A wireless-linked intelligent electronic control system for lifesaving devices

By constructing a single-device event perception vector and calculating the explanatory power of disturbance events, and combining multi-sensor collaboration and historical event correlation information, a release strategy for the intelligent electronic control system of the life-saving device is generated. This solves the problem of distinguishing between non-emergency disturbances and real emergency events in existing technologies, and achieves highly reliable and timely release of the life-saving device.

CN123113456APending Publication Date: 2026-11-17NANJING LVZHOU ELECTRIC CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611269649.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-20
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing intelligent electronic control systems for lifesaving devices struggle to distinguish between non-emergency disturbance-type triggering anomalies and real emergency-type triggering anomalies, and lack a secure and reliable local redundancy confirmation mechanism based on wireless linkage. This makes release strategies prone to mis-release, missed release, or missing the current release window.

Method used

The system employs an intelligent electronic control system for lifesaving devices based on wireless linkage, including a data construction module, an event discrimination module, a release prediction and releaseability management module, a linkage window analysis module, a risk decision-making module, and a safety interlocking module. By constructing a single device event perception vector, it calculates the explanatory power of disturbance events and the reliability value of emergency release. It combines multi-sensor collaboration and historical event correlation information to perform hierarchical judgment, generate release strategies, and use local redundancy confirmation from adjacent nodes when the central control node fails.

Benefits of technology

It improves the reliability, timeliness, and feasibility of life-saving device release strategies, reduces the risk of accidental release, ensures accurate release in real emergency situations, and avoids accidental release and missed release windows caused by non-emergency disturbances.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123113456A_ABST
    Figure CN123113456A_ABST
Patent Text Reader

Abstract

This invention relates to the field of intelligent electronic control technology for ship lifesaving equipment, and discloses an intelligent electronic control system for lifesaving devices based on wireless linkage. The system constructs a single-device event perception vector based on data from the target lifesaving device; when a triggering anomaly occurs, it extracts the triggering anomaly features, determines the explanatory relationship between the triggering anomaly and ship routine rolling, berthing impact, maintenance disturbance, or sensor anomaly, obtains the explanatory power of the disturbance event, and processes it to generate an emergency release confidence value; based on the confidence value, it adjusts the event prediction weights, generates an exponential curve and a waiting level; it combines the release success probabilities of the current window and the next release window, compares them, and obtains the incremental risk of missing the release window, reducing the risk of erroneous release caused by ship routine disturbance, berthing impact, maintenance operation, or single-point sensor anomaly. In the event of a real emergency and when the current release window may be missed, it promptly completes coordinated release, improving the reliability and executability of ship lifesaving device release decisions.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of intelligent electronic control technology for ship lifesaving equipment, and in particular to an intelligent electronic control system for lifesaving devices based on wireless linkage. Background Technology

[0002] The reliable release of shipboard lifeboats, throwable life rafts, and other lifesaving devices is directly related to the safety of crew and passengers during emergencies. Relevant international conventions, ship inspection standards, and classification society rules generally require shipboard lifesaving equipment to be reliably released, readily available in emergencies, and easily accessible to personnel. With the increasing automation and intelligence of ships, some shipboard lifesaving devices are gradually incorporating electromagnetic release devices, electrically controlled locking mechanisms, status monitoring nodes, and wireless communication nodes to enable remote triggering, linked alarms, or coordinated release in emergencies.

[0003] Existing lifesaving device release control schemes typically integrate ship attitude data, acceleration data, water pressure data, temperature data, smoke data, and device self-test data, and determine release triggers based on one or more sensor thresholds. For example, throwable life rafts can release moorings when the hull sinks to a certain depth using a hydrostatic release device; some lifeboats or life rafts' electronic control systems can determine whether to enter release preparation mode based on changes in ship heel, trim, impact acceleration, or water pressure; other schemes synchronize the status of multiple lifesaving devices to the ship's central control center via wireless links, and then the central control center issues release or prohibition commands.

[0004] However, in actual ship operation, the sensor responses on lifesaving devices are not necessarily caused by real emergencies. Normal rolling during rough seas, localized impacts during berthing or departure, human disturbances during device installation and maintenance, momentary drift of a single sensor, electromagnetic interference, and localized vibrations caused by operations on adjacent deck equipment can all cause abnormal attitude, acceleration, vibration, or short-term self-test anomalies in lifesaving devices. These non-emergency disturbances may appear similar in data manifestation to real emergencies such as capsizing, collision, fire, explosion, or sinking.

[0005] If the intelligent electronic control system directly determines release based on raw attitude values, raw acceleration values, water pressure values, or single sensor alarms, it is prone to misinterpreting recoverable anomalies caused by non-emergency disturbances as genuine emergencies, leading to false releases. False releases not only waste lifesaving devices and increase maintenance costs but may also create additional personnel risks while the ship is sailing or berthed normally. Conversely, if the system simply attributes a sensor anomaly to normal ship rolling or environmental disturbances without further determining whether the anomaly is intensifying, whether it points to the same emergency chain as multiple sensors, or whether it coincides with synchronous anomalies in multiple lifesaving devices, release may be delayed in the event of a genuine emergency, missing the optimal release opportunity.

[0006] Furthermore, existing release control schemes typically focus on whether the release threshold has been reached, whether the wireless link is online, or whether the release actuator is available. However, they lack a judgment mechanism based on the cause of the anomaly and the difference in path risk for issues such as whether the current release window is the optimal coordinated release window, whether it is possible to wait for the next release window, whether the waiting process will result in the life raft being crushed by the hull or becoming inaccessible to personnel, and whether there is release interference between multiple life rafts. For example, when a ship experiences sustained list, a life raft on one side may still be releaseable and accessible to personnel in the current window. If it waits for the next window, it may be blocked by obstacles due to the continued list of the hull. Or, after a strong impact, multiple life rafts may experience short-term acceleration anomalies. However, if the anomaly can be explained by the berthing impact and recovers after the disturbance ends, immediate release may actually create a risk of false release.

[0007] Furthermore, even when using multi-sensor data, ship attitude data, and wireless linkage data simultaneously, existing solutions often treat wireless linkage data as a normal state synchronization channel, rather than using it for triggering anomaly cause identification, coordinated release window determination, and local redundant release confirmation. Therefore, existing technologies still struggle to simultaneously address the issues of mis-release due to non-emergency disturbances, missed release due to real emergency events, and risk assessment of missed coordinated release windows within the same control chain.

[0008] Therefore, a wireless linkage-based intelligent electronic control system for life-saving devices is needed. This system should be able to first distinguish between disturbance-type triggering anomalies and real emergency-type triggering anomalies, then determine whether the current release window should be occupied based on the path risk difference between the current window immediate release path and the executable waiting path, and introduce a safety interlock and local redundancy confirmation mechanism before release execution to improve the reliability, timeliness and executability of the life-saving device release strategy. Summary of the Invention

[0009] The technical problem to be solved by the present invention is that the existing intelligent electronic control system for life-saving devices has the disadvantages of being unable to distinguish between non-emergency disturbance-type triggering anomalies and real emergency event-type triggering anomalies, having difficulty in quantifying the incremental risk between immediate release and waiting for coordinated release, and lacking a safe and reliable local redundant release confirmation mechanism when the wireless link or central control node fails. This leads to the release strategy being prone to mis-release, missed release, or missing the current release window. To address this, we propose an intelligent electronic control system for life-saving devices based on wireless linkage.

[0010] To achieve the above objectives, this application adopts the following technical solution: an intelligent electronic control system for lifesaving devices based on wireless linkage, used for the release electronic control management of shipborne lifeboats or throwable life rafts, including a data construction module, an event discrimination module, a release prediction and releaseability management module, a linkage window analysis module, a risk decision module, a safety interlock module, and a linkage feedback module.

