Intelligent driving control method and system for unmanned ship

Abstract

The invention discloses an intelligent driving control method and system for an unmanned ship, and belongs to the field of unmanned driving control. Firstly, the position and speed of the unmanned ship and the positions and speeds of surrounding obstacles are obtained through a sensor and a positioning device, then dynamic path tracking is carried out in combination with global path planning and local path planning, and finally the unmanned ship is driven to sail along an optimized path by controlling the course and an accelerator. Local path planning adopts an improved dynamic window method, and an unmanned ship trajectory prediction evaluation function is modified by introducing a dynamic risk evaluation layer and a rule compliance layer; the dynamic risk evaluation layer is used for designing a collision risk evaluation function to dynamically screen a speed sampling space and rejecting a speed vector conflicting with a predicted trajectory of a moving obstacle; and the rule compliance layer classifies the meeting situation, introduces a rule evaluation function to perform weight correction on the candidate velocity vector, and drives the unmanned ship to generate a course adjustment strategy conforming to the collision avoidance rule. The unmanned ship realizes full-autonomous cruise, high-precision obstacle avoidance and path tracking in a complex marine environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of unmanned control, in particular to an unmanned ship intelligent control method and system and a storage medium. BACKGROUND

[0002] With the continuous development of unmanned ship technology, unmanned ships are increasingly widely used in the fields of ocean exploration, marine transportation, environmental monitoring, and marine rescue. As an intelligent ship that does not require manual driving, the autonomous navigation and control capability of the unmanned ship is directly related to the application effect and safety. At present, the traditional unmanned ship control technology relies heavily on manual intervention, and there are certain technical bottlenecks in complex sea conditions, obstacle avoidance, and path tracking. How to achieve efficient autonomous cruising of unmanned ships in complex environments has become an important issue in current technology development.

[0003] In existing unmanned ship technology, the degree of intelligence and automation is low, especially in terms of navigation control and obstacle avoidance. The integration of traditional manual control and automated control systems is low, resulting in insufficient autonomous control performance of the unmanned ship. In particular, in high-speed operation or complex sea conditions, it is difficult to accurately avoid obstacles, or there is a large error in path tracking. In addition, due to the simple hardware architecture of the traditional control system, it is difficult to effectively respond to complex changes in the marine environment, and it also lacks intelligent sensing capabilities, making it difficult to autonomously optimize control strategies, resulting in insufficient safety and stability of the unmanned ship in actual application.

[0004] In existing technology, control systems based on a single algorithm framework are difficult to cope with multi-dimensional challenges. Traditional heading control schemes represented by PID (Proportional-Integral-Derivative) controllers have achieved basic applications, but have obvious defects under high-speed maneuvering or strong wave disturbance: the linear control model is difficult to adapt to the nonlinear ocean dynamics characteristics, resulting in response lag, severe overshoot, and frequent integral saturation, causing a sharp drop in heading stability; at the same time, the simple hardware architecture limits the sensor data fusion capability, making it difficult to generate an anti-disturbance control strategy in real time, further exacerbating the risk of instability of the ship's movement. In the field of obstacle avoidance, mainstream systems mostly use static threshold judgment or simple rule base strategies, which have inherent problems of broken perception-decision chain: lack of dynamic trajectory prediction capability for obstacles (such as ignoring relative speed and heading change trend), leading to misjudgment of avoidance opportunity; the rule engine cannot adapt to the complex scene adaptation needs of international maritime collision avoidance rules (COLREGs) (such as different scenarios of crossing, overtaking, and meeting), resulting in poor compliance of obstacle avoidance actions and redundant path detours. This fragmented system design makes the unmanned ship frequently fall into a vicious cycle of "obstacle avoidance failure-path deviation-manual takeover" in narrow channels or dense obstacle areas, seriously deviating from the goal of full autonomy.

[0005] In recent years, researchers have attempted to introduce intelligent algorithms such as fuzzy control, neural networks, and reinforcement learning to improve system adaptability. However, such solutions still face three constraints: first, the contradiction between algorithmic computational load and the limited hardware resources of the embedded unmanned vehicle, making it difficult to guarantee real-time performance; second, data-driven models rely on a large number of labeled samples for training, and the strong randomness of the marine environment makes it costly to build a reliable dataset; third, the integration of new algorithms with traditional control architectures lacks systematic design, often causing controller conflicts (such as the conflict between the sudden turn instruction of the obstacle avoidance module and the torque limit of the heading stabilization controller), thereby reducing overall reliability. The more fundamental problem is that existing technologies have failed to build an intelligent kernel that coordinates perception, decision-making, and control in a closed loop - obstacle avoidance systems lack the ability to quantitatively assess dynamic risks and hierarchically analyze rule logic, and heading control lacks a cooperative optimization mechanism for disturbance rejection and response speed, making it difficult for unmanned vehicles to achieve the symbiotic balance of "precise obstacle avoidance" and "stable tracking" in complex tasks.

[0006] Therefore, how to design an intelligent driving system that is efficient, stable, and capable of dealing with complex sea conditions and obstacles has become a key issue in the development of unmanned vehicle technology. This system should be able to automatically plan paths, accurately avoid obstacles, and control headings in real time, while having good adaptability and intelligence level, enabling autonomous cruising without human intervention. SUMMARY

[0007] To solve the problems of navigation accuracy, obstacle avoidance ability, and stability of the autonomous control system of unmanned vehicles in complex marine environments, the present invention proposes an innovative architecture with a double-level obstacle avoidance mechanism as the main engine and a cascaded PID heading control as the auxiliary power: In the obstacle avoidance dimension, a DWA algorithm is proposed that integrates a dynamic risk assessment layer (based on DCPA / TCPA collision probability modeling) and a rule compliance layer (embedded with COLREGs multi-scenario rule logic), implementing a double decision-making mechanism from "passive collision avoidance" to "active prediction and compliance avoidance", avoiding the problems of traditional DWA algorithms in terms of insufficient safety distance for dynamic obstacle avoidance and non-compliance with collision avoidance rules; In the control dimension, a cascaded PID structure of attitude angle-angle velocity is adopted, with inner loop damping compensation to suppress surge interference and outer loop heading precise tracking, forming a unity of response speed and control accuracy, enabling automatic path planning, precise obstacle avoidance, and real-time heading control, while having good adaptability and intelligence level, enabling autonomous cruising without human intervention, significantly improving safety, stability, and work efficiency, and meeting the increasingly complex application requirements.

