Bridge intelligent protection method and device based on active and passive combined mechanism

Abstract

The application discloses a bridge intelligent protection method and device based on an active and passive combined mechanism, and the method comprises the following steps: acquiring navigation environment data collected based on a front-end sensing device of a bridge area water area; acquiring bridge area ship passage data monitored based on a ship automatic identification system and a radar; preprocessing the navigation environment data and the ship passage data, and constructing a multi-source data fusion model based on the preprocessed data; performing data standardization fusion processing by using the multi-source data fusion model to obtain a multi-source real-time navigation data set; calculating a ship-bridge collision probability value based on the multi-source real-time navigation data set; comparing the ship-bridge collision probability value with a preset risk threshold and a preset accident threshold, and executing a corresponding hierarchical protection strategy according to the comparison result. The application can realize the organic combination of the active and passive anti-collision mechanisms, and effectively improve the intelligentialization and effectiveness of bridge anti-collision protection.

Description

Technical Field

[0001] This application relates to the field of water traffic safety technology, specifically to a bridge intelligent protection method and device based on a combined active and passive mechanism. Background Technology

[0002] As an important infrastructure of the transportation system, bridges play a key role in improving the comprehensive transportation network and promoting social and economic development. However, with the rapid development of waterway transportation, the flow of ships in the bridge area is constantly increasing and ships are becoming larger. The probability of ship-bridge collision accidents has increased significantly, posing a serious threat to the safety of bridge operation and navigation in the bridge area. Therefore, reliable bridge anti-ship collision technology is urgently needed to reduce the risk of collision.

[0003] Current bridge collision avoidance technologies are mainly divided into two categories: active collision avoidance and passive collision avoidance. Active collision avoidance monitors the dynamics of vessels in the bridge area through sensors and issues warnings when there is a collision risk, guiding vessels to avoid collisions. Passive collision avoidance, on the other hand, uses collision protection facilities around the bridge piers to absorb and dissipate collision energy to reduce bridge damage. Both technologies have been applied in practice. However, in existing technologies, active and passive collision avoidance methods operate independently, failing to form an effective synergy. Furthermore, the responses to bridge-ship collision risks are relatively simplistic, lacking tailored protective measures based on the degree of risk, making it difficult to achieve precise responses to collision risks. In summary, the existing active and passive protection systems for bridge collision avoidance are disconnected, and the protective responses to collision risks lack specificity, failing to achieve efficient bridge collision protection.

[0004] The preceding description is intended to provide general background information and does not necessarily constitute prior art. Summary of the Invention

[0005] This application provides a bridge intelligent protection method and device based on a combined active and passive collision avoidance mechanism, which can realize the organic combination of active and passive collision avoidance mechanisms and effectively improve the intelligence and effectiveness of bridge collision avoidance protection.

[0006] Based on this, embodiments of this application provide a bridge intelligent protection method based on a combined active and passive mechanism, including:

[0007] Acquire navigation environment data based on front-end sensing devices in the bridge area waters;

[0008] Obtain vessel traffic data in the bridge area based on Automatic Identification System (AIS) and radar monitoring;

[0009] The navigation environment data and the ship passage data are preprocessed, and a multi-source data fusion model is constructed based on the preprocessed data. The multi-source data fusion model is then used to perform data standardization and fusion processing to obtain a multi-source real-time navigation dataset.

[0010] Based on the multi-source real-time navigation dataset, the collision probability value between the ship and the bridge is calculated.

[0011] The collision probability value of the bridge is compared with a preset risk threshold and a preset accident threshold in a graded manner, and a corresponding graded protection strategy is executed according to the comparison result. The graded protection strategy includes: when the collision probability value of the bridge is less than the preset risk threshold, a caution driving reminder is issued to the vessel; when the collision probability value of the bridge is greater than or equal to the preset risk threshold and less than the preset accident threshold, an active collision avoidance warning is activated and auxiliary driving decision information is provided to the vessel; when the collision probability value of the bridge is greater than or equal to the preset accident threshold, a servo motor is activated to release the passive collision airbag and a tracking camera is activated to track and record the vessel's video.

[0012] Furthermore, in some embodiments of this application, the front-end sensing device includes at least one of a high-definition camera, a tracking camera, an optoelectronic monitoring device, a hydrological and meteorological sensor, a smart navigation beacon, and an audio-visual early warning device; the navigation environment data includes at least one of the water level, flow velocity, wind speed, wind direction, and visibility information of the bridge area; the vessel passage data includes at least one of the position, speed, course, length, width, and vessel type information of vessels passing under the bridge.

[0013] Furthermore, in some embodiments of this application, the preprocessing of the navigation environment data and the vessel passage data includes:

[0014] Using the data sampling time as an index, iterate through and delete duplicate sampled data to obtain the first intermediate data;

[0015] Based on the physical change characteristics and preset value range of each element data in the navigation environment data and the ship passage data, a noise reduction threshold is determined, and noise data in the first intermediate data is removed based on the noise reduction threshold to obtain the second intermediate data;

[0016] The second intermediate data is subjected to dimensionality reduction processing to obtain the third intermediate data;

[0017] The navigation environment data and vessel traffic data of the bridge area waters were obtained from secondary sampling, and the missing data in the third intermediate data were interpolated to fill in the gaps, so as to obtain the preprocessed standardized data.

[0018] Furthermore, in some embodiments of this application, the step of determining a denoising threshold based on the physical change characteristics and preset value range of each element data in the navigation environment data and the ship passage data, and removing noise data from the first intermediate data based on the denoising threshold to obtain the second intermediate data, includes:

[0019] The physical change characteristics of each element in the navigation environment data and the ship passage data are obtained. The physical change characteristics include the rate of change of water level, the range of change of current velocity, the range of change of wind speed, the range of change of ship speed, and the range of change of ship course.

[0020] Based on the preset physical value range of each element data, determine the noise reduction threshold corresponding to each element data;

[0021] Data exceeding the denoising threshold in the first intermediate data is identified as noise data, and the identified noise data is removed to obtain the second intermediate data.

[0022] Furthermore, in some embodiments of this application, the step of constructing a multi-source data fusion model based on preprocessed data, and using the multi-source data fusion model to perform data standardization and fusion processing to obtain a multi-source real-time navigation dataset, includes:

[0023] Based on the preprocessed standardized data, a multi-source data fusion model is constructed;

[0024] Using the multi-source data fusion model, the navigation environment data and ship passage data in the standardized data are spatiotemporally aligned to obtain a spatiotemporally aligned dataset;

[0025] Based on a preset association threshold, the trajectory association and matching of ship targets from different data sources in the spatiotemporally aligned dataset are performed, and a unified target identifier is assigned to the same ship target that is successfully associated, thus obtaining the associated target set.

[0026] The covariance intersection fusion algorithm is used to perform weighted fusion of multi-source data in the associated target set to generate the multi-source real-time navigation dataset.

[0027] Furthermore, in some embodiments of this application, the step of calculating the bridge collision probability value based on the multi-source real-time navigation dataset includes:

[0028] The location, speed, heading, and scale information of vessels crossing the bridge, as well as the hydrological and meteorological data of the bridge area, are extracted from the multi-source real-time navigation dataset to obtain a set of basic parameters.

[0029] Based on the relative distance between the ship and the bridge, the velocity component of the ship's speed in the direction of the bridge, the ship's yaw rate, hydro-meteorological conditions, and ship maneuvering performance parameters in the aforementioned basic parameter set, a ship-bridge collision probability calculation model is constructed.

[0030] The collision probability calculation model is used to perform real-time calculations on the multi-source real-time navigation dataset to obtain the collision probability value.

[0031] Furthermore, in some embodiments of this application, the construction of a ship-bridge collision probability calculation model based on the relative distance between the ship and the bridge, the velocity component of the ship's speed in the bridge direction, the ship's yaw rate, hydrological and meteorological conditions, and ship maneuvering performance parameters in the basic parameter set includes:

[0032] A geometric encounter factor is constructed based on the relative distance between the ship and the bridge, the nearest meeting distance, and the time to reach the nearest point.

[0033] Maneuverability factors are constructed based on the ship's turning index and following index.

[0034] Environmental impact factors are constructed based on wind speed, current velocity, and visibility.

[0035] The geometrical average of the geometric encounter factor, the manipulative factor, and the environmental impact factor is used to obtain a comprehensive risk factor.

[0036] The bridge collision probability calculation model is constructed based on the comprehensive risk factors.

[0037] Furthermore, in some embodiments of this application, activating the active collision avoidance warning and providing assisted driving decision information to the vessel includes:

[0038] Active collision avoidance warning signals are sent to vessels passing under the bridge using acoustic, optical, and electronic warning equipment.

[0039] Based on the multi-source real-time navigation dataset, a comprehensive analysis of the ship's current navigation status and the navigation environment in the bridge area is conducted to generate an assisted driving decision-making scheme that includes recommended speed, recommended course, and suggested control measures.

[0040] The assisted driving decision-making scheme is simultaneously pushed to the corresponding ship and bridge management and control center in the form of Automatic Identification System (AIS) information, high-frequency voice, and visual display.

[0041] Furthermore, in some embodiments of this application, the step of comprehensively analyzing the current navigation status of the vessel and the navigation environment in the bridge area based on the multi-source real-time navigation dataset to generate an assisted driving decision-making scheme including recommended speed, recommended course, and suggested control measures includes:

[0042] Extract the ship's current position, current speed, current course, and target pier position from the multi-source real-time navigation dataset;

[0043] Based on ship maneuvering performance parameters and navigation environment data in the bridge area, the recommended speed and recommended course for the ship to safely pass through the bridge pier are calculated.

[0044] Based on the difference between the recommended speed and the current speed, and the difference between the recommended heading and the current heading, an assisted driving decision-making scheme is generated, which includes speed adjustment suggestions and heading adjustment suggestions.

[0045] Furthermore, in some embodiments of this application, the step of activating the servo motor to release the passive anti-collision airbag and activating the tracking camera to record ship video includes:

[0046] A release command is sent to the anti-collision control unit, which inflates and deploys the anti-collision airbag installed on the ship-facing side of the bridge pier via a servo motor.

[0047] Simultaneously activate the tracking cameras in the bridge area to continuously track and record the navigation process of vessels passing under the bridge, and upload the video data with timestamps and geographic coordinates to the data storage center in real time.

[0048] Furthermore, in some embodiments of this application, sending a release command to the anti-collision control unit, wherein the anti-collision control unit drives the anti-collision airbag installed on the ship-facing side of the bridge pier to inflate and deploy via a servo motor, includes:

[0049] A release command is sent to the anti-collision control unit, which, upon receiving the release command, starts the servo motor within a preset response time.

[0050] The servo motor drives the roller to rotate, so that the anti-collision airbag can be inflated and deployed within a preset deployment time to form a buffer protection layer on the ship-facing side of the bridge pier.

[0051] Furthermore, in some embodiments of this application, the method further includes:

[0052] When a bridge collision is detected, the time of the collision event is automatically marked.

[0053] Extract video data and multi-source real-time navigation datasets within a preset time period before and after the collision event to generate an evidence package;

[0054] Based on the evidence package, an accident analysis report is generated, and the evidence package and the accident analysis report are pushed to a preset management platform.

[0055] Furthermore, in some embodiments of this application, the step of extracting video data and multi-source real-time navigation datasets within a preset time period before and after the collision event to generate an evidence package includes:

[0056] Extract video clips from the tracking camera within a first preset time period before the collision event occurs to a second preset time period after the collision event occurs;

[0057] Extract a multi-source real-time navigation dataset that is time-synchronized with the video clip. The multi-source real-time navigation dataset includes ship trajectory data, bridge collision probability value change curves, and navigation environment data.

[0058] The video clips are encapsulated with the multi-source real-time navigation dataset to generate an evidence package containing both video and data evidence.

[0059] Furthermore, in some embodiments of this application, the method further includes:

[0060] The navigation environment data and the vessel passage data are connected to the edge computing gateway, and the navigation environment data and the vessel passage data are time-synchronized in the edge computing gateway to make the time deviation between the navigation environment data and the vessel passage data less than a preset deviation threshold.

[0061] The time-synchronized navigation environment data and the ship passage data are stored in an edge cache. The edge cache adopts a ring cache structure to store the original data within a preset time period.

[0062] When a network interruption is detected, the collected navigation environment data and ship passage data are temporarily stored in the edge cache, and the temporarily stored data will be automatically uploaded after the network is restored.

[0063] Accordingly, embodiments of this application provide a bridge intelligent protection device based on a combined active and passive mechanism, comprising:

[0064] The navigation environment data acquisition module is used to acquire navigation environment data collected by the front-end sensing equipment in the bridge area waters.

[0065] The vessel data acquisition module is used to acquire vessel traffic data in the bridge area based on the Automatic Identification System (AIS) and radar monitoring.

[0066] The data fusion module is used to preprocess the navigation environment data and the ship passage data, and to build a multi-source data fusion model based on the preprocessed data. The multi-source data fusion model is then used to perform data standardization and fusion processing to obtain a multi-source real-time navigation dataset.

[0067] The probability calculation module is used to calculate the collision probability value of the ship and bridge based on the multi-source real-time navigation dataset;

[0068] The graded protection module is used to compare the bridge collision probability value with preset risk thresholds and preset accident thresholds in a graded manner, and execute corresponding graded protection strategies based on the comparison results. The graded protection strategies include: issuing a cautionary driving warning to the vessel when the bridge collision probability value is less than the preset risk threshold; activating an active collision avoidance warning and providing assisted driving decision information to the vessel when the bridge collision probability value is greater than or equal to the preset risk threshold and less than the preset accident threshold; and activating a servo motor to release passive collision airbags and activating a tracking camera to record vessel video when the bridge collision probability value is greater than or equal to the preset accident threshold.