[0011] The data construction module is used to receive multi-source sensor data, ship attitude data, sea state disturbance data, device status data and wireless linkage status data from the target life-saving device, perform time alignment according to the life-saving device number and collection time, and generate a single device event perception vector corresponding to the target life-saving device based on the installation location information of the target life-saving device.

[0012] The event discrimination module is used to determine the anomaly analysis time window when a triggering anomaly occurs in the target life-saving device, and extract the triggering anomaly features within the anomaly analysis time window. Based on the single-device event perception vector, the event discrimination module judges the degree to which the triggering anomaly features can be explained by non-emergency disturbances, thus obtaining the explanatory power of the disturbance event. The event discrimination module also combines triggering anomaly intensity, triggering anomaly duration, triggering anomaly recovery status, multi-sensor coordination, inter-device coordination, and historical event correlation information to generate an emergency release confidence value.

[0013] The release prediction and releaseability management module is used to correct the event prediction weights of the triggering abnormal characteristics based on the explanatory power of the disturbance event and the confidence value of emergency release. Based on the corrected event prediction weights, the status of the release actuator, the status of the backup power supply, the quality of the wireless linkage link, the attitude status of the target rescue device, and the personnel accessibility status, the releaseability index curve and the waiting level of the target rescue device are generated.

[0014] The linkage window analysis module is used to generate the release success probability of the current collaborative release window and the release success probability of the next collaborative release window based on the ship attitude prediction results, the coordination relationship between devices, the quality of the wireless linkage link, the status of the release actuator, the status of the backup power supply, the risk of release interference from neighboring devices, and the status of obstacles around the device.

[0015] The risk decision module is used to construct a set of immediate release paths and executable waiting paths for the current window based on the emergency release confidence value, the releaseability index curve, the release success probability of the current collaborative release window, and the release success probability of the next collaborative release window. The path risk of the immediate release path for the current window is used as the benchmark risk, and the difference between the minimum path risk in the set of executable waiting paths and the benchmark risk is determined as the release window missed incremental risk.

[0016] The safety interlock module is used to acquire the manual release prohibition signal, mechanical locking status, release actuator self-test result, backup power availability status, pre-release audible and visual alarm status, and confirmation status from at least two independent trigger sources, and generates a release interlock permission result based on the above status.

[0017] The linkage feedback module generates a release strategy for the target life-saving device based on the incremental risk of missed release window, the waiting level, and the release interlock permission result, and distributes this strategy to each intelligent electronic control node via the ship's wireless linkage network. When the central control node fails, the linkage feedback module also enables at least two adjacent intelligent electronic control nodes to collaboratively output a local redundant release strategy when the local redundant release confirmation conditions are met.

[0018] The technical effects and advantages of this invention are as follows: First, this invention does not directly trigger release with a single threshold. Instead, it first constructs a single-device event perception vector, and then calculates the explanatory power of the disturbance event and the confidence value of the emergency release. This enables ship routine swaying, berthing impact, maintenance disturbances, or single-point sensor anomalies to be identified as disturbance-type triggering anomalies, thereby reducing the risk of false release.

[0019] Second, the emergency release confidence value in this invention does not mechanically decrease as the interpretability of the disturbance event increases. Instead, it further combines the anomaly recovery state, multi-sensor synergy, inter-device synergy, and historical event correlation information for hierarchical judgment. When the anomaly is related to the disturbance but continues to increase or multi-device synergy occurs, the system still maintains or increases the emergency release confidence value, thereby reducing the risk that the real emergency event is wrongly attributed to the disturbance.

[0020] Third, this invention constructs a path risk comparison between the immediate release path of the current window and the executable waiting path by using the releaseability index curve, the success probability of the current collaborative release window, and the success probability of the next collaborative release window. It also uses the incremental risk of missing the release window to determine whether the current window should be occupied, so that the release strategy can achieve a quantitative balance between "avoiding accidental release" and "avoiding missing the release window".

[0021] Fourth, this invention introduces safety interlocking conditions such as manual release prohibition signals, mechanical locking status, release actuator self-test, backup power status, audible and visual alarms, and multi-trigger source confirmation. This makes the release strategy no longer directly triggered by algorithm scoring, but executed after the safety interlocking permission is met, thus improving the engineering feasibility in high-security scenarios.

[0022] Fifth, this invention sets up a hierarchical structure for the ship's in-ship wireless linkage network and the ship-to-shore remote emergency reporting link, and adopts a local redundancy confirmation mechanism for adjacent nodes when the central control node fails, so as to avoid the failure of a single central node or a single communication frequency band causing the life-saving device to fail to be released in a coordinated manner. Attached Figure Description

[0023] The disclosure of this invention is illustrated with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. In the drawings, the same reference numerals are used to refer to the same parts: Figure 1 This is a schematic diagram of the system topology provided for an embodiment of the present invention.

[0024] Figure 2 The overall system flowchart provided for embodiments of the present invention.

[0025] Figure 3 A schematic diagram illustrating the calculation process of perturbation event explanatory power and emergency release confidence value provided in embodiments of the present invention.

[0026] Figure 4 This is a functional flowchart of the intelligent electronic control system for ship lifesaving devices based on embodiments 1-4 of the present invention. Detailed Implementation

[0027] It is readily understood that, based on the technical solution of this invention, those skilled in the art can propose various interchangeable structural methods and implementations without altering the essential spirit of the invention. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solution of this invention.

[0028] In this invention, the lifesaving device is preferably a shipboard lifeboat or a throwable life raft, and the lifesaving device is equipped with an intelligent electronic control node. The intelligent electronic control node may include a microcontroller, a communication module, a sensor interface, a power management unit, a release actuator drive unit, a mechanical locking status detection unit, and a local storage unit. The release actuator may be one or more of an electromagnetic release device, an electronically controlled lock, a release hook controller, or a pyrotechnic release device drive. This invention can also be extended to life ring throwers, rope throwers, or emergency signal release devices according to the actual ship configuration, but it is preferably implemented in lifeboats and throwable life rafts.

[0029] The system in this embodiment can be deployed on a ship's centralized control center server, edge computing gateway, shore-based cloud server, intelligent electronic control node local controller, or cloud-edge collaborative system. To reduce the risk of communication failure, the ship's in-ship wireless linkage network preferably adopts at least two of the following: LoRa subnetwork, Sub-GHz Mesh subnetwork, Bluetooth Mesh subnetwork, in-ship Ethernet, CAN bus, or RS485 bus; the ship-to-shore or remote emergency reporting link can adopt at least one of the following: Beidou short message link, VHF link, satellite communication link, or AIS associated reporting link. In-ship release control commands are executed only after being confirmed by the local safety interlock. The ship-to-shore or remote link is mainly used for emergency reporting, status synchronization, and auxiliary confirmation.

[0030] Reference Figure 1 and Figure 4 The intelligent electronic control system for lifesaving devices based on wireless linkage includes a data construction module, an event discrimination module, a release prediction and releaseability management module, a linkage window analysis module, a risk decision-making module, a safety interlock module, and a linkage feedback module.

[0031] The system receives data including multi-source sensor data, ship attitude data, sea state disturbance data, device status data, wireless linkage status data, and historical event data.

[0032] Multi-source sensor data may include one or more of the following: triaxial acceleration, triaxial angular velocity, roll angle, pitch angle, bow angle, water pressure, temperature, smoke concentration, root mean square value of vibration, device position, device orientation, and hatch opening / closing status.

