Fire-fighting device self-checking control method and system

Abstract

The fire-fighting device self-checking control method and system adopt a multi-stage current limiting strategy composed of time-sharing wake-up, step-by-step power-on and slow start, effectively suppress the self-checking instantaneous current impact, and ensure the power supply safety; rely on CAN bus parallel communication, greatly improve the multi-node self-checking efficiency, the fire alarm response is rapid, and the blind area of protection is eliminated; through the back electromotive force waveform diagnosis, the hidden faults of the fire extinguishing circuit are effectively identified; combined with the multivariate compensation of the sensor and the bus anti-interference design, the false alarm rate under complex working conditions is reduced; the double redundant CAN architecture is adopted to realize seamless switching of the link, the relay can autonomously elect a temporary node to replace the dispatch when the main control is lost; the fire alarm instruction has the highest priority, supports self-checking instant interruption and breakpoint continuous checking; directional self-checking and hierarchical authority control are supported, the fault positioning time is shortened, the misoperation is eliminated, and most of the conventional faults can be automatically closed-loop handled by the system, effectively improving the safety and operation and maintenance efficiency of large-scale three-level networking fire-fighting devices.

Description

Technical Field

[0001] This invention relates to the field of automatic fire protection control technology, specifically to a self-test control method and system for fire protection devices. Background Technology

[0002] Electrochemical energy storage power stations, high-speed train battery compartments, and data center battery rooms are all high-risk fire scenarios. These areas have densely packed battery clusters and extremely rapid chain reactions of thermal runaway. From the initial thermal anomaly to a violent combustion and explosion, only 35 minutes are required, placing stringent demands on the real-time response and long-term operational reliability of the supporting fire protection systems.

[0003] Currently, mainstream distributed fire protection systems in the industry generally adopt a three-tier network architecture, consisting of a main control layer, a relay layer, and a field monitoring and execution layer from top to bottom. Taking a 100MW energy storage power station as an example, the entire system is configured with one main fire control device, 40 compartment-level relay devices, and 1600 cluster-level fire monitoring modules. The large number of network nodes and the complex link structure place extremely high demands on the system's self-test synchronization, power consumption control, and communication stability.

[0004] In accordance with the National Energy Administration's "Safety Regulations for Electrochemical Energy Storage Power Stations" and the mandatory requirements of the China State Railway Group's EMU onboard fire safety specifications, energy storage stations and onboard distributed fire protection devices must be compatible with three self-inspection modes: power-on self-inspection, periodic online self-inspection, and manual remote maintenance self-inspection. At the same time, strict constraints are clearly defined: accidental triggering of fire suppression spraying is strictly prohibited throughout the entire self-inspection process; dedicated fire monitoring channels must not be occupied during self-inspection operation; and core fire emergency functions such as fire alarms and emergency fire suppression linkage must not be interfered with.

[0005] Currently available traditional fire protection self-testing solutions, such as the energy storage fire protection systems disclosed by companies like Chuangwei Technology, generally employ a coexistence of dual-bus and CAN communication technologies. For example, CN220735995U discloses a relay module for a PACK-level fire protection system in an energy storage power station, whose CAN network includes both an upstream and downstream communication CAN network. However, while CAN communication is used between the control host and the relay module, and between the relay module and the detector, the space fire detection component and the fire alarm control device are still connected via a dual-bus connection. CN116392759A discloses a fire protection method for energy storage containers with self-testing capabilities, which monitors temperature and combustible gas concentration through detection devices. However, its self-testing function mainly focuses on threshold judgment at the detection level and does not involve end-to-end power consumption management or in-depth diagnosis of the fire extinguishing circuit. When these existing technical solutions are applied to large-scale multi-node fire protection network scenarios, they still have the following inherent technical defects: (i) Excessive instantaneous power consumption during synchronous self-testing can easily cause the fire protection power supply to shut down. Traditional solutions support multiple relays to perform parallel synchronous self-testing. When 40 compartment-level relays start self-testing at the same time, the instantaneous operating current of the system can reach 8A, far exceeding the rated output limit of 5A for fire protection DC power supplies. This can easily trigger the power overload protection, causing the entire fire protection system to lose power and go black, completely losing its fire protection capability.

[0006] (ii) The serial self-test takes too long and creates blind spots in fire protection. The traditional solution adopts a single-node polling serial detection mode. The main control board sends inspection pulses to the bus relay module and peripheral sensors. It takes more than 60 minutes to complete the full coverage self-test of 1,600 front-end monitoring nodes. If the fire occurs during the self-test window, the system cannot respond to the fire in time, which poses a major security vulnerability.

[0007] (III) Incomplete fire suppression link testing leads to the inability to detect hidden faults. Routine self-tests can only determine the online / offline status of equipment and cannot complete the continuity performance verification of the entire link from the main control to the end fire suppression solenoid valve and fire suppression activation circuit. According to industry operation and maintenance big data statistics, more than 25% of fire suppression failures are caused by hidden link faults that are not visible to the naked eye, such as aging lines and loose connections.

[0008] (iv) Insufficient equipment stability under complex operating conditions. The ambient temperature range of the energy storage station is 40℃ to 60℃. The operation of the train is accompanied by high-frequency mechanical vibration and strong electromagnetic interference of the whole vehicle. The traditional self-testing scheme lacks sensor adaptive calibration and communication fault tolerance protection mechanism, which is prone to sensor data drift and bus communication message errors. The overall fire false alarm rate of the equipment is as high as 15%.

[0009] (v) The hybrid bus communication has weak anti-interference capability. The existing system adopts a heterogeneous networking mode with CAN backbone bus and RS485 or two-bus branch. RS485 / two-bus has poor anti-electromagnetic interference capability, does not support multi-node synchronous response, and has a limited number of mounted nodes, which cannot meet the needs of parallel self-testing of large-scale equipment.

[0010] (vi) Risk of single point of failure in CAN bus. Most existing CAN bus fire protection systems adopt a single-channel CAN bus design. Once the main CAN bus experiences a line break or severe interference, all downstream devices will lose connection, resulting in insufficient system reliability.

[0011] (vii) The operation and maintenance logic is incomplete and cannot be adapted to the unattended operation and maintenance mode. Traditional solutions lack a highest priority preemption mechanism for fire alarms, and the system is prone to freezing when self-check commands conflict with fire alarm commands. At the same time, they lack breakpoint continuation testing, automatic closed-loop fault handling, and hierarchical operation permission control functions. Manual fault investigation can take up to 4 hours, which cannot meet the mainstream operation and maintenance needs of unattended or minimally staffed energy storage sites. Summary of the Invention

[0012] The present invention provides a self-test control method, equipment and storage medium for fire protection devices, which can at least solve one of the technical problems in the background art.

