Intelligent ship autonomous navigation system and method based on edge calculation

Abstract

The invention belongs to the technical field of ship control, and particularly relates to an intelligent ship autonomous navigation system and method based on edge calculation. Comprising a multi-source environment sensing module, an autonomous navigation decision-making module, a ship automatic control module and a man-machine interaction module. According to the invention, through a closed-loop control system constructed by the multi-source environment sensing module, the autonomous navigation decision-making module and the ship automatic control module, the whole process independence from data acquisition, model calculation, path planning to instruction execution is realized; dependence on manual observation and manual operation of sailors is reduced, fatigue misjudgment and operation errors caused by long-time monotonous operation are avoided, stable response can be achieved without relying on experience and instant response capacity of operators in complex scenes such as severe sea conditions and narrow water channels, and sailing risks are remarkably reduced.

Description

Technical Field

[0001] This invention belongs to the field of ship control technology, specifically relating to an intelligent ship autonomous navigation system and method based on edge computing. Background Technology

[0002] As the global shipping industry transforms towards automation and intelligence, maritime transport faces increasingly stringent requirements for navigation safety, economy, and autonomous operation capabilities. Traditional ship navigation heavily relies on crew members' manual lookout and control. Prolonged monotonous work can easily lead to physical and mental fatigue, resulting in a sharp increase in the risk of misjudgment and operational errors. In complex scenarios such as rough seas and narrow waterways, navigation safety relies excessively on the experience and immediate reaction capabilities of operators, which has significant limitations.

[0003] In existing technologies, some ships are equipped with autopilot systems based on simple PID control logic. Their core function is limited to maintaining course, making it difficult to cope with complex environmental interferences such as strong winds and currents, and lacking dynamic decision-making and autonomous collision avoidance capabilities. Meanwhile, although ships are now widely equipped with electronic charts (ECDIS), satellite navigation (GNSS), radar, and automatic identification systems (AIS), these devices mostly operate independently, lacking a central hardware platform capable of deeply integrating multi-source heterogeneous data, thus failing to fully realize the value of the data. Furthermore, the existing shipborne terminal hardware configuration is inadequate, unable to support complex six-degree-of-freedom ship motion model calculations, high-definition video stream hardware decoding, and real-time path planning calculations, resulting in slow system response speeds, poor situational awareness, and an inability to meet the requirements of high-precision autonomous navigation. Summary of the Invention

[0004] To address the aforementioned shortcomings of existing technologies, this invention provides an intelligent ship autonomous navigation system and method based on edge computing.

[0005] In a first aspect, the present invention provides an intelligent ship autonomous navigation system based on edge computing, comprising: The multi-source environmental sensing module is used to detect the ship's navigation status data and external environmental data in real time. The autonomous navigation decision-making module contains an industrial control computer, a data acquisition card, and an IO control card. The industrial control computer is connected to the output of the multi-source environmental perception module through the data acquisition card. Based on the industrial control computer, data fusion, six-degree-of-freedom ship motion model calculation and path planning are completed on the ship and control commands are output. The ship's automatic control module has its input end connected to the signal output end of the industrial control computer via an IO control card. The output end of the ship's automatic control module is connected to the ship's steering gear system and propulsion system to convert control commands into physical drive signals. The human-machine interaction module communicates with the autonomous navigation decision module and is used to display the navigation status and allow the crew to set navigation tasks.

[0006] Further improvements to this technical solution include a multi-source environmental perception module comprising a BeiDou / GPS dual-mode receiver, a log, and an anemometer; the BeiDou / GPS dual-mode receiver and log are used to acquire the ship's latitude and longitude coordinates, ground speed, water speed, and course information; the anemometer is used to collect real-time relative wind speed and relative wind direction data of the ship's environment; the multi-source environmental perception module is connected to the autonomous navigation decision module via a serial port or bus interface.

[0007] Further improvements to this technical solution include that the industrial control computer is equipped with a central processing unit, a running memory module, a storage module, and a graphics processing unit; the running memory module has a capacity of ≥8G, the storage module has a capacity of ≥128G, and the graphics processing unit has a video memory capacity of ≥2G.

[0008] Further improvements to this technical solution include a ship automatic control module comprising a rudder angle control converter and a main engine speed controller; the rudder angle control converter converts the digital rudder commands in the control instructions into voltage signals or PWM signals that drive the hydraulic system of the steering gear; the main engine speed controller converts the speed commands in the control instructions into on / off signals that control the throttle of the propulsion system.

[0009] Secondly, the present invention provides an intelligent ship autonomous navigation method based on edge computing, applicable to any of the aforementioned intelligent ship autonomous navigation systems based on edge computing, the method comprising: S1. Real-time collection of ship navigation status data and external environment data through multi-source environmental perception modules; S2. Preprocess the multi-source heterogeneous data collected by the multi-source environment perception module; S3. Based on the preprocessed multi-source heterogeneous data, the industrial control computer of the autonomous navigation decision module solves the six-degree-of-freedom ship motion model and obtains the real-time motion state parameters of the ship. S4. Combine motion state parameters and pre-stored electronic nautical chart data to plan the optimal navigation path for the ship and make real-time corrections. S5. Based on the planned path and the solution results of the six-degree-of-freedom ship motion model, generate rudder angle control commands and main engine speed control commands; S6. The ship's automatic control module converts control commands into physical drive signals to drive the steering gear system and propulsion system, enabling the ship to navigate autonomously.

[0010] Further improvements to this technical solution include step S1, which includes: S11. The ship's latitude and longitude coordinates are collected in real time through the Beidou / GPS dual-mode receiver of the multi-source environmental sensing module. Ground speed and track direction The data is collected and stored in binary format and marked with a corresponding collection timestamp; S12. The ship's speed relative to the water is synchronously collected by the speedometer of the multi-source environmental sensing module. Data, using the ground speed collected in step S11 With water velocity Perform data consistency verification; the verification formula is as follows: ,in A preset speed difference threshold is set, and data re-acquisition is triggered when the verification fails. S13. Collect the relative wind speed of the ship's environment using the wind speed and direction sensor of the multi-source environmental sensing module. and relative wind direction The data, combined with the track data collected in step S11, Calculate the effective wind angle of the actual ambient wind relative to the ship's sailing direction. The formula is .

