Fuzzy control method and system for unmanned ship safety autonomous berthing and leaving

Abstract

This invention relates to the field of unmanned vessel berthing and unberthing technology, specifically a fuzzy control method and system for the safe and autonomous berthing and unberthing of unmanned vessels, comprising the following steps: real-time scanning of the surrounding environment using a 3D lidar mounted on the unmanned vessel to generate point cloud data; and determining the desired longitudinal velocity u of the unmanned vessel's vector motion based on the relative pose. d Desired lateral velocity v d and desired turning angular velocity r d The system error and error change rate are input into a fuzzy controller for fuzzification. The control output is decomposed into thrust commands for each thruster of the unmanned surface vessel (USV), controlling the USV to achieve coordinated longitudinal, lateral, and bow-turning movements, enabling the USV to autonomously berth and unberth in an attitude parallel to the shore. This invention significantly improves the control accuracy and adaptability of USVs in autonomous berthing and unberthing under complex sea conditions by introducing a fuzzy adaptive PID control strategy combined with environmental perception and real-time pose calculation using 3D lidar.

Description

Technical Field

[0001] This invention relates to the field of unmanned vessel berthing and unberthing technology, specifically to a fuzzy control method and system for the safe and autonomous berthing and unberthing of unmanned vessels. Background Technology

[0002] Current research on autonomous docking technology for surface and underwater vehicles mainly focuses on intelligent docking. This docking method is complex and requires numerous sophisticated devices to complete the docking process. Related research is primarily in the development stage and is not yet mature, nor has it been put into practical use.

[0003] A research team from Harbin Engineering University has proposed an autonomous docking method for USVs based on deep learning-based target detection. The docking system consists of a small surface vehicle and a docking device equipped with a docking guidance mechanism. Both the vehicle and the docking device are equipped with sensors and positioning devices. During autonomous docking, the visual detection equipment uses an improved YOLOv3 target detection algorithm to detect and analyze the relative positions and attitudes of the vehicle and the docking device. Image processing is used to calculate the relative positional deviation of the target device. Based on the analysis results, the two devices are controlled to simultaneously deploy and adjust to achieve docking.

[0004] The State Key Laboratory of Fluid Dynamics and Electromechanical Systems at Zhejiang University has developed an optical guidance-based docking method. The docking platform of the AUV adopts a flared design. An optical four-quadrant detector mounted on the bow of the AUV uses computational light source imaging to calculate the relative angle between the AUV and the docking platform. Control commands are generated based on angle changes and input into the controller to control the AUV's attitude and the entire docking process. The guidance and control system is computationally simple, and the optical vision sensor can provide real-time input information to the control system, thus improving the overall system's real-time performance. In a relatively still water environment, after successful docking, the AUV underwent non-contact charging and data transmission, which proceeded smoothly.

[0005] The Odyssey IIB AUV underwater docking system, jointly developed by WoodsHole and MIT, uses a docking rod on its docking device as the docking target. The AUV's bow features a V-shaped mechanism that facilitates the docking rod's entry into the mechanism. After the rod enters, a locking mechanism secures it, and then the AUV is fixed in place by the rod, achieving docking between the AUV and the docking mechanism. Data transmission and power replenishment are possible after docking. Current underwater vehicle recovery technologies mainly fall into two categories: descent-based recovery and stern-slipper recovery. Both methods require manual operation and have relatively low levels of automation. Research on autonomous docking technology for underwater vehicles is still in its early stages.

[0006] Currently, the recovery of surface vessels relies on manual operation of related equipment. For example, sling-launch recovery systems require manual hooking or maneuvering of the vessel into a designated device. Similarly, stern-slipper recovery technology relies on remote control for the final docking and retrieval. During docking, both the vessel and the recovery mother ship are affected by factors such as water surface fluctuations, current direction, and wind direction changes, making docking difficult. Due to the complex docking process and relatively immature docking technology, it is not yet commercially available. In recent years, some research has emerged on autonomous docking technology for small underwater and surface vessels, but most of this research remains at the level of signal transmission, path planning, and software analysis, limited to theoretical studies. These technologies are still far from commercial application.

[0007] Because electrically powered unmanned surface vessels (USVs) have limited battery capacity, they often need to return to designated areas to replace batteries or be salvaged ashore after completing several tasks. Traditional salvage methods require coordination between the USV and other vessels, often involving manual labor. This necessitates specialized operators, resulting in low operational efficiency and significant risks associated with manual waterborne operations. Furthermore, this method of salvaging USVs is not suitable for the trend towards intelligent development of USVs. Therefore, with the future development of USV technology, USVs need to possess autonomous docking and berthing capabilities to reduce human intervention during berthing and unberthing operations and enhance the overall intelligence level of USV usage.

[0008] Current unmanned surface vessel (USV) autonomous docking technology mainly relies on USVs equipped with various sensors to process and analyze the information acquired by the sensors, and to control the USV's autonomous docking and undocking process based on the analyzed data. However, the technology in this field is not mature, and the success rate of related experiments is low. Summary of the Invention

[0009] To address the shortcomings of existing technologies, this invention provides a fuzzy control method and system for the safe and autonomous berthing and unberthing of unmanned vessels.