[0013] To achieve the above objectives, the present invention adopts the following technical solution: A self-test control method for fire protection devices involves executing the following steps via computer equipment. The power-on self-test process uses a time-sharing wake-up relay module and a step-by-step power-on peripheral method to limit the instantaneous startup current to below the preset safety threshold, and uses a bus broadcast method to perform grouped parallel self-tests on the monitoring modules of the terminals under its jurisdiction. The periodic self-inspection process dynamically calculates the self-inspection scheduling priority based on the historical operating status and risk parameters of each relay module. During the self-inspection process, the current self-inspection is interrupted in response to the fire alarm signal and the breakpoint is saved. The self-inspection is resumed from the breakpoint after the fire alarm is lifted. The system maintains a self-inspection process, receives self-inspection instructions issued manually after hierarchical authorization verification, and collects the back electromotive force waveform of the solenoid valve while performing short-term physical isolation of the fire extinguishing circuit. Based on the comparison results between the waveform and the reference waveform, the system determines the hidden faults in the fire extinguishing circuit. The bus communication in the power-on self-test process, periodic self-test process, and maintenance self-test process is all based on the unified CAN bus protocol, and the fire monitoring channel is kept active throughout the self-test process.

[0014] A self-test control system for fire protection devices, comprising: The fire control module is deployed in the central control room or electrical control cabinet, serving as the core of self-testing and dispatching and the data aggregation center. N relay modules are deployed in each compartment or carriage for data forwarding between upper and lower levels and edge control. M terminal monitoring modules are deployed at each battery cluster or monitoring point, and each module integrates a sensor and a fire extinguishing circuit drive circuit. The fire control module, the relay module, and the terminal monitoring module are connected via a dual-layer full CAN bus architecture. The dual-layer full CAN bus architecture includes at least: a main redundant CAN bus connecting the fire control module and all N relay modules, and a branch CAN bus independently led out by each relay module for mounting the terminal monitoring module within its domain. The fire control module also has a separate CAN bus for connecting to debugging equipment; all external communication interfaces in the system include only CAN bus interface and Ethernet interface, and there is no serial communication port.

[0015] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0016] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0017] As can be seen from the above technical solution, the self-test control method and system for fire-fighting devices of the present invention, through a triple current-limiting strategy consisting of a time-sharing wake-up relay module, peripheral step-by-step power-on, and a soft-start circuit, stably controls the instantaneous current of the whole machine during self-test to within 1A, far below the 5A rated threshold of the fire-fighting power supply, completely eliminating power tripping and shutdown caused by synchronous self-testing. At the same time, the device's sleep power consumption is less than 100μA, significantly reducing standby power consumption. Relying on the dual-layer CAN bus parallel communication and group parallel detection mechanism, the self-testing time of 1600 front-end nodes across the entire domain is shortened from more than 60 minutes in the traditional solution to less than 1 minute, far less than the 3 to 5 minutes golden rescue time for battery thermal runaway, thus greatly reducing the fire alarm response delay. With a latency of less than 100ms and uninterrupted monitoring throughout the entire process, the safety blind spot during the self-test window is completely eliminated. Employing back-EMF waveform diagnostic technology, combined with multi-dimensional quantitative judgment based on waveform similarity, peak attenuation rate, and time constant change rate, it can accurately identify hidden faults that traditional solutions cannot detect, such as loose connections, short circuits between coil turns, and aging lines. The reliability of the fire extinguishing system has jumped from 75% to 99.9%, and the risk of circuit failure can be predicted 6 to 12 months in advance. Through a multi-variable compensation model for sensors combined with online coefficient learning, and leveraging the inherent anti-interference characteristics of CAN bus differential signals, the sensor false alarm rate has been reduced from over 15% to less than 1%, and the bus bit error rate is controlled at 1×10⁻⁶. -9 The system is stable and adaptable to extreme operating conditions such as wide temperature range, vibration, and strong electromagnetic interference. The dual-channel CAN bus backbone provides hot backup for each other, enabling seamless millisecond-level switching in case of single-channel failure. When the main control unit loses connection, the relay module can autonomously elect a temporary management node to continue self-testing and scheduling, achieving 99.99% system availability. Fire alarm commands have the highest system priority, supporting immediate interruption and breakpoint resumption of self-tests. Background self-tests only start when the system is idle, ensuring that core emergency fire protection functions always operate with priority. It supports single-point and single-compartment directional self-tests, reducing fault location time from 4 hours to less than 10 minutes. Four-level access control prevents misoperation, and over 80% of common faults can be automatically closed-loop handled by the system without manual intervention, fully meeting the requirements of unattended operation and maintenance. This invention systematically improves the self-testing performance of large-scale three-level networked fire protection devices from six dimensions: power consumption control, detection efficiency, diagnostic depth, environmental adaptability, system reliability, and intelligent operation and maintenance. It can effectively ensure fire safety in high-risk scenarios such as electrochemical energy storage power stations and high-speed trains. Attached Figure Description

[0018] Figure 1 Figure 1 This is a schematic diagram of the three-level hardware architecture of the self-test control system for the fire protection device of the present invention; Figure 2 This is a schematic diagram of the dual-layer full CAN bus communication architecture of the present invention; Figure 3 This is a flowchart illustrating the power-on BIT self-test control method of the present invention. Figure 4 This is a flowchart illustrating the periodic BIT self-test control method of the present invention. Figure 5 This is a flowchart illustrating the maintenance BIT self-test control method of the present invention; Figure 6 This is a waveform comparison diagram for back electromotive force waveform diagnosis of the fire extinguishing circuit of the present invention. Figure 7 This is a schematic diagram of the process for self-checking the distributed autonomous system when the master controller loses connection, as described in this invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0020] The self-test control method for fire protection devices described in this embodiment involves the following steps performed via computer equipment: The power-on self-test process (power-on BIT) uses a time-sharing wake-up relay module and a step-by-step power-on peripheral method to limit the instantaneous startup current to below the preset safety threshold, and uses a bus broadcast method to perform grouped parallel self-tests on the monitoring modules of the terminals under its jurisdiction. The periodic self-test process (periodic BIT) dynamically calculates the self-test scheduling priority based on the historical operating status and risk parameters of each relay module. During the self-test, the current self-test is interrupted in response to the fire alarm signal and the breakpoint is saved. The self-test is resumed from the breakpoint after the fire alarm is lifted. The system maintains a self-inspection process (maintaining BIT), receives manual self-inspection instructions after hierarchical permission verification, and collects the back electromotive force waveform of the solenoid valve while performing short-term physical isolation of the fire extinguishing circuit. Based on the comparison result of the waveform and the reference waveform, the system determines the hidden faults in the fire extinguishing circuit.