[0069] This application provides a bridge intelligent protection method and device based on a combined active and passive mechanism. First, it acquires navigation environment data and vessel traffic data in the bridge area, preprocesses and fuses multi-source data to obtain corresponding datasets, and calculates the collision probability value between the vessel and the bridge based on this dataset. Then, it compares the collision probability value with preset risk and accident thresholds and executes appropriate protection strategies, providing a quantitative and objective basis for determining the collision risk, thus overcoming the lack of accurate data support in traditional protection methods. When the collision probability reaches the preset risk threshold, the system proactively intervenes, sending real-time warning information to the target vessel and the bridge area management department, and collaboratively guiding the vessel to adjust its speed and course in a timely manner to reduce the collision risk. To nip potential collisions in the bud; when the probability of a collision continues to rise and exceeds the preset accident threshold, the system automatically activates passive protection through active warning, implementing "in-process protection" against ship-bridge collisions by releasing inflatable airbags, minimizing the impact force and damage to the bridge structure, forming a "prevention and control" safety guarantee mechanism. The system organically connects the early warning and decision support of active collision avoidance with the physical protection of passive collision avoidance according to the collision risk level, so that the active and passive collision avoidance mechanisms no longer operate independently, but form a coordinated protection system. From risk warning and active avoidance to passive protection, a coherent protection response is formed, ultimately effectively improving the intelligence and effectiveness of bridge collision protection, and achieving efficient prevention and control of ship-bridge collision risks in the bridge area. Attached Figure Description

[0070] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0071] Figure 1 This is an application environment diagram of the bridge intelligent protection method based on the active-passive joint mechanism provided in the embodiments of this application;

[0072] Figure 2This is a flowchart illustrating the intelligent bridge protection method based on a combined active and passive mechanism provided in an embodiment of this application.

[0073] Figure 3 This is a schematic diagram of the integrated information perception and multi-source data preprocessing architecture for "bridge-ship-shore" provided in the embodiments of this application;

[0074] Figure 4 This is a flowchart illustrating the bridge protection method based on a combined active and passive mechanism provided in an embodiment of this application.

[0075] Figure 5 This is a flowchart illustrating the method for calculating the collision probability of a ship-bridge provided in an embodiment of this application;

[0076] Figure 6 This is another schematic diagram of the bridge protection method based on a combined active and passive mechanism provided in the embodiments of this application;

[0077] Figure 7 This is a schematic diagram of the structure of a bridge protection device based on a combined active and passive mechanism provided in an embodiment of this application;

[0078] Figure 8 This is a schematic diagram of the structure of the intelligent bridge protection device based on the active-passive joint mechanism provided in the embodiments of this application;

[0079] Figure 9 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0080] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of systems and methods consistent with those detailed in the appended claims or with some aspects of this application.

[0081] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover descriptions such as non-exclusive inclusion, so that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, components, features, and elements with the same names in different embodiments of this application may have the same meaning or different meanings, the specific meaning of which must be determined by its interpretation in that specific embodiment or further in conjunction with the context of that specific embodiment.

[0082] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0083] In the following description, the use of suffixes such as "module," "part," or "unit" to denote elements is solely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, "module," "part," or "unit" may be used interchangeably.

[0084] To address the aforementioned technical problems and overcome the shortcomings of existing technologies, this application provides a bridge intelligent protection method and device based on a combined active and passive collision avoidance mechanism, which can organically combine active and passive collision avoidance mechanisms and effectively improve the intelligence and effectiveness of bridge collision avoidance protection.

[0085] Figure 1 This is an application environment diagram of a bridge intelligent protection method based on a combined active and passive mechanism in one embodiment. (Refer to...) Figure 1The bridge intelligent protection method based on the active-passive joint mechanism should be based on a bridge intelligent protection system based on the active-passive joint mechanism. This bridge intelligent protection system based on the active-passive joint mechanism includes a terminal 110 and a server 120. The terminal 110 and the server 120 are connected via a network. The terminal 110 can be a desktop terminal or a mobile terminal; the mobile terminal can be at least one of a mobile phone, tablet computer, or laptop computer. The server 120 can be implemented using a standalone server or a server cluster consisting of multiple servers. Server 120 is configured to execute the aforementioned bridge intelligent protection method based on a combined active and passive mechanism, including: acquiring navigation environment data collected by front-end sensing devices in the bridge area; acquiring vessel traffic data in the bridge area monitored by the Automatic Identification System (AIS) and radar; preprocessing the navigation environment data and vessel traffic data, constructing a multi-source data fusion model based on the preprocessed data, and performing data standardization and fusion processing using the multi-source data fusion model to obtain a multi-source real-time navigation dataset; calculating the bridge-ship collision probability value based on the multi-source real-time navigation dataset; comparing the bridge-ship collision probability value with preset risk thresholds and preset accident thresholds in a graded manner, and executing corresponding graded protection strategies according to the comparison results; wherein, the graded protection strategies include: issuing a caution driving reminder to the vessel when the bridge-ship collision probability value is less than the preset risk threshold; activating an active collision avoidance warning and providing assisted driving decision information to the vessel when the bridge-ship collision probability value is greater than or equal to the preset risk threshold and less than the preset accident threshold; and activating a servo motor to release passive collision airbags and activating a tracking camera to record vessel video when the bridge-ship collision probability value is greater than or equal to the preset accident threshold.

[0086] Please see Figure 2 , Figure 2 This is a flowchart illustrating a bridge intelligent protection method based on a combined active and passive mechanism according to an embodiment of this application. This embodiment primarily uses the application of this bridge intelligent protection method based on a combined active and passive mechanism to computer equipment as an example for illustration. Specifically, the bridge intelligent protection method based on a combined active and passive mechanism provided in this embodiment may include the following steps:

[0087] S1. Obtain navigation environment data based on the front-end sensing equipment in the bridge area;

[0088] Specifically, for step S1, dedicated front-end sensing devices are deployed within the bridge area. These devices continuously and in real-time collect navigation environment information related to the bridge area. The collected navigation environment data comprehensively reflects the natural environmental state of the bridge area, providing fundamental environmental data for subsequent assessment of ship navigation risks. The front-end sensing devices can collect core environmental indicators such as water level, flow velocity, wind speed, wind direction, and visibility in the bridge area. The collection process can be set with a fixed collection frequency according to the navigation needs of the bridge area, ensuring the real-time nature and continuity of the data. For example, in the bridge area of ​​a cross-river bridge, the deployed front-end sensing devices collect water flow velocity and wind speed data every 1 second and water level data every 5 minutes, continuously acquiring navigation environment data for the bridge area and completely recording the dynamic changes in the bridge area environment.

[0089] S2. Obtain vessel traffic data in the bridge area based on Automatic Identification System (AIS) and radar monitoring;

[0090] Specifically, for step S2, comprehensive dynamic monitoring of vessels entering, leaving, and navigating within the bridge area is conducted using both Automatic Identification System (AIS) and radar. This collects various types of traffic data reflecting vessel navigation status, enabling precise capture of vessel dynamics within the bridge area and providing core vessel-level data for subsequent collision risk assessment. The AIS acquires basic traffic information such as vessel position, speed, heading, and type. Radar provides supplementary monitoring of real-time vessel navigation dynamics, correcting and refining traffic data. The combined use of these two monitoring methods ensures comprehensive and accurate acquisition of vessel traffic data within the bridge area. For example, the AIS deployed in the bridge area receives the MMSI (Maritime Mobile Service Identity) code, real-time speed, and heading information of passing cargo ships. Simultaneously, the bridge area radar monitors the actual position of the cargo ship in real time. Combining these two types of data yields complete bridge area traffic data for the cargo ship.

[0091] S3. Preprocess the navigation environment data and ship passage data, and build a multi-source data fusion model based on the preprocessed data. Use the multi-source data fusion model to perform data standardization and fusion processing to obtain a multi-source real-time navigation dataset.

[0092] Specifically, for step S3, the collected navigation environment data and acquired vessel traffic data are first processed to remove invalid and redundant parts, correct data deviations, and improve the quality and usability of both types of data. Then, based on the processed standardized data, a multi-source data fusion model adapted to the characteristics of bridge area navigation data is built. Heterogeneous navigation environment data and vessel traffic data are input into this model, and data fusion is performed according to unified standards and rules, ultimately forming a multi-source real-time navigation dataset that comprehensively reflects the navigation environment and vessel navigation status in the bridge area. This dataset integrates both environmental and vessel data, breaking down information barriers between the two types of data and providing a unified and complete data foundation for subsequent collision probability calculations. For example, duplicate visibility sampling data in the collected navigation environment data and abnormal speed data in the vessel traffic data are processed, while a small number of missing values ​​in both types of data are supplemented. The processed environmental data such as water level and wind speed, and traffic data such as vessel position and heading, are then input into the multi-source data fusion model. Through the model's standardized fusion processing, a multi-source real-time navigation dataset that can be directly used for subsequent calculations is obtained.

[0093] S4. Based on the multi-source real-time navigation dataset, the collision probability value of the ship and bridge is calculated;

[0094] Specifically, for step S4, based on the obtained multi-source real-time navigation dataset, and combined with basic information such as the fixed location and structural characteristics of the bridges in the bridge area, the risk level of collision between ships and bridges is quantitatively calculated using corresponding calculation methods, ultimately yielding a ship-bridge collision probability value. This probability value is a dimensionless numerical value that can intuitively and accurately reflect the likelihood of a specific ship colliding with a bridge in the bridge area at a given moment. It is a core quantitative indicator for subsequent risk classification and implementation of protection strategies. For example, based on the fused multi-source real-time navigation dataset, and combined with information such as the specific coordinates of a bridge pier and the width of the navigation channel of a cross-river bridge, the probability value of a passing cruise ship colliding with a bridge pier at a given moment can be calculated. This value can be directly used to determine the navigation risk level of the cruise ship.

[0095] S5. Compare the bridge collision probability value with preset risk thresholds and preset accident thresholds in a tiered manner, and execute the corresponding tiered protection strategy based on the comparison results; wherein, the tiered protection strategy includes: when the bridge collision probability value is less than the preset risk threshold, issuing a caution driving reminder to the vessel; when the bridge collision probability value is greater than or equal to the preset risk threshold and less than the preset accident threshold, activating active collision avoidance warning and providing assisted driving decision information to the vessel; when the bridge collision probability value is greater than or equal to the preset accident threshold, activating the servo motor to release the passive collision airbag and activating the tracking camera to track and record the vessel's video;

[0096] Specifically, for step S5, two thresholds for risk classification are pre-set based on the navigation characteristics of the bridge area and the bridge protection requirements: a risk threshold and an accident threshold. The calculated ship-bridge collision probability value is compared with the two preset thresholds one by one. Based on the comparison results, the collision risk level of the vessel is classified, and a pre-set, appropriate protection strategy is implemented for different risk levels. The specific protection strategies are divided into three categories: when the ship-bridge collision probability value is less than the preset risk threshold, only a caution driving warning is issued to the corresponding vessel, and no high-intensity protection measures are required; when the ship-bridge collision probability value is greater than or equal to the preset risk threshold but less than the preset accident threshold, an active collision avoidance warning is immediately activated, and auxiliary driving decision information is provided to the vessel to guide it to actively avoid collision risks; when the ship-bridge collision probability value is greater than or equal to the preset accident threshold, the servo motor is quickly activated to release the passive anti-collision airbags, and the tracking camera is activated to continuously record video of the vessel, reducing collision damage through physical protection and preserving relevant image data. For example, if a risk threshold of 0.15 and an accident threshold of 0.85 are set for a certain coastal bridge area, when the calculated probability of a ship colliding with the bridge is 0.08, a short AIS (Automatic Identification System) message reminding the ship to drive cautiously is sent to the ship; when the probability is 0.6, an active collision avoidance warning is issued through the bridge area's audio-visual equipment, and recommended course, speed, and other auxiliary driving decision information are pushed to the ship; when the probability is 0.9, the servo motors on the bridge pier facing the ship are immediately activated to release the collision avoidance airbags, and the tracking camera is activated to record the ship's video in real time.

[0097] This embodiment collects and integrates navigation environment and vessel traffic data in the bridge area, quantifies and calculates the probability of ship-bridge collision, and combines a dual-threshold tiered protection strategy to achieve an organic combination of active and passive collision avoidance mechanisms. This makes bridge protection more targeted and intelligent, and effectively improves the prevention and control of ship-bridge collision risks in the bridge area.

[0098] Furthermore, in some embodiments, the front-end sensing device includes at least one of a high-definition camera, a tracking camera, an optoelectronic monitoring device, a hydro-meteorological sensor, a smart navigation beacon, and an audible-optical-electric early warning device; the navigation environment data includes at least one of the water level, flow velocity, wind speed, wind direction, and visibility information of the bridge area; and the vessel passage data includes at least one of the position, speed, course, length, width, and vessel type information of vessels passing under the bridge.

[0099] Furthermore, in some embodiments, step S3, "preprocessing the navigation environment data and vessel passage data," may specifically include:

[0100] S31. Using the data sampling time as the index, iterate through and delete the data that was sampled repeatedly to obtain the first intermediate data;

[0101] Specifically, using the sampling time of navigation environment data and vessel traffic data as a unique index, all element data in both types of data are traversed sequentially according to time. Duplicate data collected at the same sampling time is checked and deleted during the traversal, retaining only a single valid data entry. After data deduplication, the first intermediate data is obtained. This operation, using time as an index, makes the data deduplication process more systematic, effectively avoiding data redundancy caused by duplicate sampling and reducing the amount of unnecessary work in subsequent data processing. For example, during bridge area data collection, if two water level data and three vessel speed data are collected at the same timestamp, after traversing the data using the sampling time as an index, duplicate water level and speed data are deleted, retaining only one valid data entry for each. The remaining element data are processed according to the same rules to obtain the first intermediate data.