[0011] Further improvements to this technical solution include step S3, which includes: S31. Extract the core input parameters from the preprocessed multi-source heterogeneous data, including the fused effective speed. Effective wind angle Relative wind speed It also calls upon pre-stored ship-specific parameters in the industrial control computer, including ship mass. Longitudinal rotational inertia Lateral rotational inertia Moment of inertia of sway and hydrodynamic derivatives; S32. Establish the six-degree-of-freedom ship motion dynamics equations, including the longitudinal motion equations, the lateral motion equations, and the heel motion equations: ; ; ; in, For longitudinal acceleration, For lateral acceleration, For bow angle acceleration; hydrodynamic derivatives include hydrodynamic derivatives related to the ship's longitudinal velocity. The square-related hydrodynamic derivative of the longitudinal velocity of a ship Hydrodynamic derivatives related to ship lateral velocity Hydrodynamic derivatives related to ship's bow angular velocity Derivative of the lateral velocity of a ship and the resulting pitching moment Derivative of the bow moment related to the ship's bow angular velocity ; The longitudinal speed of the ship; The lateral speed of the ship; The angular velocity of the ship's bow turn; the wind load factor includes the longitudinal wind load factor of the ship. Ship lateral wind load factor and ship's bow wind load coefficient ; S33. The fourth-order Runge-Kutta numerical integration method is used to solve the equations of motion for a six-degree-of-freedom ship. The integration formula is as follows: ; in, This is the integration step size; Let k be the motion state variable at time k; to The intermediate coefficients of the Runge-Kutta equation are used to obtain the real-time motion state parameters of the ship by solving for them.

[0012] Further improvements to this technical solution include step S4, which includes: S41. Retrieve pre-stored electronic chart data from the storage module of the autonomous navigation decision module, and extract the restricted area range and the center coordinates of obstacles. and coordinates of the left boundary of the channel and the right boundary coordinates of the channel Combined with the ship's current latitude and longitude calculated in step S3 Calculate the lateral distance between the ship and the channel boundary. and straight-line distance from the obstacle The formulas are as follows: ; ; in, This represents the arc length per degree of longitude of the ship's current latitude circle; This represents the arc length per degree of latitude along the ship's current longitude. S42, based on The path search algorithm constructs a navigation path planning space, starting from the origin. To the finish line With the optimization objectives of minimizing flight distance and obstacle avoidance cost, a path cost function is constructed as follows: ; in, Weighted by range; For obstacle avoidance weights; This represents the distance traveled on the current path segment. To avoid local minima where the denominator is zero, this function is used to search and generate an initial optimal navigation path. S43. Combining the real-time motion state parameters of the ship calculated in step S3 with those from step S41... Real-time path correction angle calculation : ; In step S3, the real-time motion parameters of the ship are obtained, including the bow angular velocity. drift angle ; To integrate effective cruising speed; This is the drift angle correction factor; This is the correction factor for the turning angular velocity; This is a correction factor for waterway distance; To preset the optimal lateral distance of the waterway; according to The initial path is adjusted in real time to ensure that the ship sails along the corrected optimal path.

[0013] Further improvements to this technical solution include step S5, which includes: S51. Extract the target trajectory from the optimal flight path corrected in step S4. Combined with the preset target speed set by the crew through the human-computer interaction module Calculate the current trajectory deviation and speed deviation The formula is: ; ; in, The current course of the ship calculated in step S3; S52. Introduce the real-time motion state parameters of the ship calculated in step S3, and construct the rudder angle control command model using an improved PID control algorithm: ; in, This is the rudder angle control command; , , These are the proportional, integral, and derivative control coefficients, respectively. This is the drift angle compensation coefficient; This refers to the acceleration due to the turning angle; The time integral of the trajectory deviation; S53, Combine the relative wind speed collected in step S1 and pre-stored ship inherent resistance coefficient Construct a host speed control command model: ; in, This is the main unit speed control command; The basic rotational speed for ship design; This is the speed adjustment coefficient; This is the wind resistance compensation coefficient, and the speed command is generated using this formula to adapt to changes in sailing resistance.

[0014] Further improvements to this technical solution include step S6, which includes: S61. The rudder angle control converter of the ship's automatic control module receives the rudder angle control command generated in step S5. The rudder angle range is constrained by a rudder angle safety limit formula, and the limited digital rudder command is converted into a voltage signal adapted to the hydraulic actuator of the rudder system. ; ; in, This is the rudder angle command after the limit is set; This is the minimum safe angle for the ship's rudder angle; This is the maximum safe angle for the ship's rudder. This is the converted servo drive voltage signal; Voltage conversion factor; This is the reference voltage corresponding to zero rudder angle; S62. The main engine speed controller of the ship's automatic control module receives the main engine speed control command generated in step S5. The rotational speed range is limited by a rotational speed safety constraint formula, and the constrained rotational speed command is converted into a switching signal adapted to the propulsion system throttle controller. ; ; in, The constrained rotational speed command; This is the minimum stable speed of the main unit; This is the highest speed of the main unit; The converted speed control switch signal; This is the rounding function; The number of bits for the switch signal; S63. Convert the servo drive voltage signal after step S61. The hydraulic actuator output to the servo system converts the speed control switching signal obtained in step S62. The output is sent to the throttle controller of the propulsion system, while simultaneously acquiring real-time feedback signals from the actuators, and verifying the signal validity using the formula. and Verify the effectiveness of the drive signal transmission to achieve autonomous navigation of the ship; among which, This refers to the actual drive voltage fed back by the servo system; To advance the actual rotational speed fed back by the system; This refers to the allowable threshold for voltage deviation. This is the allowable threshold for speed deviation.

[0015] The beneficial effects of this invention are as follows: This invention utilizes a closed-loop control system comprised of a multi-source environmental perception module, an autonomous navigation decision-making module, and a ship automatic control module to achieve full-process autonomy from data acquisition, model calculation, path planning to command execution. This reduces reliance on crew members' manual lookout and operation, avoids fatigue-induced misjudgments and operational errors caused by prolonged monotonous work, and enables stable responses in complex scenarios such as severe sea conditions and narrow waterways without relying on operators' experience and immediate reaction capabilities, significantly reducing navigation risks.

[0016] The multi-source environmental perception module in this invention integrates a Beidou / GPS dual-mode receiver, a log, an anemometer, and other equipment to comprehensively collect data on the ship's navigation status and external environment. Then, the industrial control computer of the autonomous navigation decision module completes the centralized fusion processing of multi-source heterogeneous data, breaking down the information barriers of traditional independent equipment operation. This allows data such as ship position, speed, and wind field to form a synergistic effect, providing reliable data support for accurate decision-making.

[0017] The autonomous navigation decision-making module of this invention is equipped with an industrial control computer with ≥8G of RAM, ≥128G of storage, and ≥2G of video memory. It has powerful edge computing capabilities and can efficiently complete high-intensity computing tasks such as solving six-degree-of-freedom ship motion models, hardware decoding of high-definition video streams, and real-time path planning. It solves the problems of slow response and poor situation display caused by insufficient computing power of traditional shipborne terminals, ensuring the accuracy of model solving and the real-time performance of path planning, and meeting the requirements of high-precision autonomous navigation.

[0018] The ship automatic control module in this invention uses a rudder angle control converter and a main engine speed controller to accurately convert digital control commands into hydraulic drive voltage signals, PWM signals, or throttle switching signals. Combined with an improved PID control algorithm and a wind resistance compensation mechanism, it achieves fine-tuning of rudder angle and speed. Compared with traditional simple PID autopilot, it can effectively cope with complex environmental interferences such as strong winds and large currents, and ensure the ship's heading and trajectory tracking accuracy in dynamic sea conditions.

