Shipborne overwater positioning and alarm system based on BeiDou and GPS remote sensing

Abstract

This invention relates to the field of shipborne overboard alarm systems, and discloses a shipborne overboard positioning and alarm system based on BeiDou GPS remote sensing. The system includes a remote sensing inventory unit, a ship attitude sensing unit, an exit sensor, and an alarm host. The alarm host cross-verifies the number of personnel observed and the ship's attitude to trigger an alarm. It also uses exit sensor signals to gate the absence of personnel and verifies the logical consistency between the expected number of personnel baseline and the measured data. This invention constructs a triple alarm logic of state verification, exit gating, and baseline consistency, shifting alarm triggering from reliance on individual cooperation to system platform monitoring. This solves the alarm failure problem in non-cooperative states and can distinguish between benign entry into the cabin and actual overboard incidents, avoiding systemic missed alarms due to configuration errors.

Description

Technical Field

[0001] This invention relates to a shipborne overboard positioning and alarm system based on BeiDou GPS remote sensing, belonging to the technical field of shipborne overboard alarm systems. Background Technology

[0002] Currently, a widely accepted method relies on alarm beacons worn by individual crew members. These devices, once activated, send a distress signal containing location information to initiate search and rescue, forming the basic technology for current overboard alarms. However, the entire reliability of this method depends on the uncontrollable factor of individual compliance. Under high-intensity deck work conditions, crew members may choose not to wear beacons due to operational inconvenience or the risk of equipment snagging. If a crew member falls overboard due to sudden cross waves or operational errors, all alarm systems relying on individual beacons will be silent, rendering the alarm function completely ineffective. Conversely, unexpected beacon activation, signal loss, or battery depletion can frequently trigger high-priority alarms, causing work interruptions and reducing system reliability. Reliability; meanwhile, another type of existing shipborne terminal design focuses on the ship's own positioning and communication functions, but it also fails to solve the problem of non-cooperative monitoring of deck personnel. For example, Chinese invention patent CN101329180B discloses a dual-mode shipborne terminal and method based on the Beidou satellite navigation system for positioning and communication. Although this solution integrates GPS and Beidou systems, its main function is to realize the monitoring and reporting of ship position and information interaction between ship and shore. Emergency alarm functions (such as one-click alarm) still rely on the active triggering of crew members. For non-cooperative overboard incidents involving crew members who are not wearing beacons in sudden situations, such terminals are essentially still in a passive state and cannot provide automated system-level monitoring and alarms.

[0003] The limitation of this alarm logic lies in its confusion between the physical state of the equipment and the safety status of the personnel. It attempts to solve this by improving the positioning accuracy or communication performance of individual devices, but it cannot avoid the alarm silence caused by not wearing the device. Another approach is to use complex visual recognition algorithms to track deck personnel, but due to their low reliability and high cost in harsh sea conditions such as wind, waves, rain, fog, and night, they are difficult to use as a reliable alarm basis in the shipboard environment.

[0004] Therefore, the technical problem to be solved by this invention is how to provide a system-level dual verification of personnel absence status and ship danger events that does not rely on individual wearing devices. Summary of the Invention

[0005] This invention provides a shipborne overwater positioning and alarm system based on BeiDou GPS remote sensing. Its main purpose is to solve the problems of not relying on individual wearable devices, being able to achieve dual verification of personnel absence status and ship danger events at the system level, and thus avoiding alarm failure and false alarm interference.

[0006] To achieve the above objectives, this invention provides a shipborne overboard positioning and alarm system based on BeiDou GPS remote sensing, comprising a baseline setting unit, a remote sensing inventory unit, a hull attitude sensing unit, an exit sensor, and an alarm host:

[0007] The remote sensing inventory unit includes remote sensing sensors, which are deployed in the ship's deck work area;

[0008] The exit sensors are deployed at pre-designated safety exits in the ship's deck work area;

[0009] The alarm host is communicatively connected to the baseline setting unit, remote sensing inventory unit, ship attitude perception unit, and exit sensors. The alarm host is used to acquire the expected personnel number baseline defined by the baseline setting unit; acquire the number of observed personnel collected in real time by the remote sensing inventory unit; acquire the attitude parameters characterizing the ship's motion state collected in real time by the ship attitude perception unit; generate a personnel missing status signal when the number of observed personnel is determined to be less than the expected personnel number baseline; generate a ship high-risk event signal when the attitude parameters are determined to exceed the high-risk threshold; and trigger a fall-over alarm signal when both the personnel missing status signal and the ship high-risk event signal are simultaneously determined.