[0102] S32. Based on the physical change characteristics and preset value range of each element data in the navigation environment data and ship passage data, determine the noise threshold, and remove the noise data in the first intermediate data based on the noise threshold to obtain the second intermediate data;

[0103] Specifically, the process begins by combining the physical variation characteristics of each element in the navigation environment data and ship traffic data, along with the pre-defined reasonable physical value ranges for each element, to determine a corresponding denoising threshold for each data element. Then, based on these denoising thresholds, the first intermediate data is comprehensively screened. Abnormal data exceeding the threshold range or not conforming to actual physical laws are identified as noise data and removed. After denoising, the second intermediate data is obtained. This operation sets thresholds based on the physical characteristics of the data, accurately identifying and removing invalid noise data, ensuring the authenticity and reasonableness of the data. For example, the preset physical value range for flow velocity is 0-6 m / s, and the reasonable variation range for ship speed is 0-30 knots. Based on this, denoising thresholds are set for flow velocity and ship speed respectively. If flow velocity data of 8 m / s and ship speed data of 35 knots appear in the first intermediate data, they are identified as noise data and removed. After all element data has been screened and removed, the second intermediate data is obtained.

[0104] S33. Perform dimensionality reduction on the second intermediate data to obtain the third intermediate data;

[0105] Specifically, dimensionality reduction is performed on the denoised second intermediate data. Using specialized dimensionality reduction methods, the multi-source heterogeneous data in the second intermediate data is integrated and simplified. While retaining the core features and key effective information, the dimensionality complexity of the data is reduced, making the data structure simpler and the relationships between data points stronger. This process yields the third intermediate data. This operation simplifies the computational workload of subsequent data processing, improves data processing efficiency, and avoids information redundancy or processing errors caused by excessively high data dimensionality. For example, the second intermediate data contains 12-dimensional navigation environment data vectors such as water level, current velocity, and wind speed, and 8-dimensional ship traffic data vectors such as ship position, heading, and dimensions. After dimensionality reduction, the navigation environment data vectors are reduced to 6 dimensions, and the ship traffic data vectors are reduced to 4 dimensions, while retaining more than 95% of the core effective information in each vector, ultimately yielding the third intermediate data.

[0106] S34. Obtain navigation environment data and vessel passage data from secondary sampling of the bridge area waters, and interpolate and fill in the missing data in the third intermediate data to obtain preprocessed standardized data;

[0107] Specifically, navigation environment data and vessel traffic data obtained from secondary sampling in the bridge area during the corresponding time period are first acquired. This secondary sampling data is used as a reference. To address data gaps and breaks in the third intermediate data (specific timestamps), interpolation is used to fill in the missing values, resulting in complete, formatted, and clearly defined preprocessed standardized data. This operation effectively solves the problem of localized data loss caused by equipment and environmental factors during data collection, ensuring data integrity. For example, if visibility and vessel width data are missing for a certain time period in the third intermediate data, continuous visibility and vessel width data from secondary sampling in the bridge area during that time period are retrieved, and interpolation is used to accurately fill in the missing values. After filling in all missing items in the dataset, standardized data is obtained.

[0108] This embodiment effectively solves the problems of redundancy, distortion, complexity, and missing data in the original navigation environment and ship passage data through a progressive data preprocessing approach of deduplication, noise reduction, dimensionality reduction, and interpolation. This results in complete and standardized data, improving the effectiveness and integrity of the data and providing high-quality basic data support for subsequent multi-source data fusion and collision probability calculation.

[0109] Furthermore, in some embodiments, step S32, "determining a denoising threshold based on the physical change characteristics and preset value range of each element data in the navigation environment data and ship passage data, and removing noise data from the first intermediate data based on the denoising threshold to obtain the second intermediate data," may specifically include:

[0110] S321. Obtain the physical change characteristics of each element in the navigation environment data and ship passage data. The physical change characteristics include the rate of change of water level, the range of change of current velocity, the range of change of wind speed, the range of change of ship speed, and the range of change of ship course.

[0111] Specifically, the navigation environment data collected from the bridge area and the vessel traffic data monitored are decomposed into elements. The physical change characteristics corresponding to each independent element in both types of data are extracted one by one. These characteristics are inherent to each data element and conform to objective physical laws, serving as the basis for subsequent judgments on whether the data is abnormal. The physical change characteristics of the navigation environment data include the rate of change of water level, the range of change of current velocity, and the range of change of wind speed. The physical change characteristics of the vessel traffic data include the range of change of vessel speed and the range of change of vessel heading. For example, in a certain inland river bridge area, the daily rate of change of water level, influenced by hydrological patterns, does not exceed 0.5 m / hour, and the short-term change range of vessel heading within the bridge area's navigation opening does not exceed 30°. These are the physical change characteristics of the corresponding data elements. In coastal bridge areas, the short-term change range of wind speed is mostly within 0-25 m / s, which is also the physical change characteristic corresponding to the wind speed element.

[0112] S322. Determine the noise reduction threshold for each element data based on the preset physical value range of each element data;

[0113] Specifically, based on the actual hydrological and meteorological conditions, navigation management rules, and ship navigation characteristics of the bridge area, a preset physical value range is set for each element of the navigation environment data and ship passage data, which conforms to the actual situation of the bridge area. This range is the reasonable value interval of each data element under normal navigation scenarios in the bridge area. Based on this preset physical value range, a corresponding denoising threshold is determined for each data element. The denoising threshold matches the reasonable physical value range of each element, becoming a quantitative standard for determining whether the data of that element is abnormal noise data. For example, for the flow velocity element in a certain bridge area, the preset physical value range is set to 0-6 m / s based on the water flow characteristics, and the noise reduction threshold for the flow velocity is determined to be 6 m / s; for the ship speed element in the bridge area, the preset physical value range is set to 0-15 km based on the navigation speed limit rules in the bridge area, and the noise reduction threshold for the ship speed is determined to be 15 km; for the visibility element, the preset value range is set to 0.1-10 km based on the visibility requirements for navigation in the bridge area, and the corresponding noise reduction thresholds are 0.1 km and 10 km.

[0114] S323. Identify the data in the first intermediate data that exceeds the denoising threshold as noise data, and remove the identified noise data to obtain the second intermediate data.

[0115] Specifically, using the denoising threshold corresponding to each data element as the screening standard, the first intermediate data after deduplication is checked against all elements and item by item. Values ​​in the first intermediate data that exceed the corresponding denoising threshold are accurately identified as noise data that do not conform to physical laws and the actual situation of the bridge area. All identified noise data are uniformly removed, and only valid data where each element is within the denoising threshold range are retained. The retained valid data are then integrated to obtain the second intermediate data. For example, when screening the first intermediate data at a certain moment, it is found that the flow velocity data is 7 m / s, exceeding the denoising threshold of 6 m / s, and the ship speed data is 18 knots, exceeding the denoising threshold of 15 knots. These two abnormal data are identified as noise data and removed. Other valid data at that moment, such as wind speed, ship heading, and water level, which are within the threshold range, are retained. The valid data after screening and removal at all moments are integrated to finally obtain the second intermediate data.

[0116] This embodiment sets a customized denoising threshold based on the physical change characteristics of data elements and a preset value range. It accurately identifies and removes noisy data, abandons the crude uniform threshold denoising method, improves the targeting and accuracy of noise data identification, effectively avoids the accidental deletion of valid data or the retention of noisy data, and makes the processed data more consistent with the actual situation of the bridge area, reducing the deviation of subsequent data processing and calculation.

[0117] Furthermore, in some embodiments, step S3, "constructing a multi-source data fusion model based on the preprocessed data, and using the multi-source data fusion model to perform data standardization and fusion processing to obtain a multi-source real-time navigation dataset," may specifically include:

[0118] S35. Construct a multi-source data fusion model based on the preprocessed standardized data;

[0119] Specifically, based on pre-processed standardized navigation environment data and standardized vessel traffic data, and considering the spatial and geographical characteristics of the bridge area, the operational characteristics of navigable vessels, and the heterogeneous attributes of multi-source data, a multi-source data fusion model adapted to the navigation data processing needs of the bridge area is constructed. This model pre-defines the underlying logic and execution rules for data fusion, enabling it to adapt to the fusion processing requirements of the two types of heterogeneous data. It provides dedicated model support for subsequent standardized data fusion work, ensuring the standardization and adaptability of the data fusion process. For example, based on pre-processed standardized navigation environment data such as water level, wind speed, and visibility in the bridge area, and standardized vessel traffic data such as vessel position, speed, and dimensions, and combined with the actual characteristics of the bridge area, such as the width of the navigation aperture, vessel traffic density, and data collection frequency, a targeted multi-source data fusion model is constructed. The model pre-defines relevant rules for data integration and feature matching, enabling effective processing of heterogeneous data.

[0120] S36. Using a multi-source data fusion model, the navigation environment data and ship passage data in the standardized data are spatiotemporally aligned to obtain a spatiotemporally aligned dataset;

[0121] Specifically, standardized navigation environment data and standardized vessel traffic data are input into the completed multi-source data fusion model. The model then calibrates and matches the two types of data according to a unified standard, that is, unifying heterogeneous data obtained by different collection and monitoring methods to the same time measurement standard and spatial coordinate system, eliminating deviations in the time dimension and coordinate differences in the spatial dimension, so that navigation environment data and vessel traffic data under the same spatiotemporal dimension form a precise correspondence. After completing the spatiotemporal alignment processing of all data, the spatiotemporally aligned dataset is obtained. For example, by using a multi-source data fusion model, the navigation environment data and ship passage data of the bridge area are unified to the UTC (Coordinated Universal Time) millisecond-level time base and the WGS84 (World Geodetic System 1984) spatial coordinate system. The flow velocity and wind speed data at a certain UTC millisecond-level time and a certain geographical coordinate in the bridge area are accurately matched with the speed and heading data of ships near that coordinate in the same time and space. This eliminates the slight time deviation and spatial coordinate deviation that originally existed between the two types of data. After completing the spatiotemporal alignment of all data, the corresponding dataset is obtained.

[0122] S37. Based on a preset association threshold, perform track association matching on ship targets from different data sources in the spatiotemporally aligned dataset, and assign a unified target identifier to the same ship target that is successfully associated, thereby obtaining the associated target set;

[0123] Specifically, firstly, considering the navigation characteristics of the bridge area and the accuracy features of different monitoring equipment, a matching standard, or association threshold, is set for associating ship target tracks. This threshold serves as a quantitative basis for determining whether tracks from different data sources belong to the same ship. Then, based on this preset association threshold, ship track data from different monitoring data sources in the spatiotemporally aligned dataset are compared and matched one by one. Multiple track data points that are determined to point to the same ship target are associated and integrated. At the same time, each successfully associated ship target is assigned a unique unified target identifier, thereby achieving unique management of ship targets in the bridge area. After completing the track association matching and identifier assignment for all ship targets, the associated target set is obtained. For example, for a certain bridge area, preset ship track association thresholds are set: position deviation ≤ 30m, speed deviation ≤ 1kn, and ship size deviation ≤ 20%. Based on these thresholds, after spatiotemporal alignment, the data is concentrated. The track data of a cargo ship monitored by the Automatic Identification System (AIS) and the track data of a suspected cargo ship monitored by radar are compared. The position deviation is 22m, the speed deviation is 0.6kn, and the length deviation is 16%, all of which meet the preset association thresholds. They are determined to be the same ship target and are assigned a unique unified target identifier. After all ship targets in the bridge area have completed track association matching and identifier assignment, the associated target set is obtained.

[0124] S38. The covariance intersection fusion algorithm is used to perform weighted fusion of multi-source data in the associated target set to generate a multi-source real-time navigation dataset.

[0125] Specifically, for the correlated target set, the multi-source heterogeneous data corresponding to the same ship target and the navigation environment data of the ship target in the corresponding time and space are processed using the covariance intersection fusion algorithm. This algorithm assigns corresponding weights to the information from various data sources based on their monitoring accuracy and data reliability, and then integrates the multi-source data through weighted calculation, weakening the monitoring error of a single data source and enhancing the accuracy and reliability of the data. After weighted fusion of all ship targets and their corresponding time and space navigation environment data, a multi-source real-time navigation dataset is formed, which includes the unique identifier of the ship target, the ship's full-dimensional operational data, and the corresponding time and space navigation environment data. For example, when fusing the speed data of a certain ship target in the correlated target set, different fusion weights are assigned to the speed monitoring accuracy of the Automatic Identification System (AIS) and radar data sources. The covariance intersection fusion algorithm is used to calculate more accurate ship speed data. At the same time, the ship's position, heading, and corresponding time and space water level, wind speed, and other data are subjected to the same weighted fusion process. All the fused accurate data are integrated and sorted to finally generate a multi-source real-time navigation dataset.

[0126] This embodiment builds a multi-source data fusion model to achieve spatiotemporal alignment of heterogeneous data and accurate correlation and matching of ship tracks. Then, it uses a covariance intersection fusion algorithm for weighted fusion to eliminate spatiotemporal bias and single data source error, integrating two independent types of data into a precise and unified multi-source real-time navigation dataset, providing scientific and unified core data support for collision probability calculation.