[0019] This invention is based on the A* path search algorithm and real-time correction mechanism. It combines pre-stored electronic chart data, ship motion state parameters and obstacle distance information to plan the optimal path with the shortest distance and lowest collision avoidance cost. It also dynamically adjusts the navigation trajectory through real-time path correction angle, which solves the problem that traditional autopilot lacks autonomous collision avoidance and dynamic decision-making capabilities. It improves navigation efficiency while avoiding restricted areas and obstacles. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic block diagram of a system according to an embodiment of the present invention.

[0022] Figure 2 This is a schematic flowchart illustrating a method according to an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the specific embodiments. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0025] like Figure 1 As shown, this invention provides an intelligent ship autonomous navigation system based on edge computing, including a multi-source environmental perception module 110, an autonomous navigation decision module 120, a ship automatic control module 130, and a human-machine interaction module 140. The multi-source environmental perception module 110 is used to detect the ship's navigation status data and external environmental data in real time. The autonomous navigation decision module 120 internally includes an industrial control computer, a data acquisition card, and an I / O control card. The industrial control computer is connected to the output of the multi-source environmental perception module through the data acquisition card, and performs data fusion, six-degree-of-freedom ship motion model calculation, and path planning on the ship based on the industrial control computer, and outputs control commands. The ship automatic control module 130 has its input connected to the signal output of the industrial control computer through the I / O control card. The output of the ship automatic control module is connected to the ship's steering gear system and propulsion system, used to convert control commands into physical drive signals. The human-machine interaction module 140 is communicatively connected to the autonomous navigation decision module, used to display the navigation status and allow the crew to set navigation tasks.

[0026] Specifically, the multi-source environmental perception module includes a BeiDou / GPS dual-mode receiver, a log, and an anemometer; the BeiDou / GPS dual-mode receiver and the log are used to acquire the ship's latitude and longitude coordinates, ground speed, water speed, and course information; the anemometer is used to collect real-time relative wind speed and relative wind direction data of the ship's environment; the multi-source environmental perception module is connected to the autonomous navigation decision module via a serial port or a bus interface.

[0027] The BeiDou / GPS dual-mode receiver uses a dual-frequency, dual-mode receiver supporting BDS-3 / GPS L1+L5 (model example: UBLOX ZED-F9P). This receiver integrates a multipath suppression antenna, providing anti-obstruction and anti-interference capabilities. Its working principle is as follows: by receiving carrier phase and pseudorange signals from BeiDou-3 and GPS satellites, the built-in calculation chip performs positioning calculations, outputting the ship's latitude and longitude coordinates, speed above ground (SOG), and heading (COG). The data acquisition frequency is set to 10Hz, with positioning accuracy ≤ ±0.5m (static), ≤ ±1m (dynamic), speed measurement accuracy ≤ ±0.1kN, and heading measurement accuracy ≤ ±0.1°. The acquired data is stored in binary format, with each data entry containing a 32-bit timestamp (accurate to milliseconds), latitude and longitude (in degrees, retaining 6 decimal places), speed above ground (in kN, retaining 1 decimal place), and heading (in rad, retaining 4 decimal places).

[0028] The odometer used is an electromagnetic odometer (model example: Airmar DST800), installed below the waterline on the ship's hull. It calculates the ship's speed relative to the water flow (STW) by measuring the induced electromotive force generated when the water flow cuts through the electromagnetic sensor. Its working principle is as follows: the sensor emits a constant magnetic field; when the water flows through the magnetic field, it generates an induced current. The magnitude of the current is proportional to the water flow velocity (i.e., the ship's speed relative to the water flow). After signal amplification and analog-to-digital conversion, a digital signal is output. The data acquisition frequency is synchronized with the BeiDou / GPS dual-mode receiver at 10Hz, the measurement range is 0-30kΩ, the measurement accuracy is ≤±0.1kΩ, and the data format is consistent with the receiver, including a timestamp and a speed relative to the water flow field.

[0029] An ultrasonic anemometer (example model: Gill WindMaster Pro) is used, installed on the top of the ship's mast in an unobstructed area. It calculates the relative wind speed (Vw) and relative wind direction (θw) by using the time difference of sound wave propagation between the four ultrasonic probes. Its working principle is as follows: ultrasonic waves are emitted and received by two pairs of opposite probes. The superposition effect of wind on the speed of sound wave propagation is utilized, and the relative wind speed (Vw) and relative wind direction (θw) are calculated using the formula... (in Let d be the propagation time from probe i to probe j, d be the probe spacing, and c be the standard velocity of sound. The system calculates wind speed and direction by taking the angle between the probe connection line and the wind direction. Data acquisition frequency is 10Hz. Wind speed measurement range is 0-60m / s with an accuracy of ≤±0.2m / s. Wind direction measurement range is 0-360° with an accuracy of ≤±1°. Data output includes timestamp, relative wind speed (retaining one decimal place, unit m / s), and relative wind direction (retaining one decimal place, unit rad).

[0030] All three components of the multi-source environmental perception module establish communication connections with the data acquisition card of the autonomous navigation decision module through standardized interfaces, as detailed below: The Beidou / GPS dual-mode receiver uses an RS485 serial port connection with a communication baud rate of 115200bps, 8 data bits, 1 stop bit, and no parity bit. The transmission protocol follows the NMEA-0183 standard, and the core output statements are GGA (positioning information) and VTG (speed and track information). The logger uses a CAN bus connection with a communication rate of 250kbps, following the CAN 2.0B protocol. The data frame ID is set to 0x102, and the data segment includes the water speed and parity bit. The anemometer uses an Ethernet interface (RJ45) connection, supports the TCP / IP protocol, and establishes a Socket connection through a static IP (e.g., 192.168.1.103) assigned by the autonomous navigation decision module. The data transmission cycle is synchronized with the acquisition frequency at 100ms / time.

[0031] The data acquisition card used is Advantech PCI-1712, whose input terminals are compatible with the above interface types. The driver converts the heterogeneous data output by each component into 32-bit floating-point data and then transmits it to the industrial control computer for further processing, ensuring that the data transmission delay is ≤10ms and the packet loss rate is ≤0.01%.

[0032] The three components of the multi-source environmental perception module are synchronized using a unified timestamp (based on the UTC time of the BeiDou / GPS dual-mode receiver, synchronization pulses are sent to the odometer and anemometer via the CAN bus) to ensure the time consistency of the collected data. After data acquisition, hardware-level parity checking is used to remove erroneous data, and then simple format conversion is used to encapsulate data from different protocols into a standard structure of "timestamp + parameter name + parameter value + checksum," providing standardized input data for the subsequent preprocessing stage of the autonomous navigation decision module. Specifically, the effective wind angle (θw_eff) is calculated in real time by combining the COG data collected by the BeiDou / GPS dual-mode receiver and the relative wind direction (θw) collected by the anemometer, using the following formula: ; in, Effective wind angle (unit: rad); Relative wind direction (unit: rad, with 0° directly in front of the ship as 0°, increasing clockwise); The direction of the flight path (unit: rad, with true north as 0°, increasing clockwise).

