Unmanned ship navigation state control method and system based on intelligent perception

Abstract

The invention relates to the technical field of unmanned ship navigation control, and discloses an unmanned ship navigation state control method and system based on intelligent perception. Navigation environment information of multi-dimensional data is obtained, a multi-dimensional data fusion technology is adopted for construction of an environment state matrix, and discrete environment elements are converted into structured matrix expression. In the initial path generation stage, data acquisition value evaluation is fused, obstacle avoidance requirements are considered, and a high-value detection area is actively covered. In the path correction stage, a multi-objective cost function including energy consumption, threat degree and path length is constructed, and real-time path optimization in a dynamic environment is realized. In the propulsion control link, propulsion force and rudder angle parameters are calculated by integrating the attitude, the ship speed and the ocean current speed of the unmanned ship. According to the method, the contradiction among the real-time performance of dynamic obstacle avoidance, the data acquisition efficiency and the energy consumption optimization is solved through the cooperation of global-local double-layer path planning and the control of the propulsive force and the rudder angle of the unmanned ship.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned ship navigation control, and in particular to an unmanned ship navigation state control method and system based on intelligent perception. Background Art

[0002] Digital twin technology is a digital concept and technical approach based on the integration of data and models. It accurately constructs physical objects in real time in digital space, simulating, verifying, predicting, and controlling the entire lifecycle of physical entities based on data fusion, analysis, and prediction. The development of digital twin technology offers new solutions for the integration of virtual and real waterways, real-time interaction, iterative operation and optimization, and the digital transformation and intelligent upgrade of all waterway elements. Key to this is the acquisition of data such as waterway depth measurements, sensor perception, and waterway maintenance and operation history, ultimately enabling multi-physical, multi-scale, and multi-probability simulations of waterways.

[0003] Unmanned vessels, however, possess features such as information perception, communication navigation, route planning, and autonomous navigation. They possess unparalleled advantages over traditional ships in terms of channel data collection, autonomous navigation, and positioning, providing a strong guarantee for data collection for the construction of digital twin channels. First, they can efficiently and accurately complete channel data collection tasks, providing real-time, reliable data support for the construction of digital twin channels. In addition, the unmanned vessel's real-time perception and communication capabilities enable channel data to be quickly transmitted to the digital twin platform, supporting dynamic modeling and simulation of channels and providing a scientific basis for channel management. Second, through autonomous navigation and intelligent obstacle avoidance, unmanned vessels can operate safely and stably in complex and changing channel environments, ensuring the coverage and accuracy of data collection.

[0004] Therefore, how to accurately control the navigation of unmanned ships plays an important role in the establishment of digital twin waterways. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention provides a method for controlling the navigation state of an unmanned ship based on intelligent perception, the control method comprising the following steps: S1: Acquire navigation environment information of the unmanned ship during navigation; the navigation environment information includes the position and shape of static obstacles, AIS ship dynamic information, unmanned ship posture, water depth and ocean current speed; S2: constructing an environmental state matrix according to the navigation environment information; S3: Based on the starting position, end position, static obstacle position, static obstacle shape and data collection priority of the unmanned ship, the initial navigation path of the unmanned ship is obtained through the improved weighted Voronoi diagram; S4: Based on the initial navigation path of the unmanned ship, combined with the environmental state matrix, the energy consumption and threat assessment of the unmanned ship, an improved Algorithm, obtains the corrected navigation path of the unmanned ship; S5: At the current position of the unmanned ship, the propulsion force and rudder angle of the unmanned ship are determined by integrating the unmanned ship's attitude, ship speed, and ocean current speed; S6: Control the unmanned boat to navigate along the corrected path using the propulsion force and rudder angle.

[0006] Preferably, the edge weight formula of the improved weighted Voronoi diagram is: ; Among them, d safe (e) is the safe distance from edge e to the nearest obstacle; γ is the data value weight coefficient; P(x,y) represents the data collection priority of the unmanned ship at the coordinate point (x,y).

[0007] Preferably, the data collection priority is calculated as follows: P(x,y)=α×Δh(x,y)+β×Dship(x,y); Among them, Δh(x,y) is the rate of change of the water depth of the unmanned ship at the coordinate point (x,y), D ship is the ship density, α and β are weight coefficients.