[0127] Furthermore, in some embodiments, S4 "calculating the bridge collision probability value based on multi-source real-time navigation datasets" may specifically include:

[0128] S41. Extract the position, speed, heading, and scale information of vessels crossing the bridge, as well as the hydrological and meteorological data of the bridge area, from the multi-source real-time navigation dataset to obtain the basic parameter set;

[0129] Specifically, from the integrated multi-source real-time navigation dataset, based on the core data requirements for determining ship-bridge collision risk, the location, speed, heading, and size information of vessels crossing the bridge are precisely screened and extracted. Simultaneously, hydrological and meteorological data of the bridge area are extracted. These two types of extracted key data are integrated and organized to form a basic parameter set for calculating ship-bridge collision probability. The vessel size information includes core indicators reflecting the vessel's shape characteristics, while the hydrological and meteorological data are key environmental indicators affecting the vessel's navigation status. This basic parameter set is the core data basis for subsequent collision probability calculations, and the extraction process ensures a high degree of matching between the data and the real-time navigation status of the vessels and the real-time environmental conditions of the bridge area. For example, from the multi-source real-time navigation dataset of a cross-river bridge, the latitude and longitude position, ground speed of 12 knots, heading of 45°, length of 90m, and width of 18m of a cargo ship can be extracted. At the same time, the hydrological and meteorological data of the water level of the bridge area where the cargo ship is located (4.8m), current velocity (1.5m / s), and wind speed (2.8m / s) can be extracted. After integrating these data, a basic parameter set for calculating the collision probability of the cargo ship is formed.

[0130] S42. Based on the relative distance between the ship and the bridge, the velocity component of the ship's speed in the direction of the bridge, the ship's yaw rate, hydro-meteorological conditions, and ship maneuvering performance parameters in the basic parameter set, construct a ship-bridge collision probability calculation model.

[0131] Specifically, using the obtained set of basic parameters as the core, five core influencing factors are extracted: the relative distance between the ship and the bridge, the velocity component of the ship's speed in the direction of the bridge, the ship's yaw rate, hydrological and meteorological conditions, and ship maneuvering performance parameters. Combining the influence of each factor on the risk of ship-bridge collision, a quantitative mapping relationship between each factor and the probability of ship-bridge collision is established, constructing a ship-bridge collision probability calculation model suitable for bridge-area waters. This model incorporates multi-dimensional key factors affecting collisions into the calculation system, pre-sets scientific calculation logic and quantitative formulas, and can realize the transformation from multi-dimensional parameters to collision probability values, providing core model support for subsequent quantitative calculations. For example, based on the basic parameter set of a vessel in a certain bridge area, the following parameters were extracted: the relative distance between the vessel and the bridge pier (75m), the speed component in the direction of the bridge (6kn), the yaw degree of the vessel (12°), the current speed (1.5m / s), the wind speed (2.8m / s), and the vessel's maneuverability parameters such as turning index and following index. A quantitative calculation logic was established based on the influence weight of various parameters on collision risk. These parameters were used as input variables for the model to build a ship-bridge collision probability calculation model that can adapt to the navigation characteristics of the bridge area.

[0132] S43. Using the ship-bridge collision probability calculation model, the multi-source real-time navigation dataset is calculated in real time to obtain the ship-bridge collision probability value.

[0133] Specifically, the continuously updated multi-source real-time navigation dataset of the bridge area is input into the pre-built ship-bridge collision probability calculation model at a set frequency. The model dynamically processes the input real-time data according to preset calculation logic and quantification formulas. Through comprehensive quantitative analysis of multi-dimensional influencing factors, it finally outputs a ship-bridge collision probability value that intuitively reflects the likelihood of a collision at that moment. This calculation process is synchronized with the update rhythm of the multi-source real-time navigation dataset, realizing real-time calculation of the ship-bridge collision probability. The output probability value is the core indicator for quantifying and determining the collision risk level. For example, if a multi-source real-time navigation dataset of a certain bridge area is continuously input into the collision probability calculation model every second, the model, after comprehensive calculation of the ship navigation data and bridge area environmental data input at a certain moment, obtains a probability value of 0.75 for the ship to collide with the bridge at that moment. This value accurately quantifies the degree of ship-bridge collision risk at that moment.

[0134] This embodiment accurately extracts core calculation parameters from multi-source real-time navigation datasets, constructs a calculation model around the multi-dimensional key factors affecting collisions, and realizes real-time calculation of collision probability. It overcomes the one-sidedness of judging risk by a single factor, and realizes accurate and dynamic quantification of ship-bridge collision risk, providing a scientific and reliable quantitative basis for the implementation of subsequent graded protection strategies.

[0135] Furthermore, in some embodiments, step S42, "constructing a ship-bridge collision probability calculation model based on the relative distance between the ship and the bridge, the velocity component of the ship's speed in the bridge direction, the ship's yaw rate, hydrological and meteorological conditions, and ship maneuvering performance parameters in the basic parameter set," may specifically include:

[0136] S421. Construct a geometric encounter factor based on the relative distance between the ship and the bridge, the nearest meeting distance, and the time to reach the nearest point;

[0137] Specifically, based on the spatial relationship and encounter movement trends between ships and bridges, three key parameters are extracted: relative distance, closest encounter distance (DCPA), and time to closest point (TCPA). Combined with the navigable spatial characteristics of the bridge area, a scientific quantitative calculation logic is established to map these three parameters to a geometric encounter factor, thus constructing the geometric encounter factor. This factor quantifies the potential probability of a collision between a ship and a bridge at the spatial geometric level. The factor value is positively correlated with the collision risk at the geometric level, making it a core quantitative indicator reflecting the geometric risk of ship-bridge collisions. For example, in an inland river bridge area, a cargo ship has a relative distance of 60m to a bridge pier, a closest encounter distance of 8m, and a time to closest point of arrival of 25s. Substituting these parameters into the preset quantitative calculation formula yields a corresponding geometric encounter factor of 0.85. This value directly reflects a high geometric risk of collision between the ship and the bridge.

[0138] S422. Construct maneuverability factors based on the ship's turning index and following index;

[0139] Specifically, two core parameters characterizing a ship's maneuverability are extracted: the turning index and the following index. The turning index reflects the ship's turning maneuverability, while the following index reflects the ship's responsiveness to maneuvering commands. Together, they determine the ship's actual maneuverability to avoid collisions in bridge-adjacent waters. Combining the maneuverability characteristics of different ship types, a quantitative calculation relationship is established between the turning index, the following index, and the maneuverability factor, thus constructing the maneuverability factor. This factor is used to quantify the impact of a ship's maneuverability on collision risk; the stronger the ship's maneuverability, the more the maneuverability factor reflects a reduction in collision risk. For example, a large bulk carrier has a turning index of approximately 0.8 and a following index of approximately 2.0; substituting these into a preset formula yields a maneuverability factor of 0.5. In contrast, a small speedboat has a turning index of approximately 0.3 and a following index of approximately 0.8; the calculated maneuverability factor is 0.8, directly demonstrating that the small speedboat has stronger maneuverability and a lower collision risk at the maneuverability level.

[0140] S423. Construct environmental impact factors based on wind speed, current velocity, and visibility;

[0141] Specifically, three hydro-meteorological parameters—wind speed, current speed, and visibility—that have the most significant impact on ship navigation and maneuvering in the bridge area are extracted. Combined with the navigational characteristics of the bridge area, a computational relationship between these three environmental parameters and the environmental impact factor is established through scientific quantification methods such as linear weighting, thus constructing the environmental impact factor. This factor is used to quantify the degree of influence of the natural environmental conditions in the bridge area on the risk of ship-bridge collision. The higher the wind speed, the faster the current, and the lower the visibility, the greater the difficulty of ship navigation and maneuvering, and the higher the environmental collision risk reflected by the environmental impact factor. For example, in a bridge area with a wind speed of 3 m / s, a current speed of 1.0 m / s, and visibility of 9 km, substituting these parameters into the linear weighting formula yields an environmental impact factor of 0.75. If the bridge area suddenly experiences low visibility weather, with visibility dropping to 1 km, and the other parameters remain unchanged, the recalculated environmental impact factor is 0.32, clearly demonstrating that low visibility significantly increases the collision risk for ship navigation.

[0142] S424. The geometric mean of the geometric encounter factor, the manipulative factor, and the environmental impact factor is used to obtain the comprehensive risk factor;

[0143] Specifically, the constructed geometric encounter factor, maneuverability factor, and environmental impact factor are treated as three independent risk dimensions. A geometric mean is used to comprehensively process these three factors, and the resulting comprehensive risk factor is calculated using the geometric mean formula. The geometric mean approach balances the weighting of the three different dimensions of factors on the risk of ship-bridge collisions, avoiding bias in the comprehensive risk assessment caused by the numerical deviation of a single factor. This ensures that the resulting comprehensive risk factor fully and evenly reflects the combined potential risk of ship-bridge collisions under the combined effects of geometric encounter, ship maneuverability, and environmental impact. For example, if a ship has a geometric encounter factor of 0.8, a maneuverability factor of 0.6, and an environmental impact factor of 0.7, substituting these three factors into the geometric mean formula yields a comprehensive risk factor of 0.7. This value integrates the risk characteristics of the three dimensions, objectively reflecting the ship's overall collision risk level.

[0144] S425. Construct a ship-bridge collision probability calculation model based on comprehensive risk factors;

[0145] Specifically, using the comprehensive risk factor obtained through geometric averaging as the core input variable, and combining it with the actual patterns of ship-bridge collisions in the bridge area, a scientific quantitative mapping relationship is established between the comprehensive risk factor and the ship-bridge collision probability value. Corresponding computational logic and formulas are set to complete the overall construction of the ship-bridge collision probability calculation model. This model transforms the comprehensive risk factor into a dimensionless ship-bridge collision probability value that can intuitively determine collision risk, achieving a standardized transformation from multi-dimensional risk factors to collision probability values, allowing the model's output probability value to accurately correspond to the actual degree of ship-bridge collision risk.

[0146] This embodiment constructs quantitative factors from three core dimensions of bridge collision risk: geometric encounter, ship maneuvering, and environmental impact. By balancing the weights of each dimension through geometric mean, a comprehensive risk factor is obtained. Then, a collision probability calculation model is constructed based on this factor. This ensures that the calculation logic of the model closely matches the actual collision risk formation pattern in the bridge area, avoids the judgment bias dominated by a single factor, and makes the calculated collision probability value more accurate and comprehensive in reflecting the actual risk level.

[0147] Furthermore, in some embodiments, step S5, "activating active collision avoidance warning and providing assisted driving decision information to the vessel," may specifically include:

[0148] S51. Active collision avoidance warning signals shall be issued to vessels passing under the bridge through acoustic warning equipment, optical warning equipment and electronic warning equipment;

[0149] Specifically, once the collision risk between the ship and the bridge is determined to reach the corresponding level, three types of dedicated early warning equipment—audio, optical, and electronic—deployed in the bridge area are activated simultaneously. Each device works collaboratively according to a preset early warning mode, issuing proactive collision avoidance warning signals to target vessels passing under the bridge from three dimensions: auditory, visual, and electronic signal transmission. This multi-dimensional early warning approach creates a synergistic effect, ensuring that vessel operators can quickly and accurately perceive collision risk warning information from the bridge area. The auditory warning equipment delivers warnings through voice broadcasts and horns; the optical warning equipment delivers warnings through visual means such as flashing lights and color cues; and the electronic warning equipment transmits warning information to the vessel's terminal via electronic signals. For example, in the area of ​​a cross-river bridge, the high-pitched horn of the auditory warning equipment continuously broadcasts the voice prompt, "There is a collision risk ahead; please adjust your navigation status immediately." The red LED array of the optical warning equipment flashes rapidly at a frequency of 1Hz, and the electronic warning equipment sends collision avoidance warning electronic signals to the target vessel's electronic terminal. The three types of equipment work together to issue warnings, allowing vessel operators to perceive warning information through multiple channels.

[0150] S52. Based on multi-source real-time navigation datasets, conduct a comprehensive analysis of the ship's current navigation status and the navigation environment in the bridge area, and generate an assisted driving decision-making scheme that includes recommended speed, recommended course and suggested control measures;

[0151] Specifically, using real-time updated multi-source navigation datasets as the core analytical basis, the system first extracts core navigation status information such as the target vessel's current position, real-time speed, course, and dimensions to accurately grasp the vessel's actual navigation situation. Then, it extracts navigation environment information such as water level, current speed, wind speed, wind direction, and the location and width of navigation holes in the bridge area to comprehensively assess the external environmental conditions for vessel navigation. Subsequently, combining the matching degree between the vessel's navigation status and the bridge area's navigation environment, the system determines the optimal navigation path and speed for the vessel to safely pass through the bridge area through comprehensive analysis. This generates an assisted driving decision-making scheme containing recommended speeds with specific values, recommended courses with clear bearings, and directly operable suggested control measures. All aspects of the scheme are tailored to the actual vessel situation and the bridge area environment, making it feasible to implement. For example, from a multi-source real-time navigation dataset, it is extracted that a cargo ship is currently traveling at 14 knots and heading at 28°. The current speed in the bridge area is 1.3 m / s, the main navigation channel is at a heading of 32°, and the speed limit for navigation in the bridge area is 10 knots. After comprehensive analysis, an assisted driving decision scheme is generated, which recommends a speed of 9 knots, a heading of 32°, and suggests the following control measures: "slowly decelerate to 9 knots, slightly adjust the heading to the right to 32°, and maintain a constant speed to pass through the navigation channel".

[0152] S53. The assisted driving decision-making plan will be simultaneously pushed to the corresponding ship and bridge management and control center in the form of Automatic Identification System (AIS) information, high-frequency voice and visual display.