[0021] A self-test control system for fire protection devices, comprising: The fire control module is deployed in the central control room or electrical control cabinet, serving as the core of self-testing and dispatching and the data aggregation center. N relay modules are deployed in each compartment or carriage for data forwarding between upper and lower levels and edge control. M terminal monitoring modules are deployed at each battery cluster or monitoring point, and each module integrates a sensor and a fire extinguishing circuit drive circuit. The fire control module, the relay module, and the terminal monitoring module are connected via a dual-layer full CAN bus architecture. The dual-layer full CAN bus architecture includes at least: a main redundant CAN bus connecting the fire control module and all N relay modules, and a branch CAN bus independently led out by each relay module for mounting the terminal monitoring module within its domain. The fire control module also has a separate CAN bus for connecting to debugging equipment; all external communication interfaces in the system include only CAN bus interface and Ethernet interface, and there is no serial communication port.

[0022] In practical implementation: 1. System Architecture like Figure 1 As shown, the fire protection device self-test control system upon which this invention is based includes a three-level hardware architecture and a dual-layer full CAN bus communication architecture.

[0023] 1.1 Three-tier hardware functional architecture The entire hardware system retains only two external interfaces: CAN bus and Ethernet. There are no serial communication ports. The three-level devices have clearly defined functions and hierarchical linkages. Fire protection main control module: Deployed in the central control room of the energy storage power station or the electrical control cabinet of the train, it serves as the scheduling core and data aggregation center of the entire self-testing system. The fire protection main control module is equipped with an industrial-grade processor and features at least three independent CAN bus interfaces and one Ethernet interface. Two of the CAN interfaces form a redundant backbone bus for hot backup of the communication link; the other CAN bus is used for local debugging, firmware upgrades, and log export. The fire protection main control module has a built-in large-capacity Flash storage unit, capable of retaining at least one year of self-testing logs, fault records, and fire alarm linkage data; it is equipped with a dual-power hot backup circuit, enabling seamless millisecond-level backup power switching in the event of a main power failure.

[0024] Cabin-level relay modules: One module is configured in each battery container or each train car, totaling N modules (typically N=40), responsible for data forwarding between upper and lower levels and edge control functions. Each cabin-level relay module is equipped with an industrial-grade processor and at least three independent CAN interfaces: two connected to the main redundant CAN bus and one connected to the cabin branch CAN bus. The module supports a wide voltage supply of DC 9V to 36V and operates within a temperature range of 40℃ to 85℃. It has a built-in edge computing unit capable of locally performing front-end node data preprocessing, initial fault identification, and bus fault tolerance processing; it supports switching between sleep and normal operation modes.

[0025] Cluster-level monitoring module: Deployed in a one-to-one configuration per battery cluster, with a total of M units (typical value M=1600), directly connecting to on-site fire detection equipment and end-of-line fire suppression devices. The cluster-level monitoring module integrates a high-precision temperature sensor, photoelectric smoke sensor, electrochemical carbon monoxide sensor, and fire suppression circuit drive circuit; it has a built-in backup lithium battery, allowing for independent operation for over 72 hours after a mains power outage. All command interaction and data transmission are completed via the CAN bus.

[0026] 1.2 Dual-layer full CAN bus communication architecture like Figure 2 As shown, the system adopts a dual-layer redundant CAN bus architecture, using the CAN2.0B communication protocol throughout the network. It eliminates heterogeneous branch buses, thus addressing the issues of poor compatibility and weak anti-interference in hybrid bus communication from the ground up. The bus layer design is as follows: Redundant CAN bus backbone: Connects the fire control main unit to all N compartment-level relay modules, using shielded twisted-pair cabling, with a communication baud rate of 250kbps and a standard transmission distance of 300m, which can be extended to 1000m with a fiber optic converter. The dual CAN buses provide hot backup for each other; if the main bus is interfered with or the line is broken, the system automatically switches to the backup bus, ensuring uninterrupted communication throughout the entire process.

[0027] Branch CAN bus: Each compartment relay module independently leads out 1 branch CAN bus, which is connected to the cluster-level monitoring module in this compartment. The communication baud rate is 500kbps, the stable transmission distance of a single bus is 100m, and a maximum of 110 nodes can be connected to a single bus.

[0028] Debugging CAN bus: A dedicated debugging CAN bus is led out from the fire control module to connect to debugging equipment to complete maintenance operations such as parameter configuration, firmware upgrade, log export and single-point forced self-test, which is physically isolated from the running backbone service bus.

[0029] The entire network retains only two external interfaces: CAN bus and Ethernet. There are no serial communication ports, and heterogeneous serial bus interfaces such as RS485 and RS232 have been completely eliminated.

[0030] The full CAN bus architecture has the following core advantages: (1) Excellent resistance to strong electromagnetic interference: The CAN bus adopts differential signal transmission, which can naturally suppress common-mode electromagnetic interference; (2) Support for parallel response of multiple nodes: Relying on the CAN bus broadcast communication mechanism, the relay can simultaneously send self-test commands to all front-end nodes in the cabin; (3) Perfect fault tolerance capability of the underlying communication: The bus is natively equipped with hardware CRC check, automatic error retransmission, and bus priority arbitration mechanism; (4) Strong system scalability: The number of nodes mounted on a single bus is sufficient; (5) Simplified and unified hardware structure: The entire network only retains two types of interfaces: CAN bus and Ethernet.

[0031] 1.3 Dual-mode power consumption control mechanism The main controller issues power scheduling commands to all devices at all levels of the system, which are divided into two working modes: Normal power consumption mode: All peripherals are turned on and the processor runs at full speed, suitable for full-function operation scenarios such as device self-test, real-time fire monitoring, and emergency fire fighting linkage.

[0032] Low-power sleep mode: Non-essential acquisition peripherals and processing units are turned off, the processor enters a deep sleep state, and only the CAN bus listening and wake-up circuit is retained. The device's static power consumption is less than 100μA.

[0033] 2. Core Three Types of BIT Self-Inspection Control Methods like Figure 3 As shown, this invention designs three self-test modes: Power-on BIT, Periodic BIT, and Maintenance BIT. These three modes complement each other based on their different triggering conditions, coverage, and detection depth: Power-on BIT is responsible for full power-on verification to ensure that the equipment can be put into operation without faults; Periodic BIT is responsible for routine background inspections to detect slow deterioration faults in advance; and Maintenance BIT is responsible for manual targeted investigation to meet the needs of operation and maintenance and fault location.