[0008] The application provides an unmanned ship intelligent driving control method, which comprises the following steps: firstly, obtaining the position, speed and heading data of the unmanned ship and the position and speed of surrounding obstacles by sensors and positioning devices; secondly, combining global path planning and local path planning to perform dynamic path tracking; and finally, driving the unmanned ship to sail along the optimized path by controlling the heading and throttle.

[0009] In the dynamic path tracking process, while performing global path planning path tracking, the received position information of the unmanned ship itself and the position information of surrounding obstacles are used to judge whether there is a collision risk according to a collision risk evaluation mechanism; if there is no collision risk, the global path planning path tracking is continuously performed; if there is a collision risk, local path planning is performed and local path tracking is performed; the local path planning adopts an improved dynamic window method, and the trajectory prediction evaluation function of the unmanned ship is modified by introducing a dynamic risk evaluation layer and a rule compliance layer; the dynamic risk evaluation layer dynamically screens the speed sampling space for designing the collision risk evaluation function, and removes the speed vector that conflicts with the predicted trajectory of the moving obstacle; the rule compliance layer fuses the international maritime collision avoidance rules, classifies the meeting situation, and introduces a rule evaluation function to modify the weight of the candidate speed vector, so as to drive the unmanned ship to generate a heading adjustment strategy that meets the collision avoidance rules.

[0010] Further, the collision risk evaluation mechanism is specifically as follows: according to the current position vector S O ,S T and the speed vector V O ,V T of the unmanned ship and the obstacle, the relative azimuth angle vector and the relative speed vector of the unmanned ship and the obstacle are S OT =S T -S O and V OT =V T -V O , the closest encounter distance d CPA and the minimum encounter time t CPA between the unmanned ship and the dynamic obstacle are calculated:

[0011] d CPA =||S OT ||·sin(α)

[0012] t CPA =||S OT ||·cos(α) / ||V OT ||

[0013] wherein, α is the included angle of V OT to S OT .

[0014] The collision risk evaluation mechanism is as follows:

[0015]

[0016] where d min is the collision threshold; t max is the warning time; R W is the radius of the USV; R U is the radius of the obstacle; d O is the adjustable safety margin; and R safe is the radius of the USV. U O safe

[0017] Further, the local path planning is specifically as follows:

[0018] (1) USV velocity sampling

[0019] Discretize the linear velocity u and the angular velocity r in the velocity sampling space V = V d ∩ V r ∩ V a ; V r is the velocity boundary constraint, satisfying u ∈ [0, u max ], r ∈ [-r max , r max ]; V d is the acceleration constraint, satisfying |u - u c | ≤ a u Δt, |r - r c | ≤ a r Δt, u c , r c is the current velocity, a u is the maximum acceleration of the linear velocity, and a r is the maximum acceleration of the angular velocity; V a is the obstacle distance constraint, satisfying min(d o ) > R w , d o is the relative distance.

[0020] (2) USV trajectory prediction

[0021] For each group of velocities (u, r), predict a sequence of trajectory points in the future t p = 3 seconds:

[0022]

[0023] where x0, y0 are the current position of the USV; θ0 is the current heading angle of the USV; Δt is the discrete time step; i is the time step number (i = 1, 2, 3, …, k), k = t p / Δt is the total prediction step number; x k , y​​​k represents the position of the predicted trajectory point k;

[0024] (3) Unmanned Vehicle Trajectory Evaluation

[0025] According to the evaluation function, the optimal speed group is output for trajectory evaluation; the evaluation function is:

[0026]

[0027] Where, θ g is the planned next time path point direction; θ k is the heading of the speed group (u, r) at the end point k of the trajectory; 1- is the azimuth angle evaluation function; risk(u, r) is the collision risk evaluation function; colr(u, r) is the rule evaluation function; α, β, γ are all coefficients of the evaluation function.

[0028] Further, the collision risk evaluation function is:

[0029]

[0030] Further, the collision avoidance rule function is:

[0031] According to the encounter situation, four encounter scenarios are divided according to the collision avoidance rule:

[0032] (1) When the obstacle is located at the right side of the unmanned vehicle 22.5°-112.5°, it is determined as right side crossing;

[0033] (2) When the obstacle is located at the left side of the unmanned vehicle 22.5°-112.5°, it is determined as left side crossing;

[0034] (3) When the obstacle and the unmanned vehicle have a heading difference >150° and the relative azimuth angle ≤22.5°, it is determined as meeting;

[0035] (4) When the speed of the unmanned vehicle is higher than that of the obstacle and the heading difference <30°, it is determined as overtaking;

[0036] According to the four encounter scenarios, the relative azimuth angle θ T of the predicted trajectory point of the unmanned vehicle is calculated: o k o k According to the driving rule evaluation function, the avoidance strategy is executed:

[0037] (1) Right side crossing: forced right turn, through the function colr(u, r) = max(0, θ T ) to ensure θ T >0; ​​​

[0038] (2) Portal crossing: Implement a port turn to avoid or maintain course strategy, and execute colr(u,r)=-min(0,θ) T );

[0039] (3) Encounter: Trigger the two-way right turn mechanism, and guide the heading change Δθ>0 by colr(u,r)=|Δθ|;

[0040] (4) Overtaking: Forced to pass from the stern, passing through colr(u,r)=|θ T -ψ safe |Constrain the trajectory endpoint orientation (|θ) T |<π / 2,ψ safe (The compliant and safe heading angle threshold that unmanned surface vessels must maintain when chasing and overtaking obstacle vessels).

[0041] Furthermore, a PID algorithm is used to track the global and local planned paths, controlling the heading and rudder angles of the unmanned surface vessel.

[0042] The relationship between the rudder angle β and the yaw angle Δψ is shown in the following formula:

[0043]

[0044] Where Δψ(k) is the heading angle error at this moment; ∑ i＝0 Δψ(i) represents the cumulative heading angle error; Δψ(k) - Δψ(k-1) represents the difference in heading angle error between this moment and the previous moment; K P K is the proportionality coefficient. I K is the integral coefficient; D is the differential coefficient.

[0045] The PID control adopts an inner and outer dual-loop design: the outer loop PID generates rudder angle commands based on the deviation between the desired heading angle and the real-time heading angle; the inner loop PID achieves precise control of the servo position based on the difference between the rudder angle commands and the actual rudder angle feedback.