[0008] Preferably, the improvement The cost function of the algorithm is ; Among them, L path represents the path length, T k (t) represents the threat assessment, E consumption represents the energy consumption of the unmanned ship, ω1 is the path length weight coefficient, ω2 is the threat assessment weight coefficient, and ω3 is the energy consumption weight coefficient.

[0009] Preferably, the threat level assessment is calculated using the following formula: ; Among them, d k (t) represents the real-time distance between the unmanned ship and the obstacle k, v obs,k represents the speed at which the obstacle k approaches the unmanned ship, v max Indicates the maximum speed of the unmanned vessel.

[0010] Preferably, the energy consumption of the unmanned ship is calculated based on the propulsion force and ship speed of the unmanned ship.

[0011] Preferably, in S5, the MPC controller is used to integrate the posture, speed and ocean current speed of the unmanned ship to determine the propulsion force and rudder angle of the unmanned ship.

[0012] Preferably, the objective function of the MPC controller is: ; Among them, v ship Indicates the actual sailing speed of the unmanned ship, v ref represents the expected speed of the unmanned ship, F thrust represents the propulsion force output by the unmanned ship propulsion system, and λ represents the energy consumption optimization weight; The constraints of the MPC controller are: ; Among them, (x, y) represents the current coordinate position of the unmanned ship, (x ref ,y ref ) represents the desired coordinate position of the unmanned ship, and ε represents the path tracking error threshold.

[0013] Another aspect of the present invention provides an unmanned ship navigation state control system based on intelligent perception, which executes any of the above-mentioned unmanned ship navigation state control methods based on intelligent perception, and the control system includes a navigation environment information acquisition module, an environment state matrix construction module, an unmanned ship navigation path formulation module, and an unmanned ship navigation control module; The environmental information acquisition module is used to obtain navigation environment information of the unmanned ship during navigation; the navigation environment information includes the position and shape of static obstacles, AIS ship dynamic information, unmanned ship posture, water depth and ocean current speed; The environmental state matrix construction module is used to construct an environmental state matrix according to the navigation environment information; The unmanned ship navigation path formulation module is used to obtain the unmanned ship navigation initial path through the improved weighted Voronoi diagram according to the unmanned ship's starting position, end position, static obstacle position, static obstacle shape and data collection priority; based on the unmanned ship navigation initial path, combined with the environmental state matrix, the unmanned ship's energy consumption and threat level assessment, the improved The algorithm obtains the corrected navigation path of the unmanned ship; based on the current position of the unmanned ship, the ship's attitude, speed and ocean current speed are integrated to determine the propulsion force and rudder angle of the unmanned ship; The unmanned ship navigation control module is used to control the unmanned ship to navigate along the corrected path using the propulsion force and rudder angle.

[0014] The embodiments of the present invention have the following technical effects: The unmanned ship navigation state control method provided by the present invention realizes intelligent navigation in complex environments through multi-stage path planning and dynamic adjustment mechanism. First, the navigation environment information of multi-dimensional data is obtained, and the construction of the environmental state matrix adopts multi-dimensional data fusion technology to convert discrete environmental elements into structured matrix expression, providing a unified data interface for subsequent path planning. In the initial path generation stage, an improved weighted Voronoi diagram algorithm is introduced, and data collection value evaluation is integrated to enable path planning to not only consider obstacle avoidance requirements, but also actively cover high-value detection areas. In the path correction stage, an enhanced The algorithm achieves real-time path optimization in dynamic environments by constructing a multi-objective cost function that includes energy consumption, threat level, and path length. The propulsion control link accurately calculates the propulsion force and rudder angle parameters by integrating the unmanned ship's posture, ship speed, and ocean current speed. This method coordinates the propulsion force and rudder angle control of the unmanned ship through global-local two-layer path planning, solving the contradiction between the real-time dynamic obstacle avoidance, data collection efficiency, and energy consumption optimization in traditional algorithms. The global path guides the unmanned ship to cover high-value areas, local dynamic adjustments ensure safe obstacle avoidance, and energy consumption self-balancing is achieved through the help of ocean currents, ultimately improving the overall performance of the unmanned ship in complex waterways. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 The present invention provides a flow chart of an unmanned ship navigation state control method based on intelligent perception. DETAILED DESCRIPTION

[0017] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention are described clearly and completely below. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are also within the scope of protection of the present invention.