[0010] Furthermore, the alarm control panel is also used to: when a personnel absence status signal is generated, determine whether a trigger signal from the exit sensor is received within a preset time window; if no trigger signal from the exit sensor is received within the preset time window, a high-priority warning signal is triggered; and the alarm control panel is also used to: maintain the net outflow count of deck personnel based on the trigger signal from the exit sensor; verify whether the expected personnel number baseline matches the preset logical relationship between the observed personnel number and the net outflow count of deck personnel; if they do not match, a baseline error alarm signal is triggered.

[0011] Preferably, the remote sensing sensor includes a thermal imaging sensor; the alarm host uses logic to obtain the number of observers, including: acquiring a thermal image collected by the thermal imaging sensor; performing a hot spot patch counting algorithm on the thermal image to generate the number of observers; the hot spot patch counting algorithm is an image processing algorithm used to count the number of pixel clusters in a thermal image whose brightness exceeds a preset background threshold.

[0012] Preferably, the ship attitude sensing unit includes a BeiDou global positioning system receiver with an integrated inertial measurement unit, and the attitude parameters include at least one of the ship's roll angular velocity and heave acceleration; the alarm host uses logic to generate a high-risk event signal for the ship, including: comparing the instantaneous value of the ship's roll angular velocity or the instantaneous value of the heave acceleration with a high-risk threshold, wherein the high-risk threshold is a preset value.

[0013] Preferably, the logic for generating a personnel absence status signal in the alarm host further includes: determining that the number of observed personnel is less than the expected baseline number of personnel, and continuing for more than a preset status confirmation time; the water fall alarm signal is a level two alarm signal; the high-priority warning signal is a level one warning signal, and the level one warning signal is a local audible and visual signal triggered on the control panel for verification.

[0014] Preferably, the logic used by the alarm host to verify the baseline of the expected number of personnel includes: determining whether the baseline of the expected number of personnel is equal to the sum of the observed number of personnel and the net outflow count of personnel on deck; the baseline error alarm signal is a blocking alarm used to lock the triggering of the fall-over alarm signal and the high-priority warning signal before the baseline of the expected number of personnel is corrected.

[0015] Preferably, the alarm control panel is also used to: continuously collect attitude parameters within a long preset time window, calculate the statistical average and statistical standard deviation of the attitude parameters, and dynamically determine a high-risk threshold; wherein the high-risk threshold Determined by the following rules: ,in The statistical average of attitude parameters over a long preset time window. The statistical standard deviation of attitude parameters over a long preset time window. This is the preset sensitivity coefficient.

[0016] Preferably, the alarm control panel is also used to: determine the statistical average value of attitude parameters over a long preset time window. Whether it continuously deviates from the preset safety benchmark value; if the statistical average value is determined. If the deviation from the safety benchmark value continues, an independent ship attitude anomaly alarm signal will be triggered.

[0017] Preferably, the alarm host is also used to: analyze the image quality parameters of the thermal image before executing the hot spot patch counting algorithm; the image quality parameters include the image entropy or pixel intensity standard deviation of the thermal image; when the image quality parameters are lower than a preset confidence threshold, generate a sensor blinding signal and prevent the triggering of the water fall alarm signal and the high-priority warning signal.

[0018] Preferably, the baseline setting unit includes a manual operation interface mounted on the bridge, which is used to receive the expected personnel number baseline input by the captain or watch officer.

[0019] Preferably, the exit sensor includes a photoelectric beam detector or a near-field microwave radar sensor, which generates a trigger signal when a person passes through a preset safety exit.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] 1. By establishing a dual verification mechanism for personnel absence status signals and ship high-risk event signals, the alarm triggering conditions are transformed from relying on the physical state of uncontrollable individual-worn devices to a cross-logical judgment of the ship's system platform status and dangerous events. This reconstruction of alarm logic allows the alarm system to still monitor personnel absence status even when operators are not wearing beacons as required, avoiding alarm failure due to individual non-cooperation. At the same time, by using information from the orthogonal dimension of ship high-risk events to corroborate personnel absence status, the system filters out single status signals generated by normal personnel entering the cabin or temporary sensor obstruction, distinguishing between benign personnel reductions and real dangerous events, thereby improving the confidence of the alarm.

[0022] 2. By utilizing exit sensors deployed at preset safety exits, the exit vector of the personnel absence status signal is gated and verified. When a personnel absence occurs within a preset time window and no trigger signal is received from the exit sensor, the system determines this to be a non-standard disappearance event that did not occur at a safety exit. The introduction of this mechanism constructs a parallel alarm path for the main alarm logic that does not depend on the ship's attitude parameters, enabling the system to distinguish between benign entry into the cabin and high-risk disappearance, and solving the problem of alarm omissions that may occur in calm sea conditions due to the lack of evidence of high-risk ship events.