[0046] This invention provides an intelligent driving and control system for unmanned surface vessels, comprising: a heading control module, a throttle control module, a navigation module, an obstacle avoidance module, and an intelligent guidance and control module;

[0047] The heading control module is used to control the stability of the heading, respond quickly to commands, and ensure that the unmanned surface vessel follows the predetermined route.

[0048] The throttle control module is used to adjust the speed and direction of movement of the unmanned surface vessel by controlling the throttle.

[0049] The navigation module is used to acquire the position information of the unmanned surface vessel in real time, providing information for the intelligent guidance and control module to perform path planning and heading control;

[0050] The obstacle avoidance module is used for acquiring the surrounding obstacle position information in real time, and providing information for path planning and heading control of the intelligent guidance and control module;

[0051] The intelligent guidance and control module is used for receiving and analyzing the position information of the unmanned ship and the surrounding obstacle position information of the obstacle avoidance module, and performing path planning, heading tracking and obstacle avoidance.

[0052] Further, the heading control module comprises a hydraulic cylinder, an electric hydraulic pump and a rudder angle sensor; the oil inlet of the electric hydraulic pump is connected with a hydraulic oil tank and a hand pump through a hydraulic oil pipe and a three-way valve respectively, the oil outlet of the electric hydraulic pump is connected with the hydraulic cylinder through a hydraulic oil pipe to drive the piston rod to move; the hydraulic cylinder is physically connected with a rudder machine as a power source of the rudder; the power source of the rudder also comprises a hand pump, and the hand pump and the electric hydraulic pump can both drive the hydraulic cylinder; the rudder angle sensor converts the collected position data of the electric hydraulic pump into an analog voltage signal, and after being digitized by an analog acquisition module, the intelligent guidance and control module accurately adjusts the electric hydraulic pump through a pulse width modulation signal to ensure the stability of the position of the rudder machine.

[0053] Further, the throttle control module comprises an electronic throttle controller, an electronic throttle actuator and a digital-to-analog conversion module; the digital-to-analog conversion module converts the pulse width modulation signal output by the intelligent guidance and control module into an analog voltage signal, and the electronic throttle controller controls the action of the electronic throttle actuator according to the analog voltage signal; the electronic throttle actuator adjusts the rotation of the engine gear by pushing and pulling the soft shaft to control the rotation speed and steering of the propeller of the outboard engine.

[0054] Further, the heading tracking of the intelligent guidance and control module adopts an inner-outer double-loop design: an outer loop PID generates a rudder angle instruction based on the deviation between the expected heading angle and the real-time heading angle; and an inner loop PID realizes accurate control of the position of the rudder machine according to the difference between the rudder angle instruction and the actual rudder angle feedback.

[0055] The application has the following beneficial effects:

[0056] In summary, the present application provides a new unmanned vehicle intelligent driving control method, which overcomes the problems of low navigation accuracy, poor obstacle avoidance ability and insufficient stability of the control system in dynamic sea conditions in the prior art. When the global path tracking is executed, the local path planning is introduced for the collision, the local path planning deeply reconstructs the dynamic window algorithm (DWA) framework, and a double-level evaluation mechanism is created: the dynamic risk assessment layer is based on the real-time capture of the obstacle motion vector by the millimeter wave radar, accurately calculates the closest point of approach distance (DCPA) and the minimum time of closest point of approach (TCPA), dynamically selects the speed sampling space through the collision risk evaluation function, and real-time eliminates the speed vector conflicting with the predicted trajectory of the moving obstacle; the rule compliance layer deeply integrates the international maritime collision avoidance rules (COLREGs), accurately identifies the meeting, overtaking and crossing scenarios through the encounter situation classification module, and introduces the rule evaluation function to weight correct the candidate speed vector, to drive the unmanned vehicle to generate a heading adjustment strategy with physical safety and regulatory compliance.

[0057] The double-level mechanism upgrades the traditional DWA algorithm from single physical obstacle avoidance to a "physical-rule" dual safety protection system through the collaborative decision of collision risk and rule constraints, and completely solves the problem of poor dynamic obstacle avoidance effect. At the same time, the system integrates multi-dimensional functional support algorithms: the navigation module uses the Mercator projection method to realize accurate conversion of latitude and longitude to Cartesian coordinates, improving the positioning and calculation efficiency; the cascade PID heading control architecture suppresses the rudder angle oscillation through the inner and outer double loop design; the hydraulic drive rudder and multi-thread parallel processing ensure the millisecond-level response capability. Overall, the system constructs a closed loop from instruction analysis, environment perception to dynamic decision execution, enabling the unmanned vehicle to realize full autonomous cruising, high-precision obstacle avoidance and path tracking in complex marine environments, significantly improving the navigation safety, autonomy and environmental adaptability, and providing a comprehensive solution beyond the existing technology level for marine exploration and rescue missions. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 The heading control flowchart of the unmanned vehicle intelligent driving control method of the present application;

[0059] Figure 2 The program framework diagram of the unmanned vehicle intelligent driving control method of the present application;

[0060] Figure 3 The path tracking framework diagram of the unmanned vehicle intelligent driving control method of the present application;

[0061] Figure 4 The heading control structure diagram of the unmanned vehicle intelligent driving control method of the present application;

[0062] Figure 5 The encounter situation division diagram of the unmanned vehicle intelligent driving control method of the present application;

[0063] Figure 6 This is a schematic diagram of the rule evaluation function colr(u,r) of the unmanned surface vessel intelligent driving control method of the present invention. Detailed Implementation

[0064] The present invention will now be further described with reference to the accompanying drawings.

[0065] An intelligent driving control method for unmanned surface vessels (USVs) first utilizes the latitude, longitude, speed, and heading data of the USV, along with the positions and speeds of surrounding obstacles, obtained by sensors and positioning devices. Then, it combines global path planning and local path planning to perform dynamic path tracking. Finally, it drives the USV along an optimized path by controlling the heading and throttle.

[0066] During the dynamic path tracking process, while performing global path planning and path tracking, the system utilizes the received position information of the unmanned surface vessel (USV) and surrounding obstacles to determine whether there is a collision risk based on a collision risk assessment mechanism. If there is no collision risk, global path planning and path tracking continue. If there is a collision risk, local path planning and local path tracking are performed. The local path planning adopts an improved dynamic window method, which modifies the USV trajectory prediction evaluation function by introducing a dynamic risk assessment layer and a rule compliance layer. The dynamic risk assessment layer dynamically filters the velocity sampling space by designing a collision risk evaluation function, eliminating velocity vectors that conflict with the predicted trajectory of moving obstacles. The rule compliance layer integrates international maritime collision avoidance rules, classifies the encounter situation, and introduces a rule evaluation function to adjust the weights of candidate velocity vectors, driving the USV to generate a heading adjustment strategy that conforms to the collision avoidance rules.