[0153] Specifically, the generated assisted driving decision-making schemes are converted into multiple formats, namely, Automatic Identification System (AIS) information, VHF high-frequency voice, and visual display, and simultaneously pushed to various receiving terminals of the target bridge-crossing vessels, as well as the monitoring and dispatching terminals of the bridge management and control center. This achieves synchronous transmission of decision-making schemes between the vessel and the control center, enabling vessel operators to promptly obtain and execute the schemes, and allowing the bridge management and control center to monitor and dispatch vessel navigation status in real time. Specifically, AIS information is sent to the vessel's AIS terminal in the form of electronic messages, VHF voice is transmitted to vessel operators via real-time voice broadcast through the VHF (Very High Frequency) communication channel, and visual displays are presented in an intuitive format, using text, graphics, and arrows, on the vessel's navigation screen and the large monitoring screen of the bridge control center. For example, the assisted navigation decision-making scheme of the aforementioned cargo ship can be edited into an AIS standard electronic message and sent to the cargo ship's AIS terminal; recommended speed, course, and control suggestions can be broadcast to the cargo ship operators in voice form through the bridge area VHF high-frequency communication channel; at the same time, recommended speed can be marked with numbers, recommended course can be indicated with arrows, and suggested control measures can be displayed in text on the cargo ship's navigation display screen and the monitoring screen of the bridge management and control center, so as to realize the synchronous push of the decision-making scheme.

[0154] This embodiment uses audio-visual equipment to achieve multi-dimensional active collision avoidance warning, combines multi-source data to generate scientific assisted driving decision-making schemes, and pushes them to the ship and bridge management and control centers in multiple forms simultaneously. This upgrades active collision avoidance from simple risk warning to a systematic intervention that integrates warning, decision-making, and information synchronization, effectively guiding ship operators to adjust their navigation status and reducing the probability of ship-bridge collisions at medium-risk levels.

[0155] Furthermore, in some embodiments, step S52, "based on multi-source real-time navigation datasets, comprehensively analyzes the ship's current navigation status and the navigation environment in the bridge area to generate an assisted driving decision-making scheme including recommended speed, recommended course, and suggested control measures," may specifically include:

[0156] S521. Extract the current position, current speed, current course, and target pier position of the vessel from the multi-source real-time navigation dataset;

[0157] Specifically, using the multi-source real-time navigation dataset of the bridge area as the data source, the current position, current speed, and current course of target vessels passing under the bridge are precisely filtered and extracted. Simultaneously, the specific location information of the target bridge piers that the vessels plan to pass through is extracted. All the extracted data is real-time updated and valid, serving as the foundation for subsequent calculations of recommended navigation parameters and the generation of driving decisions. The current position of the vessel is its precise spatial geographic coordinates, the current speed is the actual speed of the vessel, the current course is the actual bearing angle of the vessel, and the target bridge pier location is the precise geographic coordinates or bearing information of the bridge pier next to the corresponding navigation opening in the bridge area. For example, from the multi-source real-time navigation dataset, the current latitude and longitude coordinates of an inland river cargo ship are extracted as (118°E, 32°N), the current speed is 12 knots, and the current course is 35°. Simultaneously, the location of the target bridge pier corresponding to navigation opening No. 2 in the bridge area, which the cargo ship plans to pass through, is extracted as (118.002°E, 32.001°N), clarifying the spatial position of the vessel relative to the target bridge pier and the vessel's current navigation status.

[0158] S522. Based on ship maneuvering performance parameters and navigation environment data in the bridge area, calculate the recommended speed and recommended course for the ship to safely pass through the bridge pier;

[0159] Specifically, by combining the extracted current navigation data of the vessel with the location of the target bridge pier, and incorporating the vessel's own maneuvering performance parameters, while simultaneously retrieving bridge area navigation environment data from multi-source real-time navigation datasets, and through scientific quantitative calculation and analysis of multiple factors, the optimal navigation speed and heading that allow the vessel to avoid the bridge pier and safely and smoothly pass through the bridge area navigation opening are determined—that is, the recommended speed and recommended course. The vessel's maneuvering performance parameters determine its practical capabilities in acceleration, deceleration, and turning. The bridge area navigation environment data includes external factors affecting vessel navigation such as current speed, wind speed, and wind direction. Both serve as key bases for calculating the recommended speed and course, ensuring that the calculation results closely match the vessel's actual maneuvering capabilities and the objective navigation conditions of the bridge area. For example, considering the aforementioned cargo ship, its maneuverability parameters of 0.7 turning index and 1.8 following behavior index, as well as navigation environment data such as the current speed of 1.2 m / s and the southerly wind of 2.5 m / s in the bridge area, after comprehensive calculation, the recommended speed for the cargo ship to safely pass through the No. 2 navigation channel is determined to be 8 knots and the recommended heading is 40°. These parameters can adapt to the ship's maneuverability and offset the adverse effects of environmental factors.

[0160] S523. Based on the difference between the recommended speed and the current speed, and the difference between the recommended course and the current course, generate an assisted driving decision scheme that includes speed adjustment suggestions and course adjustment suggestions;

[0161] Specifically, the numerical difference between the recommended speed and the ship's current speed, and the angular difference between the recommended course and the ship's current course are calculated separately. Then, combined with the ship's maneuverability parameters, specific and actionable speed and course adjustment suggestions are formulated based on the magnitude and direction of these two types of differences. These two types of adjustment suggestions are then integrated to form a complete assisted navigation decision-making scheme. The adjustment suggestions fully consider the ship's practical characteristics, avoiding issuing instructions beyond the ship's maneuverability, ensuring the scheme is feasible and executable, allowing ship operators to directly adjust the navigation status according to the suggestions. For example, if the cargo ship's speed difference is 4 knots faster than the recommended speed, and its course difference is 5 degrees to the left of the recommended course, the generated assisted navigation decision-making scheme, based on its maneuverability, is: "Slowly decelerate to 8 knots at a rate of 0.4 knots / s, slightly adjust the course to the right to 40 degrees, and perform the deceleration and turning operations simultaneously and smoothly. After the adjustment is completed, maintain a constant speed and proceed straight through the passageway," clearly defining the specific operation method, magnitude, and rhythm.

[0162] This embodiment extracts core data on ship navigation and target bridge piers, combines ship maneuverability and bridge area environmental data to calculate the optimal recommended speed and course, and then generates specific and operable driving adjustment suggestions based on parameter differences. This allows the assisted driving decision-making scheme to overcome the drawbacks of general guidance, better match the ship's actual maneuverability and bridge area environment, greatly improve the feasibility and accuracy of the scheme, and effectively enhance the intervention effect of active collision avoidance.

[0163] Furthermore, in some embodiments, step S5, "activating the servo motor to release the passive anti-collision airbag and activating the tracking camera to record ship video," may specifically include:

[0164] S54. Send a release command to the anti-collision control unit, and the anti-collision control unit drives the anti-collision airbag installed on the ship-facing side of the pier to inflate and deploy via a servo motor.

[0165] Specifically, when the collision risk between the bridge and the ship is determined to reach a high-risk level, the system automatically sends a control command to the bridge's dedicated anti-collision control unit to release the anti-collision airbags. Upon receiving the command, the anti-collision control unit immediately triggers a pre-set servo motor to enter working mode. Using the servo motor as a power source, the system drives the anti-collision airbags pre-installed on the ship-facing side of the pier to inflate and deploy. This allows the airbags to quickly form a physical buffer structure covering the ship-facing side of the pier. This structure absorbs the energy generated by the ship collision, reducing the damage to the main bridge structure. The anti-collision airbags are specialized protective structures adapted to the pier's protection requirements, pre-deployed on the ship-facing side of the pier where collisions are most likely. The servo motor provides stable and efficient power support for the airbag inflation and deployment, ensuring rapid response and activation. For example, in the bridge area of ​​a cross-river bridge, when the calculated probability value of a ship collision with the bridge reaches the high-risk threshold, the system immediately sends an airbag release command to the bridge area anti-collision control unit. The control unit quickly starts the servo motor, which drives the compressed air anti-collision airbag on the ship-facing side of the bridge pier to inflate and deploy within a preset time, forming a buffer protection layer of appropriate thickness, and preparing for physical protection against ship collision.

[0166] S55. Simultaneously activate the tracking cameras in the bridge area waters to continuously track and record the navigation process of vessels passing under the bridge, and upload the video data with timestamps and geographic coordinates to the data storage center in real time.

[0167] Specifically, while sending airbag release commands to the collision avoidance control unit and activating the servo motors, the system simultaneously triggers the tracking cameras deployed in the bridge area to enter working mode. The tracking cameras automatically aim at the target vessel and continuously record high-definition video of the vessel's entire navigation process in the bridge area waters, fully capturing the vessel's navigation status and relative motion with the bridge. At the same time, the system automatically loads precise timestamps and geographic coordinates into the recorded video data, ensuring that each video segment corresponds to clear time and spatial location information, thus ensuring the traceability of the video data. Subsequently, the video data with timestamps and geographic coordinates is uploaded in real time to the bridge area's dedicated data storage center via a dedicated data transmission link for secure and long-term storage, providing complete image data for subsequent related work. For example, as the servo motor starts and the anti-collision airbags begin to inflate and deploy, the high-magnification optical zoom tracking cameras deployed in the bridge area immediately aim at the target cargo ship and continuously track and record the cargo ship's navigation trajectory and operational actions in high-definition quality. The video stream is automatically loaded with millisecond-precision timestamps and the ship's real-time geographical coordinates. This precisely labeled video data is uploaded to the bridge area's data storage center in real time via gigabit Ethernet, enabling the immediate preservation of visual evidence.

[0168] In this embodiment, under high-risk conditions, the anti-collision airbags of the bridge piers are deployed by a servo motor to form a physical buffer protection layer to reduce collision damage. At the same time, the tracking camera is activated to record and upload ship videos with spatiotemporal annotations, realizing the synchronous implementation of passive physical protection of the bridge and accident image evidence collection. This forms a closed loop of protection that reduces damage during the high-risk phase and preserves evidence afterward, improving the integrity and practical application value of the passive protection link.

[0169] Furthermore, in some embodiments, S54 "sends a release command to the anti-collision control unit, and the anti-collision control unit drives the anti-collision airbag installed on the ship-facing side of the pier to inflate and deploy via a servo motor," which may specifically include:

[0170] S541. Send a release command to the collision avoidance control unit. After receiving the release command, the collision avoidance control unit starts the servo motor within a preset response time.

[0171] Specifically, when the risk of a ship-bridge collision reaches a high-risk level, the system automatically sends a release command for the airbags to the dedicated anti-collision control unit in the bridge area. This command is the core control signal that triggers the airbag deployment. After receiving the release command, the anti-collision control unit completes a series of actions, including signal analysis and power triggering, according to a pre-set response time. Within this preset time, it precisely starts the servo motor, ensuring that the power source for airbag deployment can respond quickly, laying the foundation for the timely deployment of the anti-collision airbags. The preset response time is set based on the actual navigation characteristics of the bridge area, such as the ship's speed and the relative distance between the ship and the bridge, to ensure the timely start of the servo motor and avoid missing the protection opportunity due to response delay. For example, for a certain inland river bridge area, based on the characteristics of high ship speed and narrow waterway, the preset response time for servo motor start is 0.3 seconds. When the anti-collision control unit receives the airbag release command, it completes the internal circuit triggering and power output within 0.3 seconds, successfully starting the servo motor.

[0172] S542. The servo motor drives the roller to rotate, so that the anti-collision airbag can be fully inflated and deployed within a preset deployment time to form a buffer protection layer on the ship-facing side of the bridge pier.

[0173] Specifically, after the servo motor starts, its own power output drives the roller, which is pre-connected to the airbag, to rotate. The rotation of the roller activates the airbag's retraction and release mechanism, unlocking the airbag in its retracted state and triggering its inflation system. The airbag completes the inflation process and fully deploys within a pre-set deployment time, ultimately forming a complete and continuous buffer layer on the pier's facing side. This layer directly faces the collision direction of the ship, providing physical cushioning protection for the pier. The roller is the key transmission structure for the airbag's retraction and release. The preset deployment time is set in conjunction with the estimated time of the ship's arrival at the pier, ensuring that the airbag can be fully deployed before the ship's collision, and that the deployed buffer layer can cover the vulnerable collision area on the pier's facing side. For example, in the aforementioned inland river bridge, the servo motor drives the roller to rotate clockwise at a constant speed after startup, unlocking the compressed air anti-collision airbag on the pier's ship-facing surface and triggering the inflation valve. The airbag completes inflation within a preset deployment time of 0.8 seconds, forming a 1.2m thick buffer protective layer covering all easily collided areas of the pier's ship-facing surface, thus ensuring complete protection before the ship arrives.

[0174] This embodiment achieves rapid and smooth unlocking and inflation of the anti-collision airbag by pre-setting the response time of the anti-collision control unit to the release command and the deployment time of the anti-collision airbag, and using a servo motor to drive the roller. This ensures that the airbag can be deployed before the ship collision and forms an effective buffer protective layer covering the collision-prone area on the pier's facing side. This improves the timeliness, stability and effectiveness of the passive anti-collision airbag deployment and minimizes the damage to the pier from the collision.

[0175] Furthermore, in some embodiments, the method further includes:

[0176] S61. When a bridge collision is detected, the time of the collision event is automatically marked;

[0177] Specifically, the system continuously monitors collision-related signals in and around the bridge area. When various sensors deployed in the bridge area detect characteristic signals of an actual ship-bridge collision, an accident marking mechanism is triggered. The system automatically captures and precisely marks the specific time of the collision event, which serves as the core time reference for accident analysis, defining the time range for subsequent data extraction. The collision detection is based on collision characteristic signals captured by dedicated sensors in the bridge area, and the marked time is a precise, standardized time to ensure the consistency of time in subsequent data extraction. For example, if an accelerometer deployed at the bridge pier captures a collision impact characteristic signal of ≥2g, the system determines that a ship-bridge collision has occurred and immediately and automatically marks the UTC millisecond time at that moment as the collision event time t0, serving as the time origin for accident data extraction.