[0033] Ship attitude data may include one or more of the following: overall roll angle, pitch angle, bow angle, overall acceleration, overall angular velocity, draft, speed, heading, ship position, and ship inclination.

[0034] Sea state disturbance data may include one or more of the following: wind speed, wind direction, gusts, wave height, wave period, tides, ocean currents, visibility, rainfall, lightning, typhoon track, weather forecast confidence level, and sea state forecast confidence level.

[0035] The device's own status data may include one or more of the following: main control self-test status, sensor self-test status, release actuator self-test status, power management self-test status, backup battery capacity, backup battery health, release actuator action history, mechanical locking status, manual release prohibition signal status, device unlocking status, and device release status.

[0036] Wireless linkage status data may include one or more of the following: LoRa link received signal strength, Sub-GHz Mesh link quality, Bluetooth Mesh link received signal strength, shipboard Ethernet status, CAN bus status, RS485 bus status, link bit error rate, link latency, neighbor node list, and Mesh routing cost.

[0037] Historical event data can include real historical emergency events, historical accidental release events, historical missed release events, simulation events, expert rule bases, test data of the same type of device and corresponding sensor data fragments, decision process logs and post-verification results.

[0038] The main data processing logic of this implementation is as follows: the data construction module generates a single-device event perception vector; the event discrimination module calculates the explanatory power of the disturbance event based on the single-device event perception vector and the triggering anomaly characteristics, and generates an emergency release confidence value; the release prediction and releaseability management module corrects the event prediction weight based on the emergency release confidence value and generates a releaseability index curve; the linkage window analysis module generates the release success probability of the current collaborative release window and the next collaborative release window; the risk decision module calculates the release window miss incremental risk based on the path risk difference between the current window immediate release path and the executable waiting path; the safety interlock module determines the release interlock permission; and the linkage feedback module outputs the release strategy based on the release window miss incremental risk, the waiting level, and the release interlock permission result.

[0039] The aforementioned continuous processing chain enables ship attitude data, sea state disturbance data, and wireless linkage status data to no longer be merely ordinary inputs, but to participate in the determination of trigger anomaly causes, feasibility analysis of release windows, and confirmation of local redundancy release, respectively.

[0040] In one implementation, the data construction module performs data preprocessing on multi-source sensor data, ship attitude data, sea state disturbance data, device status data, and wireless linkage status data before constructing the single device event perception vector.

[0041] Data preprocessing includes data cleaning, outlier removal, missing value completion, sampling frequency standardization, and time alignment. The system establishes a unified index based on the life-saving device number, acquisition time, data source, and data type, enabling the target life-saving device to correspond to multi-source sensor data, ship attitude data, sea state disturbance data, device status data, and wireless linkage status data on the same timeline.

[0042] For example, accelerometers and vibration sensors can provide high-frequency sampling data, attitude sensors can provide medium-frequency sampling data, water pressure sensors, temperature sensors, and smoke sensors can provide second-level sampling data, and sea state disturbance data can be minute-level or forecast time-step data. The system can aggregate, interpolate, or compress the above data according to a preset anomaly analysis time window, enabling comparisons within the same time window.

[0043] When some sensor data is missing, the system can use data from adjacent sensors, adjacent devices, historical statistics, or preset default values ​​to complete it. When a sensor has a constant value, exceeds its range, experiences a short-term zero-crossing jump, or fails a self-test, the system can mark the data corresponding to that sensor as abnormal and reduce or set the weight of that data to zero when calculating the explanatory power of disturbance events or the confidence value of emergency release.

[0044] The data construction module converts ship-level or section-level data into a single-device event perception vector corresponding to the target life-saving device based on at least two of the following: the installation location coordinates of the target life-saving device, its deck level, its position relative to the ship's center of gravity, its lever arm relative to the ship's roll and pitch axes, the layout of adjacent devices, the ship's dominant motion direction, the ship's draft, the direction of sea waves, the degree of device exposure, and the data acquisition time.

[0045] A single device event perception vector is a set of state data corresponding to a target life-saving device, which may include effective roll angle, effective pitch angle, effective impact acceleration, effective water pressure, effective temperature rise, effective smoke concentration, effective root mean square value of vibration, device attitude deviation value, mechanical locking status, backup power availability status, wireless link confidence, and trigger event label.

[0046] For example, when the target liferaft is installed on the outer side of the upper deck on the port side, the overall ship roll will manifest as a more pronounced effective roll angle and local acceleration at that device; when the target lifeboat is installed on the starboard pylon, continued starboard listing will reduce its personnel accessibility and increase the risk of the device being crushed by the hull. Through single-device event perception vectors, the system can transform ship-level attitude and sea state disturbances into interpretive inputs for the target lifeboat, rather than simply using overall ship attitude thresholds for release decisions.

[0047] Trigger event tags can include short-duration strong impact events, sustained capsizing events, water ingress events, fire events, explosion impact events, typhoon outer band impact events, prolonged swaying events, berthing impact events, or maintenance disturbance events. For example, when the acceleration modulus of the target life-saving device exceeds a preset impact threshold within a short period of time, and the ship's berthing operation signal is present, a berthing impact event tag can be generated; when the water pressure of the target life-saving device continues to rise, and the ship's heel angle continues to increase, a water ingress event tag or a sustained capsizing event tag can be generated.

[0048] When a triggering anomaly occurs in the target life-saving device, the event discrimination module determines the anomaly analysis time window based on the occurrence time of the triggering anomaly and extracts the triggering anomaly features within the anomaly analysis time window.

[0049] Triggering anomaly characteristics include one or more of the following: attitude anomaly, impact anomaly, water entry anomaly, temperature rise anomaly, smoke anomaly, vibration anomaly, and device self-test anomaly. Each type of triggering anomaly characteristic may include the time of anomaly occurrence, duration of anomaly, intensity of anomaly, trend of anomaly change, recovery time of anomaly, and number of times anomaly recurs.

[0050] An abnormal attitude can be determined based on the magnitude and duration of changes in roll and pitch angles; an abnormal impact can be determined based on the peak value of the triaxial acceleration modulus, the duration of the impact, and the frequency of the impact; an abnormal water immersion can be determined based on the water pressure sensor data continuously exceeding a preset threshold; an abnormal temperature rise can be determined based on the rate of temperature change and the magnitude of the continuous temperature increase; an abnormal smoke can be determined based on the sudden increase in smoke concentration and its duration; an abnormal vibration can be determined based on the root mean square value of vibration continuously exceeding a reference value; and an abnormal device self-test can be determined based on the number and frequency of self-test failures of the power supply, release actuator, sensor, or communication module.

[0051] The anomaly analysis time window can be set according to the type of anomaly triggered. For example, for impact anomalies, the anomaly analysis time window can be from 2 seconds before the trigger to 10 seconds after the trigger; for attitude anomalies, the anomaly analysis time window can be from 30 seconds before the trigger to 120 seconds after the trigger; for temperature rise anomalies and smoke anomalies, the anomaly analysis time window can be further extended. The above time windows are only examples and can be adjusted according to the ship type, device model, and sampling frequency.

[0052] Disturbance event explanatory power characterizes the extent to which triggering anomalies can be explained by non-urgent disturbances. A higher disturbance event explanatory power indicates that the triggering anomaly is more likely to be caused by routine ship rolling, berthing shocks, maintenance disturbances, or other non-urgent disturbances; a lower disturbance event explanatory power indicates that the triggering anomaly is less likely to be explained by non-urgent disturbances.

[0053] The event discrimination module determines at least two of the following: time synchronization score, trend consistency score, duration matching score, recovery score, historical disturbance similarity score, and neighboring device response score.