[0067] In the above method, the collision risk assessment mechanism is specifically as follows: based on the current position vector S of the unmanned surface vessel and the obstacle... O ,S T and velocity vector V O V T The relative azimuth vector and relative velocity vector between the unmanned surface vessel and the obstacle are S. OT =S T -S O V OT =V T -V O Calculate the nearest encounter distance d between the unmanned surface vessel and dynamic obstacles. CPA and minimum meeting time t CPA :

[0068] d CPA =||S OT ||·sin(α)

[0069] t CPA =||S OT ||·cos(α) / ||V OT||

[0070] Where α is V OT To S OT The included angle;

[0071] The collision risk assessment mechanism is as follows:

[0072]

[0073] Where, d min The collision threshold; t max Warning duration; unmanned surface vessel (USV) radius R W =R U +R O +d safe R U R is the radius of the unmanned surface vessel. O d is the radius of the obstacle; safe For adjustable safety margin;

[0074] In the above method, the local path planning specifically refers to:

[0075] (1) Unmanned surface vessel speed sampling

[0076] In the velocity sampling space V = V d ∩V r ∩V a Discretized sampling linear velocity u and angular velocity r; V r For velocity boundary constraints, satisfying u∈[0,u max ], r∈[-r max ,r max ];V d For acceleration constraints, satisfy |uu c |≤a u Δt, |rr c |≤a r Δt, u c r c Given the current speed, a u It is the maximum acceleration of linear velocity, a r It is the maximum acceleration of angular velocity; V a For obstacle distance constraints, satisfy min(d) o R w d o It is a relative distance;

[0077] (2) Unmanned surface vessel trajectory prediction

[0078] Predict the future t for each set of velocities (u,r) p =A sequence of trajectory points over 3 seconds:

[0079]

[0080] Where x0 and y0 are the current positions of the unmanned surface vessel (USV); θ0 is the current heading angle of the USV; Δt is the discrete time step; i is the time step number (i = 1, 2, 3, ..., k), k = t p / Δt represents the total number of prediction steps; x k y k This indicates the position of the predicted trajectory point k;

[0081] (3) Unmanned surface vessel trajectory evaluation

[0082] The optimal velocity set is output based on the evaluation function; the evaluation function is:

[0083]

[0084] Where, θ g It is the planned path point direction at the next moment; θ k It is the bow direction of the velocity group (u,r) at the end point k of the trajectory; 1- α is the azimuth evaluation function; risk(u,r) is the collision risk evaluation function; colr(u,r) is the rule evaluation function; α, β, and γ are all coefficients of the evaluation functions.

[0085] In the above method, the collision risk assessment function is:

[0086]

[0087] In the above method, the collision avoidance rule function is:

[0088] Based on the encounter situation and collision avoidance rules, there are four encounter scenarios:

[0089] (1) When the obstacle is located at 22.5° to 112.5° on the starboard side of the unmanned surface vessel, it is determined to be a starboard crossing;

[0090] (2) When the obstacle is located at 22.5° to 112.5° on the port side of the unmanned surface vessel, it is determined to be a port side crossing;

[0091] (3) When the difference in heading between the obstacle and the unmanned surface vessel is greater than 150° and the relative azimuth angle is less than or equal to 22.5°, it is considered a head-on encounter;

[0092] (4) When the speed of the unmanned surface vessel is higher than that of the obstacle and the heading difference is less than 30°, it is judged as overtaking;

[0093] Based on four encounter scenarios, calculate the relative azimuth angle θ of the predicted trajectory of the unmanned surface vessel. T =arctan2(y o -y k ,x o -x kThe driving rule evaluation function executes the avoidance strategy:

[0094] (1) Starboard crossing: Forced right turn is compliant, via the function colr(u,r) = max(0,θ) T Ensure θ T >0;

[0095] (2) Portal crossing: Implement a port turn to avoid or maintain course strategy, and execute colr(u,r)=-min(0,θ) T );

[0096] (3) Encounter: Trigger the two-way right turn mechanism, and guide the heading change Δθ>0 by colr(u,r)=|Δθ|;

[0097] (4) Overtaking: Forced to pass from the stern, passing through colr(u,r)=|θ T -ψ safe |Constrain the trajectory endpoint orientation (|θ) T |<π / 2,ψ safe (The compliant and safe heading angle threshold that unmanned surface vessels must maintain when chasing and overtaking obstacle vessels).

[0098] In the above method, the PID algorithm is used to track the global and local planning paths and control the heading and rudder angles of the unmanned surface vessel.

[0099] The relationship between the rudder angle β and the yaw angle Δψ is shown in the following formula:

[0100]

[0101] Where Δψ(k) is the heading angle error at this moment; Σ i＝0 Δψ(i) represents the cumulative heading angle error; Δψ(k) - Δψ(k-1) represents the difference in heading angle error between this moment and the previous moment; K P K is the proportionality coefficient. I K is the integral coefficient; D is the differential coefficient.

[0102] Example 1

[0103] An intelligent control system for unmanned surface vessels (USVs) includes: a heading control module, a throttle control module, a navigation module, an obstacle avoidance module, and an intelligent guidance and control module. These modules cooperate to complete the autonomous cruising and control tasks of the USV.

[0104] The heading control module is hydraulically driven and includes a servo motor, hydraulic lines, hydraulic cylinders, an electro-hydraulic pump, a three-way valve, a manual pump, and a rudder angle sensor. The hydraulic lines of this module are connected to the hydraulic cylinders, and the electro-hydraulic pump controls the hydraulic cylinders via the hydraulic lines to change the position of the servo motor. Figure 1As shown, the inlet of the electric hydraulic pump is connected to the hydraulic tank and the manual pump via hydraulic pipes and a three-way valve, respectively. Its outlet is connected to the hydraulic cylinder via hydraulic pipes, driving the piston rod. The hydraulic pipe design allows both the manual and electric hydraulic pumps to drive the hydraulic cylinder. This hydraulic closed-loop control not only improves response speed (e.g., adjusting the rudder angle in milliseconds under wave interference) but also enhances system fault tolerance through redundant design. The rudder angle sensor converts the collected electric hydraulic pump position data into an analog voltage signal, which is then digitized by the analog acquisition module and transmitted to the intelligent guidance and control module. The pulse width modulation signal output by the intelligent guidance and control module drives the electric hydraulic pump, enabling the rudder to rotate forward and backward, ensuring precise and stable heading control.

