each detection device are recorded. After receiving data, the system determines which functional area the detection device falls within based on its coordinates, thus identifying its assigned area.

[0028] S3. Based on the functional area to which the detection equipment belongs and real-time environmental data, identify fire events in the functional area to which the detection equipment belongs.

[0029] Among them, when the functional area is a Class I functional area, if only the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, the smoke concentration is less than the smoke concentration threshold, and the ambient temperature is less than the temperature threshold, it is identified as a non-fire event; the Class I functional area is an area where there is interference from routine operations. When the functional area is a second type of functional area, if the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, it is identified as a fire warning event; the second type of functional area is the lithium battery storage area. When the functional area is a third type of functional area, if the smoke concentration is greater than or equal to the smoke concentration threshold and the ambient temperature is greater than or equal to the temperature threshold, it is identified as a fire warning event; the third type of functional area is the area other than the first type of functional area and the second type of functional area.

[0030] In one example, when the functional area is classified as a Class I functional area, i.e., an area with routine operational disturbances, such as forklift operations, welding, or cutting, momentary carbon monoxide or smoke may be generated, but this is not a fire. In this case, if only the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, the smoke concentration is less than the smoke concentration threshold, and the ambient temperature is less than the temperature threshold, it is identified as a non-fire event. This means that even if the carbon monoxide concentration reaches the warning level, as long as the smoke and temperature do not rise synchronously, the system judges it as operational disturbance rather than a fire. For example, in a production workshop, due to forklift exhaust emissions, the carbon monoxide concentration briefly rises to 50 ppm (carbon monoxide concentration threshold), but the smoke concentration remains below 1% OBS / m (smoke concentration threshold), and the ambient temperature remains below 25°C (temperature threshold). In this case, the system would judge it as a non-fire event.

[0031] In one example, when the functional area is a Class II functional area, namely a lithium battery storage area, lithium battery fires are characterized by the initial release of toxic gases (such as carbon monoxide), while smoke and temperature increases may be delayed. Therefore, for this type of area, a fire warning event is identified if the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold. For example, in a lithium battery warehouse, even without obvious smoke or temperature increases, the system will trigger a fire warning as long as the carbon monoxide concentration reaches 20 ppm (the carbon monoxide concentration threshold).

[0032] In one example, when the functional area is a third-class functional area, that is, an area other than the first and second-class functional areas, the system typically follows traditional fire detection logic. If the smoke concentration is greater than or equal to the smoke concentration threshold and the ambient temperature is greater than or equal to the temperature threshold, a fire warning event is identified. For example, in a general cargo storage area, when the smoke concentration reaches 5% OBS / m³ (smoke concentration threshold) and the ambient temperature reaches 60°C (temperature threshold), the system triggers a fire warning.

[0033] S4. Based on the fire incident identification results, trigger the corresponding level of alarm.

[0034] The triggering method and level of the alarm can be configured according to the preset emergency plan.

[0035] As one possible implementation, for "non-fire events," the system may not trigger any alarms, only recording them in the background. For "fire warning events," the system can trigger a Level 1 alarm, such as sounding an audible and visual alarm, sending SMS or app push notifications to designated personnel, and displaying the information on the monitoring screen. Higher-level fire events (e.g., when multiple sensor data continue to deteriorate or reach higher thresholds) can trigger Level 2 or Level 3 alarms, such as automatically activating the fire sprinkler system, activating smoke exhaust fans, and automatically alerting the fire department.

[0036] Understandably, in the first type of functional area where routine operations cause interference, the system pays special attention to the coordinated changes in carbon monoxide concentration, smoke concentration, and ambient temperature. Only when the carbon monoxide concentration reaches a threshold, while the smoke concentration and ambient temperature are both below their respective thresholds, will the system classify it as a non-fire event, effectively avoiding false alarms caused by operational activities. In the second type of functional area, such as lithium battery storage areas, due to the unique nature of lithium battery fires (initially only gas release may occur), the system uses carbon monoxide concentration as the primary criterion. As long as the carbon monoxide concentration reaches a preset threshold, even if smoke and temperature have not yet significantly increased, the system will immediately identify it as a fire warning event, thus achieving early warning for this type of special fire. For the third type of functional area, excluding the two types mentioned above, the system adopts a traditional, more universal fire judgment logic: only when the smoke concentration and ambient temperature simultaneously reach or exceed their respective thresholds will it be identified as a fire warning event. This regionalized and differentiated judgment logic allows the system to flexibly adjust its warning strategy according to the characteristics of the actual scenario, significantly improving the accuracy and reliability of the warnings.

[0037] In one possible design, such as Figure 2 As shown, in order to identify fire events in the functional area to which the detection equipment belongs, this application may further include the following steps: S101. When the identified functional area is the second type of functional area and a work mode start signal is received, adjust the signal processing logic of the carbon monoxide detector in the second type of functional area.

[0038] The adjustments include raising the carbon monoxide concentration threshold and activating monitoring of the rate of change of carbon monoxide concentration.

[0039] The start signal for the work mode can be triggered manually, by a preset schedule, or by linkage with the work equipment.