[0178] S62. Extract video data and multi-source real-time navigation datasets within a preset time period before and after the collision event, and generate an evidence package;

[0179] Specifically, based on the collision event time automatically marked by the system, video data from bridge tracking cameras within a specified time range before and after that time point is extracted. Simultaneously, multi-source real-time navigation datasets completely synchronized with the video data are extracted. These two types of data are integrated and packaged to form a complete evidence package containing both visual and data evidence. The preset time range is set according to the navigation characteristics of the bridge area and the needs of accident analysis, ensuring that the extracted data fully covers the entire accident process, the vessel navigation status before the incident, and the on-site situation after the incident. The evidence package is packaged in a standardized format to ensure data integrity and readability. For example, using the collision time t0 as a benchmark, high-definition video data from tracking cameras from 60 seconds before t0 to 180 seconds after t0 is extracted. Simultaneously, synchronized vessel trajectory data, bridge collision probability value change curves, and multi-source real-time navigation datasets such as bridge area hydrological and meteorological data are extracted during this time period. These two types of data are packaged into a standardized evidence package in ZIP format, completely preserving all original data related to the accident.

[0180] S63. Based on the evidence package, generate an accident analysis report and push the evidence package and accident analysis report to the preset management platform;

[0181] Specifically, the system incorporates a standardized accident analysis template. Based on the extracted and generated complete evidence package, the system automatically analyzes, organizes, and summarizes the video data and multi-source real-time navigation datasets within the evidence package, generating a standardized accident analysis report containing key accident information. Subsequently, the system uses a preset transmission method and link to synchronously push the accident analysis report and the original evidence package to a pre-defined management platform in the bridge area, ensuring that relevant management parties can promptly obtain complete accident data and analysis results. The accident analysis report contains core information about the accident. The preset management platform is a dedicated platform responsible for navigation safety and bridge operation management in the bridge area, ensuring accurate and timely transmission of accident data. For example, the system uses a built-in template to generate a PDF accident analysis report based on the data in the evidence package. The report includes a dynamic thumbnail of the accident process, a vessel navigation track table, collision probability trends, navigation environment data for the bridge area, and preliminary recommendations for accident liability. This PDF report is then synchronously pushed via SMTP protocol to the preset management platforms and designated work email addresses of the maritime administration department and the bridge operating unit.

[0182] This embodiment automatically marks the time of the accident after detecting a ship-bridge collision, extracts videos and multi-source data before and after the accident to generate an evidence package, and then automatically generates an analysis report based on the evidence package and pushes it to a preset management platform. This realizes an automated and standardized closed-loop process for post-accident evidence collection, analysis and reporting, which solves the problems of difficulty in accident evidence collection and untimely data transmission, and provides comprehensive and objective core evidence for accident cause analysis and responsibility determination.

[0183] Furthermore, in some embodiments, step S62, "extracting video data and multi-source real-time navigation datasets within a preset time period before and after the collision event, and generating an evidence package," may specifically include:

[0184] S621. Extract the video clip from the tracking camera within the first preset duration before the collision event occurs to the second preset duration after the collision event occurs;

[0185] Specifically, using the automatically marked time of the ship-bridge collision event as the time reference point, and based on the navigation characteristics of the bridge area and the actual evidence collection needs of the accident analysis, a first preset duration before the collision and a second preset duration after the collision are pre-set to define the complete time range for accident evidence collection. From the video storage data of the bridge area tracking cameras, the full-process tracking video clips of the target vessel within this time range are accurately extracted. These video clips completely cover the entire process of the vessel approaching the bridge pier, the collision occurring, and its post-collision navigation, while retaining the original high-definition image quality and relevant annotation information, and are the core video data for accident evidence collection. For example, for a collision accident in the area of ​​a coastal bridge, the first preset duration is set to 60 seconds and the second preset duration to 180 seconds. Using the collision time t0 as the reference, high-definition tracking video clips of the offending vessel from the tracking cameras between t0-60 seconds and t0+180 seconds are extracted. These clips completely record the entire process of the vessel from the emergence of collision risk, the actual collision, to its departure after the collision.

[0186] S622. Extract a multi-source real-time navigation dataset that is time-synchronized with the video clip. The multi-source real-time navigation dataset includes ship trajectory data, bridge collision probability value change curves, and navigation environment data.

[0187] Specifically, using the timeline of the extracted video clip as the sole benchmark, a multi-source real-time navigation dataset perfectly synchronized with the video clip's time is precisely extracted from the multi-source real-time navigation data repository of the bridge area. This ensures that every data item in the dataset matches the corresponding time point of the video clip, achieving spatiotemporal synchronization between the video footage and the quantitative data. This multi-source real-time navigation dataset contains three core types of data: ship track data reflecting changes in ship navigation trajectories, a bridge-ship collision probability value change curve reflecting dynamic changes in collision risk, and navigation environment data showing the environmental conditions at the time of the accident. These are the core quantitative data for accident evidence collection. For example, the multi-source real-time navigation dataset synchronized with the aforementioned 60-180 second video clip includes the ship's latitude, longitude, speed, and heading for each second, the bridge-ship collision probability value change curve gradually increasing from 0 to 1, and navigation environment data such as water level, current speed, wind speed, and visibility in the bridge area during that time period. The time points of all data completely correspond to the video footage.

[0188] S623. Encapsulate the video clips with the multi-source real-time navigation dataset to generate an evidence package containing video evidence and data evidence;

[0189] Specifically, the extracted full-process accident tracking video clips and the extracted time-synchronized multi-source real-time navigation dataset are integrated. Following standardized encapsulation rules and a common file format, both types of data are uniformly packaged to form a single accident evidence package. This evidence package contains both visual video evidence and quantitative numerical data evidence. The two types of evidence are interconnected and mutually corroborative. During the encapsulation process, the evidence package undergoes standardized naming and information labeling to ensure it is searchable, readable, and traceable. For example, the aforementioned high-definition tracking video clips are converted into a common video file format. Vessel track data, collision probability change curves, and navigation environment data are organized into easily readable file formats such as tables, line graphs, and text. All files are then integrated and packaged into a ZIP format accident evidence package, named "Collision Accident Time - Attack Vessel Identification - Bridge Area Name," facilitating subsequent retrieval, viewing, and analysis.

[0190] This embodiment defines a specific time range for evidence collection before and after a collision accident, extracts ship tracking video clips within this range, and simultaneously extracts time-matched multi-source real-time navigation datasets. After standardized packaging, it forms an accident evidence package in which video evidence and data evidence corroborate each other, constructing a complete and traceable accident evidence chain, improving the objectivity and persuasiveness of accident evidence, and providing sufficient and effective data support for subsequent accident handling.

[0191] Furthermore, in some embodiments, the method further includes:

[0192] S71. Connect the navigation environment data and ship passage data to the edge computing gateway, and perform time synchronization processing on the navigation environment data and ship passage data in the edge computing gateway to ensure that the time deviation between the navigation environment data and ship passage data is less than the preset deviation threshold.

[0193] Specifically, navigation environment data collected from the bridge area and vessel traffic data monitored are uniformly connected to a dedicated edge computing gateway in the bridge area. This gateway serves as the core node for data processing, performing professional time synchronization processing on both types of data. The synchronization process follows preset time calibration rules to unify the time measurement benchmarks of the two types of data, ultimately controlling the time deviation between navigation environment data and vessel traffic data within a preset deviation threshold. This ensures that data from different types and collection channels maintain a high degree of consistency in the time dimension. For example, navigation environment data such as water level and flow velocity collected by hydrological and meteorological sensors in the bridge area, as well as traffic data such as vessel position and speed monitored by AIS and radar, are all connected to the industrial-grade edge computing gateway. The gateway completes time calibration through its built-in multi-system timing module. The preset time deviation threshold is 50ms. After processing, the actual time deviation of both types of data is controlled within 50ms, achieving precise time synchronization.

[0194] S72. Store the time-synchronized navigation environment data and ship passage data in the edge cache. The edge cache adopts a circular cache structure and is used to store the original data within a preset time period.

[0195] Specifically, store the time-synchronized navigation environment data and ship passage data in the edge cache supporting the edge computing gateway. The edge cache adopts a dedicated data storage structure of circular cache. According to the actual requirements of bridge area data application and accident backtracking, preset the storage duration of the original data. The edge cache will continuously retain the latest navigation environment and ship passage original data according to this duration requirement. The circular cache structure will automatically cycle and overwrite the historical data that exceeds the preset storage duration, ensuring the integrity of the original data within the preset time period while guaranteeing the data storage efficiency. For example, store the time-synchronized navigation data of the bridge area in the NVMe RAID1 edge cache with a capacity of 2TB. Preset the storage duration of the original data to be 7 days. This circular cache will continuously retain the navigation environment and ship passage data of the most recent 7 days, automatically cycle and overwrite the old data that exceeds 7 days, and always keep the cache with the latest 7-day original data to meet the basic data requirements for accident backtracking.

[0196] S73. When a network interruption is detected, temporarily store the collected navigation environment data and ship passage data in the edge cache, and automatically upload the temporarily stored data after the network is restored.

[0197] Specifically, continuously monitor the network communication status between the edge computing gateway and the remote data center. When a network interruption occurs and data cannot be normally uploaded, the system will immediately trigger the local temporary storage mechanism, and continue to temporarily store the newly collected navigation environment data and ship passage data in the above-mentioned circular edge cache. The cache will reserve a dedicated storage space for the temporarily stored data to avoid data loss. When the system detects that the network communication status returns to normal, it will automatically start the data upload program and upload all the data temporarily stored in the edge cache during the network interruption to the remote data center in chronological order to complete the upload of the interrupted data and ensure the integrity of data transmission. For example, when a sudden network failure occurs in the bridge area, after the system detects the network interruption, it will continuously temporarily store the newly collected data such as wind speed and ship heading in the circular edge cache. After 2 hours, when the network is restored, the system will immediately and automatically upload all the temporarily stored data in the cache during the network interruption to the provincial maritime cloud platform through Gigabit Ethernet without any data omission or loss.

[0198] This embodiment unifies navigation environment and vessel traffic data into an edge computing gateway to achieve time synchronization. It achieves efficient storage of raw data and retention for a preset duration through a ring caching structure. At the same time, it establishes a mechanism for local temporary storage of data when the network is interrupted and automatic retransmission after the network is restored. This comprehensively ensures the consistency, integrity and transmission continuity of navigation data in the bridge area, laying a solid foundation for the stable operation of the entire bridge intelligent protection system from the perspective of data storage and transmission.

[0199] To facilitate understanding of the bridge intelligent protection method based on a combined active and passive mechanism provided in this embodiment, the following will be explained in conjunction with specific application scenarios.

[0200] See Figure 3 This embodiment provides an integrated information perception and multi-source data preprocessing architecture for "bridge-ship-shore" systems, used to provide a real-time dataset with high precision, low latency, and a unified spatiotemporal reference for subsequent collision probability calculations. The architecture is designed in five levels: "acquisition-aggregation-preprocessing-fusion-output," as detailed below:

[0201] The data acquisition layer comes from equipment on the bridge side, ship side, and shore side.

[0202] Bridge-side sensing node equipment includes:

[0203] High-definition visible light PTZ camera: 4K@50fps, covering 150m to the left and right of the navigation hole, used for ship type identification and initial track extraction;

[0204] Tracking camera: 1080p@50fps, 30x optical zoom, used for high-risk continuous close-ups of ships;

[0205] Photoelectric monitoring equipment: 640×512 infrared thermal imager, used for nighttime and low-visibility blind spot coverage;

[0206] Hydrological and meteorological sensors: water level gauges, flow meters, anemometers, etc., with a sampling frequency of 1 hour;

[0207] Intelligent navigation beacon: powered by solar energy and lithium battery, with a range of ≥15 days;

[0208] Sound, light, and electricity warning light: red / yellow / green tri-color LED array, sound pressure level ≥120dB@1m, used for active warning output.

[0209] Ship-side sensing node equipment includes:

[0210] AISClassA: Default interval is 3 minutes. This architecture remotely triggers a 30-second interval via the AISApp-spec message to increase trajectory density.

[0211] Radar target: X-band solid-state radar (12W), range resolution 0.6m, azimuth resolution 1°;

[0212] Shipboard camera: 1080p@25fps, transmitted via 5G network, used for crew behavior verification.

[0213] The shore-side sensing node equipment includes:

[0214] Shore-based radar station: S-band pulse radar (2MW), maximum range 12nmile, forming a "bistatic" redundancy with the bridge-side radar;

[0215] Hydrological station: Acoustic Doppler current profiler, float-type water level gauge, laser visibility meter;

[0216] Data aggregation layer. Node data from the bridge side, ship side, and shore side are all connected to the industrial-grade edge gateway.

[0217] Time synchronization: The gateway has a built-in UBLOXZED-F9T timing module that supports GPS / BeiDou / GLONASS systems with a PPS accuracy of ±20ns. All sensor data is encapsulated after being timestamped by FPGA hardware (UTC accuracy ±1ms) to ensure that the time deviation of data from the bridge to the ship to the shore is less than 50ms.

[0218] Protocol conversion: FPGA hard core implements parallel parsing of multiple protocols including RS-422, Ethernet, CAN, and UART, with zero-copy writing to DDR;

[0219] Secure encryption using TLS 1.3 + SM4 national cryptographic algorithm, hardware encryption of the original video stream with bandwidth overhead <3%;

[0220] Edge caching, 2TB NVMeRAID1, and ring caching of 7 days of raw data meet the needs of incident backtracking.

[0221] Data preprocessing layer. Containerized "preprocessing microservices" are deployed on the gateway to clean the raw data.