[0105] The throttle control module includes an electronic throttle controller, an actuator, and a digital-to-analog converter (DAC). The DAC converts the pulse-width modulation (PWM) signal output from the intelligent guidance and control module into an analog voltage signal, which the electronic throttle controller uses to control the electronic throttle actuator. The electronic throttle actuator adjusts the rotation of the engine gears by pushing and pulling a flexible shaft, thereby controlling the propeller speed and direction of the outboard engine. Forward gear rotates the propeller forward, reverse gear rotates it backward, and neutral gear rotates the propeller at zero speed. The actuator drives the gear mechanism via the flexible shaft, achieving linear adjustment of power output and avoiding the lag problem of traditional mechanical systems. For example, during high-speed navigation, the system can fine-tune the throttle opening according to instructions from the host computer, optimizing energy efficiency and extending the unmanned surface vessel's endurance. This system can effectively regulate the speed and direction of motion of the unmanned surface vessel through precise throttle control.

[0106] The navigation module includes sensors and positioning equipment. It acquires real-time information such as the unmanned surface vessel's (USV) position, current heading, and speed through these sensors and equipment, and performs calculations and calibrations using high-precision positioning algorithms. This provides accurate data support for path planning and heading control. Furthermore, to further improve the system's accuracy and stability, since the navigation module calculates positioning data based on the WGS-84 coordinate system, a Mercator projection method is used for coordinate transformation. This converts the latitude and longitude information of the navigation data to a Cartesian coordinate system, improving navigation accuracy and simplifying path calculations. This module employs high-precision positioning technology to ensure the USV can accurately locate itself in complex environments and continuously track the target's heading.

[0107] The Mercator projection method for coordinate transformation is as follows:

[0108] x=(N+H)cosBcosL

[0109] y = (N + H)cosBsinL

[0110] Where L, B, and H represent longitude, latitude, and altitude in the WGS-84 coordinate system, respectively; x and y are the x-axis and y-axis coordinates of the point to be transformed in the local rectangular coordinate system; and N is the radius of curvature of the ramusoidal circle.

[0111]

[0112] Where e is the first eccentricity of the Earth; a is the semi-major axis of the Earth, taken as 6,378,137 meters; and b is the semi-minor axis of the Earth, taken as 6,356,782.3142 meters.

[0113] The obstacle avoidance module uses a high-performance millimeter-wave radar as its core sensing unit. Leveraging its all-weather operation and centimeter-level ranging accuracy, it achieves multi-dimensional real-time perception of dynamic obstacles. The system continuously scans the surrounding environment using millimeter-wave radar, collecting obstacle distance, azimuth, and relative velocity data 10 times per second. This data is then integrated with the unmanned surface vessel's (USV) navigation information to construct a dynamic nautical chart. Furthermore, a circular obstacle model is introduced to dynamically calculate the vessel's domain radius based on the USV's radius and safety margin, providing a quantitative spatial basis for obstacle avoidance decisions. The millimeter-wave radar array also collects real-time polar coordinate data of obstacles, including relative distance d. o Azimuth θ o and relative velocity vector V o =(V ox V oy The sampling frequency is 10Hz. Raw data is transmitted to the intelligent guidance and control module via serial communication, with a transmission delay of less than 50ms. Simultaneously, a spatial parameter quantization model is established based on the unmanned surface vessel's length L. u Equivalent radius R of the obstacle o Define the dynamic safety radius:

[0114]

[0115] Where, d safe This is an adjustable safety margin (default value is 2 times the length of the unmanned surface vessel), used to characterize the urgency of collision avoidance.

[0116] The navigation module and obstacle avoidance module work together in this invention to ensure that the unmanned surface vessel can accurately obtain its own position information, heading and speed, and effectively avoid obstacles according to their position and speed.

[0117] The intelligent guidance and control module, consisting of a host computer and a slave computer, is primarily responsible for receiving and parsing sensor data, and performing tasks such as path planning, heading tracking, and obstacle avoidance. The host computer receives position information from the navigation system and radar, and performs path planning, analysis, and optimization, as well as obstacle avoidance decisions and execution. The slave computer communicates with the host computer via serial port, and executes specific tasks such as heading control, rudder angle adjustment, and speed regulation according to the instructions from the host computer, ensuring that the unmanned surface vessel can efficiently execute control commands throughout its journey.

[0118] The host computer program employs a multi-threaded design, enabling it to receive data in real time while simultaneously performing control processing. The flowchart of the entire program's operation is as follows: Figure 2 As shown, the system performs global path planning based on the starting and ending point positions of the unmanned surface vessel (USV). Following the cruise path commands issued by the system, the USV tracks the planned path. Navigation data and radar perception information are then processed synchronously via parallel threads. Based on a real-time collision risk assessment mechanism, the system dynamically switches between collision avoidance mode and global path tracking mode. Both modes drive the actuators through a cascaded PID heading control algorithm. While cruising before reaching the destination, the system continuously acquires new path commands, forming an adaptive closed-loop circuit from command parsing, environmental perception, dynamic decision-making to servo / throttle execution, ultimately achieving autonomous cruising and safe obstacle avoidance from the starting point to the target point. The path tracking method optimizes path tracking performance through a cascaded PID heading control architecture. This cascaded PID control employs an inner and outer double-loop design: the outer PID generates rudder angle commands based on the deviation between the desired heading angle and the real-time heading angle; the inner PID dynamically adjusts the PWM duty cycle based on the difference between the rudder angle command and the actual rudder angle feedback to achieve precise servo position following. This hierarchical control structure effectively suppresses rudder angle oscillations caused by wind and waves, improving the anti-disturbance capability and response speed of heading control. The lower-level machine, based on the cascaded PID control commands generated by the upper-level machine, executes servo drive, rudder angle sensor data acquisition, and PWM speed regulation functions, ensuring the real-time performance and stability of the inner-loop rudder angle tracking. By adopting a multi-threaded mode, the upper-level and lower-level machines can process data streams in parallel: the upper-level thread focuses on path planning and cascaded PID calculations, while the lower-level thread responds to and executes commands in real time. The dual-loop PID control cycles are synchronized through a time-slice synchronization mechanism, thereby ensuring the system's rapid response capability to path deviations and obstacle threats under complex sea conditions, maintaining stable navigation of the unmanned surface vessel along the optimized path.