[0023] 3. Using the number of personnel observed by the remote sensing inventory unit and the net outflow count maintained by the exit sensor signal, an internal logical self-consistency check is performed on the baseline of the expected number of personnel set by the operator. When the check finds that the baseline does not match the preset logical relationship between the two measured values, a baseline error alarm signal is triggered. This mechanism uses the measured data of the sensor to audit the human input premise of the alarm logic in reverse. Before the main alarm function is started, it locks the systemic functional failure that may be caused by the initial configuration error, transforming the alarm device from a passive response to events to an active protection of the integrity of its own operating logic. Attached Figure Description

[0024] Figure 1 This is a diagram of the triple logic verification architecture of the alarm system of the present invention;

[0025] Figure 2This is a graph showing the relationship between the ship attitude parameters and the dynamic high-risk threshold of this invention.

[0026] Figure 3 This is a system physical deployment diagram of the bridge and deck work area of ​​the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the scope of protection of the invention.

[0028] This invention provides a shipborne overboard positioning and alarm system based on BeiDou GPS remote sensing. The system, as a whole, mainly includes a baseline setting unit, a remote sensing inventory unit, a ship attitude sensing unit, an exit sensor, and an alarm host as the logic center. The alarm host communicates with each unit. Its core function is to no longer rely on individual beacons, but instead cross-validate deck personnel status information obtained by the remote sensing inventory unit and ship event information obtained by the ship attitude sensing unit, combined with the gating signals from the exit sensor and baseline self-consistency verification logic, to construct a multi-dimensional alarm decision system. The baseline setting unit is used to obtain the expected personnel number baseline defined in the shipborne deck work area. In the specific deployment of this technical solution, this unit can be installed on the ship's bridge using a manual operation interface, allowing the captain or watch officer to manually input and confirm the expected personnel number baseline value when deck work shifts change. This value can be represented as follows: The manual setting method avoids systemic risks introduced by automated scheduling systems due to logical complexity or avoidance, ensuring the reliability of the baseline source; the remote sensing inventory unit is used to obtain the number of observers in real time, which can be expressed as follows: In shipboard deck work areas, which are susceptible to interference from wind, waves, rain, fog, strong light, or nighttime conditions, the remote sensing inventory unit preferably includes a wide-angle thermal imaging sensor to eliminate monitoring blind spots. The alarm host acquires thermal images collected by the thermal imaging sensor and applies a hotspot patch counting algorithm to them. The algorithm is strictly limited to counting the number of pixel clusters in the image whose brightness exceeds a preset background threshold, rather than performing human figure recognition in the field. This allows for the determination of the number of observers under adverse sea conditions with extremely low computational overhead. To address the issue of sensor physical blinding, the alarm control unit also analyzes the image quality parameters of the thermal image before executing the hot spot patch counting algorithm. These parameters can be the image entropy or pixel intensity standard deviation of the thermal image. When the image contrast drops sharply under dense fog or water mist, causing the image quality parameters to fall below the preset confidence threshold, the alarm control unit generates a sensor blinding signal to prevent the subsequent triggering of water fall alarm signals and high-priority warning signals. Instead, it triggers a sensor blinding alarm signal that indicates a degraded system function.

[0029] The ship attitude sensing unit is used to acquire attitude parameters characterizing the ship's motion state in real time. This unit can be specifically implemented as a BeiDou global positioning system receiver with an integrated inertial measurement unit (IMU). In this technical solution, its use is strictly limited to acquiring the ship attitude data it generates, rather than positioning data specific to this field. The attitude parameters may specifically include at least one of the ship's roll rate and heave acceleration. To address the failure of static thresholds in the alarm logic under varying sea conditions, the alarm host dynamically determines high-risk thresholds. The procedure includes: continuously collecting attitude parameters within a long preset time window, such as the past 10 minutes, and calculating the statistical average value of the attitude parameters. with statistical standard deviation These two values ​​together characterize the current average sea state; high-risk threshold Then it is determined through adaptive rules, such as ,in, This represents the statistical average of the attitude parameters over a long preset time window. This represents the statistical standard deviation of the attitude parameters over a long preset time window. The preset sensitivity coefficient; this dynamic threshold This ensures that event signals are triggered only in response to instantaneous shocks exceeding the current mean sea state, thereby suppressing false alarms in high sea states; simultaneously, the alarm control unit is also used to determine statistical averages. Whether there is a continuous deviation from the preset safety benchmark value to identify slow hull attitude anomalies. If such slow deviation is determined, an independent ship attitude anomaly alarm signal is triggered. Exit sensors are deployed at each preset safety exit in the ship's deck work area, such as doors or passages leading to the cabin or bridge. The implementation of the exit sensor can use components commonly used in the art, such as photoelectric beam probes or near-field microwave radar sensors. Its only function is to generate a binary trigger signal when personnel pass through the preset safety exit.