[0033] In addition, industrial control computers are equipped with a central processing unit (CPU), a RAM module, a storage module, and a graphics processing unit (GPU); the RAM module has a capacity of ≥8G, the storage module has a capacity of ≥128G, and the GPU's video memory capacity is ≥2G.

[0034] The central processing unit (CPU) uses a multi-core high-performance x86 architecture CPU (supporting industrial-grade wide-temperature operating environment, temperature range -40℃~85℃). Core parameter requirements: ≥8 physical cores, ≥16 threads, base clock speed ≥2.5GHz, maximum turbo frequency ≥4.0GHz, cache capacity ≥24MB (L3 cache), and support for DDR4 / DDR5 memory controllers and PCIe 4.0 bus. Its working principle is as follows: through a multi-threaded parallel processing mechanism, it simultaneously schedules core tasks such as data fusion algorithms, six-degree-of-freedom ship motion model solutions, and path planning algorithms. It utilizes the CPU's built-in vector processing unit (AVX2) to accelerate floating-point operations and reduce the latency of solving complex mathematical models.

[0035] The RAM module uses DDR4 / DDR5 industrial-grade memory, with a single module capacity of ≥8GB, supporting dual-channel architecture (total capacity ≥16GB, with reserved expansion slots), a memory frequency of ≥3200MHz, timings ≤CL22, and an operating voltage of 1.2V (low-power design). The memory module must have ECC (Error Correction and Correction) functionality to automatically detect and correct single-bit errors, preventing system anomalies caused by data transmission errors. Its configuration is based on the following: six-DOF ship motion model calculation requires approximately 2-3GB of memory, multi-source data fusion cache requires 1-2GB, electronic chart loading requires 2-4GB, and the remaining memory is used for system processes, algorithm temporary variable storage, and redundancy backup, ensuring no memory bottleneck during multi-task parallelism.

[0036] The storage module uses an industrial-grade NVMe M.2 interface solid-state drive (SSD) with a capacity of ≥128GB (512GB recommended), read / write speeds of ≥2000MB / s (sequential read) and ≥1000MB / s (sequential write), supporting TRIM commands and power-loss protection (built-in supercapacitor to ensure data is not lost in case of accidental power failure). The storage module is functionally partitioned as follows: ① System partition (≥64GB): Installs an industrial-grade operating system (e.g., Windows 10 IoT Enterprise or Linux Ubuntu 22.04 LTS) and drivers; ② Chart data partition (≥64GB): Stores high-precision electronic chart data (including information on channels, water depth, obstacles, etc.), supporting incremental updates; ③ Log partition (≥32GB): Records navigation data, system operating status, and fault information, with a storage period of ≥90 days; ④ Algorithm partition (≥32GB): Stores six-DOF ship motion model parameters, path planning algorithm library, and control strategy program.

[0037] The graphics processing unit (GPU) uses a professional-grade industrial graphics card with ≥2GB of video memory (GDDR6 memory), ≥384 CUDA cores, ≥128-bit memory bus width, and ≥64GB / s memory bandwidth. It supports DirectX 12 and OpenGL 4.6 standards and has hardware video decoding capabilities (supporting H.264 / H.265 formats). The core functions of the GPU are: ① Accelerating hardware decoding of radar scan lines and visual perception video streams through CUDA cores, reducing CPU computational load; ② Performing hardware-accelerated rendering of electronic charts, navigation instrument data, and obstacle overlay layers to ensure the smoothness of the human-machine interface; ③ Assisting the CPU in matrix operations during multi-source data fusion, improving data processing efficiency.

[0038] All hardware components are integrated through an industrial-grade motherboard, utilizing a PCIe 4.0 bus to achieve high-speed data transfer between the CPU, GPU, and storage modules. The memory and CPU are directly connected via a dual-channel memory controller, with a data transfer latency of ≤5ns. The specific collaborative process is as follows: Data reception stage: The CPU receives heterogeneous data from the multi-source environmental perception module through the data acquisition card and temporarily stores it in the cache area of ​​the running memory; During the computation scheduling phase, the CPU allocates the data fusion task to some threads, while simultaneously activating the GPU's CUDA cores to accelerate matrix operations. The six-DOF ship motion model solution task is processed in parallel by the CPU's main core. Storage and rendering stage: The calculated motion state parameters and path planning results are written to the temporary partition of the storage module. At the same time, the GPU reads the electronic chart data and real-time calculation results to complete the rendering of the comprehensive situation interface and output it to the display of the human-computer interaction module. Redundancy protection: The ECC function of the running memory verifies data in real time, the power loss protection mechanism of the storage module ensures that critical data is not lost, and the CPU temperature monitoring module dynamically adjusts the main frequency to avoid performance throttling under high temperature environment.

[0039] In addition, the ship's automatic control module includes a rudder angle control converter and a main engine speed controller; the rudder angle control converter converts the digital rudder commands in the control instructions into voltage signals or PWM signals that drive the hydraulic system of the steering gear; the main engine speed controller converts the speed commands in the control instructions into on / off signals that control the throttle of the propulsion system.

[0040] The rudder angle control converter uses an industrial-grade high-precision D / A conversion module (example model: Advantech PCI-1723), integrating signal isolation, amplification, and calibration functions, suitable for the strong electromagnetic interference environment of ships. Its workflow is as follows: Command reception: Receives digital steering commands output by the autonomous navigation decision module via the IO control card. The command format is a 16-bit binary number, corresponding to a rudder angle range of -0.6109 rad (-35°) to 0.6109 rad (35°). Safety limit: The rudder angle range is constrained by a formula to prevent damage to the rudder hydraulic system due to overtravel. The limit formula is as follows: ; in, The rudder angle command after the limit is set (unit: rad); ; ; Signal conversion: Supports both voltage signal and PWM signal output modes, which can be manually switched via the human-machine interface module. Voltage signal conversion: The rudder angle command is converted into a 0~5V voltage signal using a linear calibration formula, which is as follows: ; in, Drive voltage (unit: V); (Voltage conversion factor); (Zero rudder angle reference voltage, corresponding to 0 rad rudder angle); PWM signal conversion: Outputs a 1kHz PWM signal with a duty cycle linearly related to the rudder angle command. The duty cycle formula is: ; in, This refers to the PWM duty cycle (range 25%~75%); the corresponding rudder angle is from... arrive Linear change; Signal output: The signal is output to the driver of the servo hydraulic system through an isolation amplifier circuit. The output impedance is ≤50Ω and the load capacity is ≥50mA to ensure stable signal transmission.