[0040] S102. In the operating mode, compare the rate of change of carbon monoxide concentration with the preset gradual increase threshold and the sharp jump threshold.

[0041] The gradual increase threshold is used to define the slow accumulation of carbon monoxide that may be caused by normal work activities, while the sharp jump threshold is used to identify abnormal, rapid spikes in carbon monoxide concentration, which are often an early sign of a fire.

[0042] S103. If the rate of change of carbon monoxide concentration is lower than the gradual growth threshold, the smoke concentration is lower than the smoke concentration threshold, and the ambient temperature is lower than the temperature threshold, then it is determined to be a non-fire event.

[0043] The gradual growth threshold, smoke concentration threshold, and temperature threshold can be set as needed and are not restricted here.

[0044] S104. If the rate of change of carbon monoxide concentration is greater than or equal to the threshold for drastic change, it is judged as a fire warning event.

[0045] S105. After the operation mode ends, restore the signal processing logic of the carbon monoxide detector to the default warning state.

[0046] As one possible implementation, the system can monitor real-time environmental data within the second type of functional area; determine whether the monitored real-time environmental data is within the normal environmental baseline range for a continuous period of time, and determine whether there is any operating equipment in the second type of functional area; if the monitored real-time environmental data is within the normal environmental baseline range and no operating equipment is in the second type of functional area, then the signal processing logic of the carbon monoxide detector is restored to the default warning state.

[0047] The normal environmental baseline range can be understood as the range of environmental parameter fluctuations under conditions of no work activities and no abnormalities.

[0048] Determining whether there is operating equipment is to confirm whether the work activity has completely stopped, because some operating equipment may still produce residual interfering gases or heat even after the work mode has ended.

[0049] As a specific implementation method, a concrete example is given below. Assume a lithium battery storage area where daily operations include using forklifts to move batteries. During charging, the forklifts may generate trace amounts of carbon monoxide. Without the solution described in this application, if the carbon monoxide concentration reaches a preset low threshold, the system may falsely report a fire. However, with the solution described in this application, when the forklift begins charging or moving operations, the system receives a work mode activation signal. At this time, the system automatically increases the carbon monoxide concentration threshold of the carbon monoxide detector in that area, for example, from 50 ppm to 100 ppm, and simultaneously activates carbon monoxide concentration change rate monitoring. During operation, if the carbon monoxide concentration slowly rises to 80 ppm, but its rate of change is lower than the preset gradual increase threshold, and the smoke concentration and ambient temperature are normal, the system will determine it as a non-fire event and will not trigger an alarm. However, if a fire occurs due to a battery short circuit or other reasons, causing the carbon monoxide concentration to rapidly surge from 50 ppm to 150 ppm in a short period, exceeding the drastic change threshold, even if the smoke and temperature have not yet significantly increased, the system will immediately determine it as a fire warning event and trigger an alarm. When the operation is completed and the forklift leaves the charging area, the system will automatically restore the signal processing logic of the carbon monoxide detector to the default warning state, that is, restore the carbon monoxide concentration threshold to 50ppm, to ensure timely response to any abnormal situation during non-operation periods.

[0050] In one possible design, if the job mode times out or no job mode termination instruction is received, this application further includes the following steps: S201. Send a verification reminder to the designated personnel.

[0051] The verification reminder includes the real-time environmental parameters of the second type of functional area, the duration of the operation mode, and the last recorded type of operation activity.

[0052] The duration of the work mode refers to the time that has been ongoing since the work mode was initiated. The last recorded work activity type can be, for example, forklift operation, equipment maintenance, or material handling. This information helps designated personnel determine whether the current abnormal environmental parameters are related to the work activity.

[0053] S202. While sending verification reminders, continuously monitor the changing trends of real-time environmental data within the second type of functional area.

[0054] This monitoring is not simply about reading data, but about analyzing the patterns of data changes over time, such as whether it is continuously rising, falling, fluctuating, or remaining stable.

[0055] As one possible implementation, the system can continuously monitor the changing trends of real-time environmental data within the second type of functional area based on the following steps: S2021. Monitor real-time environmental data within the second type of functional area.

[0056] S2022. For the monitored real-time environmental data, calculate the average rate of change and the magnitude of change within the preset time window.

[0057] S2023. Based on the average rate of change and the amplitude of change, identify and filter out short-term, high-frequency instantaneous fluctuations.

[0058] As one possible implementation, the system can identify and filter out short-term, high-frequency instantaneous fluctuations by setting a threshold or using filtering algorithms (such as moving average filtering, low-pass filtering, etc.).

[0059] S2024. Identify persistent deviations in real-time environmental data.

[0060] Among them, persistent deviation refers to the average rate of change of real-time environmental data being consistently higher or lower than the preset normal fluctuation threshold over multiple consecutive monitoring periods.

[0061] Such persistent deviations usually indicate a stable and abnormal situation, rather than a short-lived, accidental fluctuation.

[0062] S2025. Identify collaborative changes in real-time environmental data.

[0063] Among them, coordinated change refers to at least two parameters in real-time environmental data that exhibit synchronous and abnormal change trends within the same time period.