[0034] 2.1 Power-on BIT Forced Self-Test Method The power-on BIT self-test is applicable to the initial power-on of the system, automatic startup after power failure recovery and restart, and automatic startup after firmware upgrade restart. Its core objective is to comprehensively verify the integrity of the entire network hardware, communication links, and fire suppression circuits upon power-on.

[0035] The specific steps are as follows: Step 1: Main Control Self-Test and Soft Start Control. After powering on, the fire protection main control module first completes self-tests of its own CPU, storage unit, various CAN interfaces, dual backup power supplies, and fire extinguishing linkage motherboard. At the same time, through the front-end soft start circuit, it limits the rise rate of the overall current to within a preset rise rate threshold (preferably 0.1A / ms) to eliminate the instantaneous inrush current caused by capacitor charging.

[0036] Step Two: Time-Sharing Wake-up and Scheduling of Relays. The fire control module prioritizes the relay modules of each compartment according to the fire risk of the compartment, and adopts a time-sharing wake-up strategy. Only one compartment-level relay module is woken up and put into working state at the same time, while the remaining compartment-level relay modules remain dormant, thus avoiding current overload caused by multiple devices working simultaneously from the source.

[0037] Step 3: Relay-level self-test. The awakened cabin-level relay module completes self-tests of its own power supply, dual main CAN interfaces, branch CAN interfaces, internal relays, and circuit insulation performance.

[0038] Step Four: Hierarchical Power-On Control for Peripherals. Peripherals of the cabin-level relay module are powered on sequentially in the following order: power module → communication interface → audible and visual alarm peripherals → fire extinguishing detection circuit. The next level is activated only after the previous level has passed the test. Instantaneous operating current must meet the following requirements: ; Among them, This refers to the instantaneous operating current of the cabin-level relay module. Let n be the operating current of the k-th peripheral device in operation, and n be the total number of peripheral devices that are powered on. This is the power supply safety threshold. Through this strategy, the instantaneous current during the system's self-test is stably controlled within 1A.

[0039] Step 5: Front-end nodes perform parallel self-tests in groups. The cabin-level relay module divides its cluster-level monitoring modules into several groups (preferably 4 groups) for parallel testing. Self-test tasks are uniformly issued based on CAN broadcast commands. Each cluster-level monitoring module uploads data in time-sharing according to its own CANID, without message conflicts.

[0040] Step Six: Bus Status Monitoring and Anomaly Isolation. The cabin-level relay module calculates the bus packet loss rate in real time using the following formula: ; in The current bus packet loss rate, This represents the number of erroneous or lost messages per unit of time. This represents the total number of messages that should be received per unit of time. When the packet loss rate exceeds the preset fault threshold (preferably 10%), the abnormal node is locally logically isolated.

[0041] Preferably, the packet loss rate fault threshold is a dynamic threshold, calculated according to the following formula: ; in Based on the basic packet loss rate threshold, The standard deviation of the recent packet loss rate. The adjustment coefficient is 23 (preferably).

[0042] Step 7: Single-compartment self-test completion. After the self-test of a single compartment-level relay module is completed, reset the fire extinguishing isolation relay, upload the self-test data, and switch to sleep mode.

[0043] Step 8: Comprehensive Self-Inspection Summary. The fire control module summarizes the self-inspection data of all compartment-level relay modules and generates a comprehensive power-on self-inspection report; when a high-risk fault is detected, an audible and visual alarm is immediately triggered and the fire protection system is locked. The total power-on self-inspection time for the entire network is controlled within 1 minute. 2.2 Periodic BIT Online Self-Test Method like Figure 4As shown, the periodic BIT self-test is suitable for background operation without human intervention during normal system maintenance, and continuously monitors slow deterioration faults such as line aging, hidden bus errors, and increased battery internal resistance.

[0044] The specific steps are as follows: Step 1: Self-test Start Authorization. After the self-test cycle is reached, the fire control module sends a background self-test request to the superior fire platform, continuously monitors the bus fire intelligence messages and operation and maintenance instructions, and only starts the self-test when the system is completely idle.

[0045] Step 2: Dynamic Calculation of Relay Self-Check Priority. The fire control module dynamically calculates the relay self-check priority by combining three parameters: the historical number of failures of each compartment's relay module, the compartment's risk level, and the time interval since the last self-check.

[0046] Preferably, the self-test priority $P_i$ is calculated according to the following nonlinear mapping formula: ; in, This is the Sigmoid function, used to map priority values ​​to the (0,1) interval; The number of historical failures for the i-th relay module; The risk level of the area where the i-th relay module is located; The time interval between the i-th relay module and its last self-test; Basic self-inspection cycle; This is the set of relay modules adjacent to the i-th relay module; The spatial correlation coefficient represents the degree of fire spread risk correlation between the i-th relay module and the adjacent j-th relay module. is the historical number of failures of the j-th adjacent relay module; w1, w2, w3, and w4 are the weighting coefficients of the corresponding parameters.

[0047] Step 3: Differentiated Testing by Grade. Testing is performed according to the equipment's health status: Cluster-level monitoring modules with stable health status only undergo heartbeat detection, voltage detection, and bit error rate detection; cluster-level monitoring modules with occasional anomalies undergo additional impedance scanning and battery internal resistance detection; cluster-level monitoring modules with a history of faults undergo a full-parameter deep self-test.

[0048] Step 4: Sensor Environmental Adaptive Calibration. Simultaneously collect on-site temperature, vibration, and humidity data, and correct sensor zero-point drift using an adaptive compensation algorithm.

[0049] Collect ambient temperature T, vibration frequency f, and ambient humidity H at the site; Correct sensor readings using a multivariable compensation model : ; in, These are the original values ​​collected by the sensor. Reference temperature For reference humidity, , , These are the compensation coefficients for temperature, vibration, and humidity, respectively. The temperature compensation coefficient is updated in real time using the online gradient descent method. : ; in, For learning rate, For calibration reference values, The ambient temperature collected at the current moment; the vibration compensation coefficient Humidity compensation coefficient Update synchronously in the same way.

[0050] Step 5: Fire Alarm Emergency Response and Resumption of Inspection. The self-test monitors fire alarm messages in real time throughout the process. Upon detecting a fire, the self-test immediately terminates, switches to emergency mode, and records the current progress of the self-test slice. After the fire is dealt with, the test resumes from the interrupted point. There is a response delay in the self-test if the fire alarm signal is interrupted. satisfy: ; in, Due to delays in fire detection, The delay between self-test task interruption and fire alarm task switching is preferably ≤10ms. Allow for a delay in saving self-test breakpoints (ideally ≤5ms).

[0051] The self-test task is broken down into the smallest task slices of no more than 200ms, and only the current slice is interrupted when a fire alarm occurs.