[0222] Data fusion layer. The preprocessed multi-source tracks enter the "fusion engine":

[0223] For spatiotemporal alignment, UTM projection and high-precision digital elevation model are used, and the coordinates are unified to WGS84 coordinate system.

[0224] Track association, using nearest neighbor (NN) + JPDA joint probabilistic data association, with thresholds of 30m for position, 2kn for velocity, and 20% for scale; success rate of association between AIS and radar ≥98%;

[0225] State estimation employs IMM-CKF (Interactive Multi-Model-Volume Kalman Filter), with model sets including uniform velocity (CV), uniform turning (CT), and uniform acceleration (CA). The model switching probability matrix is ​​calibrated offline based on the bridge control characteristics. Position estimation accuracy (95% elliptic error) is improved from 6.7m before fusion to 2.1m.

[0226] Covariance intersection fusion conservatively merges the variances of heterogeneous sensors to ensure consistent estimation under unknown correlations.

[0227] Data output layer. The fusion engine outputs one JSON record every second, with the following key fields:

[0228] timestamp_utc, UTC time in milliseconds;

[0229] track_id, a unique identifier (UUID), is shared across bridges, ships, and shore for the same target;

[0230] vel:[Vx,Vy] (m / s), ship speed data;

[0231] dimension:[L,B,d] (m), ship size data;

[0232] env:[h_water,v_current,w_wind,dir_wind,Vis] (m / m / s / m / s / ° / km), navigation environment data;

[0233] The dataset is pushed to the local computing module via Gigabit Ethernet + MQTT, and simultaneously uploaded to the maritime cloud platform via 5G network, with bandwidth usage of <8Mbps.

[0234] See Figure 4 This embodiment provides a bridge protection method based on a combined active and passive mechanism, including the following steps:

[0235] Step S201: Obtain navigation environment data collected by the front-end sensing device in the bridge area waterway.

[0236] The front-end sensing equipment includes at least high-definition cameras, tracking cameras, photoelectric monitoring equipment, hydrological and meteorological sensors, intelligent navigation marks, and audible and optical early warning equipment. The collected navigation environment data includes, but is not limited to, water level, current velocity, wind speed, wind direction, visibility, obstacle information, and current traffic rules for the bridge area. In this embodiment, all sensors are connected to the edge computing node via industrial Ethernet or 4G / 5G wireless links, with a uniform sampling frequency of 1Hz. Timestamps are synchronized using an NTP network to ensure that the time synchronization error of multi-source data is <50ms.

[0237] Step S202: Obtain bridge area vessel traffic data monitored by AIS and radar.

[0238] AIS data is parsed via a VHF interface, and radar data uses raw point clouds from millimeter-wave radar, which are then internally detected before outputting the target track. In this embodiment, the radar and AIS target are first jointly correlated in a spatiotemporal and scale manner:

[0239] Spatial threshold: Horizontal distance ≤ 30m;

[0240] Speed ​​threshold: Speed ​​difference ≤ 1 kn;

[0241] Dimensional threshold: The difference between ship length and beam ≤ 20%.

[0242] Targets that are successfully associated are assigned a unified ID, while unassociated targets are retained independently, forming a "fusion track pool".

[0243] Step S203: Preprocess the navigation environment data and ship passage data, and construct a multi-source data fusion model to obtain a multi-source real-time navigation dataset.

[0244] The preprocessing steps are as follows:

[0245] Deduplication is performed by using the sampling time as an index to remove duplicate samples;

[0246] Noise reduction is performed by removing outliers based on the physical thresholds of each element (e.g., flow velocity |v|≤6m / s, wind speed |w|≤25m / s).

[0247] Dimensionality reduction was achieved by using principal component analysis (PCA) to reduce the 12-dimensional environment vector to 6 dimensions while retaining 95% of the variance.

[0248] Interpolation is used to fill in missing values ​​using cubic spline interpolation, with a maximum interpolation time window of ≤5s.

[0249] After preprocessing, a multi-source data fusion algorithm based on covariance intersection (CI) is used to weightedly fuse heterogeneous sensor data, outputting a standardized dataset D. k ={X k ,P k}, where X k Let P be the state vector. k This is the covariance matrix after fusion.

[0250] Step S204: Based on dataset D k Calculate the collision probability value P between the ship and the bridge.

[0251] This embodiment uses a comprehensive model that combines "short-term encounter geometry, maneuverability, and environmental impact":

[0252] P=1 exp( λ)

[0253] λ=(λ g ·λ m ·λ e ) / 3

[0254] Where, λ g The geometric encounter factor (λ) is exponentially related to the nearest meeting distance (DCPA) and the time to reach the nearest point (TCPA); m λ is a maneuverability factor, correlated with the ship's turning index K and following index T; e It is an environmental impact factor, linearly weighted with wind speed, current velocity, and visibility.

[0255] The dimensionless probability value of 0≤P≤1 is calculated in real time using λ, with an update frequency of 1Hz.

[0256] Steps S205-S207 involve comparing P with the preset risk threshold Pc and accident threshold Pa in a hierarchical manner, and then executing the corresponding protection strategy.

[0257] In this embodiment, Pc=0.15 and Pa=0.85 are set, and the threshold is calibrated by combining bridge area historical collision statistics and Monte Carlo simulation.

[0258] Step S205: When P < Pc, only send the "Drive with Caution" AIS short message and VHF voice reminder to the vessel to avoid excessive intervention;

[0259] Step S206: When Pc ≤ P < Pa, activate active collision avoidance warning, including active warning and auxiliary decision-making:

[0260] A sound and light warning is issued, with the LED red lights on the side of the bridge pier flashing rapidly (frequency 1Hz), and a high-pitched horn broadcasting the voice message "Bridge ahead, be careful and avoid it";

[0261] Assisted decision-making, recommending flight speed V via AIS message push. rec With recommended heading ψ rec At the same time, a "speed limit / direction limit" icon is overlaid on the VTS center graphical interface;

[0262] Step S207: When P≥Pa, passive protection is activated, including airbag release and video recording: The servo motor drives the reel to release the compressed air anti-collision airbag (working pressure 0.7MPa, unfolded thickness 1.2m) within 0.8s; the tracking camera (30x optical zoom, 1080p@50fps) is activated simultaneously to continuously record video of the target ship. The video stream is pushed to the data storage center via RTMP (Real-Time Messaging Protocol) and overlaid with UTC (Coordinated Universal Time) time and WGS84 (World Geodetic System 1984) coordinates for post-event evidence collection.

[0263] See Figure 5 This embodiment is used in the bridge protection method of the first embodiment to transform multi-source real-time navigation datasets into dimensionless probability values ​​P ranging from 0 to 1, providing a quantitative basis for subsequent graded protection strategies. This method can be embedded in an FPGA+ARM heterogeneous chip, with a single-frame computation time of <20ms, meeting the 1Hz real-time update requirement. The specific steps are as follows:

[0264] Step S301: Input the multi-source real-time navigation dataset.

[0265] The output standardized dataset Dk, in JSON (JavaScript Object Notation) format, is pushed to compute module 540 via the PCIe (Peripheral Component Interconnect Express) bus at a frequency of 1 Hz. Dk contains the following fields:

[0266] Vessel information, including: MMSI, length L (m), beam B (m), draft d (m), turning index K, and following index T;

[0267] Motion information, including: longitude λ, latitude φ, ground speed V (Kn), heading ψ (°), and heading θ (°);

[0268] Bridge information, including: latitude and longitude of the pier center (λ) b φ b ), pier width W b (m), Navigation aperture width W n (m);

[0269] Hydrological and meteorological data, including: water level h (m) and flow velocity V. c (m / s), wind speed V w (m / s), wind direction ψ w (°), Visibility Ve (km).

[0270] Step S302: Extract ship position, speed, course, ship dimensions, etc.

[0271] Perform coordinate transformation and feature calculation using programming:

[0272] Using the WGS84 coordinate system, perform UTM (Universal Transverse Mercator) to transform (λ,φ) into planar coordinates (x,y).

[0273] Calculate the relative distance D (m) between the ship and the nearest bridge pier, and the relative velocity vector (V). x V y );

[0274] The following values ​​are derived: DCPA (Distance to Closest Point of Approach) and TCPA (Time to Closest Point of Approach):

[0275] Calculate the effective navigable width of a ship: We = L·sin|ψ θ|+B·cos|ψ θ| is used for subsequent yaw assessment.

[0276] Step S303: Extract hydrological and meteorological data and bridge navigation dimensions data to construct a three-dimensional risk factor based on geometry, manipulation, and environment.

[0277] Performing floating-point operations yields the following results:

[0278] Calculate the geometric encounter factor λ g =exp( α·DCPA β / TCPA), where α = 0.08m -1 β=0.05s -1 Through Monte Carlo simulation calibration, λ g ∈(0,1).

[0279] Calculate the manipulation factor λ m =1 / (1+0.1·K+0.05·T), where K and T are the ship maneuverability indices, with K≈0.8 and T≈2.0 for large ships.

[0280] Calculate the environmental impact factor λ e =1 0.15·(Vw / 25) 0.10 (Vc / 3) 0.05·(10 Ve) / 10, where V w Wind speed (m / s), V c For flow velocity (m / s), V e Visibility (km)

[0281] Step S304: Construct a ship collision probability calculation model.

[0282] P=1 exp( λ), where λ = (λ) g ·λ m ·λ e )^(1 / 3), P∈[0,1].

[0283] It has been verified that when DCPA=0m and TCPA→0, P→1;

[0284] When DCPA>200m or TCPA>600s, P<0.01, which is consistent with the crew's subjective risk perception.

[0285] Step S305: Output the bridge collision probability value.

[0286] After the calculation is completed, it is written to the ARM shared DDR via the AXI4-Stream interface for the threshold comparison module to read; at the same time, it is written to the local circular buffer (depth 3600 frames, ≥1h) for post-accident backtracking analysis.

[0287] See Figure 6 This embodiment further includes an automatic evidence collection and report generation step after an incident, which can occur after a medium-risk level warning or a high-risk level warning:

[0288] When P≥Pa and the system detects a collision alarm (accelerometer ≥2g), or when a collision occurs after an active warning, the airbag is deployed and video recording begins, and the event time t0 is automatically marked.

[0289] Extract t0 The video clips from 60s to t0+180s and all multi-source data are used to generate a ZIP format evidence package.

[0290] The system uses a built-in accident analysis template to output an accident analysis report in PDF format, which includes: a GIF thumbnail of the accident process; a comparison table of vessel tracks, probability curves, and thresholds; a navigation environment data table; and preliminary liability determination suggestions.

[0291] By using SMTP (Simple Mail Transfer Protocol), reports and evidence packages are pushed to the email addresses of management departments such as maritime authorities or bridge operators, thus achieving a closed loop of "post-event evidence collection - rapid liability determination".

[0292] See Figure 7 , Figure 7 This is a structural diagram of a bridge protection device based on a combined active and passive mechanism. The bridge area monitoring unit inputs the collected navigation environment and ship traffic data of the bridge area into the active collision avoidance module. After multi-dimensional risk monitoring of over-height, yaw, overspeed, and loss of control, the data is transmitted to the risk warning unit. The risk warning unit determines whether the ship collision risk exceeds the preset risk threshold. If it does not exceed the threshold, it issues a caution driving reminder to the ship. If it does exceed the threshold, it triggers the auxiliary decision-making unit to conduct in-depth risk assessment. The auxiliary decision-making unit further determines whether the risk exceeds the preset accident threshold. If it does not exceed the threshold, it sends an active collision avoidance warning to the ship and bridge control terminal through multiple channels such as audible and visual alarms, VHF broadcasts, and AIS short messages. If it exceeds the threshold, it activates the passive collision avoidance module, sequentially completing protective preparations such as starting the air compressor, inflating the collision airbag, and calibrating the orientation controller. After overall scheduling by the collision avoidance control unit, the collision airbag is finally released by the servo motor, forming a physical buffer protection layer on the ship-facing surface of the bridge pier, realizing full-chain intelligent bridge protection.

[0293] In summary, the bridge intelligent protection method based on a combined active and passive mechanism provided in this embodiment calculates the collision probability between a ship and a bridge by integrating navigation environment data and ship traffic data in the bridge area, and executes a graded protection strategy based on the comparison results of the probability value and a preset threshold. This achieves an organic combination of active and passive collision avoidance mechanisms, effectively improves the intelligence and effectiveness of bridge collision protection, and realizes efficient prevention and control of ship-bridge collision risks in the bridge area.

[0294] To facilitate better implementation of the bridge intelligent protection method based on the active-passive combined mechanism of the embodiments of this application, the embodiments of the present invention also provide a bridge intelligent protection device based on the above-mentioned bridge intelligent protection method based on the active-passive combined mechanism. The meanings of the terms used are the same as in the above-mentioned bridge intelligent protection method based on the active-passive combined mechanism, and specific implementation details can be found in the description of the method embodiments.

[0295] Please see Figure 8 , Figure 8 The schematic diagram of the bridge intelligent protection device based on the active-passive combined mechanism provided in the embodiments of this application shows that the bridge intelligent protection device based on the active-passive combined mechanism may specifically include:

[0296] The navigation environment data acquisition module 201 is used to acquire navigation environment data based on the front-end sensing equipment in the bridge area waters.

[0297] The vessel data acquisition module 202 is used to acquire vessel traffic data in the bridge area based on the Automatic Identification System (AIS) and radar monitoring.

[0298] The data fusion module 203 is used to preprocess navigation environment data and ship passage data, and to build a multi-source data fusion model based on the preprocessed data. The multi-source data fusion model is used to perform data standardization and fusion processing to obtain a multi-source real-time navigation dataset.