[0054] The time synchronization score characterizes the degree of synchronization between the occurrence time of the triggering anomaly and the occurrence time of routine or operational disturbances on the ship. The trend consistency score characterizes the degree of consistency between the trend of the triggering anomaly and the trend of changes in ship attitude disturbances, sea state disturbances, or berthing impacts. The duration matching score characterizes the degree of matching between the duration of the triggering anomaly and the duration of the disturbance event. The recovery score characterizes whether the triggering anomaly recovers to the baseline range after the disturbance event ends. The historical disturbance similarity score characterizes whether the target lifesaving device has experienced similar triggering anomalies under similar historical disturbance events, simulated disturbance events, or test events of the same type of device, but without confirming a real emergency. The adjacent device response score characterizes whether other lifesaving devices at adjacent installation locations, the same deck level, or the same orientation on the same ship exhibit similar and recoverable triggering responses under the same disturbance event.

[0055] When determining the explanatory power of a disturbance event, the event discrimination module first obtains the aforementioned explanatory scores, and then determines the participation weight of each explanatory score based on the data reliability, sensor self-test status, and trigger anomaly type corresponding to each score. Subsequently, the event discrimination module fuses the participating explanatory scores according to their corresponding participation weights to obtain the explanatory power of the disturbance event. The participation weights can be determined based on historical samples, simulation samples, expert rules, or device model configurations. For example, for impact anomalies, the participation weights of time synchronization scores, duration matching scores, and neighboring device response scores can be relatively high; for temperature rise anomalies and smoke anomalies, the participation weights of recovery scores, trend consistency scores, and multi-sensor synergy correlation scores can be relatively high.

[0056] When any explanatory score has missing data or a corresponding sensor malfunctions, the event discrimination module will not use that explanatory score or will reduce its weight, while increasing the weight of the remaining valid explanatory scores. When all explanatory scores are unavailable, the event discrimination module will set the explanatory power of the disturbance event to a preset default value, or mark the triggered anomaly as an anomaly to be confirmed, thus avoiding making a disturbance-type anomaly judgment directly when data is insufficient.

[0057] In one specific implementation, the time synchronization score can be determined based on the time difference between the peak occurrence time of the triggered anomaly and the peak occurrence time of the disturbance event; the closer the two times are, the higher the time synchronization score. The trend consistency score can be determined based on the consistency of the changing directions of the triggered anomaly sequence and the disturbance event sequence within the anomaly analysis time window. The duration matching score can be determined based on the closeness between the duration of the triggered anomaly and the duration of the disturbance event. The recovery score can be determined based on the degree of recovery of the triggered anomaly relative to the anomaly peak and the pre-anomaly baseline value after the disturbance event ends.

[0058] The proximity device response score utilizes multi-device synchronization information obtained through a wireless linkage network. When multiple adjacent lifesaving devices on the same vessel exhibit similar but brief trigger responses under the same disturbance event, and each device returns to its baseline state after the disturbance ends, the proximity device response score is higher, indicating that the trigger response is more likely to originate from a general disturbance to the vessel rather than a localized emergency.

[0059] The emergency release confidence value characterizes the degree of confidence that the triggered anomaly is caused by a genuine emergency and necessitates the release of life-saving equipment. The emergency release confidence value differs from a single sensor alarm level and also from the probability of hardware failure. A single sensor alarm level only indicates the degree to which a particular sensor exceeds a threshold, and the probability of hardware failure only indicates the likelihood of hardware malfunction. The emergency release confidence value, however, indicates whether the current triggered anomaly is more likely to originate from a genuine ship-level emergency and whether a release decision process should be initiated.

[0060] The event discrimination module uses a hierarchical judgment logic to generate an emergency release confidence value. First, it determines whether the triggering anomaly has a non-emergency disturbance explanation basis based on the explanatory power of the disturbance event. Second, it determines whether the triggering anomaly recovers after the disturbance event ends, whether it continues to intensify, whether it points to the same emergency event chain as multiple sensor triggering anomalies, whether it points to the same emergency event chain as multiple life-saving device triggering anomalies, and whether there is a correlation with historical events. Finally, it combines the above results to generate an emergency release confidence value.

[0061] When the triggered anomaly has a high interpretability of the disturbance event, and the triggered anomaly recovers after the disturbance event ends, there is a lack of coordination between multiple sensors, a lack of coordination between multiple life-saving devices, and a weak correlation between historical events, the event discrimination module reduces the reliability of emergency release.

[0062] When the triggered anomaly has a high explanatory power for the disturbance event, but continues to exist or intensify after the disturbance event ends, or when there is synergy among multiple sensors, or synergy among multiple life-saving devices, or strong correlation among historical events, the event discrimination module maintains or increases the emergency release confidence value.

[0063] The emergency release confidence value is jointly determined by the basic anomaly contribution, the synergy enhancement contribution, the historical event correlation contribution, and the disturbance deduction contribution. The basic anomaly contribution is determined based on the intensity and duration of the triggered anomaly; the synergy enhancement contribution is determined based on multi-sensor synergy and inter-device synergy; the historical event correlation contribution is determined based on historical real events, simulated events, expert rule bases, or test data from the same type of device; the disturbance deduction contribution is activated when the triggered anomaly meets the non-emergency disturbance interpretation conditions and the triggered anomaly recovers.

[0064] For example, if the target life raft experiences a short-term acceleration peak, but this peak is synchronized with the berthing impact height, and neighboring life rafts all exhibit similar short-term responses and recover after the disturbance ends, the event discrimination module activates disturbance deduction contribution, reducing the emergency release confidence value. If the target lifeboat exhibits abnormal attitude under the influence of the outer periphery of a typhoon, although this abnormality is synchronized with sea state disturbances, but the heel continues to increase, the water pressure continues to rise, and adjacent devices also exhibit coordinated water ingress anomalies, the event discrimination module does not activate disturbance deduction contribution or reduces disturbance deduction contribution, while increasing coordination enhancement contribution, keeping the emergency release confidence value high.

[0065] The event discrimination module can also generate at least two of the following: attitude trigger confidence values, impact trigger confidence values, water ingress trigger confidence values, fire trigger confidence values, vibration trigger confidence values, and self-test trigger confidence values. Based on the correlation between different trigger anomaly characteristics and emergency event types, it merges these to obtain the emergency release confidence value of the target life-saving device. For example, the attitude trigger confidence value is associated with ship capsizing or severe list events; the impact trigger confidence value is associated with ship collision or explosion impact events; the water ingress trigger confidence value is associated with ship sinking events; the fire trigger confidence value is associated with ship fire or explosion events; the vibration trigger confidence value is associated with structural damage or severe impact events; and the self-test trigger confidence value is associated with the device's own malfunction events.

[0066] The release prediction and releaseability management module adjusts the event prediction weights of triggering anomaly features based on the explanatory power of the disturbance event and the confidence value of emergency release. The event prediction weights represent the contribution of a particular triggering anomaly feature to event confidence calculation, event level determination, and releaseability index curve generation.

[0067] For triggering anomaly features where the explanatory power of the disturbance event meets the preset high explanatory power condition and the emergency release confidence value meets the preset low confidence value condition, the system determines that the triggering anomaly feature is more likely to be caused by a non-emergency disturbance, and therefore reduces the event prediction weight of the triggering anomaly feature. For triggering anomaly features where the explanatory power of the disturbance event meets the preset low explanatory power condition and the emergency release confidence value meets the preset high confidence value condition, the system determines that the triggering anomaly feature is more likely to be caused by a real emergency event, and therefore increases the event prediction weight of the triggering anomaly feature.