[0052] Step Six: Self-Inspection Cycle Adaptive Adjustment. The next self-inspection cycle will be dynamically adjusted based on the results of this self-inspection. Next Self-Inspection Cycle Calculate using the following formula with a sliding window and a rate of change limit: ; ; in, This is the current self-inspection cycle; This is the maximum self-test cycle threshold. The minimum self-test cycle threshold; Basic self-inspection cycle; This represents the maximum adjustment percentage in a single instance, with a range of 5% to 10%. This represents the number of faults in the W most recent self-test results within the sliding window; The number of normal self-checks within the sliding window; The fault penalty coefficient and >1, For health reward coefficient and <1; The width W of the sliding window is a preset positive integer (W is preferably 1020).

[0053] 2.3 Maintenance BIT-oriented self-testing method like Figure 5 As shown, the maintenance BIT self-test is suitable for manual triggering during equipment installation and acceptance, quarterly operation and maintenance, on-site fault diagnosis, and fire extinguishing circuit repair. It supports three detection modes: full domain, single compartment, and single node.

[0054] The specific steps are as follows: Step 1: Operation Permission Verification. All maintenance self-test commands are issued via the debugging CAN bus or an encrypted and authenticated remote channel, physically isolated from the running backbone service bus. The system performs four levels of pre-access permission verification: observer read-only data permission, maintenance personnel partial self-test permission, engineer deep loop detection permission, and administrator device configuration permission.

[0055] Step Two: Detection Mode Selection. Maintenance personnel select the detection mode as needed: full network inspection for system acceptance, single-compartment inspection for regional anomalies, and precise node inspection for single-point alarms.

[0056] Step 3: Full Data Capture. During the self-test, capture all original interaction messages on the bus and simultaneously detect underlying hardware parameters such as chip load, power ripple, and system clock.

[0057] Step 4: Non-destructive testing of the fire extinguishing circuit. A hardware isolation relay is used to briefly disconnect the actual fire extinguishing circuit (disconnection time not exceeding 50ms). The back electromotive force waveform generated at the moment of de-energization of the fire extinguishing solenoid valve is collected and compared with a reference waveform.

[0058] Calculate the similarity S between the measured back electromotive force waveform and the reference waveform: ; in, To measure the amplitude of the back electromotive force waveform at the t-th sampling point, The amplitude of the reference waveform at the t-th sampling point is... The mean of the measured waveform. K represents the mean of the reference waveform, and K is the number of sampling points.

[0059] Based on the similarity Peak attenuation deviation and the rate of change of decay time constant The fault type is identified according to the following fault determination matrix:

[0060] Calculate the diagnostic confidence level C: ; in, , , These are the standard deviations of measurement noise for similarity, peak attenuation rate deviation, and rate of change of attenuation time constant, respectively.

[0061] Step 5: Data Verification and Report Generation. The fire control module performs a second CRC check on the uploaded messages and automatically generates an operation and maintenance report containing the fault location, fault cause, and standardized repair plan. The data is encrypted, stored, and cannot be tampered with.

[0062] 3. Distributed autonomous self-test when the main controller is out of contact. like Figure 7 As shown, this invention also provides a distributed autonomous self-checking mechanism for scenarios where the master controller is disconnected: The cabin-level relay module periodically listens to the heartbeat messages broadcast by the fire control module (preferably once every 5 seconds); When the cabin-level relay module does not receive a heartbeat message for M consecutive heartbeat cycles (M is preferred to be 35), it is determined that the fire control module is out of contact. Each cabin-level relay module calculates the temporary management node score according to the following formula. The node with the highest score is then elected as the temporary management node. ; in, Let be the processing power of the i-th relay module. This represents the maximum processing power of the processors across all relay modules. Let be the remaining backup power capacity of the i-th relay module. The maximum value of the remaining backup power capacity among all relay modules. Let be the communication link quality of the i-th relay module. is the maximum value of the communication link quality among all relay modules; a, b, and c are the weighting coefficients of the corresponding parameters, respectively; The temporary management node takes over the self-test scheduling function from the fire control module. The self-test data during offline periods is cached in the local Flash memory and automatically re-uploaded after the fire control module is restored.

[0063] Example 1: Application of a 100MW electrochemical energy storage power station This embodiment uses a 100MW electrochemical energy storage power station as the standard application scenario. The system configuration includes: 1 fire-fighting main control module, 40 compartment-level relay modules, and 1600 cluster-level monitoring modules. Its hardware architecture is as follows: Figure 1 As shown, the communication bus topology is as follows: Figure 2 As shown.

[0064] The fire control module is deployed in the central control room of the energy storage power station, equipped with an industrial-grade processor and three independent CAN bus interfaces. Two of these CAN interfaces form a redundant backbone CAN bus, connecting all 40 compartment-level relay modules, with a communication baud rate of 250kbps; the remaining CAN interface serves as a debugging CAN bus for local debugging and maintenance.

[0065] Each compartment-level relay module is deployed within its corresponding battery container, equipped with an industrial-grade processor and three independent CAN interfaces. Two of these interfaces connect to the main redundant CAN bus, and one interface leads out to a branch CAN bus, connecting to 40 cluster-level monitoring modules within the compartment, with a communication baud rate of 500kbps.

[0066] Each cluster-level monitoring module is deployed in a corresponding battery cluster, integrating a temperature sensor, a photoelectric smoke sensor, an electrochemical carbon monoxide sensor, and a fire extinguishing circuit drive circuit, and has a built-in backup lithium battery.

[0067] After the system is powered on, the power-on bit self-test process is automatically triggered (see reference). Figure 3 The fire control module first completes its self-test, limiting the current rise rate to within 0.1A / ms through a soft-start circuit. Then, it wakes up 40 compartment-level relay modules sequentially according to the compartment's risk level. After each relay completes its own self-test, it powers on in the following order: power module → communication interface → audible and visual alarm peripherals → fire extinguishing detection circuit. The 40 cluster-level monitoring modules are then divided into four groups for parallel testing, with self-test tasks uniformly distributed via CAN broadcast commands. Each node uploads data in a time-sharing manner based on its CAN ID. Actual measurements show that the instantaneous current during the overall self-test is stably controlled within 1A, and the total self-test time for the entire area is controlled within 1 minute.

[0068] After the system enters normalized operation, the periodic BIT self-check will be automatically triggered once every 24 hours (reference). Figure 4The fire control module dynamically calculates the self-test sequence based on the historical fault count of each relay, the risk level of the compartment, and the time interval since the last self-test, using a non-linear priority formula. During the self-test, sensors simultaneously collect ambient temperature, vibration, and humidity data, and the collected values ​​are corrected in real time using a multivariate compensation model. The compensation coefficients are continuously updated using an online gradient descent method. If a fire is detected during the self-test, the system immediately terminates the self-test and saves the breakpoint. The fire alarm response delay is less than 100ms, and the self-test continues from the breakpoint after the fire is dealt with. After the self-test is completed, the next self-test cycle is dynamically adjusted based on the results.