[0064] For example, the simultaneous abnormal increase in carbon monoxide concentration, smoke concentration, and ambient temperature is often a typical characteristic of fire incidents, while an abnormality in a single parameter may be caused by other non-fire factors.

[0065] As one possible implementation, the system can determine whether at least two parameters in the real-time environmental data exhibit synchronous, abnormal changing trends within the same time period. If at least two parameters exhibit synchronous, abnormal changing trends within the same time period, the system determines whether the coordinated changes match a specific coordinated pattern of multiple preset independent, non-hazardous interference sources. If the coordinated changes match the specific coordinated pattern of multiple preset independent, non-hazardous interference sources, the system determines it to be a non-fire event. If the coordinated changes do not match the specific coordinated pattern of multiple preset independent, non-hazardous interference sources, the system determines it to be a fire warning event.

[0066] It should be noted that the pre-defined specific synergistic patterns of multiple independent, non-hazardous interference sources refer to synergistic patterns within a specific functional area that are caused by known non-fire sources (such as industrial production processes, vehicle exhaust emissions, and the start-up of heating equipment), and have specific parameter combinations and variation patterns. For example, in a Class I functional area where routine operational interference exists, forklift operation may lead to an increase in carbon monoxide concentration, while the start-up of nearby heat treatment equipment may lead to an increase in ambient temperature. Both exhibit a specific synergistic change pattern, but this is not a fire.

[0067] In some preferred embodiments, a specific example is given below. Assume a lithium battery storage area (a second-class functional area) where battery charging and forklift handling operations are simultaneously taking place. Charging operations may cause a slight increase in local temperature, while forklift operations generate a certain amount of carbon monoxide. At a certain moment, the system detects that the carbon monoxide concentration and ambient temperature in this area exhibit a synchronous, abnormal trend, meaning both increase within a short period, but have not yet reached the fire warning threshold for a single parameter.

[0068] At this point, the system will initiate cooperative change pattern recognition. The preset interference pattern library may contain a specific cooperative pattern of "charging operation + forklift operation," characterized by a slow increase in carbon monoxide concentration, a slight increase in ambient temperature, and smoke concentration remaining at a normal level. The system will compare the currently monitored cooperative change with this preset pattern. If the comparison shows that the characteristics of the current cooperative change (such as the rate of change, amplitude, and parameter combination) highly match the "charging operation + forklift operation" pattern, the system will determine that this event is not a fire event and will not trigger a fire alarm.

[0069] Conversely, if the system detects a simultaneous and rapid increase in carbon monoxide concentration and ambient temperature, and a rise in smoke concentration, and this coordinated change pattern does not match any preset non-hazardous disturbance pattern, the system will immediately identify it as a fire warning event and trigger the corresponding level of alarm.

[0070] Thus, when the system detects synchronous and abnormal trends in real-time environmental data such as carbon monoxide concentration, smoke concentration, and ambient temperature, it no longer simply interprets these as fire warnings. Instead, it further analyzes the "fingerprint" or "features" of these coordinated changes. By comparing these coordinated changes with pre-established specific coordinated patterns generated by known non-hazardous disturbances (such as welding fumes and localized heating, vehicle starting and exhaust), the system can intelligently distinguish between genuine fire signs and complex operational disturbances. This pattern recognition-based judgment mechanism enables the system to more accurately identify potential fire risks while significantly reducing the false alarm rate caused by environmental complexity or the superposition of multiple non-fire factors.

[0071] S2026. Based on the real-time environmental data that has been filtered to remove instantaneous fluctuations and identified as persistent deviations and collaborative changes, the changing trend of real-time environmental data within the second type of functional area is obtained.

[0072] In some preferred embodiments, a specific example is given below. Assume that the operating mode times out in the lithium battery storage area (second functional area). The system will initiate continuous monitoring of real-time environmental data for this area. Specifically, the system will monitor carbon monoxide concentration, smoke concentration, and ambient temperature every 5 seconds. During monitoring, the system will calculate the average rate of change and maximum amplitude of these parameters per minute. For example, if the carbon monoxide concentration suddenly increases at a certain moment, but only lasts for 10 seconds before rapidly decreasing, and its rate of change and amplitude do not reach the threshold for sustained deviation, then this fluctuation will be identified as an instantaneous fluctuation and filtered out. Simultaneously, the system will continuously determine whether the average rate of change of carbon monoxide concentration, smoke concentration, and ambient temperature is consistently higher than the preset normal fluctuation threshold within 5 consecutive monitoring cycles (e.g., 25 seconds) to identify sustained deviations. Furthermore, the system will analyze whether these three parameters exhibit a synchronous, abnormal upward trend within the same time period, such as a simultaneous rapid increase in carbon monoxide concentration, smoke concentration, and temperature, which will be identified as a coordinated change. By comprehensively analyzing this processed data, the system can obtain a more realistic and reliable real-time environmental data change trend, and send it as auxiliary information to designated personnel for verification.