[0030] The alarm control panel, as the logical hub of the system, is used to execute state-event dual verification logic; firstly, the alarm control panel determines the number of observers. Less than the expected number of personnel baseline Furthermore, if the missing person status persists for more than a preset status confirmation time, such as 10 seconds, to filter out brief obstructions, a missing person status signal is generated; secondly, the alarm host simultaneously compares the real-time attitude parameter instantaneous values ​​with the dynamic high-risk threshold. Comparison: When the attitude parameters exceed the threshold, a high-risk event signal is generated. Ultimately, the alarm host only triggers the highest-priority overboard alarm signal, which can be a ship-wide secondary alarm signal, when both the missing person status signal and the high-risk event signal are simultaneously confirmed. To handle overboard scenarios in calm seas, the alarm host also executes parallel exit vector gating logic. That is, when a missing person status signal is generated, the alarm host immediately determines whether a trigger signal from the exit sensor has been received within a preset time window, such as 3 seconds before and after the missing person status is confirmed. If so... If no trigger signal is received from any exit sensor within the preset time window, the system determines this to be a non-standard disappearance event that did not occur at a safe exit and triggers a high-priority warning signal. This high-priority warning signal can be implemented as a local audible and visual signal triggered on the bridge to prompt immediate verification, i.e., a level one warning signal. To ensure the reliability of the alarm logic's operational premise, the alarm host is also used to execute a logic self-consistency verification mechanism. Based on the trigger signal from the exit sensor, the host internally maintains the net outflow count of deck personnel, for example, personnel exiting the cabin is counted as +1, and entering the cabin as -1. The system also monitors the baseline of expected personnel numbers set or updated by the operator. At that time, the host immediately performs a verification to determine the baseline of the expected number of personnel. Is it equal to the number of observers? The sum of the net outflow count of deck personnel; if the verification result is inconsistent, it indicates that there is a conflict between the logical premise of the system (human input) and the physical measurement (sensor input), and the host immediately triggers the baseline error alarm signal; the baseline error alarm signal is a blocking alarm used to lock the triggering of the fall-over alarm signal and the high-priority warning signal before the expected personnel number baseline is corrected.

[0031] Example 1: In a nighttime deck operation scenario with high sea states, the shipborne overboard positioning and alarm system is deployed and operational. The alarm host obtains the expected personnel number baseline from the baseline setting unit. The system is set to operate with a capacity of 5 people by the bridge operator. During continuous operation, the ship's attitude sensing unit, as in the specific implementation, collects attitude parameters within a long preset time window and calculates the statistical average value reflecting the average wind and wave conditions at that time. with statistical standard deviation This allows for the determination of dynamic high-risk thresholds. At a certain moment, the ship's hull was suddenly hit by an abnormal cross wave, causing the instantaneous value of the ship's roll angular velocity detected by the ship's attitude sensing unit to exceed the dynamic high-risk threshold. The alarm host thus generates a high-risk event signal for the ship; almost at the same time, a worker (person A) who was not wearing an individual beacon as required was hit by the cross wave and fell into the water, while another worker (person B) quickly entered the cabin through a pre-set safety exit equipped with an exit sensor in order to avoid the rocking.

[0032] Following this sequence of events, the remote sensing inventory unit deployed in the deck work area, using thermal imaging sensors, reported the number of observers via a hotspot patch counting algorithm. The number of people decreased from 5 to 3; the alarm control panel received this. The value is determined, and it is determined to be less than 1. The value (3<5) and this state continues for more than the preset state confirmation time, thus generating a personnel absence status signal; at the same time, the alarm host also receives a trigger signal from the exit sensor of the preset safety exit (triggered by personnel B). The alarm host enters the logic decision, and the system executes the state-event dual verification logic to determine whether the two conditions (personnel absence status signal is true) and (ship high-risk event signal is true) are met simultaneously; in this scenario, both conditions are met, therefore, the alarm host immediately triggers the highest priority overboard alarm signal, i.e., the second-level alarm signal, notifying the entire ship to take emergency measures; the execution of this decision demonstrates the core logic of this solution, that is, the system uses the temporal coupling of personnel absence and ship danger as the basis for triggering the highest priority alarm. This logic takes precedence over other parallel logics, so that even when the operator is not cooperating (not wearing a beacon), the actual overboard event of personnel A can still be captured; in another parallel logic of this scenario, if the ship does not encounter the cross wave impact (ship high-risk event signal is false), and only personnel B enters the cabin, then The number of people is reduced to 4 (generating a personnel missing status signal), but when the alarm host executes the exit vector gating logic, it will receive a trigger signal from the exit sensor. The system determines that this is a benign entry activity and therefore will not trigger a high-priority warning signal.