[0041] Main unit speed controller: An industrial-grade PLC module (example model: Siemens S7-1200) is selected, integrating analog input, digital output, and speed feedback acquisition functions. Its workflow is as follows: Command reception: Receives digital speed commands output by the autonomous navigation decision module via the I / O control card. The instruction format is a 16-bit binary number, corresponding to a speed range. (Minimum stable speed of the main unit, pre-stored in the industrial control computer) (The maximum speed of the main unit is pre-stored in the industrial control computer); Speed ​​constraint: The speed range is limited by a formula to prevent the main engine from over-speeding or stalling at low speeds. The constraint formula is as follows: ; in, The constrained rotational speed command (unit: r / min); Switching: The constrained speed command is converted into an N-bit switch signal (N=8~16, configurable) to control the stepper motor of the propulsion system throttle controller. The conversion formula is as follows: ; Where, is an N-bit binary switch signal; This is a rounding function to ensure a linear correspondence between the rotational speed and the switching input; Feedback acquisition: The actual rotational speed of the propulsion system is acquired through a rotational speed sensor. The feedback is sent to the autonomous navigation decision-making module for closed-loop adjustment.

[0042] Component collaborative working mechanism: The rudder angle control converter and the main engine speed controller achieve synchronous control through the IO control card. The collaborative process is as follows: Command synchronization: The autonomous navigation decision module outputs rudder angle command and speed command simultaneously. The IO control card marks the two sets of commands with the same timestamp to ensure that the actuator actions are synchronized. Signal isolation: Both integrate opto-isolation circuits with an isolation voltage ≥2500V to prevent electromagnetic interference between the servo hydraulic system and the propulsion system from being conducted to each other; Fault monitoring: Real-time monitoring of the integrity of the output signal; if the output voltage deviation of the rudder angle control converter exceeds... If the main engine speed controller fails to transmit the switch signal, it will immediately output a fault signal to the autonomous navigation decision module, triggering an audible and visual alarm and switching to manual control mode. Closed-loop calibration: The actuator feedback signal (actual servo angle of the servo motor and actual speed of the main engine) is collected every 100ms, and the control command is corrected through the autonomous navigation decision module to ensure control accuracy.

[0043] Figure 2 This is a schematic flowchart illustrating a method according to an embodiment of the present invention. Wherein, Figure 2 The implementing entity can be an intelligent ship autonomous navigation system based on edge computing. Depending on different needs, the order of the steps in this flowchart can be changed, and some can be omitted.

[0044] like Figure 2 As shown, the method includes: S1. Real-time collection of ship navigation status data and external environment data through multi-source environmental perception modules; S2. Preprocess the multi-source heterogeneous data collected by the multi-source environment perception module; S3. Based on the preprocessed multi-source heterogeneous data, the industrial control computer of the autonomous navigation decision module solves the six-degree-of-freedom ship motion model and obtains the real-time motion state parameters of the ship. S4. Combine motion state parameters and pre-stored electronic nautical chart data to plan the optimal navigation path for the ship and make real-time corrections. S5. Based on the planned path and the solution results of the six-degree-of-freedom ship motion model, generate rudder angle control commands and main engine speed control commands; S6. The ship's automatic control module converts control commands into physical drive signals to drive the steering gear system and propulsion system, enabling the ship to navigate autonomously.

[0045] To facilitate understanding of the present invention, the following description further illustrates the autonomous navigation method for intelligent ships based on edge computing, using the principle of this invention and the process of autonomous navigation of intelligent ships based on edge computing in the embodiments.

[0046] First, step S1 includes: S11. The ship's latitude and longitude coordinates are collected in real time at a frequency of 10-20Hz using a Beidou / GPS dual-mode receiver from a multi-source environmental sensing module. Ground speed and track direction The data is collected and stored in binary format and marked with a corresponding collection timestamp; S12. The ship's speed relative to the water is synchronously collected by the speedometer of the multi-source environmental sensing module. Data, using the ground speed collected in step S11 With water velocity Perform data consistency verification; the verification formula is as follows: ,in The preset speed difference threshold is set to 2kn. Data re-acquisition is triggered when the verification fails. S13. Collect the relative wind speed of the ship's environment using the wind speed and direction sensor of the multi-source environmental sensing module. and relative wind direction The data, combined with the track data collected in step S11, Calculate the effective wind angle of the actual ambient wind relative to the ship's sailing direction. The formula is .

[0047] Secondly, step S2 includes: For each type of data collected in step S1 (ground speed) , speed of water Relative wind speed Effective wind angle (etc.), perform outlier detection respectively: Data grouping: Grouped by data type (e.g., speed, wind field), each group contains continuous data within the same time window (default 1 second, i.e., 15 data points, corresponding to the 15Hz sampling frequency of S1); Statistical calculation: For each data set, calculate the mean μ and sample standard deviation σ, using the following formula: ; ; in, For the i-th collected value of a single type of data, N is the number of data points within the time window (N=15). Outlier identification and removal: If a data point satisfies If the value is abnormal, it is identified as an outlier and removed directly; at the same time, the timestamp of the abnormal data is recorded to provide a basis for subsequent troubleshooting. Additional notes: Latitude and longitude data are subject to additional rationality checks. If the data exceeds the preset navigation area (such as the coastal areas of China, 117°E-124°E, 30°N-38°N), it will be directly identified as an outlier.

[0048] For outlier removal or missing values ​​present in the original collected data (marked as...) Linear interpolation is used to complete the data and ensure the continuity of the time series. Missing data point location: finding missing data points The two most recent valid data points and Required time interval (To avoid excessive interpolation errors); Linear interpolation calculation: The complete formula is: ; in, The timestamp of the missing data point; These are the missing values ​​after completion; Handling extreme cases: If the interval between valid data points before and after a missing point exceeds 300ms, use forward padding. The system temporarily completes the task and triggers a low-priority alarm, prompting the crew to check the corresponding sensors.

[0049] Due to the differences in the units of measurement of data collected by different sensors (e.g., air speed is measured in kilometers per second, while wind speed is measured in meters per second), standardization is required to unify the data range. Normalization formula: ; in, The value is the standardized value (range 0-1). The original data (after anomaly removal and missing data completion); , For this type of data, a preset reasonable range of extreme values ​​(e.g.) of , ; of ); Basis for setting the scope: Based on the ship's design performance and the conventional range of the marine environment, which are pre-stored in the industrial control computer, adjustments can be made through the human-computer interaction module.