[0073] Thus, this application significantly improves the accuracy and robustness of judging real-time environmental data change trends within the second-class functional area. Compared to continuous monitoring alone, this solution effectively filters out environmental noise and non-hazardous transient interference, reducing false alarm rates. Simultaneously, by identifying persistent deviations and multi-parameter coordinated changes, it can more sensitively and comprehensively capture potential fire risks, avoiding missed alarms. This enables the system to provide more reliable trend analysis results even when the work mode timeout occurs or no work mode termination instruction is received, providing more accurate auxiliary information for verification reminders to designated personnel, thereby improving the intelligence level and response efficiency of the entire fire early warning system.

[0074] S203. Send the trend analysis results as supplementary information to the designated recipient.

[0075] In some preferred embodiments, a specific example is given below. Assume that in a lithium battery storage area (a second-class functional area), a forklift is performing battery handling operations. The system has activated its operation mode and adjusted the signal processing logic of the carbon monoxide detector. The operation was originally scheduled to end at 5 PM, with the operator manually sending an end-of-operation command. However, due to operator negligence or equipment malfunction, by 5:30 PM, the system has still not received the end-of-operation command, and the preset maximum duration of the operation mode (e.g., 4 hours) has also expired.

[0076] At this point, the proposed solution will be triggered. The system will automatically send a verification notification to the mobile phone or management platform of the pre-designated fire safety supervisor (designated personnel). The notification may display: "[Emergency Verification] Operation mode of Category II functional area (lithium battery storage area) has been abnormally terminated! Current time: [current time], operation duration: [actual duration, e.g., 4 hours 30 minutes], last recorded activity: forklift handling. Real-time environmental parameters: carbon monoxide concentration [current carbon monoxide value] ppm, smoke concentration [current smoke value] %, ambient temperature [current temperature] °C." Meanwhile, the system continuously monitors real-time environmental data for the area, including trends in carbon monoxide concentration, smoke concentration, and ambient temperature. For example, the system might detect that while carbon monoxide concentration decreases after operations, the rate of decrease is far lower than normal recovery levels, or that a slight but sustained increase occurs at some point in time. The system will analyze these trends (e.g., "carbon monoxide concentration slowly decreased followed by a slight increase," "smoke concentration remained stable at a low level") and send the analysis results as supplementary information to the fire safety manager through the same channel or in a separate report.

[0077] After receiving alerts and supplementary information, the fire safety supervisor can determine whether to immediately dispatch personnel to the site for verification or to further review real-time video through a remote monitoring system. For example, if supplementary information shows that carbon monoxide concentration continues to rise slowly after the operation is completed, even if it has not reached the fire warning threshold, the supervisor may still consider it a potential risk and take immediate action, such as dispatching inspectors to check for initial signs of battery spontaneous combustion or confirming whether any equipment is still in operation, causing continuous emissions. This mechanism prevents the system from falling into a "blind spot" after an abnormal termination of the operation mode, ensuring that potential fire safety hazards can be detected and addressed promptly even under abnormal circumstances.

[0078] Through the aforementioned technical solution, this application effectively addresses the special circumstances of abnormal termination of work modes, significantly improving the robustness and security of the fire alarm system. Specifically, by promptly sending verification reminders containing detailed environmental parameters, work duration, and activity type to designated personnel, timely human intervention is facilitated, preventing the system from remaining in an uncertain or non-default warning state for extended periods due to abnormal termination of work modes. Furthermore, continuous monitoring and provision of real-time environmental data trends as supplementary information provides a more scientific and comprehensive basis for human judgment, helping designated personnel to more accurately identify whether abnormal situations are due to residual effects of work, equipment malfunction, or genuine fire hazards. This reduces false alarm rates and ensures timely response to real fire incidents, greatly enhancing the fire safety management level of warehouses, especially high-risk areas such as lithium battery storage areas.

[0079] However, in practical applications, some new non-hazardous interference patterns may emerge that are not pre-set. The cooperative variation characteristics of these patterns may not be consistent with existing patterns, thus being misjudged as fire warning events, leading to unnecessary alarms and waste of resources.

[0080] To address the aforementioned technical problems, this application further includes the following steps: S301. If a cooperative change does not conform to any specific cooperative pattern, then the cooperative change is marked as an unknown cooperative pattern.

[0081] S302. Record the characteristic information of the unknown collaboration mode and submit the unknown collaboration mode to the manual review queue.

[0082] Among them, recording the characteristic information of unknown collaborative modes means that the system will capture and store the key environmental parameters when the unknown collaborative mode occurs, such as the trend, rate of change, and duration of changes in ambient temperature, as well as the types of related work activities, forming a detailed data snapshot.

[0083] The manual review queue can be a software interface that displays a list of events to be reviewed and their detailed characteristics, so that reviewers can process them efficiently.

[0084] S303. Receive the analysis results of unknown collaboration modes from human reviewers.

[0085] S304. If the analysis results indicate that the unknown cooperative mode is a new type of non-dangerous interference mode, the feature information of the new type of non-dangerous interference mode is added to the preset interference mode library, and the judgment logic of the new type of non-dangerous interference mode is set to determine whether the cooperative change of the unknown cooperative mode is consistent with the preset interference mode.