[0033] Example 2: To objectively verify the reliability of the alarm logic of the present invention in dealing with complex shipboard deck operation scenarios, especially its performance in avoiding missed alarms and false alarms, the following comparative test was set up; the test was conducted on the deck of an engineering vessel in an anchored state. The deck operation area included a preset safety exit leading to the cabin and was equipped with all the sensing units of the system of the present invention, including a wide-angle thermal imaging sensor, a Beidou GPS receiver with integrated IMU, and an exit sensor; the test set up three system sample groups: control group A, which adopted the alarm method of relying on individual beacons known in the art, specifically the standard AIS-MOB beacon; control group B, which adopted part of the scheme of the present invention, that is, only the personnel absence status signal logic of the remote sensing inventory unit (hot spot patch counting) and the alarm host was enabled, but the logic gating of the ship attitude perception unit and the exit sensor was not enabled; the present invention sample group adopted the complete technical solution disclosed in the specific implementation, that is, all logics of state-event dual verification, exit vector gating, and logic self-consistency verification were enabled simultaneously.

[0034] The experiment simulated five key diagnostic scenarios faced by the system during shipboard operations; Scenario S1 (benign entry into the cabin): 5 people working on deck ( Scenario S2 (Non-cooperative fall into the water): 5 people on deck. One person, C, enters the cabin normally through the preset safety exit. No high-risk events occur during this process. (Set to 5), a person D not wearing an AIS-MOB beacon is involved in a simulated high-risk ship event (the ship's attitude sensing unit detects that the roll rate instantaneously exceeds 5). When the threshold occurs, simulate a fall into the water (i.e., leaving the deck monitoring area and not passing through the exit); Scenario S3 (Missing configuration error): Before the start of the operation, the bridge operator mistakenly... The scenario is set to have 4 people, but there are actually 5 workers on deck. One of them, E, simulates falling overboard (without passing through the exit) when no high-risk ship event occurs; Scenario S4 (false alarm of individual beacon): 5 people on deck ( Setting it to 5), there were no missing personnel or high-risk ship events, but the AIS-MOB beacon in control group A was unexpectedly activated on deck; Scenario S5 (Individual beacon miss): There were 5 people on deck ( (Set to 5), a person F wearing an AIS-MOB beacon is involved in a high-risk ship incident (the roll rate instantaneously exceeds 5). The system simulated falling into water when the threshold was reached, but the beacon it was wearing failed to activate due to a malfunction.

[0035] Table 1: Alarm Response Results of Different System Samples in Various Simulation Scenarios

[0036]

[0037] Referring to Table 1, the experimental results are analyzed as follows: In scenario S1 (benign entry into the cabin), control group B, due to its reliance solely on personnel absence ( The single logic (changing from 5 to 4) triggers false alarms; while the alarm host of the present invention, because it receives the trigger signal from the exit sensor within the preset time window for generating the personnel absence status signal, its exit vector gating logic takes effect, determines this as a benign exit, and thus suppresses the alarm. The S1 result shows the necessity of the exit sensor gating logic for distinguishing between benign and high-risk disappearances; in the S2 (non-cooperative fall into water) and S5 (beacon failure fall into water) scenarios, the control group A, because its logic completely depends on the cooperation or functional integrity of individual beacons, experienced false alarms in both cases; although the control group B detected personnel absence and triggered an alarm, its alarm level could not distinguish the severity of the event; while the state-event dual verification logic of the present invention's sample group was activated, and the alarm host simultaneously determined the personnel absence status signal ( (Change to 4) and ship high-risk event signals (attitude parameters exceeding threshold) When both conditions are met simultaneously, the highest priority level 2 alarm (water fall alarm signal) is triggered. The results of S2 and S5 show the effectiveness of the state-event dual verification logic in capturing non-cooperative and beacon failure water fall events. In the S3 (configuration error missed report) scenario, due to... It was mistakenly set to 4. After person E fell into the water, The logic of changing from 5 to 4 in control group B ( The alarm host of the present invention receives a false alarm (4 < 4 is false), thus a false alarm occurs; When the set value is reached, a logical self-consistency check is immediately performed to determine... (4) Is it equal to (5)+ (0), the verification result is inconsistent, therefore a baseline error alarm signal is triggered at the start of the operation to prevent the main alarm function from running under the incorrect configuration. The result of S3 shows the necessity of the logical self-consistent verification mechanism to avoid systemic functional failure; in the S4 (individual beacon false alarm) scenario, the control group A triggers a false alarm; while the remote sensing inventory unit of the sample group of this invention is measured It is 5, and (5) Matched, no personnel missing status signal was generated, the logic remained silent, and the result of S4 showed that the logic of this solution is immune to abnormal physical status of individual equipment.