[0050] For speed-related ( ) and wind field type ( The data is weighted and fused separately to improve data reliability. Weight allocation principle: Assign fusion weights based on sensor acquisition accuracy. The higher the precision, the greater the weight, and the weight satisfies... (M represents the number of sensors participating in the fusion). Weight Calculation: Weights and Sensor Acquisition Accuracy They are inversely proportional, and the formula is: ; in, Let be the acquisition accuracy of the i-th sensor; Fusion Computation: Speed ​​Fusion (outputs fused effective speed) ): ; In addition, step S3 includes: S31. Extract the core input parameters from the preprocessed multi-source heterogeneous data, including the fused effective speed. Effective wind angle Relative wind speed It also calls upon pre-stored ship-specific parameters in the industrial control computer, including ship mass. Longitudinal rotational inertia Lateral rotational inertia Moment of inertia of sway and hydrodynamic derivatives; S32. Establish the six-degree-of-freedom ship motion dynamics equations, including the longitudinal motion equations, the lateral motion equations, and the heel motion equations: ; ; ; in, For longitudinal acceleration, For lateral acceleration, For bow angle acceleration; hydrodynamic derivatives include hydrodynamic derivatives related to the ship's longitudinal velocity. The square-related hydrodynamic derivative of the longitudinal velocity of a ship Hydrodynamic derivatives related to ship lateral velocity Hydrodynamic derivatives related to ship's bow angular velocity Derivative of the lateral velocity of a ship and the resulting pitching moment Derivative of the bow moment related to the ship's bow angular velocity ; This is the longitudinal speed of the ship (the positive direction is forward along the bow and stern line). This is the ship's lateral speed (positive direction is to the right along the port and starboard sides of the ship); The angular velocity of the ship's bow (clockwise around the vertical axis (z-axis) is positive); the wind load factor includes the ship's longitudinal wind load factor. Ship lateral wind load factor and ship's bow wind load coefficient ; S33. The fourth-order Runge-Kutta numerical integration method is used to solve the equations of motion for a six-degree-of-freedom ship. The integration formula is as follows: ; in, This is the integration step size; Let k be the motion state variable at time k; to The Runge-Kutta intermediate coefficients are used to obtain the ship's real-time motion parameters, including longitudinal velocity, by solving for them. lateral velocity angular velocity of rotation drift angle and attitude angle (roll angle) Pitch angle Head rocking angle ).

[0051] S31. Extraction and Recall of Core Parameters: ① Extraction of core input parameters: From the standardized data after preprocessing in step S2, the core parameters necessary for model solution are selected. The parameter details are as follows: Combined effective speed Values ​​range from 0 to 30 kN, with an accuracy of 0.01 kN, derived from S2 multi-source speed fusion results; Effective wind angle Values ​​range from 0 to 2π rad, with a precision of 0.001 rad, derived from calculation results in S13; relative wind speed Values ​​range from 0 to 60 m / s, with an accuracy of 0.01 m / s, and are data after S2 anomaly removal and standardization. Parameter format: All input parameters are double-precision floating-point type, encapsulated in the structure of "timestamp + parameter name + parameter value", and synchronized with the ship's inherent parameter calling sequence (time deviation ≤ 1ms).

[0052] ② Calling up inherent ship parameters: The pre-stored inherent ship parameters are retrieved from the industrial control computer's storage module (path: / data / ship / parameter / ). These parameters are obtained through ship design drawings or tank tests, and specifically include: Basic physical parameters: Ship mass m: unit t (tons), example value is 5000t; Moment of inertia: Longitudinal moment of inertia (unit: t) m², Example 5.2 × 10 6 ), lateral rotational inertia (unit: t) m², example 1.8 × 10 8 Vertical moment of inertia (unit: t) m², example 1.9 × 10 8 ), Roll-Hop Coupled Inertia (unit: t) m², example 1.5 × 10 5 ); Hydrodynamic derivative (unitless, based on ship maneuverability test calibration): Vertical correlation: ; Horizontal correlation: ; First lottery related: ; Roll-related: ; Related to pitch and roll: ; Related to drooping: ; Wind load factor (unitless, based on wind tunnel test calibration): .

[0053] S32. Establishment of the six-degree-of-freedom ship motion dynamics equations: Based on the theory of ship maneuvering hydrodynamics, a complete six-degree-of-freedom (longitudinal, transverse, vertical, roll, pitch, and bow) kinematic equation is established, covering all translational and rotational motion states of the ship. The equations are as follows: ① Translation equations (balance between acceleration and force): Longitudinal (x-axis, head-to-tail direction): ; Lateral (y-axis, port and starboard): ; Vertical (z-axis, up and down): ; ② Equations of rotation (balance between angular acceleration and torque): Roll (around the x-axis): ; Pitch (around the y-axis): ; First rocking (around the z-axis): ; in, These represent longitudinal, lateral, and vertical accelerations (unit: m / s²). These are the rates of change of roll angular velocity, pitch angular velocity, and tumble angular velocity (angular acceleration, unit: rad / s²). Wave interference load (unit: N or N) m), calculated in real time based on sea state data pre-stored on electronic nautical charts, using the formula: (ζ is the significant wave height in meters; ρ is the seawater density, taken as 1025 kg / m³; L is the ship length in meters;) (where is the acceleration due to gravity, taken as 9.8 m / s²).

[0054] S33, Solving using the fourth-order Runge-Kutta numerical integration method: ① Integral parameter settings: Integration step size h: 0.02s (can be adjusted by the industrial control computer according to real-time requirements, ranging from 0.01 to 0.05s) to ensure a balance between solution accuracy and system response speed; State variable vector: definition (Six-degree-of-freedom motion state variables) (Rate of change of the state variable, i.e., acceleration / angular acceleration), the integration objective is to pass Solve Changes over time.

[0055] ② Calculation of Runge-Kutta intermediate coefficients: The fourth-order Runge-Kutta method approximates the integral result using four intermediate coefficients. The calculation process is as follows: ; ; ; ; in, The timestamp at time k; The state variable value at time k; The dynamic equations established for S32 (inputs are time t and state variable y, output is the rate of change) ).

[0056] ③ State variable update and motion parameter derivation: State variable update: Calculate the state variables at time k+1 using the integral formula: ; Calculation of derived motion parameters: Drift angle β: The angle between the actual direction of the ship's motion and its longitudinal axis, expressed by the formula: (Unit: rad) Attitude angle (roll angle) (Pitch angle θ, heave angle ψ): These are obtained by integrating angular acceleration, and the formula is: (Unit: rad) actual speed The resultant speed of a ship is given by the formula: (Unit: m / s, which can be converted to kn: 1kn≈0.5144m / s).

[0057] ④ Output of solution results: After the calculation is completed, the set of real-time motion state parameters of the ship is output, including The data format is double-precision floating point, and the output frequency is synchronized with the S1 acquisition frequency (10-20Hz). It is stored in the calculation result area of ​​the industrial control computer (path: / data / calculation / state / ) for S4 and S5 to call.

[0058] Solution stability guarantee mechanism: Numerical stability verification: After every 100 integration steps, the solution results (ship kinetic energy) are verified using the law of conservation of energy. If the rate of change is ≤5%, and exceeds the threshold, the integration step size h is automatically adjusted (decreased by 0.005s) and the calculation is recalculated. Parameter boundary constraints: Set reasonable ranges for motion state parameters (e.g., roll angle). ≤±0.785rad (±45°), pitch angle θ≤±0.349rad (±20°)), to avoid solution divergence; Real-time optimization: The matrix operations in the dynamic equations are accelerated by using the CUDA core of the GPU, and the calculation time for a single set of parameters is ≤5ms, which meets the real-time requirements at a sampling frequency of 15Hz.