[0086] As one possible implementation, the system can add the feature information of novel non-dangerous interference modes to a preset interference mode library based on the following steps, and set the judgment logic for novel non-dangerous interference modes: S3041. Standardize the characteristic parameters of novel non-hazardous interference modes.

[0087] The standardization process includes formatting parameter ranges, rates of change, and collaborative relationships.

[0088] For example, parameters with different dimensions can be unified to the same numerical range, or time series data can be converted into a unified feature vector to eliminate the impact of differences in dimensions and scales on subsequent comparisons and ensure the comparability of features from different patterns.

[0089] S3042. The standardized new pattern features are compared with the features of existing interference patterns in the preset interference pattern library to obtain the feature similarity.

[0090] As one possible implementation, the system can use Euclidean distance, cosine similarity, or machine learning-based feature matching algorithms to compare the standardized new pattern features with the features of existing interference patterns in a preset interference pattern library to obtain feature similarity.

[0091] S3043. If the feature similarity is greater than or equal to the similarity threshold, then prompt the manual reviewers to optimize or merge the existing patterns.

[0092] S3044. If the feature similarity is less than the similarity threshold, the feature information of the new pattern will be added as an independent entry to the preset interference pattern library.

[0093] S3045. Based on the characteristic parameters of the new pattern, generate a set of judgment rules based on thresholds and logical relationships.

[0094] The judgment rules set the initial judgment range and the initial confidence weight.

[0095] For example, for a new type of operational disturbance, if the carbon monoxide concentration is within a specific range, the rate of change is below a certain threshold, and the smoke concentration and ambient temperature remain normal, then it can be judged as a new type of non-hazardous disturbance mode.

[0096] S3046. Start the verification period. During the verification period, continuously monitor real-time environmental data similar to the new mode, and record the number of times the new mode judgment is triggered and whether false alarms are caused.

[0097] When a new mode judgment is triggered, the system displays an input box on the display interface and prompts the operator to enter whether there is a false alarm in the input box on the display interface, so as to record the number of times the new mode judgment is triggered and whether a false alarm is caused.

[0098] S3047. Adjust the confidence weight and the threshold of the judgment rule based on the false alarm rate during the verification period.

[0099] The adjustment process can be found in the following sections, and will not be elaborated upon here.

[0100] In some preferred embodiments, a specific example is given below. Suppose that a novel automated handling robot is introduced into the lithium battery storage area (second type of functional area). During charging or operation, it generates trace amounts of carbon monoxide, insufficient to pose a fire risk. Traditional systems might misinterpret this as a fire warning.

[0101] According to the scheme of this application, when a human reviewer confirms that the specific environmental data pattern generated by this robot operation is a novel non-hazardous interference pattern: First, the characteristic parameters of the model, such as peak carbon monoxide concentration, duration, and temperature variation range, will be standardized to facilitate effective comparison with other existing interference models in the model library (such as forklift exhaust and battery discharge thermal effects).

[0102] Next, the system calculates the feature similarity between the new robot operation mode and existing modes. If the similarity is found to be low, it is added as a separate "robot operation mode" entry to the preset interference mode library.

[0103] Subsequently, the system will generate a set of initial judgment rules based on the characteristics of the mode, such as: "If the carbon monoxide concentration is between 5-10 ppm for 5-15 minutes and the smoke concentration and temperature are within the normal range, it will be judged as robot operation mode (non-fire event)", and set an initial confidence weight.

[0104] Then, the system will initiate a one-week verification period. During this period, the system will continuously monitor real-time environmental data within the lithium battery storage area. Whenever data matching the "robot operation mode" judgment rules is detected, the system will record it and simultaneously observe whether it triggers a false alarm. For example, if this mode is triggered 100 times, and 2 of them are manually confirmed as false alarms (i.e., actually fire warnings), the false alarm rate is 2%.

[0105] Finally, based on this false alarm rate, the system will adjust the threshold or confidence weight of the judgment rule. For example, if the false alarm rate is high, the judgment range for carbon monoxide concentration may be tightened, or the confidence weight of the pattern may be reduced, making it more likely to trigger a fire warning when it conflicts with other potential fire signals, thereby further optimizing the accuracy of the judgment.

[0106] S305. If the analysis results indicate a fire event, a fire warning will be triggered.

[0107] In one possible design, in order to adjust the confidence weight and the threshold of the judgment rule based on the false alarm rate during the verification period, this application also includes: S401. During the verification period, false alarm events are classified according to the classification criteria.

[0108] The classification criteria include the environmental parameter characteristics at the time of the false alarm, the type of work activity, and the duration of the false alarm.

[0109] For example, environmental parameter characteristics can include the specific values ​​or trends of carbon monoxide concentration, smoke concentration, and ambient temperature at the time of the false alarm; the type of work activity can refer to the specific work that caused the false alarm, such as welding, cutting, or forklift operation; and the duration records the length of time the false alarm signal lasts. Through these classification criteria, a clearer "profile" of the false alarm event can be drawn.

[0110] S402. Based on the classification results of false alarm events, identify the dominant factors that lead to false alarms.