[0038] Example 3: This example combines Figures 1 to 3 A description of a shipborne overwater positioning and alarm system based on BeiDou and GPS remote sensing, such as... Figure 1As shown, the architecture includes a remote sensing inventory unit, a hull attitude sensing unit, an exit sensor, and a baseline setting unit. All of these units are communicatively connected to the alarm host, which serves as the logic hub. The remote sensing inventory unit is designed to use thermal imaging to collect real-time data on the number of personnel. The hull attitude sensing unit is designed to use BeiDou GPS and IMU to collect attitude parameters, such as roll or heave, in real time. The exit sensor is deployed at the safety exit to monitor personnel passage and generate an exit trigger signal. The baseline setting unit is manually operated through the bridge to generate a baseline for the expected number of personnel. The alarm host executes a triple alarm logic: first, a status check indicating missing personnel and high ship risk; second, an exit gate control indicating missing personnel and no exit trigger; and third, a baseline self-consistency baseline that does not match the actual measurement. Based on the above logic, a secondary overwater alarm signal, a primary high-priority warning signal, and a baseline error alarm signal are triggered respectively.

[0039] like Figure 2 As shown, this graph uses time (in minutes) on the horizontal axis and attitude parameters (in degrees / second) on the vertical axis. The graph illustrates this relationship through the relationship of three curves: the instantaneous value of the ship's roll rate, the moving average value representing the current mean sea state, and the other two curves. and based on the mean and statistical standard deviation Determined dynamic high-risk threshold A high-risk event signal for a ship is only triggered when the instantaneous value of the solid line crosses above the threshold of the short dashed line; for example... Figure 3 As shown, the system is physically divided into two main nodes: the bridge and the deck work area. The two are connected by communication. The bridge has a core logical hub that performs triple verification: the alarm host, and a baseline setting unit with a manual operation interface. The deck work area has a remote sensing inventory unit, specifically a thermal imaging sensor, and a hull attitude sensing unit, specifically a Beidou global positioning system receiver with an integrated inertial measurement unit. The system also has exit sensors deployed at preset safety exits, such as photoelectric beam detectors.

[0040] Example 4: The reliability of the alarm system of this invention depends on the precise matching of its core logic parameters with the physical environment, sensor layout, and operational characteristics of a specific ship. These parameters are not universally fixed values ​​but need to be determined during system deployment or maintenance through standardized engineering calibration procedures. These procedures are used to determine the logic parameters to improve the accuracy of alarm triggering. The first step of these calibration procedures is to determine the sensor layer parameters of the remote sensing inventory unit, namely the background threshold and confidence threshold of thermal imaging. Under the initial condition of no personnel present in the ship's docking and deck work area, the calibration procedure is initiated. The alarm host continuously collects raw thermal images from the thermal imaging sensor at a preset frequency, such as once per minute, throughout the complete day-night cycle to obtain a background image dataset. To calibrate the background threshold, the alarm host performs a hot spot patch counting algorithm on all images in the dataset to count the maximum pixel cluster brightness value that can be generated by all non-personnel heat sources. The background threshold is set to a value higher than the maximum background noise value, for example, a setting of... To calibrate the confidence threshold, maintenance personnel simulated blinding conditions that the system might encounter in real sea conditions, such as using high-pressure water mist to simulate heavy rain or giant wave spray in front of the sensor lens, and collected a dataset of blinding images. The alarm host then executed the image quality parameter calculation path mentioned in the specific implementation for each frame in the blinding image dataset. Taking image entropy calculation as an example, the algorithmic steps include: calculating the pixel intensity histogram of the thermal image; and secondly, calculating the intensity level of each pixel based on the histogram. probability of occurrence Finally, through The image entropy of this frame is calculated. The alarm control panel calculates the maximum entropy of all images in the blinding image dataset. The confidence threshold is then set to a value higher than the maximum approximate blind entropy, for example, a setting of... .

[0041] The second step of this calibration procedure is to determine the logic layer timing parameters of the alarm host, namely the status confirmation duration and the preset time window. This step requires the cooperation of personnel in the deck work area. To calibrate the status confirmation duration, test personnel walk along the normal work route in the deck work area, deliberately passing behind large obstructions in the ship's deck work area multiple times. The remote sensing inventory unit continuously monitors, and the alarm host records the complete duration from when a person disappears from the thermal image due to brief obstruction until they reappear, and calculates the maximum obstruction duration among all test samples. The status confirmation time is set to a safety value greater than this maximum value; for example, it is set as follows: Seconds; to calibrate the preset time window, testers passed through the preset safety exit multiple times at different speeds; during each passage, the alarm control panel simultaneously recorded two time points: the moment the exit sensor was triggered. The number of personnel confirmed by the hotspot patch counting algorithm of the remote sensing inventory unit (and the number of personnel) The moment when the decrease occurs The alarm control panel calculates the maximum absolute time difference between these two moments across all test samples. The preset time window is set to a tolerance value that can cover the maximum time difference, for example, the setting is as follows: Seconds; By executing the above-mentioned standardized engineering calibration procedures, the four core thresholds and timing parameters of the alarm system are all determined to be calibration engineering values ​​based on specific ship physical measurement data; This calibration process makes the alarm logic of the system compatible with the specific environment of the deployed ship, including background thermal noise, obstruction layout and exit channel delay, thereby helping to reduce the probability of false alarms and missed alarms before the system is put into actual operation.