[0299] The probability calculation module 204 is used to calculate the collision probability value of the ship and bridge based on the multi-source real-time navigation dataset;

[0300] The graded protection module 205 is used to compare the collision probability value of the ship-bridge with preset risk thresholds and preset accident thresholds in a graded manner, and execute the corresponding graded protection strategy according to the comparison results. The graded protection strategy includes: when the collision probability value of the ship-bridge is less than the preset risk threshold, issuing a caution driving reminder to the ship; when the collision probability value of the ship-bridge is greater than or equal to the preset risk threshold and less than the preset accident threshold, activating active collision avoidance warning and providing auxiliary driving decision information to the ship; when the collision probability value of the ship-bridge is greater than or equal to the preset accident threshold, activating the servo motor to release the passive collision airbag and activating the tracking camera to track and record the ship's video.

[0301] Specific limitations regarding the bridge intelligent protection device based on the active-passive combined mechanism can be found in the limitations of the bridge intelligent protection method based on the active-passive combined mechanism mentioned above, and will not be repeated here. Each module in the aforementioned bridge intelligent protection device based on the active-passive combined mechanism can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0302] The bridge intelligent protection device based on the active-passive joint mechanism provided in this embodiment calculates the collision probability between the bridge and the ship by integrating the navigation environment and ship traffic data in the bridge area, and executes a graded protection strategy based on the comparison results of the probability value and the preset threshold. This realizes the organic combination of active and passive anti-collision mechanisms, effectively improves the intelligence and effectiveness of bridge anti-collision protection, and achieves efficient prevention and control of the risk of ship-bridge collision in the bridge area.

[0303] Furthermore, embodiments of this application also provide an electronic device, such as... Figure 9 As shown, it illustrates a structural schematic diagram of the electronic device involved in the embodiments of this application, specifically:

[0304] The electronic device may include components such as a processor 301 with one or more processing cores, a memory 302 with one or more computer-readable storage media, a power supply 303, and an input unit 304. Those skilled in the art will understand that... Figure 9The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein:

[0305] The processor 301 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines, and performs various functions and processes data by running or executing software programs and / or modules stored in the memory 302, and by calling data stored in the memory 302, thereby providing overall monitoring of the electronic device. Optionally, the processor 301 may include one or more processing cores; preferably, the processor 301 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 301.

[0306] The memory 302 can be used to store software programs and modules. The processor 301 executes various functional applications and a bridge intelligent protection method based on a combined active and passive mechanism by running the software programs and modules stored in the memory 302. The memory 302 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 302 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 302 may also include a memory controller to provide the processor 301 with access to the memory 302.

[0307] The electronic device also includes a power supply 303 that supplies power to various components. Preferably, the power supply 303 can be logically connected to the processor 301 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 303 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0308] The electronic device may also include an input unit 304, which can be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0309] Although not shown, the electronic device may also include a display unit, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 301 in the electronic device loads the executable files corresponding to the processes of one or more applications into the memory 302 according to the following instructions, and the processor 301 runs the applications stored in the memory 302 to realize various functions, as follows:

[0310] The system acquires navigation environment data collected by front-end sensing devices in the bridge area; it also acquires vessel traffic data monitored by the Automatic Identification System (AIS) and radar in the bridge area. The navigation environment data and vessel traffic data are preprocessed, and a multi-source data fusion model is constructed based on the preprocessed data. This model is then used for data standardization and fusion to obtain a multi-source real-time navigation dataset. Based on this dataset, the probability value of a ship-bridge collision is calculated. This probability value is then compared with preset risk thresholds and preset accident thresholds, and corresponding graded protection strategies are implemented based on the comparison results. These graded protection strategies include: issuing a cautionary driving warning to the vessel when the probability value is less than the preset risk threshold; activating an active collision avoidance warning and providing assisted driving decision information to the vessel when the probability value is greater than or equal to the preset risk threshold and less than the preset accident threshold; and activating a servo motor to release passive collision airbags and activating a tracking camera to record video of the vessel when the probability value is greater than or equal to the preset accident threshold.

[0311] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0312] This application embodiment calculates the probability of ship-bridge collision by integrating bridge area navigation environment and ship traffic data, and executes a graded protection strategy based on the comparison results of probability values ​​and preset thresholds, realizing the organic combination of active and passive collision avoidance mechanisms, effectively improving the intelligence and effectiveness of bridge collision protection, and achieving efficient prevention and control of ship-bridge collision risks in bridge areas.

[0313] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0314] To this end, embodiments of this application provide a storage medium storing multiple instructions that can be loaded by a processor to execute steps in any of the bridge intelligent protection methods based on a combined active and passive mechanism provided in embodiments of this application. For example, the instructions can execute the following steps:

[0315] The system acquires navigation environment data collected by front-end sensing devices in the bridge area; it also acquires vessel traffic data monitored by the Automatic Identification System (AIS) and radar in the bridge area. The navigation environment data and vessel traffic data are preprocessed, and a multi-source data fusion model is constructed based on the preprocessed data. This model is then used for data standardization and fusion to obtain a multi-source real-time navigation dataset. Based on this dataset, the probability value of a ship-bridge collision is calculated. This probability value is then compared with preset risk thresholds and preset accident thresholds, and corresponding graded protection strategies are implemented based on the comparison results. These graded protection strategies include: issuing a cautionary driving warning to the vessel when the probability value is less than the preset risk threshold; activating an active collision avoidance warning and providing assisted driving decision information to the vessel when the probability value is greater than or equal to the preset risk threshold and less than the preset accident threshold; and activating a servo motor to release passive collision airbags and activating a tracking camera to record video of the vessel when the probability value is greater than or equal to the preset accident threshold.

[0316] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0317] The storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0318] Since the instructions stored in the storage medium can execute the steps of any of the bridge intelligent protection methods based on the active-passive joint mechanism provided in the embodiments of this application, the beneficial effects that any of the bridge intelligent protection methods based on the active-passive joint mechanism provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0319] The above provides a detailed description of a bridge intelligent protection method and device based on a combined active and passive mechanism provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A bridge intelligent protection method based on a combined active and passive mechanism, characterized in that, include: Acquire navigation environment data based on front-end sensing devices in the bridge area waters; Obtain vessel traffic data in the bridge area based on Automatic Identification System (AIS) and radar monitoring; The navigation environment data and the ship passage data are preprocessed, and a multi-source data fusion model is constructed based on the preprocessed data. The multi-source data fusion model is then used to perform data standardization and fusion processing to obtain a multi-source real-time navigation dataset. Based on the multi-source real-time navigation dataset, the collision probability value between the vessel and the bridge is calculated, including: extracting the position, speed, heading, and scale information of the vessel passing under the bridge, as well as hydrological and meteorological data of the bridge area, from the multi-source real-time navigation dataset to obtain a basic parameter set; constructing a geometric encounter factor based on the relative distance between the vessel and the bridge, the nearest encounter distance, and the time to reach the nearest point; constructing a maneuvering factor based on the vessel's turning index and following index; constructing an environmental impact factor based on wind speed, current speed, and visibility; performing geometric averaging on the geometric encounter factor, the maneuvering factor, and the environmental impact factor to obtain a comprehensive risk factor; constructing a collision probability calculation model between the vessel and the bridge based on the comprehensive risk factor; and using the collision probability calculation model to perform real-time calculation on the multi-source real-time navigation dataset to obtain the collision probability value between the vessel and the bridge. The collision probability value of the bridge is compared with a preset risk threshold and a preset accident threshold in a graded manner, and a corresponding graded protection strategy is executed according to the comparison result. The graded protection strategy includes: when the collision probability value of the bridge is less than the preset risk threshold, a caution driving reminder is issued to the vessel; when the collision probability value of the bridge is greater than or equal to the preset risk threshold and less than the preset accident threshold, an active collision avoidance warning is activated and auxiliary driving decision information is provided to the vessel; when the collision probability value of the bridge is greater than or equal to the preset accident threshold, a servo motor is activated to release the passive collision airbag and a tracking camera is activated to track and record the vessel's video.

2. The bridge intelligent protection method based on a combined active and passive mechanism according to claim 1, characterized in that, The preprocessing of the navigation environment data and the vessel passage data includes: Using the data sampling time as an index, iterate through and delete duplicate sampled data to obtain the first intermediate data; Based on the physical change characteristics and preset value range of each element data in the navigation environment data and the ship passage data, a noise reduction threshold is determined, and noise data in the first intermediate data is removed based on the noise reduction threshold to obtain the second intermediate data; The second intermediate data is subjected to dimensionality reduction processing to obtain the third intermediate data; The navigation environment data and vessel traffic data of the bridge area waters were obtained from secondary sampling, and the missing data in the third intermediate data were interpolated to fill in the gaps, so as to obtain the preprocessed standardized data.

3. The bridge intelligent protection method based on a combined active and passive mechanism according to claim 2, characterized in that, The multi-source data fusion model is constructed based on the preprocessed data, and the data is standardized and fused using the multi-source data fusion model to obtain a multi-source real-time general aviation dataset, including: Based on the preprocessed standardized data, a multi-source data fusion model is constructed; Using the multi-source data fusion model, the navigation environment data and ship passage data in the standardized data are spatiotemporally aligned to obtain a spatiotemporally aligned dataset; Based on a preset association threshold, the trajectory association and matching of ship targets from different data sources in the spatiotemporally aligned dataset are performed, and a unified target identifier is assigned to the same ship target that is successfully associated, thus obtaining the associated target set. The covariance intersection fusion algorithm is used to perform weighted fusion of multi-source data in the associated target set to generate the multi-source real-time navigation dataset.

4. The bridge intelligent protection method based on a combined active and passive mechanism according to claim 1, characterized in that, The activation of active collision avoidance warning and the provision of assisted navigation decision information to the vessel include: Active collision avoidance warning signals are sent to vessels passing under the bridge using acoustic, optical, and electronic warning equipment. Based on the multi-source real-time navigation dataset, a comprehensive analysis of the ship's current navigation status and the navigation environment in the bridge area is conducted to generate an assisted driving decision-making scheme that includes recommended speed, recommended course, and suggested control measures. The assisted driving decision-making scheme is simultaneously pushed to the corresponding ship and bridge management and control center in the form of Automatic Identification System (AIS) information, high-frequency voice, and visual display.

5. The bridge intelligent protection method based on a combined active and passive mechanism according to claim 1, characterized in that, The activation of the servo motor to release the passive anti-collision airbag and the activation of the tracking camera to record video of the ship include: A release command is sent to the anti-collision control unit, which inflates and deploys the anti-collision airbag installed on the ship-facing side of the bridge pier via a servo motor. Simultaneously activate the tracking cameras in the bridge area to continuously track and record the navigation process of vessels passing under the bridge, and upload the video data with timestamps and geographic coordinates to the data storage center in real time.

6. The bridge intelligent protection method based on a combined active and passive mechanism according to claim 1, characterized in that, The method further includes: When a bridge collision is detected, the time of the collision event is automatically marked. Extract video data and multi-source real-time navigation datasets within a preset time period before and after the collision event to generate an evidence package; Based on the evidence package, an accident analysis report is generated, and the evidence package and the accident analysis report are pushed to a preset management platform.

7. The bridge intelligent protection method based on a combined active and passive mechanism according to claim 1, characterized in that, The method further includes: The navigation environment data and the vessel passage data are connected to the edge computing gateway, and the navigation environment data and the vessel passage data are time-synchronized in the edge computing gateway to make the time deviation between the navigation environment data and the vessel passage data less than a preset deviation threshold. The time-synchronized navigation environment data and the ship passage data are stored in an edge cache. The edge cache adopts a ring cache structure to store the original data within a preset time period. When a network interruption is detected, the collected navigation environment data and ship passage data are temporarily stored in the edge cache, and the temporarily stored data will be automatically uploaded after the network is restored.

8. A bridge intelligent protection device based on a combined active and passive mechanism, characterized in that, include: The navigation environment data acquisition module is used to acquire navigation environment data collected by the front-end sensing equipment in the bridge area waters. The vessel data acquisition module is used to acquire vessel traffic data in the bridge area based on the Automatic Identification System (AIS) and radar monitoring. The data fusion module is used to preprocess the navigation environment data and the ship passage data, and to build a multi-source data fusion model based on the preprocessed data. The multi-source data fusion model is then used to perform data standardization and fusion processing to obtain a multi-source real-time navigation dataset. The probability calculation module is used to calculate the collision probability value of the ship-bridge based on the multi-source real-time navigation dataset. This includes: extracting the position, speed, heading, and scale information of the vessel crossing the bridge, as well as hydrological and meteorological data of the bridge area, from the multi-source real-time navigation dataset to obtain a basic parameter set; constructing a geometric encounter factor based on the relative distance between the vessel and the bridge, the nearest encounter distance, and the time to reach the nearest point; constructing a maneuvering factor based on the vessel's turning index and following index; constructing an environmental impact factor based on wind speed, current speed, and visibility; performing geometric averaging on the geometric encounter factor, the maneuvering factor, and the environmental impact factor to obtain a comprehensive risk factor; constructing the ship-bridge collision probability calculation model based on the comprehensive risk factor; and using the ship-bridge collision probability calculation model to perform real-time calculations on the multi-source real-time navigation dataset to obtain the ship-bridge collision probability value. The graded protection module is used to compare the bridge collision probability value with preset risk thresholds and preset accident thresholds in a graded manner, and execute corresponding graded protection strategies based on the comparison results. The graded protection strategies include: issuing a cautionary driving warning to the vessel when the bridge collision probability value is less than the preset risk threshold; activating an active collision avoidance warning and providing assisted driving decision information to the vessel when the bridge collision probability value is greater than or equal to the preset risk threshold and less than the preset accident threshold; and activating a servo motor to release passive collision airbags and activating a tracking camera to record vessel video when the bridge collision probability value is greater than or equal to the preset accident threshold.

Images