[0018] In order to accurately control the navigation of unmanned vessels, this application provides a method for controlling the navigation state of unmanned vessels based on intelligent perception. This method realizes efficient and safe navigation of unmanned vessels in dynamic waterway environments through multi-level perception, decision-making and control coordination. Its core lies in building a closed-loop system of "perception-decision-execution". Figure 1 , specifically including the following steps: S1: Acquire navigation environment information of the unmanned ship during navigation; the navigation environment information includes the position and shape of static obstacles, AIS ship dynamic information, unmanned ship posture, water depth and ocean current speed; Unmanned vessels collect real-time information about the navigation environment through multiple sensors. The location and shape of static obstacles (such as reefs) are determined through LiDAR point cloud data analysis. The real-time position, speed, and heading of dynamic obstacles (such as other ships) are acquired through the Automatic Identification System (AIS). The vessel's attitude (heading angle, acceleration) is measured by an IMU (Inertial Measurement Unit). Water depth is measured in real time using an onboard depth gauge, and current velocity is collected by a current sensor. This data is fused through spatiotemporal alignment and Kalman filtering to form a unified foundation for environmental perception.

[0019] S2: constructing an environmental state matrix according to the navigation environment information; The environmental state matrix E(t) is a digital abstraction of the navigation environment, and its structure is: ; Among them, the static obstacle position and static obstacle shape are stored in the form of coordinate sets. static ; Dynamic obstacle O dynamic Contains AIS ship dynamic information real-time position, speed and heading; water depth h(x,y) is modeled according to grid distribution; ocean current speed v current (x,y) is represented in the form of a vector field. This matrix provides dynamically updated environment input for subsequent path planning and control.

[0020] S3: Based on the starting position, end position, static obstacle position, static obstacle shape and data collection priority of the unmanned ship, the initial navigation path of the unmanned ship is obtained through the improved weighted Voronoi diagram; Based on the UAV's starting and ending positions, as well as the locations of static obstacles, an improved weighted Voronoi diagram algorithm is used to generate a global initial path—the initial navigation path for the UAV. While traditional Voronoi diagrams only consider obstacle distances to generate safe corridors, this new method incorporates data collection priority as part of the edge weights, prioritizing the path through areas with significant water depth variations or high vessel density, thus balancing data collection value with safety. Once the global path is generated, it serves as a benchmark for local dynamic adjustments.

[0021] Preferably, the edge weight formula of the improved weighted Voronoi diagram is: ; Among them, d safe(e) is the safe distance from edge e to the nearest obstacle, ensuring the UAV stays clear of obstacles; γ is the data value weight coefficient, 0.1≤γ≤0.5, which regulates the impact of P(x,y) on path selection; P(x,y) represents the data collection priority of the UAV at the coordinate point (x,y). This formula balances path safety and data collection value through a dual-weight design. For example, in areas where waterway depth fluctuates frequently, the P(x,y) value is high, and the path tends to pass through such areas to improve data collection efficiency. By dynamically adjusting γ, it can flexibly adapt to different task requirements. For example, in data collection tasks, increasing γ to 0.5 will bias the path toward high-value areas; in emergency obstacle avoidance scenarios, reducing γ to 0.1 will prioritize safety. This method significantly improves the adaptability and task-orientation of the global path.

[0022] Traditional Voronoi diagrams use only obstacle distance as weight, potentially causing routes to detour around low-value areas. This application introduces data prioritization, integrating waterway monitoring requirements into path planning, allowing unmanned vessels to maximize data value while ensuring safety. This design is particularly suitable for digital twin waterway construction scenarios, where both data coverage and navigation efficiency must be balanced.