[0069] When quarterly maintenance and repair are required, maintenance personnel issue maintenance BIT self-test commands via the CAN bus or an encrypted and authenticated remote channel (see reference). Figure 5 After the system completes four-level permission verification, the maintenance personnel select the single-compartment testing mode. During the self-test, the actual fire extinguishing circuit is briefly disconnected using an isolation relay (disconnection time ≤ 50ms), and the back electromotive force waveform at the moment of solenoid valve de-energization is collected (waveform comparison reference). Figure 6 The method calculates waveform similarity and, combined with peak attenuation rate and time constant change rate, identifies circuit status according to the fault judgment matrix. Actual testing shows that this method can accurately identify latent faults such as loose connections and inter-turn short circuits in coils, achieving a fault detection rate of 99.9% for fire extinguishing circuits.

[0070] When the fire control module loses connection due to an extreme malfunction, each compartment's relay module triggers a distributed autonomous self-test via a heartbeat detection mechanism (every 5 seconds; failure to receive a heartbeat after 3 consecutive heartbeats indicates a loss of connection). Figure 7 Each relay is scored based on its processor computing power, remaining backup power capacity, and communication link quality, and a temporary management node is elected to take over the main control scheduling function. Offline self-test data is cached in local Flash and automatically retransmitted after the main control recovers.

[0071] Example 2: Application of power battery compartment in high-speed trains This embodiment uses the power battery compartment of a high-speed train as the application scenario, and the system configuration is as follows: 1 on-board fire-fighting main control module, 8 carriage-level relay modules, and 32 cluster-level monitoring modules.

[0072] Because high-frequency mechanical vibrations and strong electromagnetic interference occur during the operation of high-speed trains, the system's anti-interference capabilities are extremely high. This invention addresses this by... Figure 2 A dual-layer full CAN bus architecture (CAN bus uses differential signal transmission, naturally suppressing common-mode electromagnetic interference) combined with sensor environment adaptive calibration ( Figure 4 Step 4) allows for stable operation under high-frequency vibration and strong electromagnetic interference environments. Actual measurements show that the sensor false alarm rate is ≤1%, and the bus communication false judgment rate is ≤0.5%.

[0073] The remaining implementation methods are basically the same as those in Example 1, and will not be described again here.

[0074] Example 3: Specific implementation of back EMF waveform diagnosis for fire extinguishing circuits, and the corresponding waveforms, for example... Figure 6 As shown This embodiment details the specific implementation method for diagnosing the back electromotive force waveform of the fire extinguishing circuit.

[0075] During the fire suppression circuit detection phase of the BIT self-test, the system briefly disconnects the actual fire suppression circuit via a hardware isolation relay (disconnection duration set to 50ms). At the moment of disconnection, the fire suppression solenoid valve coil generates a back electromotive force due to a sudden change in current. This back electromotive force waveform is acquired by the sampling circuit and sent to the processor.

[0076] The processor first performs filtering and normalization preprocessing on the acquired waveform, and then performs diagnosis according to the following steps: The first step is to calculate the difference between the measured waveform and the reference waveform stored at the factory. Figure 6 The similarity S (similar to the solid green line in the middle) is used. If S ≥ 0.98, the loop is considered intact; if S < 0.98, proceed to the next step.

[0077] The second step is to extract the peak attenuation rate of the waveform. and the rate of change of decay time constant .

[0078] The third step is to identify the fault type according to the fault determination matrix, such as... Figure 6 As shown, the orange dashed line represents the waveform of a line connection fault (slow oscillation decay), and the red dotted line represents the waveform of an inter-turn short circuit fault (severe distortion, rapid decay). The specific judgment rules are as follows: If 0.85≤S<0.90 and peak attenuation rate>15%, it is determined to be a loose connection in the line; If S < 0.85 and the decay time constant changes by more than 15%, it is determined to be an inter-turn short circuit in the coil; If S < 0.60 and peak attenuation rate > 80%, the coil is considered completely open-circuited. If the time constant change is greater than 8% and the peak attenuation rate is less than 10%, the line is considered to be aging.

[0079] The fourth step is to calculate the diagnostic confidence level C. When C ≥ 0.95, the fault determination result is reliably output; when C < 0.95, the system automatically repeats the test once to confirm the result.

[0080] Actual testing has shown that this method can predict the risk of fire extinguishing circuit failure 612 months in advance without spraying any extinguishing medium, and improve the reliability of the fire extinguishing system from 75% to 99.9% compared to traditional methods.

[0081] Experimental Example To verify the technical effectiveness of this invention, a comparative test was conducted in a 100MW energy storage power station simulation environment. The system was configured with one main controller, 40 relays, and 1600 front-end nodes. The test results are as follows: Experiment Example 1: Self-test Instantaneous Current Comparison Test

[0082] Test conditions: Fire protection power supply rated output 5A, ambient temperature 25℃. In traditional solutions, when 40 compartment-level relays simultaneously start self-test, the system's instantaneous operating current reaches 8A, far exceeding the 5A rated limit, triggering power overload protection and causing a system power outage. This invention employs a triple current-limiting strategy—relay time-sharing wake-up, step-by-step power-on of peripherals, and soft-start of the front-end power supply—to stably control the instantaneous self-test current of the entire unit to within 1A.

[0083] Experiment Example 2: Comparison Test of Global Self-Test Time

[0084] Test conditions: All 1600 cluster-level monitoring modules are online. Traditional solutions use a single-node polling serial detection mode, requiring more than 60 minutes to complete a full-coverage self-test. This invention relies on a dual-layer CAN bus parallel communication and group parallel detection strategy, achieving a full-domain self-test time of ≤1 minute, far shorter than the 35-minute golden rescue time for battery thermal runaway.

[0085] Experiment Example 3: Comparative Test of Fault Detection Rate of Fire Extinguishing Circuits

[0086] Test conditions: Twenty instances each of three types of latent faults were artificially introduced in 1600 fire extinguishing circuits: loose wiring (contact resistance 0.52Ω), inter-turn short circuit (520 turns of inter-turn short circuit), and aging wiring (20%-50% increase in cable resistance). Traditional continuity testing methods cannot identify these latent faults. This invention, by acquiring the back electromotive force waveform at the moment of solenoid valve de-energization, calculating waveform similarity, and combining it with a multi-dimensional fault judgment matrix, can accurately identify all three types of latent faults, improving the reliability of the fire extinguishing system from 75% to 99.9%.