[0042] Example 5: After the alarm system is installed on a specific vessel, a systematic engineering debugging and parameter calibration procedure must be performed to ensure that its logic matches the vessel's operational characteristics. This procedure clarifies the hotspot patch counting algorithm for the remote sensing inventory unit. The algorithm is defined as follows: First, the thermal image is binarized, with pixels above the background threshold assigned a value of 1; Second, an eight-neighbor connectivity algorithm is used to scan all pixels with a value of 1, forming one or more connected components; Third, all connected components with an area greater than a preset minimum area, such as 30 pixels, are counted to filter out small thermal noise. The total number of these connected components represents the number of observers. ; calibrate dynamic high-risk threshold Sensitivity coefficient in The calibration was conducted under moderate sea states. Maintenance personnel simulated ten instances on deck of normal human activities, such as running or carrying heavy objects, to capture the most dramatic instantaneous changes in ship attitude. Simultaneously, the alarm control unit recorded the peak values ​​of instantaneous attitude parameters reported by the ship attitude sensing unit during these ten actions, and calculated these peak values ​​relative to the time of the incident. The statistical standard deviation multiples, take the maximum value. To ensure that event signals are triggered only by genuine ship-related hazards and not by violent human activity, the sensitivity coefficient is... It is set to a safety value greater than this maximum value; for example, it is set to... The calibration procedure also includes the initial setting of safety baseline values ​​for ship attitude anomaly alarms. This setting is performed under baseline operating conditions: the ship is at its standard operating draft, balanced load, and moored in calm waters. Under these baseline conditions, the alarm host continuously collects attitude parameters from the ship's attitude sensing unit, such as heel and pitch angles, for at least 30 minutes, and calculates the statistical average value over this period. This statistical average value is not assumed to be zero but is stored as a ship-specific safety baseline value. During subsequent system operation, the statistical average value calculated by the alarm host... The ship's attitude will be compared with this safety benchmark value. Once the continuous deviation between the two exceeds the preset slow deviation threshold, an abnormal ship attitude alarm signal will be triggered.

[0043] Example 6: After the alarm system is initially installed on a specific vessel, to ensure that the logic of the remote sensing inventory unit is logically adapted to the specific geometric layout of the ship's deck, a calibration procedure for a preset minimum area needs to be performed. This procedure requires maintenance personnel to mark the monitoring boundary of the ship's deck work area in the system according to the ship's drawings and operational requirements, especially the work point farthest from the thermal imaging sensor. A tester stands at this farthest work point, and the alarm host acquires its thermal image, performs the binarization and eight-neighbor connectivity steps of the publicly available hotspot patch counting algorithm, calculates the area of ​​the connected pixel region presented by the person at that location, and records it as... To ensure the system can reliably detect personnel at the furthest point in the work area, while filtering out thermal noise in images smaller than the personnel to the maximum extent possible, a preset minimum area is set to [value missing]. In the system calibration procedure, it is also necessary to determine the slow deviation threshold for ship attitude anomaly alarms. This parameter is determined not by dynamic testing, but based on the ship's existing safety regulations. Maintenance personnel consult the ship's stability calculations or operating manual for the maximum permissible heel angle, which represents the ship's attitude safety boundary under non-instantaneous, slow-speed changes. The preset slow deviation threshold in the alarm host is set to this maximum safe heel angle value. During actual system operation, when the alarm host calculates the average attitude parameters... When the deviation from the calibrated safety benchmark value continues to exceed the preset slow deviation threshold, the system will trigger an independent ship attitude anomaly alarm signal.