[0068] When the explanatory power of a disturbance event is high but the confidence value of emergency release has not decreased, the release prediction and releaseability management module does not directly filter out the triggering anomaly feature, but retains its contribution to the event prediction results and the releaseability index curve. Therefore, the system can avoid incorrectly excluding genuine emergency event triggering anomalies that are "disturbance-related but not recoverable".

[0069] The releaseability index curve characterizes the change in the release feasibility of a target lifesaving device at the current moment and under different waiting actions. The horizontal axis of the releaseability index curve can be the prediction time or the sequence of waiting actions; the vertical axis is the releaseability index. The releaseability index is determined based on at least two of the following: emergency release confidence value, release actuator availability factor, backup power availability factor, radio link reliability factor, device attitude releaseability factor, personnel accessibility factor, and release interference risk correction factor.

[0070] When determining the releaseability index, the release prediction and releaseability management modules first determine whether the target rescue device is necessary to release, and then determine whether the target rescue device is feasible to release. Release necessity is mainly determined by the emergency release confidence value and event prediction weight, while release feasibility is mainly determined by the status of the release actuator, the status of the backup power supply, the reliability of the wireless link, personnel accessibility, device attitude, and release interference risk. Only when both release necessity and release feasibility meet their respective conditions will the system evaluate the target rescue device as currently releaseable.

[0071] The release prediction and releaseability management module can generate at least two of the following: immediate release releaseability curve, continuous monitoring and waiting releaseability curve, encrypted monitoring and waiting releaseability curve, release releaseability curve after attitude stabilization, and release releaseability curve after nearby device release. If the releaseability index continues to decrease after waiting, the waiting level of the target life-saving device decreases; if the emergency release confidence value decreases after waiting but the releaseability index does not decrease, the waiting level of the target life-saving device increases.

[0072] The waiting levels can include non-waiting, short-term waiting, cooperative waiting, and full-scale waiting. Non-waiting means that the target rescue device should take priority in occupying the current cooperative release window; short-term waiting means that the target rescue device can wait for a short time under encrypted monitoring conditions; cooperative waiting means that the target rescue device can wait for attitude stabilization or for a nearby device to release first; full-scale waiting means that the target rescue device is currently mainly experiencing a disturbance-type anomaly and can continue to be monitored.

[0073] The coordinated release window analysis module is used to generate the release success probability of the current coordinated release window and the release success probability of the next coordinated release window. The coordinated release window refers to the time interval in which preset conditions are met for ship attitude, personnel accessibility, wireless link, release actuator, release interference from nearby devices, and obstacle status.

[0074] When determining the probability of successful release, the linkage window analysis module first determines at least two of the following factors: ship attitude stability factor, release actuator availability factor, backup power availability factor, wireless linkage link reliability factor, personnel accessibility factor, nearby device release interference risk factor, and device surrounding obstacle risk factor. Then, based on the positive contribution or negative impact of each factor on the success of release, it comprehensively judges the possibility of the target life-saving device completing the release within the corresponding coordinated release window.

[0075] The higher the stability of the ship's attitude, the more reliable the self-test results of the release actuator, the more sufficient the backup power supply, the more reliable the wireless linkage link, and the higher the personnel accessibility, the higher the probability of successful release; the higher the risk of interference from nearby devices and the higher the risk of obstacles around the device, the lower the probability of successful release.

[0076] The success probability of release in the current coordinated release window is determined based on the aforementioned factors within the current timeframe or the current predicted time period; the success probability of release in the next coordinated release window is determined based on the aforementioned factors within the predicted time period after the waiting period. For example, if the target lifeboat is still within the personnel reachable area at the current time, the release actuator passes its self-test, and the wireless link quality is good, then the success probability of release in the current coordinated release window is high; if the ship continues to tilt after the waiting period, the target lifeboat may be suppressed by the ship's structure or personnel may be unable to approach, then the success probability of release in the next coordinated release window decreases.

[0077] The risk decision module is used to construct a set of immediate release paths and executable waiting paths for the current window, and to calculate the incremental risk of missing the release window.

[0078] The "Current Window Immediate Release Path" indicates that the release actuator of the target life-saving device will be triggered immediately within the current coordinated release window. The set of executable waiting paths may include at least one of the following: continuous monitoring waiting path, encrypted monitoring waiting path, attitude stabilization release waiting path, nearby device release before release waiting path, and manual confirmation release waiting path.

[0079] The "Continuous Monitoring Waiting Path" indicates that after abandoning the current coordinated release window, the target lifesaving device maintains its current monitoring state and waits for the next coordinated release window. The "Encrypted Monitoring Waiting Path" indicates that after increasing the sensor sampling frequency and wireless linkage frequency, it waits for the next coordinated release window. The "Attitude Stabilization Release Waiting Path" indicates that it waits until the ship's attitude enters a preset stable range before triggering the release. The "Adjacent Device Release Before Release Waiting Path" indicates that it waits for adjacent devices to complete their releases before triggering the release, in order to reduce interference between devices. The "Manual Confirmation Release Waiting Path" indicates that it waits for manual or remote confirmation before releasing.

[0080] Before calculating the risk of waiting paths, the risk decision module assesses the feasibility of each waiting path. If a waiting path does not meet safety constraints, equipment control conditions, manual release prohibition status, release mechanism status, or ship lifesaving rules, then that waiting path will not be included in the selection of the minimum path risk. For example, if the side of the target lifeboat is expected to be suppressed by the ship's hull in a short time, the waiting path for release after attitude stabilization will not be included in the calculation; if the wireless link quality cannot support the synchronization of release status of nearby devices, the waiting path for release after nearby devices release first will not be included in the calculation; and if the manual release prohibition signal is valid, non-emergency release paths will not be included in the calculation.

[0081] The path risk of the immediate release path and the path risk of the waiting path in the current window can be determined based on at least two of the following: risk of accidental release, risk of missed release, risk of release failure, risk of personnel inaccessibility, risk of device failure escalation, risk of failure to reach the next coordinated release window, and the release success probability of the corresponding release window.

[0082] The risk decision-making module first determines the path risk of the immediate release path in the current window and uses this path risk as the baseline risk. Next, it determines the path risk of each executable waiting path and selects the waiting path with the lowest path risk from the set of executable waiting paths as the optimal waiting path. Then, it compares the path risk of the optimal waiting path with the baseline risk. When the path risk of the optimal waiting path is higher than the baseline risk, the difference between the two is determined as the incremental risk of missing the release window; when the path risk of the optimal waiting path is not higher than the baseline risk, the incremental risk of missing the release window is determined as a low-risk level or zero-incremental-risk.

[0083] The risk of missing a release window differs from the risk of a normal delayed release. The risk of a normal delayed release only assesses whether postponing the release itself is dangerous, while this implementation assesses the increased security risk of abandoning the current coordinated release window and waiting for the next coordinated release window compared to immediately releasing using the current coordinated release window. This risk difference is calculated by combining the emergency release confidence value, the releaseability index curve, the probability of successful release in the current coordinated release window, and the probability of successful release in the next coordinated release window.

[0084] The safety interlock module is used to determine the release interlock permission result before the linkage feedback module outputs the release strategy. The safety interlock module acquires the manual release prohibition signal, mechanical locking status, release actuator self-test result, backup power availability status, pre-release audible and visual alarm status, and confirmation status from at least two independent trigger sources.

[0085] When the manual release prohibition signal is valid and the preset extreme emergency conditions are not met, the safety interlock module outputs a release prohibition result. When the mechanical locking state does not meet the release conditions, the release actuator self-test fails, or the backup power supply is unavailable, the safety interlock module outputs a release prohibition result or an alternative release path result. When the release strategy is immediate release within the current window or local redundant release, the safety interlock module preferably requires an audible and visual alarm before release; in cases of ship sinking, fire and explosion, or when personnel safety is highly urgent and alarm exemption conditions are met, the audible and visual alarm can be skipped, but the reason must be recorded.