[0111] For example, if a large number of false alarms occur during a specific operational activity, then that activity is likely the dominant factor; if false alarms are mostly manifested as short-term, high-frequency fluctuations in environmental parameters, then instantaneous environmental fluctuations are likely the dominant factor. Identifying the dominant factor is key to achieving precise adjustments.

[0112] S403. For the dominant factors, adjust the confidence weight and the threshold of the judgment rule in segments.

[0113] The segmented adjustments include: for false alarms caused by specific work activities, adjusting the parameter thresholds related to those work activities; for false alarms caused by instantaneous environmental fluctuations, adjusting the confidence weights.

[0114] Segmented adjustment means that not all parameters are adjusted uniformly, but rather differentiated based on the root cause of false alarms. For example, for false alarms caused by specific work activities, the system will adjust the threshold values ​​of parameters related to that activity, such as increasing the carbon monoxide concentration threshold for that work mode, to avoid misjudging gases generated during normal operations. For false alarms caused by instantaneous environmental fluctuations, the system will adjust the confidence weights, such as reducing sensitivity to short-term, small-amplitude fluctuations, thereby improving the system's robustness.

[0115] S404. After the adjustment, start a new verification period and compare the false alarm rate in the new verification period with the false alarm rate before the adjustment.

[0116] In practical applications, the corrected effective latent heat of volatilization parameter can be input into the heat balance model or empirical formula along with the curing characteristic parameter and the material characteristic parameter of the prepreg (such as specific heat capacity, thermal conductivity, etc.) to accurately calculate the heat absorption per unit time required by the prepreg during the vaporization stage of the volatiles, that is, the target heat absorption rate during the vaporization stage of the volatiles.

[0117] S405. If the false alarm rate decreases less than the magnitude threshold or a new false alarm type appears during the new validation period, the parameter configuration before the adjustment should be reviewed and the adjustment strategy should be re-evaluated.

[0118] The amplitude threshold can be set as needed, and will not be elaborated here.

[0119] In some preferred embodiments, a specific example is given below. Assume that within the lithium battery storage area (second type of functional area), the system, through manual review, confirms a novel non-hazardous interference mode: the "trace gas release mode during the initial battery charging phase." This mode is characterized by a gradual increase in carbon monoxide concentration shortly after battery charging begins, but it does not reach the drastic threshold for fire warnings, while smoke concentration and ambient temperature remain normal. The system adds the characteristic information of this mode to a preset interference mode library and sets initial judgment logic and confidence weights.

[0120] During the subsequent verification period, the system detected multiple false alarms caused by this "trace gas release pattern during the initial stage of battery charging." According to the scheme in this application, the system categorizes these false alarm events. For example, false alarm event A occurred at 10:00 AM, with environmental parameters showing a carbon monoxide concentration rising from 5 ppm to 15 ppm for 5 minutes, without smoke or abnormal temperature; the work activity type was "battery charging." False alarm event B occurred at 3:00 PM, with environmental parameters showing a carbon monoxide concentration rising from 8 ppm to 18 ppm for 7 minutes, without smoke or abnormal temperature; the work activity type was also "battery charging." Through classification, the system identified the dominant factor causing the false alarms as "trace gas release during the initial stage of battery charging."

[0121] To address this primary factor, the system will make phased adjustments. Specifically, since false alarms are caused by a specific operational activity (battery charging), the system will adjust the threshold parameters related to "battery charging" accordingly. For example, the threshold for a gradual increase in carbon monoxide concentration under this operational mode will be increased from 10 ppm to 20 ppm to tolerate normal trace gas releases during the initial charging phase. Simultaneously, the system may fine-tune the confidence weights related to smoke concentration and ambient temperature to ensure that a gradual increase in carbon monoxide alone will not trigger an alarm when these parameters are normal.

[0122] After the adjustment is completed, the system starts a new verification period. During this period, the system continues to monitor and compare the new false alarm rate with the false alarm rate before the adjustment. If, during the new verification period, the false alarms caused by the "trace gas release mode in the early stage of battery charging" are significantly reduced, and no new false alarm types appear, it indicates that the adjustment is effective. Conversely, if the false alarm rate does not decrease significantly or new false alarms appear, the system will revert to the parameter configuration before the adjustment and re-evaluate the adjustment strategy. For example, it may further refine the characteristic parameters of the "trace gas release mode in the early stage of battery charging" or consider introducing more auxiliary judgment conditions to ensure that the system can continuously optimize and accurately identify various events.

[0123] like Figure 3 As shown in the figure, this embodiment of the invention also provides a multi-sensor fire early warning data processing system. The system includes: The area information preset module is used to receive real-time environmental data from the detection devices in the preset warehouse. The real-time environmental data includes carbon monoxide concentration, smoke concentration, and ambient temperature. The preset warehouse includes different functional areas. The real-time data receiving module is used to identify the functional area to which the detection device belongs based on the detection device identifier; The functional area identification module is used to identify fire events in the functional area to which the detection device belongs, based on the functional area to which the detection device belongs and real-time environmental data. Among them, when the functional area is a Class I functional area, if only the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, the smoke concentration is less than the smoke concentration threshold, and the ambient temperature is less than the temperature threshold, it is identified as a non-fire event; the Class I functional area is an area where there is interference from routine operations. When the functional area is a second type of functional area, if the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, it is identified as a fire warning event; the second type of functional area is the lithium battery storage area. When the functional area is a third type of functional area, if the smoke concentration is greater than or equal to the smoke concentration threshold and the ambient temperature is greater than or equal to the temperature threshold, it is identified as a fire warning event; the third type of functional area is the area other than the first type of functional area and the second type of functional area. The alarm triggering module is used to trigger alarms of the corresponding level based on the identification results of fire events.