[0023] In some embodiments, the data collection priority is calculated as follows: P(x,y)=α×Δh(x,y)+β×D ship (x,y); Among them, Δh(x,y) is the rate of change of the water depth at the coordinate point (x,y) of the unmanned ship, which reflects the degree of fluctuation of the bottom terrain of the channel and is obtained by differential calculation of historical water depth data. Areas with high rate of change may have reefs or siltation and require special monitoring. ship (x, y) represents the ship density. The frequency of ship traffic per unit area is calculated based on historical AIS data. High-density areas are typically key waterway nodes, requiring enhanced data collection to support traffic management. α and β are weighting coefficients, with 0.6 ≤ α ≤ 0.8, 0.2 ≤ β ≤ 0.4, and α + β = 1, used to balance the contributions of the two factors. For example, in a dredging operation scenario, α can be increased to focus on water depth changes, setting α = 0.8 and β = 0.2; in a port monitoring scenario, β can be increased to focus on ship dynamics, setting α = 0.6 and β = 0.4. By adjusting the weights, priority calculations can be adapted to the needs of different tasks.

[0024] This formula converts the actual needs of waterway management into quantifiable path planning parameters, enabling unmanned ships to autonomously select high-value routes and improve the pertinence and efficiency of data collection.

[0025] S4: Based on the initial navigation path of the unmanned ship, combined with the environmental state matrix, the energy consumption and threat assessment of the unmanned ship, an improved Algorithm, obtains the corrected navigation path of the unmanned ship; Based on the global path, combined with the real-time environment state matrix E(t), the improved The algorithm performs local path optimization. The algorithm quantifies the obstacle risk by threat assessment and dynamically adjusts the path in combination with the energy consumption model. The threat degree is calculated by the real-time distance and approach speed between the obstacle and the unmanned vessel, and the improved The algorithm's energy consumption model is constructed based on the relationship between propulsion force and ship speed. It ultimately outputs a corrected path to achieve the dual goals of dynamic obstacle avoidance and energy consumption optimization.

[0026] In some embodiments, the improvement The cost function of the algorithm is ; Among them, L path Indicates the path length, which directly affects the navigation time; T k (t) represents the threat assessment, which quantifies the real-time threat of all obstacles to the unmanned ship and is calculated by the obstacle distance and approach speed; E consumption The ω1 represents the energy consumption of the unmanned vessel. Based on an integral model of propulsion force and speed, it reflects the energy consumption during navigation. ω1 is the path length weighting coefficient, ω2 is the threat assessment weighting coefficient, and ω3 is the energy consumption weighting coefficient. ω1 + ω2 + ω3 = 1, which is used to adjust the priority of each target. For example, in high-threat environments, ω2 is increased to prioritize obstacle avoidance, with settings of ω1 = 0.3, ω2 = 0.5, and ω3 = 0.2. For long-distance missions, ω3 is increased to optimize energy consumption, with settings of ω1 = 0.3, ω2 = 0.2, and ω3 = 0.5.

[0027] Tradition The algorithm primarily optimizes path length, but in dynamic navigation, it must simultaneously address obstacle threats and energy constraints. This application incorporates the uncertainty of dynamic environments and the economics of missions into path decision-making by introducing threat level and energy consumption terms. The threat level assessment updates obstacle status in real time, and the energy consumption model dynamically adjusts propulsion requirements based on ocean currents, making path planning more adaptable and significantly improving the robustness of local path planning.

[0028] When replanning a local path, the algorithm traverses candidate nodes near the global path and calculates the cost value of each potential path. For example, if a path is short but close to a high-speed approaching ship (∑T k (t) is higher), its total cost may be higher than that of a slightly longer but safer path. At the same time, the algorithm predicts the energy consumption E of the path consumption If a path requires frequent upstream navigation, resulting in a surge in energy consumption, it will be suppressed by the weight ω3. Finally, the path with the minimum total cost is selected as the correction result.

[0029] In some embodiments, the threat level assessment is calculated using the following formula: ; Among them, d k (t) represents the real-time distance between the unmanned ship and the obstacle k, which is calculated in real time through LiDAR or AIS data; v obs,k represents the speed at which the obstacle k approaches the unmanned ship, which is obtained from the speed and heading vector difference provided by AIS; v max It represents the maximum speed of the unmanned ship, that is, the maximum speed allowed by the unmanned ship's power system, which is used to normalize the impact of the approach speed.

[0030] The core of threat assessment is to identify high-risk obstacles that are “close and approaching quickly”. In the formula, Gives close obstacles a higher threat value, while Reflects the motion threat of obstacles. For example, a ship approaching from the side at high speed may be identified as a high-risk target due to its high approach speed, even if the current distance is far.