[0086] When the central control node fails, the local redundancy release strategy is only permitted if the following conditions are met: a manual release prohibition signal is absent; the release actuator passes its self-test; the mechanical locking state meets the release conditions; backup power is available; at least two adjacent intelligent electronic control nodes generate high emergency release confidence values ​​within the same anomaly analysis time window; and at least two independent trigger sources point to the same emergency event chain. These independent trigger sources may include attitude trigger sources, water ingress trigger sources, fire trigger sources, impact trigger sources, inter-device coordinated trigger sources, or manually confirmed trigger sources.

[0087] Through safety interlocking and local redundancy confirmation mechanisms, the system avoids single-node misjudgment or a single wireless command directly causing the group to release, while retaining the local release capability in extreme emergency situations when the central control node fails.

[0088] The linkage feedback module generates a release strategy based on the incremental risk of missing the release window, the waiting level, and the release interlock permission result. The release strategy may include a continuous monitoring strategy, an encrypted monitoring strategy, a coordinated release strategy after attitude stabilization, a nearby device-first release strategy, an immediate release strategy within the current window, and a local redundant release strategy.

[0089] When the risk of missing the release window is at a low risk level, and the explanatory power of the disturbance event meets the preset high explanatory power condition and the trust value of the emergency release meets the preset low trust value condition, the linkage feedback module generates a continuous monitoring strategy or an encrypted monitoring strategy to avoid accidental release.

[0090] When the incremental risk of missing the release window is at a medium risk level, and the emergency release confidence value under the release waiting path after attitude stabilization does not continue to increase, and the release success probability of the next collaborative release window meets the preset requirements, the linkage feedback module generates a collaborative release strategy after attitude stabilization or a nearby device release strategy.

[0091] When the risk of missing the incremental release window is at a high risk level, and the emergency release confidence value meets the preset high confidence condition, the current collaborative release window release success probability meets the preset requirements, and the release interlocking permission result allows release, the linkage feedback module generates an immediate release strategy for the current window.

[0092] When the risk of missing the release window is at an extremely high level, and continuing to wait will lead to an expansion of device failure or a significant increase in the risk of personnel being trapped, and at the same time the central control node fails but meets the local redundancy release confirmation conditions, the linkage feedback module generates a local redundancy release strategy.

[0093] When multiple lifesaving devices simultaneously meet the release conditions, the linkage feedback module generates a release sequence according to the accessibility index of each lifesaving device from high to low, and controls the release interval between adjacent lifesaving devices to be no less than a preset safety interval to avoid interference between releases. The release strategy may also include the target lifesaving device number, suspected emergency event type, emergency release confidence value, disturbance event interpretability, release window missed incremental risk level, recommended release method, release sequence, release safety interval between devices, backup release path, and information to be simultaneously reported to the emergency broadcast system, VDR system, or shore-based rescue system.

[0094] After the actual release results are transmitted back, the linkage feedback module can reverse-label the triggered anomalies. If no real emergency event occurs after release, or if the response is verified to be a disturbance, the corresponding triggered anomaly is marked as a disturbance-type anomaly, and the disturbance event interpretability calculation parameters are updated. If a real emergency event is confirmed after release, the corresponding triggered anomaly is marked as an emergency-type anomaly, and the emergency release confidence value calculation parameters are updated. The system can also update the release success probability calculation parameters based on the difference between the actual and predicted ship attitude stabilization duration; update the releaseability index curve parameters based on changes in the releaseability index after waiting; and update the release window missed incremental risk calculation parameters based on whether an event escalation or personnel inaccessibility occurs after missing the coordinated release window.

[0095] Example 1: This embodiment illustrates how the system identifies non-urgent disturbance-type triggering anomalies and avoids unnecessary accidental releases.

[0096] In this embodiment, the target life-saving device is port-side throwable life raft #3. During the berthing process, the ship experiences a short-term impact, resulting in a significant peak in the three-axis acceleration modulus of port-side throwable life raft #3, a short-term increase in attitude deviation, and a short-term increase in the root mean square value of vibration. The system simultaneously receives the ship's berthing operation signal, and the overall roll angle of the ship fluctuates within a short period. Adjacent port-side throwable life rafts #1, #2, and #4 all experience similar impact responses within the same time window, but all recover to the baseline range after the impact ends.

[0097] The data construction module generates a single-device event perception vector based on the installation location, deck level, relative position to the ship's center of gravity, and berthing impact direction of the port-side drop-off life raft #3. The event discrimination module extracts impact anomalies, attitude anomalies, and vibration anomalies within the anomaly analysis time window.

[0098] The event discrimination module evaluates each interpretation score. The system confirms that the impact anomaly and the berthing impact event are highly synchronized in time; the trend of the impact anomaly is consistent with the trend of the ship's berthing impact; the duration of the anomaly matches the duration of the berthing impact; after the impact, the attitude and vibration anomalies of the target life rafts return to the baseline range; similar short-term anomalies have occurred in similar historical berthing events or test events of the same type of device, but no real emergency event has been confirmed; neighboring life rafts show similar and recoverable responses within the same time window. Based on the above judgments, the system determines that the triggering anomaly has a high degree of explanatory power for the disturbance event.

[0099] The system further generates an emergency release confidence value. Since the impact anomaly was synchronous with the berthing impact and the anomaly recovered after the disturbance ended, the coordination of multiple sensors did not jointly point to sinking, fire, or continuous capsizing events. Although the adjacent devices responded synchronously, they all recovered. The correlation of historical events was low. Therefore, the event discrimination module reduced the contribution of impact anomaly and attitude anomaly to the emergency release confidence value and obtained a lower emergency release confidence value.

[0100] The release prediction and releaseability management module determines that the anomaly meets the conditions of high perturbation interpretability and low emergency release confidence value, thus reducing the event prediction weights for impact anomalies and attitude anomalies. Although the release actuator of target life raft #3 is available at the current moment, due to its low emergency release confidence value, the system determines its waiting level as "fully available for waiting".

[0101] The linkage window analysis module determines that neither the current nor the next coordinated release window presents a significant risk of being unreachable. The risk decision module constructs an immediate release path, a continuous monitoring waiting path, and an encrypted monitoring waiting path for the current window. The immediate release path for the current window carries a high risk of false release. Under the waiting path, the releaseability index of the target life raft is not expected to decrease significantly. Therefore, the path risk of the optimal waiting path is no higher than that of the immediate release path for the current window. The system classifies the incremental risk of missing the release window as low risk or zero incremental risk.

[0102] The safety interlock module detected the absence of a manual release prohibition signal and the release actuator's self-test passed. However, due to the low reliability of the emergency release and the low-risk level of the incremental risk of missing the release window, the linkage feedback module did not output a release command but instead output an encrypted monitoring strategy. The system increased the sensor sampling frequency of port-side drop-off life raft #3 and shortened the wireless linkage status reporting cycle. Subsequently, the target life raft #3 did not exhibit any continuously amplifying anomalies for a period of time, and the system marked this event as a berthing disturbance type anomaly.

[0103] As can be seen from this embodiment, the present invention can avoid erroneous release caused by berthing impact by using the explanatory power of the disturbance event, the confidence value of emergency release, and the incremental risk of missing the release window to jointly determine when the original acceleration and attitude data encounter abnormal conditions.

[0104] Example 2: This embodiment illustrates how the system identifies real emergency-type triggered anomalies and promptly occupies the current collaborative release window.