[0087] Experiment Example 4: Comparative Test of False Alarm Rates under Complex Operating Conditions

[0088] Test conditions: 168 hours of continuous operation were conducted under various environmental conditions, including a 40℃ low-temperature environment, an 85℃ high-temperature environment, a 10g high-frequency vibration platform, and a 100V / m strong electromagnetic interference field. Traditional solutions lack a sensor-based adaptive calibration mechanism, resulting in a false alarm rate as high as 15%-25% under various operating conditions. The proposed solution utilizes a multivariate compensation model (three-dimensional compensation for temperature, vibration, and humidity) and online gradient descent coefficient learning, achieving a false alarm rate ≤1% under all operating conditions.

[0089] Experiment Example 5: Comparative Test of Bus Communication Reliability

[0090] Test conditions: The bus is manually disconnected at a midpoint in the bus transmission path, applying strong electromagnetic interference of 100V / m. In a traditional single-channel CAN solution, all downstream devices lose connection after the bus is disconnected. In the dual-redundant CAN bus architecture of this invention, the system automatically switches to the backup bus within milliseconds when the main bus is disconnected, ensuring uninterrupted communication throughout the entire process. The differential signal transmission of the CAN bus naturally suppresses common-mode interference, and with hardware CRC verification, the bit error rate under strong interference is ≤1×10⁻⁶. -9 .

[0091] Experiment Example 6: System Availability Comparison Test under Master Control Loss Scenario

[0092] Test conditions: Simulating a complete power outage of the fire control module due to a power failure. In traditional solutions, all equipment becomes paralyzed and self-testing is completely interrupted after the main control module loses contact. In this invention, each compartment-level relay module triggers a distributed autonomous self-test through a heartbeat detection mechanism (once every 5 seconds, a loss of contact is determined after 3 consecutive heartbeats). A temporary management node is elected based on processor computing power, remaining backup power capacity, and communication link quality. Offline self-test data is cached in local Flash memory and automatically retransmitted after the main control module recovers, achieving a system availability of 99.99%.

[0093] In summary, this application has the following beneficial effects: 1. Tiered current limiting and power consumption control completely solves the problem of power overload and shutdown. Through a triple current limiting strategy of relay time-sharing wake-up, peripheral step-by-step power-up, and front-end power supply soft start, the instantaneous current of the whole machine during self-test is stably controlled within 1A, far below the 5A rated threshold of the fire protection power supply, completely eliminating power tripping and system black screen failures caused by synchronous self-test; the static power consumption of the equipment during sleep is less than 100μA.

[0094] 2. Parallel self-testing significantly speeds up the process, eliminating blind spots in fire protection time. Relying on dual-layer CAN bus parallel communication and task slicing scheduling algorithms, the self-testing time of 1600 front-end nodes is ≤1 minute, far shorter than the 3-5 minute golden rescue time for battery thermal runaway; fire alarm commands unconditionally preempt self-testing resources, fire alarm response latency is less than 100ms, and fire monitoring is uninterrupted throughout the process.

[0095] 3. Non-destructive testing of the entire fire extinguishing circuit eliminates hidden fire hazards. By using coil back electromotive force waveform diagnostic technology, combined with waveform similarity calculation, multi-dimensional fault judgment matrix and diagnostic confidence assessment, hidden faults that traditional self-inspection cannot detect, such as loose wiring, aging wiring, and deterioration of solenoid valve coils, can be accurately identified, increasing the reliability of the fire extinguishing system from 75% to 99.9%.

[0096] 4. Adaptable to extreme and complex working conditions, significantly improving system stability. Equipped with adaptive sensor calibration (multivariable compensation model + online coefficient learning) and a dual-layer CAN fault-tolerant communication mechanism, it adapts to wide-temperature, high-frequency vibration, and strong electromagnetic interference field conditions, with a sensor false alarm rate ≤1%.

[0097] 5. Dual-redundant CAN bus architecture eliminates the risk of single-point failure. The dual CAN bus backbones serve as hot backups for each other. When a single bus is interfered with or disconnected, the system automatically and seamlessly switches over, ensuring uninterrupted communication and significantly improving system availability.

[0098] 6. Fire protection business priority closed loop, self-test and emergency functions do not conflict. The fire command system is clearly defined as the highest priority, and supports immediate interruption and resume of self-test; background self-test can only be started with idle authorization, ensuring that the core fire protection emergency functions always operate with priority.

[0099] 7. Hierarchical access control + targeted self-inspection significantly improves maintenance and troubleshooting efficiency. Supports precise self-inspection of single points and single compartments, eliminating the need for whole-machine retesting, and reducing manual fault location time from 4 hours to less than 10 minutes.

[0100] 8. Adapts to unattended operation mode, reducing the total lifecycle maintenance cost. More than 80% of common faults can be automatically repaired by the system in a closed loop (including distributed autonomous self-testing in the case of main control failure), without the need for manual on-site handling.

[0101] 9. A unified CAN bus architecture across the entire network, resulting in simplified hardware and enhanced scalability. The heterogeneous serial bus 27 is completely eliminated, ensuring unified interfaces and standardized cabling across the entire network; the CAN bus reserves ample node capacity, eliminating the need to modify existing cabling for future expansion of fire protection equipment.

[0102] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0103] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0104] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the fire-fighting device self-test control methods described in the above embodiments.

[0105] It is understood that the systems, devices, and storage media provided in the embodiments of the present invention correspond to the methods provided in the embodiments of the present invention, and the explanations, examples, and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

[0106] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state disk (SSD)).

[0107] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0108] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0109] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A self-inspection control method for fire-fighting devices, characterized in that, Perform the following steps using a computer device: The power-on self-test process uses a time-sharing wake-up relay module and a step-by-step power-on peripheral method to limit the instantaneous startup current to below the preset safety threshold, and uses a bus broadcast method to perform grouped parallel self-tests on the monitoring modules of the terminals under its jurisdiction. The periodic self-inspection process dynamically calculates the self-inspection scheduling priority based on the historical operating status and risk parameters of each relay module. During the self-inspection process, the current self-inspection is interrupted in response to the fire alarm signal and the breakpoint is saved. The self-inspection is resumed from the breakpoint after the fire alarm is lifted. The system maintains a self-inspection process, receives self-inspection instructions issued manually after hierarchical authorization verification, and collects the back electromotive force waveform of the solenoid valve while performing short-term physical isolation of the fire extinguishing circuit. Based on the comparison results between the waveform and the reference waveform, the system determines the hidden faults in the fire extinguishing circuit. The bus communication in the power-on self-test process, periodic self-test process, and maintenance self-test process is all based on the unified CAN bus protocol, and the fire monitoring channel is kept active throughout the self-test process.