[0031] Traditional threat models often ignore the movement trend of obstacles, resulting in delayed obstacle avoidance decisions. This invention introduces the approaching speed term to predict the future position of obstacles and achieve proactive obstacle avoidance. For example, a ship is moving sideways, and its v obs,k If it is a negative value, the threat level is low, and the algorithm can allow the unmanned ship to maintain the original path and avoid unnecessary detours.

[0032] During the local path planning cycle, the algorithm traverses all detected obstacles and calculates their T k (t) values ​​and sum them. For static obstacles (such as reefs), v obs,k = 0, the threat level is determined only by the distance; for dynamic obstacles (such as ships), the two act together. k When (t) exceeds a preset threshold, path replanning is triggered to ensure that the unmanned vessel avoids risks in advance. The real-time threat level of dynamic obstacles is quantified through a linear combination of obstacle distance and approach speed.

[0033] In some embodiments, the energy consumption of the unmanned vessel is calculated based on the propulsion force and speed of the unmanned vessel.

[0034] For example, the calculation of the energy consumption of the unmanned vessel is based on the dynamic relationship between propulsion force and vessel speed, and its model is: ; The energy consumption of the entire voyage is quantified by integrating the product of propulsion force and ship speed.

[0035] Propulsion force F thrust Indicates the thrust output by the propeller, calculated by the MPC controller. Ship speed v shipIndicates the actual navigation speed of the unmanned vessel, measured by integrating IMU and GPS.

[0036] The greater the propulsion force or the higher the ship speed, the more energy is consumed per unit time. However, in actual navigation, the ship speed is significantly affected by the ocean current. For example, when sailing downstream, v ship= v prop+ v current , v prop is the propeller propulsion speed), the propulsion force F required at the same ship speed thrust Smaller, thus reducing energy consumption. In the path planning stage, the algorithm predicts the energy consumption of candidate paths: first, the v of each segment is estimated based on the path length and ocean current distribution. ship , and then inversely deduce the required F through the dynamic model thrust, Finally, the total energy consumption is obtained by integrating over time. For example, a downstream path may be preferred due to its low propulsion force requirement, even though its actual length is slightly longer.

[0037] This application incorporates ocean current support into energy consumption calculations, allowing route planning to consider not only geometric length but also energy efficiency. For example, in areas with strong ocean currents, the algorithm might generate a zigzag path to maximize the proportion of downstream segments, thereby reducing overall energy consumption.

[0038] S5: At the current position of the unmanned ship, the propulsion force and rudder angle of the unmanned ship are determined by integrating the unmanned ship's attitude, ship speed, and ocean current speed; Based on the current target point of the corrected path, the propulsion force and rudder angle are calculated using an MPC (Model Predictive Control) controller. MPC comprehensively considers ship speed, current speed, and attitude parameters, dynamically adjusting power output with the optimization goals of minimizing path tracking error and minimizing propulsion energy consumption. For example, when the current direction aligns with the direction of navigation, the propulsion force is reduced to utilize the kinetic energy of the ocean current; when the current is against the current, the propulsion force is moderately increased to maintain speed. For unmanned vessels, path tracking accuracy, power system constraints, and energy consumption optimization must be met simultaneously. For example, when a sudden change in the current causes the ship's speed to deviate from expectations, MPC adjusts the propulsion force and rudder angle to return the ship to the target path, while suppressing propulsion force fluctuations to reduce energy consumption.

[0039] Exemplarily, MPC performs the following steps: State prediction: Based on current ship speed v ship 、Ocean Currents current and dynamics models to predict the trajectory of the ship several seconds into the future.

[0040] Optimization solution: Based on the criterion of minimizing the objective function, the propulsion force and rudder angle sequence in the future time domain are calculated.

[0041] Command execution: Only the first-step control command (thrust force and rudder angle) is used, and the remaining commands are used as reference.

[0042] Feedback update: Measure the actual ship position through IMU and GPS, correct the prediction model error, and enter the next cycle.

[0043] MPC incorporates environmental disturbances into control decisions through online rolling optimization, achieving a closed-loop control system of "predict-execute-correct." For example, when a lateral current is detected, MPC proactively adjusts the rudder angle to compensate for drift, rather than waiting for errors to accumulate before reacting.