[0059] Next, step S4 includes: S41. Retrieve pre-stored electronic chart data from the storage module of the autonomous navigation decision module, and extract the restricted area range and the center coordinates of obstacles. and coordinates of the left boundary of the channel and the right boundary coordinates of the channel Combined with the ship's current latitude and longitude calculated in step S3 Calculate the lateral distance between the ship and the channel boundary. and straight-line distance from the obstacle The formulas are as follows: ; ; in, This represents the arc length per degree of longitude of the ship's current latitude circle; This represents the arc length per degree of latitude along the ship's current longitude. S42, based on The path search algorithm constructs a navigation path planning space, starting from the origin. To the finish line With the optimization objectives of minimizing flight distance and obstacle avoidance cost, a path cost function is constructed as follows: ; in, Weighted by range; For obstacle avoidance weights; This represents the distance traveled on the current path segment. To avoid local minima where the denominator is zero, this function is used to search and generate an initial optimal navigation path. S43. Combining the real-time motion state parameters of the ship calculated in step S3 with those from step S41... Real-time path correction angle calculation : ; In step S3, the real-time motion parameters of the ship are obtained, including the bow angular velocity. drift angle ; To integrate effective cruising speed; This is the drift angle correction factor; This is the correction factor for the turning angular velocity; This is a correction factor for waterway distance; To preset the optimal lateral distance of the waterway; according to The initial path is adjusted in real time to ensure that the ship sails along the corrected optimal path.

[0060] Additionally, step S5 includes: S51. Extract the target trajectory from the optimal flight path corrected in step S4. Combined with the preset target speed set by the crew through the human-computer interaction module Calculate the current trajectory deviation and speed deviation The formula is: ; ; in, The current course of the ship calculated in step S3; S52. Introduce the real-time motion state parameters of the ship calculated in step S3, and construct the rudder angle control command model using an improved PID control algorithm: ; in, This is the rudder angle control command; , , These are the proportional, integral, and derivative control coefficients, respectively. This is the drift angle compensation coefficient; This refers to the acceleration due to the turning angle; The time integral of the trajectory deviation; S53, Combine the relative wind speed collected in step S1 and pre-stored ship inherent resistance coefficient Construct a host speed control command model: ; in, This is the main unit speed control command; The basic rotational speed for ship design; This is the speed adjustment coefficient; This is the wind resistance compensation coefficient, and the speed command is generated using this formula to adapt to changes in sailing resistance.

[0061] Finally, step S6 includes: S61. The rudder angle control converter of the ship's automatic control module receives the rudder angle control command generated in step S5. The rudder angle range is constrained by a rudder angle safety limit formula, and the limited digital rudder command is converted into a voltage signal adapted to the hydraulic actuator of the rudder system. ; ; in, This is the rudder angle command after the limit is set; This is the minimum safe angle for the ship's rudder angle; This is the maximum safe angle for the ship's rudder. This is the converted servo drive voltage signal; Voltage conversion factor; This is the reference voltage corresponding to zero rudder angle; S62. The main engine speed controller of the ship's automatic control module receives the main engine speed control command generated in step S5. The rotational speed range is limited by a rotational speed safety constraint formula, and the constrained rotational speed command is converted into a switching signal adapted to the propulsion system throttle controller. ; ; in, The constrained rotational speed command; This is the minimum stable speed of the main unit; This is the highest speed of the main unit; The converted speed control switch signal; This is the rounding function; The number of bits for the switch signal; S63. Convert the servo drive voltage signal after step S61. The hydraulic actuator output to the servo system converts the speed control switching signal obtained in step S62. The output is sent to the throttle controller of the propulsion system, while simultaneously acquiring real-time feedback signals from the actuators, and verifying the signal validity using the formula. and Verify the effectiveness of drive signal transmission to ensure that the steering gear system adjusts the rudder angle according to instructions and the propulsion system adjusts the speed according to instructions, thereby achieving autonomous navigation of the ship; among other things... This refers to the actual drive voltage fed back by the servo system; To advance the actual rotational speed fed back by the system; This refers to the allowable threshold for voltage deviation. This is the allowable threshold for speed deviation.

[0062] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the present invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the present invention by those skilled in the art without departing from the spirit and essence of the invention, and such modifications or substitutions should all be within the scope of the present invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should also be covered within the protection scope of the present invention.

Claims

1. An intelligent ship autonomous navigation system based on edge computing, characterized in that, include: The multi-source environmental sensing module is used to detect the ship's navigation status data and external environmental data in real time. The autonomous navigation decision-making module contains an industrial control computer, a data acquisition card, and an IO control card. The industrial control computer is connected to the output of the multi-source environmental perception module through the data acquisition card. Based on the industrial control computer, data fusion, six-degree-of-freedom ship motion model calculation and path planning are completed on the ship and control commands are output. The ship's automatic control module has its input end connected to the signal output end of the industrial control computer via an IO control card. The output end of the ship's automatic control module is connected to the ship's steering gear system and propulsion system to convert control commands into physical drive signals. The human-machine interaction module communicates with the autonomous navigation decision module and is used to display the navigation status and allow the crew to set navigation tasks.

2. The intelligent ship autonomous navigation system based on edge computing according to claim 1, characterized in that, The multi-source environmental perception module includes a BeiDou / GPS dual-mode receiver, a log, and an anemometer; the BeiDou / GPS dual-mode receiver and log are used to acquire the ship's latitude and longitude coordinates, ground speed, water speed, and course information; the anemometer is used to collect real-time relative wind speed and relative wind direction data of the ship's environment; the multi-source environmental perception module is connected to the autonomous navigation decision module via serial port or bus interface.

3. The intelligent ship autonomous navigation system based on edge computing according to claim 1, characterized in that, The industrial control computer is equipped with a central processing unit, a running memory module, a storage module, and a graphics processing unit; the running memory module has a capacity of ≥8G, the storage module has a capacity of ≥128G, and the graphics processing unit has a video memory capacity of ≥2G.

4. The intelligent ship autonomous navigation system based on edge computing according to claim 1, characterized in that, The ship's automatic control module includes a rudder angle control converter and a main engine speed controller; the rudder angle control converter converts the digital rudder commands in the control instructions into voltage signals or PWM signals that drive the hydraulic system of the steering gear; the main engine speed controller converts the speed commands in the control instructions into on / off signals that control the throttle of the propulsion system.

5. A method for autonomous navigation of intelligent ships based on edge computing, characterized in that, The method applicable to the edge computing-based intelligent ship autonomous navigation system according to any one of claims 1-4 includes: S1. Real-time collection of ship navigation status data and external environment data through multi-source environmental perception modules; S2. Preprocess the multi-source heterogeneous data collected by the multi-source environment perception module; S3. Based on the preprocessed multi-source heterogeneous data, the industrial control computer of the autonomous navigation decision module solves the six-degree-of-freedom ship motion model and obtains the real-time motion state parameters of the ship. S4. Combine motion state parameters and pre-stored electronic nautical chart data to plan the optimal navigation path for the ship and make real-time corrections. S5. Based on the planned path and the solution results of the six-degree-of-freedom ship motion model, generate rudder angle control commands and main engine speed control commands; S6. The ship's automatic control module converts control commands into physical drive signals to drive the steering gear system and propulsion system, enabling the ship to navigate autonomously.