[0119] This invention optimizes heading stability through a cascaded PID algorithm, performing PID adjustments on both the heading angle and rudder angle separately. The outer loop outputs the heading control signal to the inner loop rudder angle adjuster, improving control accuracy and damping, and smoothing heading changes. The position-based PID executes control based on the difference between the actual and target positions, utilizing the proportional coefficient K. P Integral coefficient KI Differential coefficient K D Adjustments are made to enable rapid response, reduce steady-state errors and oscillations, improve adjustment speed, and shorten adjustment time.

[0120] The relationship between the rudder angle β and the yaw angle Δψ is shown in the following formula:

[0121]

[0122] Where Δψ(k) is the error at this moment, Σ i＝0 Δψ(i) represents the cumulative error, and Δψ(k) - Δψ(k-1) represents the error at this moment minus the error at the previous moment, because we have the error integral Σ. i＝0 Δψ(i), continuously accumulated, is the current output β d (k) is associated with all states in history.

[0123] During unmanned surface vessel (USV) operations, the distance between the current point and the target point needs to be calculated in real time and compared with a preset distance threshold. If the actual distance is lower than the threshold, the USV is considered to have successfully reached the predetermined target point; otherwise, it is considered not to have reached the target point, and the USV continues to move, thus achieving path tracking capability. Furthermore, it is necessary to determine whether the current target point is the last point in the path point array. If it is the last target point, the host computer will send a stop command, and the slave computer will stop operating upon receiving the command, thus completing the USV's task.

[0124] The path tracking function of the unmanned surface vessel (USV) also needs to ensure that the cruise mission is completed in a specified order. When the USV approaches the current waypoint, the system calculates and switches to the next target waypoint. This process continues until the entire cruise mission is completed. To ensure the accuracy of path tracking, a LOS guidance algorithm can be used, such as... Figure 3 As shown, once the unmanned surface vessel reaches a certain distance, it switches to the next waypoint.

[0125] During path tracking (using the LOS guidance algorithm to switch path points), the unmanned surface vessel (USV) receives real-time position information from the navigation and obstacle avoidance modules, along with information on the positions of surrounding obstacles. Based on a collision risk assessment mechanism, it determines whether a collision risk exists. If no collision risk exists, global path tracking is executed; if a collision risk exists, global path tracking is immediately interrupted, and obstacle avoidance mode is executed. The original route is resumed after the threat is eliminated and the rule evaluation function confirms compliance. A time-slice synchronization mechanism ensures seamless integration of cascaded PID heading control and obstacle avoidance decision-making, guaranteeing track continuity. The obstacle avoidance mode initiates local path planning and performs local path tracking. Local path planning, based on the DWA algorithm, introduces a two-tiered evaluation mechanism: a dynamic risk assessment layer, which accurately calculates the nearest encounter distance (DCPA) and minimum encounter time (TCPA) based on obstacle motion vectors acquired in real-time by millimeter-wave radar. The system dynamically filters the velocity sampling space through a collision risk assessment function, eliminating velocity vectors that conflict with the predicted trajectory of moving obstacles. At the rule compliance layer, it deeply integrates the International Maritime Collision Avoidance Codes (COLREGs), accurately identifying scenarios such as face-to-face encounters, overtaking, and cross-traffic collisions through an encounter situation classification module. A rule assessment function is introduced to adjust the weights of candidate velocity vectors, driving the unmanned surface vessel (USV) to generate a heading adjustment strategy that complies with collision avoidance rules (e.g., a forced right turn during a starboard cross-traffic collision, or passing behind the obstacle in an overtaking scenario). This mechanism superimposes the logical constraints of maritime regulations onto the physical constraints perceived by millimeter-wave radar, forming a safety guarantee. This two-layer mechanism is embedded in the trajectory evaluation stage of the DWA algorithm, ensuring that avoidance actions are both physically safe and legally compliant through collaborative decision-making on collision risk and rule compliance.

[0126] The collision risk assessment mechanism is as follows:

[0127] Based on the current position vector S of the unmanned surface vessel and obstacles O S T and velocity vector V O V T The relative azimuth vector and relative velocity vector between the unmanned surface vessel and the obstacle are S. OT =S T -S O V OT =V T -V O Calculate the nearest encounter distance d between the unmanned surface vessel and the obstacle. CPA and minimum meeting time t CPA :

[0128] d CPA =||S OT ||·sin(α)

[0129] t CPA =||S OT||·cos(α) / ||V OT ||

[0130] Among them, S OT V is the relative displacement vector between the boat and the obstacle; OT Let α be the relative velocity vector, and α be the angle between the two vectors.

[0131] Collision risk assessment mechanism:

[0132]

[0133] Where, d min The collision threshold; t max Warning duration; unmanned surface vessel (USV) radius R W =R U +R O +d safe R U R is the radius of the unmanned surface vessel. O d is the radius of the obstacle; safe For adjustable safety margin;

[0134] The obstacle avoidance module is triggered, and a modified dynamic window method (DWA) is used for local path planning and tracking, specifically as follows:

[0135] S1: Unmanned surface vessel speed sampling

[0136] In the velocity sampling space V = V d ∩V r ∩V a Discretized sampling linear velocity u and angular velocity r; V r For velocity boundary constraints, satisfying u∈[0,u max ], r∈[-r max ,r max ];V d For acceleration constraints, satisfy |uu c |≤a u Δt, |rr c |≤a r Δt, u c r c Given the current speed, a u It is the maximum acceleration of linear velocity, a r It is the maximum acceleration of angular velocity; V a For obstacle distance constraints, satisfy min(d) o R w d o It is a relative distance;

[0137] S2: Unmanned Surface Vessel Trajectory Planning

[0138] Predict the future t for each set of velocities (u,r)p =A sequence of trajectory points over 3 seconds:

[0139]

[0140] Where x0 and y0 are the current positions of the unmanned surface vessel (USV); θ0 is the current heading angle of the USV; Δt is the discrete time step; i is the time step number (i = 1, 2, 3, ..., k), k = t p / Δt represents the total number of prediction steps; x k y k This indicates the position of the predicted trajectory endpoint k;

[0141] (3) Unmanned surface vessel trajectory evaluation

[0142] The optimal velocity set is output based on the evaluation function; the evaluation function is:

[0143]

[0144] Where, θ g It is the planned path point direction at the next moment; θ k It is the bow direction of the velocity group (u,r) at the end point k of the trajectory; The azimuth angle evaluation function is used; the smaller the angle difference, the larger the evaluation function value. `risk(u,r)` is the collision risk evaluation function; `colr(u,r)` is the rule evaluation function; α, β, and γ are coefficients of the evaluation functions. This fusion decision-making mechanism enables the unmanned surface vessel to simultaneously meet dynamic obstacle avoidance safety and COLREGs rule compliance (e.g., 100% right turn avoidance rate in CR scenario, 100% stern-passing rate in OT scenario), significantly improving navigation stability. The execution layer generates rudder angle commands based on the output (u,r), driving the hydraulic servo system to achieve precise heading tracking.