[0044] In some embodiments, in S5, the MPC controller is used to integrate the posture, speed and ocean current speed of the unmanned ship to determine the propulsion force and rudder angle of the unmanned ship.

[0045] In some embodiments, the objective function of the MPC controller is: ; Among them, v ship Indicates the actual sailing speed of the unmanned ship, v ref represents the expected speed of the unmanned ship, F thrust represents the propulsion force output by the unmanned ship propulsion system, λ represents the energy consumption optimization weight, 0.01≤λ≤0.1; The constraints of the MPC controller are: ; Among them, (x, y) represents the current coordinate position of the unmanned ship, (x ref ,y ref ) represents the expected coordinate position of the unmanned ship, ε represents the path tracking error threshold, that is, the allowable hull position deviation threshold, ε=5.

[0046] In the objective function, the first term Forces the actual ship speed to track the reference speed to avoid path deviation due to speed mismatch; the second term λF 2 thrust Suppress propulsion fluctuations and reduce energy consumption. The energy consumption optimization weight λ adjusts the priority of the two: a high λ focuses on energy conservation, set to λ=0.08; a low λ focuses on tracking accuracy, set to λ=0.02.

[0047] When optimizing, MPC must satisfy both the dynamic constraints and the path error constraints. For example, if the ocean current causes the ship speed v ship Lower than v ref , MPC may moderately increase F thrust To compensate, but subject to F 2 thrust Limits are set to avoid excessive energy consumption. At the same time, the path error constraint ensures that the adjusted ship position is always within the safety corridor.

[0048] The present invention introduces energy consumption items and constraints to achieve Pareto optimality between control accuracy and energy saving. For example, in the downstream section, MPC allows the ship speed to be slightly higher than v ref This allows the use of ocean current energy while reducing propulsion output, while still satisfying the ε constraint overall.

[0049] S6: Control the unmanned boat to navigate along the corrected path using the propulsion force and rudder angle.

[0050] The propulsion force and rudder angle commands output by the MPC are sent to the UAV actuators, driving the propellers and steering gear to steer the UAV along the corrected path. Simultaneously, the navigation status is fed back to the perception layer in real time, forming a closed-loop control system.

[0051] Another aspect of the present invention provides an unmanned ship navigation state control system based on intelligent perception, which executes any of the above-mentioned unmanned ship navigation state control methods based on intelligent perception, and the control system includes a navigation environment information acquisition module, an environment state matrix construction module, an unmanned ship navigation path formulation module, and an unmanned ship navigation control module; The environmental information acquisition module is used to obtain navigation environment information of the unmanned ship during navigation; the navigation environment information includes the position and shape of static obstacles, AIS ship dynamic information, unmanned ship posture, water depth and ocean current speed; The environmental state matrix construction module is used to construct an environmental state matrix according to the navigation environment information; The unmanned ship navigation path formulation module is used to obtain the unmanned ship navigation initial path through the improved weighted Voronoi diagram according to the unmanned ship's starting position, end position, static obstacle position, static obstacle shape and data collection priority; based on the unmanned ship navigation initial path, combined with the environmental state matrix, the unmanned ship's energy consumption and threat level assessment, the improved The algorithm obtains the corrected navigation path of the unmanned ship; based on the current position of the unmanned ship, the ship's attitude, speed and ocean current speed are integrated to determine the propulsion force and rudder angle of the unmanned ship; The unmanned ship navigation control module is used to control the unmanned ship to navigate along the corrected path using the propulsion force and rudder angle.

[0052] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A method for controlling the navigation state of an unmanned ship based on intelligent perception, characterized in that: The control method comprises the following steps: S1: Acquire navigation environment information of the unmanned ship during navigation; the navigation environment information includes the position and shape of static obstacles, AIS ship dynamic information, unmanned ship posture, water depth and ocean current speed; S2: constructing an environmental state matrix according to the navigation environment information; S3: Based on the starting position, end position, static obstacle position, static obstacle shape and data collection priority of the unmanned ship, the initial navigation path of the unmanned ship is obtained through the improved weighted Voronoi diagram; S4: Based on the initial navigation path of the unmanned ship, combined with the environmental state matrix, the energy consumption and threat assessment of the unmanned ship, an improved Algorithm, obtains the corrected navigation path of the unmanned ship; S5: At the current position of the unmanned ship, the propulsion force and rudder angle of the unmanned ship are determined by integrating the unmanned ship's attitude, ship speed, and ocean current speed; S6: Control the unmanned boat to navigate along the corrected path using the propulsion force and rudder angle.