[0124] This application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be executed by a computer program instructing related hardware. This program can be stored in the computer-readable storage medium, and when executed, it can include the processes of the above method embodiments. The computer-readable storage medium can be an internal storage unit of the task execution device (including a data sending end and / or a data receiving end) of any of the foregoing embodiments, such as the hard disk or memory of the task execution device. The computer-readable storage medium can also be an external storage device of the terminal device, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the terminal device. Further, the computer-readable storage medium can include both the internal storage unit of the task execution device and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by the task execution device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0125] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0126] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, essentially, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0127] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be covered within the scope of protection of this application.

Claims

1. A multi-sensor fire early warning data processing method, characterized in that, Includes the following steps: Receive real-time environmental data from detection devices within a preset warehouse, wherein the preset warehouse includes multiple functional areas, and the real-time environmental data includes carbon monoxide concentration, smoke concentration, and ambient temperature; Identify the functional area to which the detection device belongs based on the device's identifier; Based on the functional area to which the detection device belongs and the real-time environmental data, fire events in the functional area to which the detection device belongs are identified; wherein... When the functional area is a first type of functional area, if only the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, the smoke concentration is less than the smoke concentration threshold, and the ambient temperature is less than the temperature threshold, then it is identified as a non-fire event; the first type of functional area is an area where there is interference from routine operations. When the functional area is a second type of functional area, if the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, it is identified as a fire warning event; the second type of functional area is a lithium battery storage area. When the functional area is a third type of functional area, if the smoke concentration is greater than or equal to the smoke concentration threshold and the ambient temperature is greater than or equal to the temperature threshold, it is identified as a fire warning event; the third type of functional area is the area other than the first type of functional area and the second type of functional area. Based on the identification results of the fire event, an alarm of the corresponding level is triggered.

2. The multi-sensor fire early warning data processing method according to claim 1, characterized in that, The process of identifying fire events in the functional area to which the detection device belongs, based on the functional area to which the detection device belongs and the real-time environmental data, includes: When the identified functional area is the second type of functional area and a work mode start signal is received, the signal processing logic of the carbon monoxide detector in the second type of functional area is adjusted; the adjustment includes raising the carbon monoxide concentration threshold and activating carbon monoxide concentration change rate monitoring. In the operating mode, the rate of change of the carbon monoxide concentration is compared with preset gradual increase thresholds and drastic jump thresholds; If the rate of change of the carbon monoxide concentration is lower than the gradual growth threshold, the smoke concentration is lower than the smoke concentration threshold, and the ambient temperature is lower than the temperature threshold, then it is determined to be a non-fire event. If the rate of change of the carbon monoxide concentration is greater than or equal to the drastic change threshold, it is determined to be a fire warning event; After the operation mode ends, the signal processing logic of the carbon monoxide detector is restored to the default warning state.

3. The multi-sensor fire early warning data processing method according to claim 2, characterized in that, After the operation mode ends, restoring the signal processing logic of the carbon monoxide detector to the default warning state includes: Monitor real-time environmental data within the second type of functional area; Determine whether the monitored real-time environmental data are within the normal environmental baseline range over a continuous period of time, and determine whether there is operating equipment in the second type of functional area; If the monitored real-time environmental data are all within the normal environmental baseline and no operating equipment is running in the second type of functional area, then the signal processing logic of the carbon monoxide detector is restored to the default warning state.

4. The multi-sensor fire early warning data processing method according to claim 2, characterized in that, If the job mode times out or no job mode termination instruction is received, the method further includes: Send a verification reminder to the designated personnel. The verification reminder includes the real-time environmental parameters of the second type of functional area, the duration of the operation mode, and the last recorded type of operation activity. While sending the verification reminder, continuously monitor the changing trend of real-time environmental data within the second type of functional area; The trend analysis results will be sent as supplementary information to the designated recipient.

5. The multi-sensor fire early warning data processing method according to claim 4, characterized in that, The continuous monitoring of the changing trends of real-time environmental data within the second type of functional area includes: Monitor real-time environmental data within the second type of functional area; For the monitored real-time environmental data, calculate the average rate of change and magnitude of change within a preset time window; Based on the average rate of change and the magnitude of change, short-term, high-frequency instantaneous fluctuations are identified and filtered out; Identify persistent deviations in real-time environmental data; the persistent deviation is defined as the average rate of change of the real-time environmental data being consistently higher or lower than a preset normal fluctuation threshold over multiple consecutive monitoring periods. Identify the coordinated changes in the real-time environmental data; the coordinated changes are at least two parameters in the real-time environmental data that exhibit synchronous, abnormal change trends within the same time period; Based on real-time environmental data that has been filtered to remove instantaneous fluctuations and identified persistent deviations and co-variations, the changing trend of real-time environmental data within the second type of functional area is obtained.