2. The self-test control method for a fire-fighting device according to claim 1, characterized in that, The dynamic calculation of self-test scheduling priority includes: The self-test priority value Pi of the i-th relay module is calculated according to the following nonlinear mapping formula: ； in, This is the Sigmoid function, used to map priority values ​​to the (0,1) interval; This represents the historical number of failures of the i-th relay module; The risk level of the area where the i-th relay module is located; The time interval between the i-th relay module and its last self-test; Basic self-inspection cycle; This is the set of relay modules adjacent to the i-th relay module; The spatial correlation coefficient represents the degree of fire spread risk correlation between the i-th relay module and the adjacent j-th relay module. is the historical number of failures of the j-th adjacent relay module; w1, w2, w3, and w4 are the weighting coefficients of the corresponding parameters.

3. The self-test control method for a fire-fighting device according to claim 1, characterized in that, The periodic self-test process also includes: The next self-inspection cycle will be dynamically adjusted based on the current self-inspection results. The calculation formula is as follows: ； ； in, This is the current self-inspection cycle; This is the maximum self-test cycle threshold. The minimum self-test cycle threshold; Basic self-inspection cycle; This represents the maximum adjustment percentage in a single instance, with a range of 5% to 10%. This represents the number of faults in the W most recent self-test results within the sliding window; The number of normal self-checks within the sliding window; The fault penalty coefficient and >1, For health reward coefficient and <1; The width W of the sliding window is a preset positive integer.

4. The self-test control method for a fire-fighting device according to claim 1, characterized in that, The method of determining latent faults in the fire extinguishing circuit based on the comparison results of the waveform and the reference waveform specifically includes: Calculate the similarity S between the measured back electromotive force waveform and the reference waveform: ； in, To measure the amplitude of the back electromotive force waveform at the t-th sampling point, The amplitude of the reference waveform at the t-th sampling point is... The mean of the measured waveforms. K represents the mean of the baseline waveform, and K is the number of sampling points. Extract the peak attenuation rate deviation of the measured waveform and the rate of change of decay time constant ; Based on the similarity Peak attenuation deviation and the rate of change of decay time constant The fault type is identified by comparing it with the preset fault determination matrix; Calculate the diagnostic confidence level C: ； in, , , These are the measurement noise standard deviations for similarity, peak attenuation rate deviation, and attenuation time constant change rate, respectively; when the diagnostic confidence C is higher than the preset confidence threshold, the fault determination result is output.

5. The self-test control method for a fire-fighting device according to claim 1, characterized in that, The power-on self-test process also includes: Real-time bus packet loss rate Ploss = Nerr / Ntotal, where Nerr is the number of erroneous or lost packets per unit time, and Ntotal is the total number of packets that should be received per unit time. The packet loss rate fault threshold Pth(t) is dynamically adjusted according to the following formula: ； in, Based on the basic packet loss rate threshold, This represents the standard deviation of the recent packet loss rate. This is the preset adjustment coefficient; When the bus packet loss rate Ploss exceeds the dynamic threshold When this happens, the corresponding abnormal nodes will be logically isolated.

6. The self-test control method for a fire-fighting device according to claim 1, characterized in that, The periodic self-test process also includes: Perform tiered and differentiated detection based on the health status of the terminal monitoring module; Specifically, heartbeat detection, voltage detection, and bit error rate detection are performed on terminal monitoring modules with stable health status; impedance scanning and battery internal resistance detection are added for terminal monitoring modules with occasional abnormal health status; and full-parameter deep self-test is performed on terminal monitoring modules with fault history in their health status.

7. A self-test control method for a fire-fighting device according to claim 5, characterized in that, The periodic self-test process also includes a sensor environment adaptive calibration step: Collect ambient temperature T, vibration frequency f, and ambient humidity H at the site; Correct sensor readings using a multivariable compensation model : ； in, These are the original values ​​collected by the sensor. Reference temperature For reference humidity, , , These are the compensation coefficients for temperature, vibration, and humidity, respectively. The temperature compensation coefficient is updated in real time using the online gradient descent method. : ； in, For learning rate, For calibration reference values, The ambient temperature collected at the current moment; the vibration compensation coefficient Humidity compensation coefficient Update synchronously in the same way.

8. The self-test control method for a fire-fighting device according to claim 1, characterized in that, In the aforementioned periodic self-test process, the response delay of the self-test interrupted by the fire alarm signal... satisfy: ； in, Due to delays in fire detection, The self-test task was interrupted and the fire alarm task switching was delayed. Delay in saving self-test breakpoints; The self-check process is divided into multiple minimum task slices with a duration of no more than 200ms. When a fire alarm occurs, only the currently executing slice is interrupted, and completed slices do not need to be executed again.

9. A self-test control method for a fire-fighting device according to claim 1, characterized in that, It also includes the self-inspection steps for when the main controller loses contact: Each relay module periodically listens to the heartbeat messages broadcast by the fire control module; If no heartbeat message is received for M consecutive heartbeat cycles, the fire control module is determined to be disconnected, where M is a preset positive integer; Each relay module calculates the score of the temporary management node according to the following formula. The node with the highest score is then elected as the temporary management node. ; in, Let be the processing power of the i-th relay module. This represents the maximum processing power of the processors across all relay modules. Let be the remaining backup power capacity of the i-th relay module. This represents the maximum remaining backup power capacity across all relay modules. Let be the communication link quality of the i-th relay module. is the maximum value of the communication link quality among all relay modules; a, b, and c are the weighting coefficients of the corresponding parameters, respectively; The temporary management node takes over the self-test scheduling function from the fire control module. The self-test data during offline periods is cached locally and automatically retransmitted after the fire control module recovers.

10. A self-test control system for fire-fighting equipment, comprising the self-test control method for fire-fighting equipment as described in any one of claims 19, characterized in that, include: The fire control module is deployed in the central control room or electrical control cabinet, serving as the core of self-testing and dispatching and the data aggregation center. N relay modules are deployed in each compartment or carriage for data forwarding between upper and lower levels and edge control. M terminal monitoring modules are deployed at each battery cluster or monitoring point, and each module integrates a sensor and a fire extinguishing circuit drive circuit. The fire control module, the relay module, and the terminal monitoring module are connected via a dual-layer full CAN bus architecture. The dual-layer full CAN bus architecture includes at least: a main redundant CAN bus connecting the fire control module and all N relay modules, and a branch CAN bus independently led out by each relay module for mounting the terminal monitoring module within its domain. The fire control module also has a separate CAN bus for connecting to debugging equipment; all external communication interfaces in the system include only CAN bus interface and Ethernet interface, and there is no serial communication port.

Images