[0145] Dynamic risk assessment layer: Generates quantifiable collision risk values

[0146]

[0147] This function represents the dynamic obstacle threat level in real time (positive values ​​indicate spatial safety margin);

[0148] Rule Compliance Layer: In the rule compliance layer, through the encounter situation classification module ( Figure 5 Based on the International Collision Avoidance Regulations (COLREGs), four encounter scenarios are precisely identified: when the obstructing vessel is 22.5°-112.5° to starboard, it is considered a starboard crossing (CR); when it is 22.5°-112.5° to port, it is considered a port crossing (CL); when the difference in heading between the obstructing vessel and the vessel is >150° and the relative bearing is ≤22.5°, it is considered a head-on encounter (HO); if the vessel's speed is higher than that of the obstructing vessel (|Vu |>|V o If the heading difference is less than 30°, it is identified as an overtaking maneuver (OT). Based on this scenario classification result, in the coordinate system of the predicted trajectory endpoint of the obstacle vessel ( Figure 6 Calculate the relative azimuth angle θ T =arctan2(y o -y k ,x o -x k The driving rule evaluation function generates differentiated avoidance strategies:

[0149] 1. CR Scenario: Enforcing right turn compliance, using the function colr(u,r) = max(0,θ) T Ensure θ T >0.

[0150] 2. OT Scenario: Forced passage from the stern, passing through colr(u,r)=|θ T -ψ safe |Constrain the trajectory endpoint orientation (|θ) T |<π / 2).

[0151] 3. CL Scenario: Implement a left-turn avoidance or direction-preserving strategy, and execute colr(u,r) = -min(0,θ). T ).

[0152] 4. HO scenario: Trigger the bidirectional right turn mechanism, guiding the heading change Δθ>0 through colr(u,r)=|Δθ|;

[0153] Because the rudder angle control method of the unmanned surface vessel (USV) in this invention is a closed-loop control method, when the target rudder angle is greater than the current rudder angle, the rudder should turn to the left; conversely, when the target rudder angle is less than the current rudder angle, the rudder should turn to the right. Given that the rudder angle of the USV is limited to a certain angle, the output of the heading control must be limited to prevent the rudder angle output from reaching saturation.

[0154] Through the above implementation scheme, this invention achieves autonomous cruising, precise path tracking, and intelligent obstacle avoidance functions for unmanned surface vessels (USVs). USVs can complete tasks such as path planning, course adjustment, and obstacle avoidance without human intervention, significantly improving the system's safety, stability, and reliability. This intelligent control system has broad application prospects and can be used in fields such as marine environmental monitoring, logistics transportation, and marine rescue, enhancing the autonomy and intelligence level of USVs.

[0155] The content of the method embodiments of the present invention is applicable to the system embodiments. The specific functions implemented in the system embodiments are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above methods.

[0156] This invention provides a storage medium storing a computer-executable program, which, when executed by a processor, is used to implement an intelligent driving control method for unmanned surface vessels.

[0157] This invention also provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform an unmanned surface vessel intelligent driving control method.

[0158] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments, and various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention. Furthermore, the embodiments of the present invention and the features thereof can be combined with each other unless otherwise specified.

Claims

1. An intelligent control method for unmanned surface vessels, characterized in that, First, the position, speed, and heading data of the unmanned surface vessel (USV) are obtained using sensors and positioning devices, along with the position and speed of surrounding obstacles. Then, dynamic path tracking is performed by combining global path planning and local path planning. Finally, the USV is driven to navigate along the optimized path by controlling the heading and throttle. During the dynamic path tracking process, while performing global path planning and path tracking, the system uses the received position information of the unmanned surface vessel itself and the position information of surrounding obstacles to determine whether there is a collision risk based on the collision risk assessment mechanism; if there is no collision risk, the system continues to perform global path planning and path tracking. If there is a risk of collision, perform local path planning and local path tracking; The local path planning adopts an improved dynamic window method, which modifies the trajectory prediction evaluation function of the unmanned surface vessel by introducing a dynamic risk assessment layer and a rule compliance layer. The dynamic risk assessment layer is designed to dynamically screen the velocity sampling space by designing a collision risk evaluation function, and eliminate velocity vectors that conflict with the predicted trajectory of moving obstacles. The rule compliance layer integrates international maritime collision avoidance rules, classifies the encounter situation, and introduces a rule evaluation function to adjust the weights of candidate velocity vectors, driving the unmanned surface vessel to generate a heading adjustment strategy that conforms to the collision avoidance rules.

2. The intelligent driving and control method for unmanned surface vessels according to claim 1, characterized in that, The collision risk assessment mechanism is specifically based on the current position vector S of the unmanned surface vessel and the obstacle. O S T and velocity vector V O V T The relative azimuth vector and relative velocity vector between the unmanned surface vessel and the obstacle are S. OT =S T -S O V OT =V T -V O Calculate the nearest encounter distance d between the unmanned surface vessel and dynamic obstacles. CPA and minimum meeting time t CPA : d CPA =‖S OT ‖·sin(a) t CPA =‖S OT ‖·cos(α) / ‖V OT ‖ Where α is V OT To S OT The included angle; The collision risk assessment mechanism is as follows: Where, d min The collision threshold; t max Warning duration; unmanned surface vessel (USV) radius R W =R U +R O +d safe R U R is the radius of the unmanned surface vessel. O d is the radius of the obstacle; safe This provides an adjustable safety margin.