6. The multi-sensor fire early warning data processing method according to claim 5, characterized in that, The identification of coordinated changes in the real-time environmental data includes: Determine whether at least two parameters in the real-time environmental data exhibit synchronous, abnormal trends within the same time period; If at least two parameters exhibit a synchronous, abnormal trend of change within the same time period, it is determined whether the coordinated change is consistent with a specific coordinated pattern of multiple independent, non-dangerous interference sources preset. If the cooperative change matches a specific cooperative pattern of multiple preset independent, non-hazardous interference sources, it is determined to be a non-fire event. If the cooperative change does not match the specific cooperative pattern of multiple independent, non-hazardous interference sources preset, it is judged as a fire early warning event.

7. The multi-sensor fire early warning data processing method according to claim 6, characterized in that, The method further includes: If the cooperative change does not conform to any specific cooperative pattern, then the cooperative change is marked as an unknown cooperative pattern; Record the feature information of the unknown collaboration mode and submit the unknown collaboration mode to the manual review queue; Receive the analysis results of the unknown collaboration mode from human reviewers; If the analysis results indicate that the unknown cooperative mode is a novel non-dangerous interference mode, then the feature information of the novel non-dangerous interference mode is added to the preset interference mode library, and the judgment logic of the novel non-dangerous interference mode is set to determine whether the cooperative change of the unknown cooperative mode is consistent with the preset interference mode. If the analysis results indicate a fire event, a fire warning will be triggered.

8. The multi-sensor fire early warning data processing method according to claim 7, characterized in that, The step of adding the feature information of the novel non-dangerous interference mode to a preset interference mode library and setting the judgment logic of the novel non-dangerous interference mode includes: The characteristic parameters of the novel non-hazardous interference mode are standardized; the standardization process includes formatting the parameter range, rate of change, and cooperative relationship. The standardized new pattern features are compared with the features of existing interference patterns in the preset interference pattern library to obtain the feature similarity. If the feature similarity is greater than or equal to the similarity threshold, the human reviewer will be prompted to optimize or merge the existing patterns. If the feature similarity is less than the similarity threshold, the feature information of the new pattern is added as an independent entry to the preset interference pattern library. Based on the feature parameters of the novel pattern, a set of judgment rules based on thresholds and logical relationships are generated, wherein the judgment rules set an initial judgment range and an initial confidence weight; During the initial verification period, real-time environmental data similar to the novel mode are continuously monitored, and the number of times the novel mode judgment is triggered and whether false alarms are caused are recorded. Based on the false alarm rate during the verification period, adjust the confidence weight and the threshold of the judgment rule.

9. A multi-sensor fire early warning data processing method according to claim 8, characterized in that, The step of adjusting the confidence weight and the threshold of the judgment rule based on the false alarm rate during the verification period includes: During the verification period, false alarm events are classified according to classification criteria, which include environmental parameter characteristics at the time of the false alarm, type of work activity, and duration. Based on the classification results of the false alarm events, identify the dominant factors leading to the false alarms; For the aforementioned dominant factors, the confidence weights and the thresholds of the judgment rules are adjusted in segments. The segmented adjustment includes: for false alarms caused by specific work activities, adjusting the thresholds of parameters related to the work activities; for false alarms caused by instantaneous environmental fluctuations, adjusting the confidence weights. After the adjustment, a new validation period is started, and the false alarm rate during the new validation period is compared with the false alarm rate before the adjustment. If the false alarm rate decreases less than the magnitude threshold during the new verification period or a new false alarm type appears, the parameter configuration before the adjustment is reviewed and the adjustment strategy is re-evaluated.

10. A multi-sensor fire early warning data processing system, characterized in that, The system includes: The area information preset module is used to receive real-time environmental data from the detection devices in the preset warehouse. The real-time environmental data includes carbon monoxide concentration, smoke concentration, and ambient temperature. The preset warehouse includes different functional areas. The real-time data receiving module is used to identify the functional area to which the detection device belongs based on the detection device identifier; The functional area identification module is used to identify fire events in the functional area to which the detection device belongs, based on the functional area to which the detection device belongs and the real-time environmental data. Wherein, when the functional area is a first type of functional area, if only the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, the smoke concentration is less than the smoke concentration threshold, and the ambient temperature is less than the temperature threshold, then it is identified as a non-fire event; the first type of functional area is an area where there is interference from routine operations. When the functional area is a second type of functional area, if the carbon monoxide concentration is greater than or equal to the carbon monoxide concentration threshold, it is identified as a fire warning event; the second type of functional area is a lithium battery storage area. When the functional area is a third type of functional area, if the smoke concentration is greater than or equal to the smoke concentration threshold and the ambient temperature is greater than or equal to the temperature threshold, it is identified as a fire warning event; the third type of functional area is the area other than the first type of functional area and the second type of functional area. The alarm triggering module is used to trigger an alarm of the corresponding level based on the identification result of the fire event.

Images