[0105] In this embodiment, the target lifesaving device is starboard lifeboat #2. After the ship suffers a side collision, it continues to list to starboard. The effective roll angle of starboard lifeboat #2 continues to increase, the effective water pressure continues to rise, the device attitude deviation continues to increase, and the root mean square value of vibration continues to be higher than the reference range. Starboard lifeboat #1 and starboard throw-out life raft #3 also exhibit continuous attitude anomalies and water pressure rise anomalies within the same anomaly analysis time window.

[0106] The data construction module generates a single-device event perception vector based on the davit position of starboard lifeboat #2, its position relative to the ship's center of gravity, and the ship's starboard list direction. The event discrimination module extracts persistent attitude anomalies, water entry anomalies, and vibration anomalies.

[0107] The event discrimination module assesses each interpretation score. The system confirms that the anomaly has a weak temporal synchronization with ordinary rolling or berthing disturbances; the duration of the anomaly is significantly longer than that of short-term disturbances; the anomaly shows no recovery trend within the observation time window; multiple sensor anomalies collectively point to the starboard sinking event chain; and multiple lifesaving devices on the starboard side exhibit coordinated anomalies in the same direction. Based on the above assessments, the system determines that the disturbance event triggering the anomaly has low explanatory power and is difficult to explain by non-emergency disturbances.

[0108] The system further generates an emergency release confidence value. Since the attitude anomaly of starboard lifeboat #2 has not recovered and continues to intensify, and the water ingress anomaly continues to rise, the coordination of multiple devices points to the starboard sinking event chain. The event discrimination module does not enable disturbance deduction and increases the contribution of continuous attitude anomaly, water ingress anomaly and inter-device coordination anomaly to the emergency release confidence value, resulting in a higher emergency release confidence value.

[0109] The release prediction and releaseability management module increases the event prediction weights for persistent attitude anomalies, water entry anomalies, and vibration anomalies, generating a suspected emergency event type of severe list and sinking. The system-generated releaseability index curve shows that: the target lifeboat within the current window still has high releaseability; if continuous monitoring and waiting continue, the target lifeboat may be obstructed by the structure due to continued hull listing, reducing personnel accessibility; the release path after attitude stabilization is unexecutable because the attitude stabilization prediction results do not meet the requirements; the release path after adjacent devices are released first is unexecutable because the interference from the release of adjacent devices cannot be resolved.

[0110] The linkage window analysis module determines that the current collaborative release window has a high probability of successful release; however, due to the ship's continued list to starboard, the starboard lifeboat #2 may be suppressed by the ship's structure or inaccessible to personnel in the next collaborative release window, resulting in a low probability of successful release in the next collaborative release window.

[0111] The risk decision-making module constructs three paths: an immediate release path within the current window, a continuous monitoring waiting path, and an encrypted monitoring waiting path. The path risk of the immediate release path within the current window is lower than that of the waiting path. The optimal waiting path significantly increases the risks of personnel inaccessibility, release failure, and expanded device failure compared to the immediate release path within the current window. Therefore, the incremental risk of missing the release window is classified as either high-risk or extremely high-risk.

[0112] The safety interlock module detected that the manual release prohibition signal was not present, the mechanical locking state met the release conditions, the release actuator passed the self-test, the backup power supply was available, the audible and visual alarms before release were completed, and the attitude trigger source, water entry trigger source and inter-device coordination trigger source of starboard lifeboat #2 all pointed to the same emergency event chain, and the release interlock permission result was output.

[0113] Based on this, the linkage feedback module generates an immediate release strategy for the current window and sends it to the intelligent electronic control node of starboard lifeboat #2 via the ship's wireless linkage network. At the same time, it sends a coordinated release sequence to starboard lifeboat #1 and starboard throw-out life raft #3. The release sequence is arranged from high to low according to the accessibility index, and the release interval between adjacent life-saving devices is not less than the preset safety interval to reduce the risk of release interference.

[0114] As can be seen from this embodiment, when a real emergency event causes the releasability of the target life-saving device to decrease rapidly and the probability of successful release in the next coordinated release window to decrease significantly, the present invention can promptly determine whether the current window should be occupied by the incremental risk of missing the release window, thereby avoiding the delay in release that would cause the life-saving device to become unreachable or fail to be released.

[0115] Example 3: This embodiment is used to illustrate the confirmation of local redundancy release when the central control node fails.

[0116] In this embodiment, a fire occurs in the ship's engine room, causing a partial communication outage. The main link between the ship's central control server and some starboard lifeboats fails. The intelligent electronic control nodes of starboard throw-out life raft #5, starboard throw-out life raft #6, and starboard lifeboat #4 can still maintain communication with neighboring nodes through the Sub-GHz Mesh subnetwork.

[0117] The abnormal temperature rise of starboard throwable liferaft #5 persisted, with smoke concentration significantly higher than the baseline range; starboard throwable liferaft #6 exhibited abnormal smoke concentration and abnormal backup power supply temperature rise within the same time window; starboard lifeboat #4 showed an increased fire trigger confidence value and abnormal attitude. The event discrimination module determined locally at each node that: multiple adjacent nodes generated high emergency release confidence values, the interpretability of the disturbance event was low, and the trigger anomalies collectively pointed to the fire extension event chain.

[0118] Due to the failure of the central control node, the linkage feedback module enters the local redundant release confirmation process. The safety interlock module determines that the manual release prohibition signal does not exist, the mechanical locking state meets the release conditions, the release actuator self-test passes, the backup power supply is available, and at least two adjacent intelligent electronic control nodes generate high emergency release confidence values ​​within the same anomaly analysis time window.

[0119] The system further assesses the incremental risk of missing the release window. While the current coordinated release window remains feasible, the probability of successful release in the next coordinated release window may be significantly reduced due to fire spread, smoke obstruction, inaccessibility of personnel, or deterioration in communication quality. The risk decision module determines that waiting for the release window to proceed significantly increases the risk of inaccessibility of personnel and the risk of release failure compared to releasing immediately within the current window; therefore, the incremental risk of missing the release window is classified as high-risk.

[0120] After the local redundancy release confirmation conditions are met, starboard-side drop-out life rafts #5 and #6 sequentially enter the local redundancy release strategy according to their accessibility index. The system first issues a pre-release alarm via audible and visual devices, followed by the exchange of confirmation frames between adjacent nodes, and then the release actuators are driven by their respective intelligent electronic control nodes. The release result is transmitted back to available nodes via the Sub-GHz Mesh subnetwork and synchronized to the shore-based rescue system after the ship-to-shore remote emergency reporting link is restored.

[0121] As can be seen from this embodiment, when the central control node fails, the present invention does not allow a single node to release independently. Instead, it requires that at least two adjacent nodes, at least two independent triggering sources, a safety interlocking state, and the incremental risk of missing the release window all meet the conditions, thereby taking into account both local redundant release capability and the safety of preventing accidental release.

[0122] Example 4: This embodiment illustrates that system parameters can be updated through historical events, simulation events, and experimental data.

[0123] In this embodiment, after the system is running, it continuously records the triggering abnormal segments of each lifesaving device, the single device event perception vector, the explanatory power of the disturbance event, the confidence value of emergency release, the releaseability index curve, the incremental risk of missing the release window, the final release strategy, and the post-verification results.

[0124] If a historical anomaly segment recovers after a disturbance event ends, and verification confirms no actual emergency event, then that historical anomaly segment is marked as a disturbance-type anomaly sample. If a historical anomaly segment persists during normal ship rolling and sea state fluctuations, or if verification confirms the existence of an actual emergency event such as a collision, fire, capsizing, or sinking, then that historical anomaly segment is marked as an emergency-type anomaly sample.