2. The method for controlling the navigation state of an unmanned ship based on intelligent perception according to claim 1, characterized in that: The edge weight formula of the improved weighted Voronoi diagram is: ； Among them, d safe (e) is the safe distance from edge e to the nearest obstacle; γ is the data value weight coefficient; P(x,y) represents the data collection priority of the unmanned ship at the coordinate point (x,y).

3. The method for controlling the navigation state of an unmanned ship based on intelligent perception according to claim 2, characterized in that: The calculation method of the data collection priority is: P(x,y)=α×Δh(x,y)+β×D ship (x,y); Among them, Δh(x,y) is the rate of change of the water depth of the unmanned ship at the coordinate point (x,y), D ship is the ship density, α and β are weight coefficients.

4. The method for controlling the navigation state of an unmanned ship based on intelligent perception according to claim 1, characterized in that: The improvements The cost function of the algorithm is: ； Among them, L path represents the path length, T k (t) represents the threat assessment, E consumption represents the energy consumption of the unmanned ship, ω1 is the path length weight coefficient, ω2 is the threat assessment weight coefficient, and ω3 is the energy consumption weight coefficient.

5. The method for controlling the navigation state of an unmanned ship based on intelligent perception according to claim 4 is characterized in that: The threat level assessment is calculated using the following formula: ； Among them, d k (t) represents the real-time distance between the unmanned ship and the obstacle k, v obs,k represents the speed at which the obstacle k approaches the unmanned ship, v max Indicates the maximum speed of the unmanned vessel.

6. The method for controlling the navigation state of an unmanned ship based on intelligent perception according to claim 1, characterized in that: The energy consumption of the unmanned ship is calculated based on the propulsion force and ship speed of the unmanned ship.

7. The method for controlling the navigation state of an unmanned ship based on intelligent perception according to claim 1, characterized in that: In S5, the MPC controller is used to integrate the unmanned ship's attitude, ship speed, and ocean current speed to determine the propulsion force and rudder angle of the unmanned ship.

8. The method for controlling the navigation state of an unmanned ship based on intelligent perception according to claim 7, characterized in that: The objective function of the MPC controller is: ； Among them, v ship Indicates the actual sailing speed of the unmanned ship, v ref represents the expected speed of the unmanned ship, F thrust represents the propulsion force output by the unmanned ship propulsion system, and λ represents the energy consumption optimization weight; The constraints of the MPC controller are: ； Among them, (x, y) represents the current coordinate position of the unmanned ship, (x ref ,y ref ) represents the desired coordinate position of the unmanned ship, and ε represents the path tracking error threshold.

9. An unmanned ship navigation state control system based on intelligent perception, characterized in that: Execute the unmanned ship navigation state control method based on intelligent perception according to any one of claims 1 to 8, the control system includes a navigation environment information acquisition module, an environment state matrix construction module, an unmanned ship navigation path formulation module, and an unmanned ship navigation control module; The environmental information acquisition module is used to obtain navigation environment information of the unmanned ship during navigation; the navigation environment information includes the position and shape of static obstacles, AIS ship dynamic information, unmanned ship posture, water depth and ocean current speed; The environmental state matrix construction module is used to construct an environmental state matrix according to the navigation environment information; The unmanned ship navigation path formulation module is used to obtain the unmanned ship's initial navigation path through an improved weighted Voronoi diagram based on the unmanned ship's starting position, end position, static obstacle position, static obstacle shape and data collection priority; Based on the initial navigation path of the unmanned ship, combined with the environmental state matrix, the energy consumption and threat assessment of the unmanned ship, the improved The algorithm obtains the corrected navigation path of the unmanned ship; based on the current position of the unmanned ship, the ship's attitude, speed and ocean current speed are integrated to determine the propulsion force and rudder angle of the unmanned ship; The unmanned ship navigation control module is used to control the unmanned ship to navigate along the corrected path using the propulsion force and rudder angle.

Images