[0044] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0045] Finally, it should be noted that 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 preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A shipborne fall-over positioning alarm system based on Beidou GPS remote sensing, characterized in that, The system comprises a baseline setting unit, a remote sensing inventory unit, a ship body posture sensing unit, an exit sensor and an alarm host; The remote sensing inventory unit comprises a remote sensing sensor, which is arranged in a shipboard deck operation area; The exit sensor is arranged at a preset safety exit of the shipboard deck operation area; The alarm host is in communication connection with the baseline setting unit, the remote sensing inventory unit, the ship body posture sensing unit and the exit sensor; the alarm host is used to acquire the expected personnel quantity baseline defined by the baseline setting unit; acquire the observed personnel quantity collected by the remote sensing inventory unit in real time; acquire the posture parameter representing the ship body motion state collected by the ship body posture sensing unit in real time; and generate a personnel missing state signal when it is determined that the observed personnel quantity is less than the expected personnel quantity baseline; Generate a ship high-risk event signal when it is determined that the posture parameter exceeds a high-risk threshold; and trigger a falling into water alarm signal when the personnel missing state signal and the ship high-risk event signal are determined at the same time; Moreover, the alarm host is further used to: when the personnel missing state signal is generated, judge whether a trigger signal from the exit sensor is received within a preset time window; If the trigger signal from the exit sensor is not received within the preset time window, a high-priority early warning signal is triggered; and the alarm host is further used to: based on the trigger signal of the exit sensor, maintain a deck personnel net outflow count; and verify whether the expected personnel quantity baseline is consistent with a preset logical relationship of the observed personnel quantity and the deck personnel net outflow count; If not, a baseline error alarm signal is triggered; Moreover, the logic of the alarm host for generating the personnel missing state signal further comprises: the state that the observed personnel quantity is less than the expected personnel quantity baseline lasts for more than a preset state confirmation duration; the falling into water alarm signal is a secondary alarm signal; and the high-priority early warning signal is a primary early warning signal, which is a local sound and light signal triggered at the bridge for prompting verification; Moreover, the logic of the alarm host for verifying the expected personnel quantity baseline comprises: judging whether the expected personnel quantity baseline is equal to the sum of the observed personnel quantity and the deck personnel net outflow count; and the baseline error alarm signal is a blocking alarm, which is used to lock the trigger of the falling into water alarm signal and the high-priority early warning signal before the expected personnel quantity baseline is corrected.

2. The shipborne fall-over positioning and warning system based on Beidou GPS remote sensing according to claim 1, characterized in that, The remote sensing sensor comprises a thermal imaging sensor; The logic of the alarm host for acquiring the observed personnel quantity comprises: acquiring a thermal image collected by the thermal imaging sensor; and performing a hot spot patch counting algorithm on the thermal image to generate the observed personnel quantity; the hot spot patch counting algorithm is an image processing algorithm for counting the number of pixel clusters with brightness exceeding a preset background threshold in the thermal image.

3. The shipborne fall-over positioning and warning system based on Beidou GPS remote sensing according to claim 1, characterized in that, The ship body posture sensing unit comprises a Beidou global positioning system receiver integrated with an inertial measurement unit; the posture parameter comprises at least one of a ship body roll angular velocity and a heave acceleration; and the logic of the alarm host for generating the ship high-risk event signal comprises: comparing an instantaneous value of the ship body roll angular velocity or an instantaneous value of the heave acceleration with a high-risk threshold, which is a preset value.

4. The shipborne fall-over positioning and warning system based on Beidou GPS remote sensing according to claim 1, characterized in that, The alarm host is further configured to: continuously collect the attitude parameter and calculate a statistical mean value and a statistical standard deviation of the attitude parameter within a long preset time window, and dynamically determine a high-risk threshold value; wherein the high-risk threshold value is determined by the following rules: wherein is a statistical mean value of the attitude parameter within the long preset time window, is a statistical standard deviation of the attitude parameter within the long preset time window, is a preset sensitivity coefficient.

5. The shipborne fall-over positioning and warning system based on Beidou GPS remote sensing according to claim 4, characterized in that, The alarm control panel is also used to: determine the statistical average value of attitude parameters over a long preset time window. Whether it continuously deviates from the preset safety benchmark value; if the statistical average value is determined. If the deviation from the safety benchmark value continues, an independent ship attitude anomaly alarm signal will be triggered.

6. The shipborne fall-over positioning and warning system based on Beidou GPS remote sensing according to claim 2, characterized in that, The alarm host is further configured to: analyze an image quality parameter of the thermal image before executing the hot spot patch counting algorithm; the image quality parameter comprises an image entropy or a pixel intensity standard deviation of the thermal image; when the image quality parameter is lower than a preset confidence threshold, a sensor blindness signal is generated, and triggering of the falling into water alarm signal and the high-priority early warning signal is prevented.

7. The shipborne fall-over positioning and warning system based on Beidou GPS remote sensing according to claim 1, characterized in that, The baseline setting unit comprises a manual operation interface installed on the bridge, and the manual operation interface is used for receiving an expected number of personnel baseline input by the captain or the duty officer.

8. The shipborne fall-over positioning and warning system based on Beidou GPS remote sensing according to claim 1, characterized in that, The exit sensor comprises a photoelectric opposite-acting probe or a near-field microwave radar sensor, and the exit sensor is used for generating a trigger signal when personnel pass through a preset safety exit.

Images