[0125] The system can utilize the participation weights of the above-mentioned sample update time synchronization score, trend consistency score, duration matching score, recovery score, historical disturbance similarity score, and neighboring device response score. It can also update the participation weights of basic anomaly contribution, synergy enhancement contribution, historical event correlation contribution, and disturbance deduction contribution during the emergency release confidence value generation process.

[0126] When historical samples are insufficient, the system can incorporate simulated events, expert rule bases, and test data from devices of the same model to form initial samples. For example, it can generate continuous capsizing events through ship attitude simulation, berthing impact events through vibration table tests, water ingress events through hydrostatic tests, and fire events through temperature and smoke tests. The system initially operates using a rule-based calculation method, and then gradually adjusts the parameters based on actual verification results.

[0127] As can be seen from this embodiment, the historical event association information of the present invention does not rely solely on a large sample of real accidents, but can be derived from historical real events, simulated events, expert rule bases, and test data of the same type of device, thereby improving the feasibility of the system.

[0128] The technical scope of this invention is not limited to the content described above. Those skilled in the art can make various modifications and variations to the above embodiments without departing from the technical concept of this invention, and all such modifications and variations should fall within the protection scope of this invention.

Claims

1. A wireless linkage-based lifesaving device intelligent electric control system, characterized in that, The system is used for the release control management of shipborne lifeboats or throwable life rafts. The data construction module receives multi-source sensor data, ship attitude data, sea state disturbance data, device status data, and wireless linkage status data from the target life-saving device, and generates a single-device event perception vector based on the installation location information of the target life-saving device. When a triggering anomaly occurs in the target life-saving device, the event discrimination module extracts the triggering anomaly features, determines the degree to which the triggering anomaly can be explained by non-emergency disturbances based on the single-device event perception vector, generates a disturbance event explanatory power, and generates an emergency release confidence value by combining anomaly recovery status, multi-sensor coordination, inter-device coordination, and historical event correlation information. The release prediction and releaseability management module adjusts the event prediction weights according to the disturbance event explanatory power and the emergency release confidence value, and generates a releaseability index curve and a waiting level. The linkage window analysis module generates the current collaborative release window and the probability of successful release in the next collaborative release window based on the ship attitude prediction results, the coordination relationship between devices, the quality of the wireless linkage link, the status of the release actuator, the status of the backup power supply, the risk of interference with the release of adjacent devices, and the status of obstacles around the device. The risk decision module compares the path risk of the immediate release path within the current window with the set of executable waiting paths to obtain the incremental risk of missing the release window; the safety interlock module acquires the manual prohibition signal, mechanical locking status, release actuator self-test results, backup power status, pre-release audible and visual alarms, and confirmation status from at least two independent trigger sources to generate a release interlock permission result; the linkage feedback module generates a release strategy based on the incremental risk of missing the release window, the waiting level, and the release interlock permission result, and distributes it to each intelligent electronic control node through the ship's wireless linkage network; when the central control node fails, adjacent nodes that meet the local redundant release confirmation conditions collaboratively output the local redundant release strategy.

2. The wireless link-based lifesaving device intelligent control system according to claim 1, wherein, The data construction module converts ship-level or section-level data into single-device event perception vectors based on the target life-saving device's installation location, deck level, relative to the ship's center of gravity, lever arm relative to the ship's rocking axis, adjacent device layout, ship's dominant motion direction, ship's draft, sea wave direction, and acquisition time. These vectors include at least two of the following: effective roll angle, effective pitch angle, effective impact acceleration, effective water pressure, effective temperature rise, effective smoke concentration, effective root mean square value of vibration, device attitude deviation, release mechanism availability flag, backup power availability flag, wireless link confidence, and trigger event label.

3. The wireless link-based lifesaving device intelligent control system according to claim 1 or 2, characterized in that, When determining the explanatory power of a disturbance event, the event discrimination module first obtains at least two of the following: time synchronization score, trend consistency score, duration matching score, recovery score, historical disturbance similarity score, and neighboring device response score. It then sets participation weights based on the reliability of each score data, sensor status, and trigger anomaly type, and finally fuses the scores according to the weights to generate the explanatory power of the disturbance event. When the explanatory score is missing or abnormal, the weights are adjusted or a preset default value is used.

4. The intelligent electronic control system for wireless link-based lifesaving device according to claim 1 or 3, characterized in that, The event discrimination module determines whether the triggered anomaly has a non-urgent disturbance basis based on the explanatory power of the disturbance event, and generates an emergency release confidence value by combining the anomaly recovery status, multi-sensor synergy, inter-device synergy, and historical event correlation. The emergency release confidence value is jointly determined by the basic anomaly contribution, the synergy enhancement contribution, the historical event correlation contribution, and the disturbance deduction contribution.

5. The intelligent electronic control system for a life-saving device based on wireless linkage according to claim 1, characterized in that, The release prediction and releaseability management module adjusts the event prediction weights using the emergency release confidence value, and generates a releaseability index curve and a waiting level based on the event prediction weights, the status of the release actuator, the status of the backup power supply, the quality of the wireless link, the device attitude status, and the personnel accessibility status. The waiting levels include no waiting, short-term waiting, cooperative waiting, and full waiting.

6. The wireless link-based lifesaving device intelligent control system according to claim 1, wherein, The linkage window analysis module generates the success probability of the current collaborative release window and the next collaborative release window based on factors such as ship attitude stability factor, release actuator availability factor, backup power availability factor, wireless linkage link reliability factor, personnel accessibility factor, nearby device release interference risk factor, and obstacle risk factor around the device.

7. The wireless link-based lifesaving device intelligent control system according to claim 1, wherein, The risk decision-making module constructs a set of immediate release paths and executable waiting paths for the current window, compares the risks of each path, and determines the difference between the risk of the optimal waiting path and the risk of the immediate release path as the incremental risk of missing the release window. The waiting paths include continuous monitoring waiting paths, encrypted monitoring waiting paths, release waiting paths after attitude stabilization, release waiting paths after nearby devices release first, and release waiting paths after manual confirmation.

8. The wireless link-based lifesaving device intelligent control system according to claim 1, wherein, Before the linkage feedback module outputs the release strategy, the safety interlock module judges the release interlock permission result, including manual release prohibition signal, mechanical locking status, release actuator self-test, backup power status, pre-release audible and visual alarm and confirmation from at least two independent trigger sources. When the central control node fails, only adjacent nodes that meet the local redundant release confirmation conditions can execute the local redundant release strategy.

9. The wireless link-based lifesaving device intelligent control system according to claim 1, wherein, The linkage feedback module generates a release strategy based on the incremental risk of missing the release window, the waiting level, and the release interlock permission result. This strategy includes a continuous monitoring strategy, an encrypted monitoring strategy, a coordinated release strategy after attitude stabilization, a release strategy prior to nearby devices, an immediate release strategy for the current window, and a local redundant release strategy. The strategy is then distributed to each intelligent electronic control node via the ship's wireless linkage network. When multiple devices simultaneously meet the release conditions, a release sequence is generated according to the reachability index, with the release interval between adjacent devices not less than a preset safety interval.

10. The intelligent electronic control system for a life-saving device based on wireless linkage according to any one of claims 1 to 9, characterized in that, The target life-saving device is a shipboard lifeboat or a throwable life raft. The intelligent electronic control node includes a microcontroller, a communication module, a sensor interface, a power management unit, a release actuator drive unit, a mechanical locking status detection unit, and a local storage unit. The release actuator can be an electromagnetic release device, an electronically controlled lock device, or a release hook controller. The system is suitable for collaborative operation of a ship's centralized control center server, an edge computing gateway, a shore-based cloud server, and a local controller of the intelligent electronic control node.