6. The intelligent ship autonomous navigation method based on edge computing according to claim 5, characterized in that, Step S1 includes: S11. The ship's latitude and longitude coordinates are collected in real time through the Beidou / GPS dual-mode receiver of the multi-source environmental sensing module. Ground speed and track direction The data is collected and stored in binary format and marked with a corresponding collection timestamp; S12. The ship's speed relative to the water is synchronously collected by the speedometer of the multi-source environmental sensing module. Data, using the ground speed collected in step S11 With water velocity Perform data consistency verification; the verification formula is as follows: ,in A preset speed difference threshold is set, and data re-acquisition is triggered when the verification fails. S13. Collect the relative wind speed of the ship's environment using the wind speed and direction sensor of the multi-source environmental sensing module. and relative wind direction The data, combined with the track data collected in step S11, Calculate the effective wind angle of the actual ambient wind relative to the ship's sailing direction. The formula is .

7. The intelligent ship autonomous navigation method based on edge computing according to claim 6, characterized in that, Step S3 includes: S31. Extract the core input parameters from the preprocessed multi-source heterogeneous data, including the fused effective speed. Effective wind angle Relative wind speed It also calls upon pre-stored ship-specific parameters in the industrial control computer, including ship mass. Longitudinal rotational inertia Lateral rotational inertia Moment of inertia of sway and hydrodynamic derivatives; S32. Establish the six-degree-of-freedom ship motion dynamics equations, including the longitudinal motion equations, the lateral motion equations, and the heel motion equations: ； ； ； in, For longitudinal acceleration, For lateral acceleration, For bow angle acceleration; hydrodynamic derivatives include hydrodynamic derivatives related to the ship's longitudinal velocity. The square-related hydrodynamic derivative of the longitudinal velocity of a ship Hydrodynamic derivatives related to ship lateral velocity Hydrodynamic derivatives related to ship's bow angular velocity Derivative of the lateral velocity of a ship and the resulting pitching moment Derivative of the bow moment related to the ship's bow angular velocity ; The longitudinal speed of the ship; The lateral speed of the ship; The angular velocity of the ship's bow turn; the wind load factor includes the longitudinal wind load factor of the ship. Ship lateral wind load factor and ship's bow wind load coefficient ; S33. The fourth-order Runge-Kutta numerical integration method is used to solve the equations of motion for a six-degree-of-freedom ship. The integration formula is as follows: ； in, This is the integration step size; Let k be the motion state variable at time k; to The intermediate coefficients of the Runge-Kutta equation are used to obtain the real-time motion state parameters of the ship by solving for them.

8. The intelligent ship autonomous navigation method based on edge computing according to claim 6, characterized in that, Step S4 includes: S41. Retrieve pre-stored electronic chart data from the storage module of the autonomous navigation decision module, and extract the restricted area range and the center coordinates of obstacles. and coordinates of the left boundary of the channel and the right boundary coordinates of the channel Combined with the ship's current latitude and longitude calculated in step S3 Calculate the lateral distance between the ship and the channel boundary. and straight-line distance from the obstacle The formulas are as follows: ； ； in, This represents the arc length per degree of longitude of the ship's current latitude circle; This represents the arc length per degree of latitude along the ship's current longitude. S42, based on The path search algorithm constructs a navigation path planning space, starting from the origin. To the finish line With the optimization objectives of minimizing flight distance and obstacle avoidance cost, a path cost function is constructed as follows: ； in, Weighted by range; For obstacle avoidance weights; This represents the distance traveled on the current path segment. To avoid local minima where the denominator is zero, this function is used to search and generate an initial optimal navigation path. S43. Combining the real-time motion state parameters of the ship calculated in step S3 with those from step S41... Real-time path correction angle calculation : ； In step S3, the real-time motion parameters of the ship are obtained, including the bow angular velocity. drift angle ; To integrate effective cruising speed; This is the drift angle correction factor; This is the correction factor for the turning angular velocity; This is a correction factor for waterway distance; To preset the optimal lateral distance of the waterway; according to The initial path is adjusted in real time to ensure that the ship sails along the corrected optimal path.

9. The intelligent ship autonomous navigation method based on edge computing according to claim 8, characterized in that, Step S5 includes: S51. Extract the target trajectory from the optimal flight path corrected in step S4. Combined with the preset target speed set by the crew through the human-computer interaction module Calculate the current trajectory deviation and speed deviation The formula is: ； ； in, The current course of the ship calculated in step S3; S52. Introduce the real-time motion state parameters of the ship calculated in step S3, and construct the rudder angle control command model using an improved PID control algorithm: ； in, This is the rudder angle control command; , , These are the proportional, integral, and derivative control coefficients, respectively. This is the drift angle compensation coefficient; This refers to the acceleration due to the turning angle; The time integral of the trajectory deviation; S53, Combine the relative wind speed collected in step S1 and pre-stored ship inherent resistance coefficient Construct a host speed control command model: ； in, This is the main unit speed control command; The basic rotational speed for ship design; This is the speed adjustment coefficient; This is the wind resistance compensation coefficient, and the speed command is generated using this formula to adapt to changes in sailing resistance.

10. The intelligent ship autonomous navigation method based on edge computing according to claim 9, characterized in that, Step S6 includes: S61. The rudder angle control converter of the ship's automatic control module receives the rudder angle control command generated in step S5. The rudder angle range is constrained by a rudder angle safety limit formula, and the limited digital rudder command is converted into a voltage signal adapted to the hydraulic actuator of the rudder system. ； ； in, This is the rudder angle command after the limit is set; This is the minimum safe angle for the ship's rudder angle; This is the maximum safe angle for the ship's rudder. This is the converted servo drive voltage signal; Voltage conversion factor; This is the reference voltage corresponding to zero rudder angle; S62. The main engine speed controller of the ship's automatic control module receives the main engine speed control command generated in step S5. The rotational speed range is limited by a rotational speed safety constraint formula, and the constrained rotational speed command is converted into a switching signal adapted to the propulsion system throttle controller. ； ； in, The constrained rotational speed command; This is the minimum stable speed of the main unit; This is the highest speed of the main unit; The converted speed control switch signal; This is the rounding function; The number of bits for the switch signal; S63. Convert the servo drive voltage signal after step S61. The hydraulic actuator output to the servo system converts the speed control switching signal obtained in step S62. The output is sent to the throttle controller of the propulsion system, while simultaneously acquiring real-time feedback signals from the actuators, and verifying the signal validity using the formula. and Verify the effectiveness of the drive signal transmission to achieve autonomous navigation of the ship; among which, This refers to the actual drive voltage fed back by the servo system; To advance the actual rotational speed fed back by the system; This refers to the allowable threshold for voltage deviation. This is the allowable threshold for speed deviation.

Images