3. The intelligent driving and control method for unmanned surface vessels according to claim 1, characterized in that, The local path planning specifically refers to: (1) Unmanned surface vessel speed sampling In the velocity sampling space V = V d ∩V r ∩V a Discretized sampling linear velocity u and angular velocity r; V r For velocity boundary constraints, satisfying u∈[0,u max ], r∈[-r max ,r max ];V d For acceleration constraints, satisfy |uu c |≤a u Δt, |rr c |≤a r Δt, u c r c Given the current speed, a u It is the maximum acceleration of linear velocity, a r It is the maximum acceleration of angular velocity; V a For obstacle distance constraints, satisfy min(d) o R w d o It is a relative distance; (2) Unmanned surface vessel trajectory prediction Predict the future t for each set of velocities (u,r) p =A sequence of trajectory points over 3 seconds: Where x0 and y0 are the current positions of the unmanned surface vessel (USV); θ0 is the current heading angle of the USV; Δt is the discrete time step; i is the time step number (i = 1, 2, 3, ..., k), k = t p / Δt represents the total number of prediction steps; x k y k This indicates the position of the predicted trajectory endpoint k; (3) Unmanned surface vessel trajectory evaluation The optimal velocity set is output based on the evaluation function; the evaluation function is: Where, θ g It is the planned path point direction at the next moment; θ k It is the bow direction of the velocity group (u,r) at the end point k of the trajectory; α is the azimuth evaluation function; risk(u,r) is the collision risk evaluation function; clor(u,r) is the rule evaluation function; α, β, and γ are all coefficients of the evaluation functions.

4. The intelligent driving and control method for unmanned surface vessels according to claim 3, characterized in that, The collision risk assessment function is:

5. The intelligent driving and control method for unmanned surface vessels according to claim 3, characterized in that, The collision avoidance rule function is: Based on the encounter situation and collision avoidance rules, there are four encounter scenarios: (1) When the obstacle is located at 22.5° to 112.5° on the starboard side of the unmanned surface vessel, it is determined to be a starboard crossing; (2) When the obstacle is located at 22.5° to 112.5° on the port side of the unmanned surface vessel, it is determined to be a port side crossing; (3) When the difference in heading between the obstacle and the unmanned surface vessel is greater than 150° and the relative azimuth angle is less than or equal to 22.5°, it is considered a head-on encounter; (4) When the speed of the unmanned surface vessel is higher than that of the obstacle and the heading difference is less than 30°, it is judged as overtaking; Based on four encounter scenarios, calculate the relative azimuth angle θ of the predicted trajectory of the unmanned surface vessel. T =arctan2(y o -y k ,x o -x k The driving rule evaluation function executes the avoidance strategy: (1) Starboard crossing: Forced right turn is compliant, via the function colr(u,r) = max(0,θ) T Ensure θ T >0; (2) Portal crossing: Implement a port turn to avoid or maintain course strategy, and execute colr(u,r)=-min(0,θ) T ); (3) Encounter: Trigger the two-way right turn mechanism, and guide the heading change Δθ>0 by colr(u,r)=|Δθ|; (4) Overtaking: Forced to pass from the stern, passing through colr(u,r)=|θ T -ψ safe |Constrain the trajectory endpoint orientation (|θ) T |<π / 2,ψ safe (The compliant and safe heading angle threshold that unmanned surface vessels must maintain when chasing and overtaking obstacle vessels).

6. The intelligent driving and control method for unmanned surface vessels according to claim 1, characterized in that, The PID algorithm is used to track the global and local planned paths and control the heading and rudder angles of the unmanned surface vessel. The relationship between the rudder angle β and the yaw angle Δψ is shown in the following formula: Where Δψ(k) is the heading angle error at this moment; ∑ i＝0 Δψ(i) represents the cumulative heading angle error; Δψ(k) - Δψ(k-1) represents the difference in heading angle error between this moment and the previous moment; K P K is the proportionality coefficient. I K is the integral coefficient; D is the differential coefficient.

7. An intelligent driving and control system for unmanned surface vessels, executing the intelligent driving and control method for unmanned surface vessels as described in any one of claims 1 to 6, characterized in that, include: The system includes a heading control module, a throttle control module, a navigation module, an obstacle avoidance module, and an intelligent guidance and control module. The heading control module is used to control the stability of the heading, respond quickly to commands, and ensure that the unmanned surface vessel follows the predetermined route. The throttle control module is used to adjust the speed and direction of movement of the unmanned surface vessel by controlling the throttle. The navigation module is used to acquire the position information of the unmanned surface vessel in real time, providing information for the intelligent guidance and control module to perform path planning and heading control; The obstacle avoidance module is used to acquire the location information of surrounding obstacles in real time, providing information for the intelligent guidance and control module for path planning and heading control; The intelligent guidance and control module is used to receive and parse the position information of the unmanned surface vessel from the navigation module and obstacle avoidance module, as well as the position information of surrounding obstacles, and to perform path planning, heading tracking and obstacle avoidance.

8. The unmanned surface vessel intelligent control system according to claim 7, characterized in that, The heading control module includes a hydraulic cylinder, an electric hydraulic pump, and a rudder angle sensor. The inlet of the electric hydraulic pump is connected to a hydraulic tank and a manual pump via hydraulic pipes and a three-way valve, respectively. Its outlet is connected to the hydraulic cylinder via hydraulic pipes, driving the piston rod. The hydraulic cylinder is physically connected to the rudder and serves as the power source for the rudder. The rudder's power source also includes a manual pump, and both the manual pump and the electric hydraulic pump can drive the hydraulic cylinder. The rudder angle sensor converts the collected position data of the electric hydraulic pump into an analog voltage signal. After being digitized by the analog acquisition module, the intelligent guidance and control module precisely adjusts the electric hydraulic pump through pulse width modulation signals to ensure the stability of the rudder's position.

9. The unmanned surface vessel intelligent control system according to claim 7, characterized in that, The throttle control module includes an electronic throttle controller, an electronic throttle actuator, and a digital-to-analog converter module. The digital-to-analog converter module converts the pulse width modulation signal output by the intelligent guidance and control module into an analog voltage signal. The electronic throttle controller controls the action of the electronic throttle actuator according to the analog voltage signal. The electronic throttle actuator adjusts the rotation of the engine gear by pushing and pulling the flexible shaft, thereby controlling the propeller speed and direction of the outboard engine.

10. The unmanned surface vessel intelligent control system according to claim 7, characterized in that, The intelligent guidance and control module adopts a dual-loop design for heading tracking: the outer loop PID generates rudder angle commands based on the deviation between the desired heading angle and the real-time heading angle; the inner loop PID achieves precise control of the servo motor position based on the difference between the rudder angle commands and the actual rudder angle feedback.

Images