[0010] To achieve the above objectives, the present invention is implemented through the following technical solution: A fuzzy control method for the safe and autonomous berthing and unberthing of unmanned vessels, comprising the following steps: (1) scanning the surrounding environment in real time using a three-dimensional lidar mounted on the unmanned vessel to generate point cloud data, using a point cloud processing algorithm to identify the structured contour of the unmanned dock, constructing a three-dimensional geometric model, and calculating the relative pose between the unmanned vessel and the dock, including position and attitude; (2) determining the desired longitudinal velocity u of the unmanned vessel's vector motion based on the relative pose. d Desired lateral velocity v d and desired turning angular velocity r dThe system obtains the actual longitudinal velocity u, actual lateral velocity v, and actual turning angular velocity r of the unmanned surface vessel's vector motion, and calculates the system error e(t) and the error rate of change ec(t), where: The systematic error e(t) is defined as:

[0011] The error change rate ec(t) is defined as ec(t) = de(t) / dt; (3) Input the system error e(t) and the error change rate ec(t) into the fuzzy controller and perform fuzzification processing to obtain fuzzy linguistic variables E and Ec; Based on the preset fuzzy rule base, perform fuzzy reasoning on E and Ec to output the adjustment amounts ΔKp, ΔKi, ΔKd of the PID parameters; Through defuzzification processing, convert the adjustment amounts into precise values ​​and calculate the adjusted PID parameters according to the following formula: ,in, , , This is the baseline value for the PID parameters; Using the adjusted PID parameters ΔKp, ΔKi, and ΔKd, calculate the control output u(t) according to the PID controller output formula:

[0012] (4) Decompose the control output u(t) into thrust commands for each propeller of the unmanned vessel, and control the unmanned vessel to achieve coordinated operation of longitudinal, lateral and bow movements, so that the unmanned vessel can complete autonomous berthing and unberthing in an attitude parallel to the shore.

[0013] Preferably, in the environmental perception step, the point cloud processing algorithm includes point cloud segmentation, feature extraction, and model matching to accurately identify the dock outline and calculate the relative pose.

[0014] Preferably, the fuzzy rule base includes multiple rules, each rule being in the form of: if E is A and Ec is B, then ΔKp is C, ΔKi is D, and ΔKd is E, where A, B, C, D, and E are fuzzy sets.

[0015] Preferably, the defuzzification process employs the centroid method or the maximum membership method.

[0016] Preferably, the PID controller is an incremental PID controller, which avoids integral saturation by calculating the increment of the control quantity.

[0017] Preferably, the thruster control steps include intelligently decomposing and distributing the total thrust and torque requirements to multiple independent thrusters, including the main thruster and the side thrusters, to achieve six degrees of freedom maneuverability of the unmanned vessel.

[0018] The present invention also discloses a system for implementing the method, characterized in that, 3D LiDAR is used to scan the environment and generate point cloud data; The processing unit is used to run point cloud processing algorithms and fuzzy adaptive PID control algorithms. Control unit, used to generate thruster commands; Multiple thrusters, controlled by the control unit, perform berthing and unberthing operations.

[0019] Preferably, the processing unit includes a fuzzy controller and a PID controller, wherein the fuzzy controller dynamically adjusts the parameters of the PID controller.

[0020] The beneficial effects of this invention are as follows: By introducing a fuzzy adaptive PID control strategy and combining it with environmental perception and real-time pose calculation using 3D lidar, this invention significantly improves the control accuracy and adaptability of unmanned vessels in autonomous berthing and unberthing under complex sea conditions. It effectively overcomes the response lag, overshoot, and model mismatch problems caused by fixed parameters in traditional PID systems, enabling unmanned vessels to berth smoothly and safely parallel to the shoreline. This greatly reduces the need for manual intervention and operational risks, providing reliable technical support for the long-term autonomous and intelligent operation of unmanned vessels. Attached Figure Description

[0021] Figure 1 This is an overall framework diagram of the present invention; Figure 2 This is an application diagram of the present invention. Detailed Implementation

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0023] The present invention will be further described in detail below with reference to the accompanying drawings. Those skilled in the art should understand that these descriptions are exemplary and not intended to limit the scope of protection of the present invention.

[0024] This invention provides a fuzzy control method and system for the safe and autonomous berthing and unberthing of unmanned surface vessels (USVs). First, the USV, equipped with a 3D LiDAR, can scan its surrounding environment in real time, generating high-density point cloud data. Using point cloud processing algorithms, the system can accurately identify the structured contours of the unmanned dock, construct its 3D geometric model, and calculate the relative pose (position and attitude) between the USV and the dock in real time. This provides crucial distance and angle information for precise close-range docking.

[0025] Secondly, traditional PID (proportional-integral-derivative) controllers are widely used due to their simple structure and good robustness. However, to cope with the uncertainty and strong nonlinearity of the USV model, this system integrates more advanced control algorithms, which can predict the future system state and optimize the control sequence online. It performs excellently in handling constraints and disturbances and has extremely strong robustness to parameter changes and external disturbances.

[0026] At the same time, the algorithm intelligently decomposes and distributes the total thrust and torque requirements output by the controller to each individual thruster (including the main thruster and the side thrusters). By adjusting the speed and / or angle of each thruster, it achieves the six-degree-of-freedom precise maneuvering of the USV, including complex movements such as translation, rotation and lateral movement, thereby completing the final berthing and unberthing operations.

[0027] Unmanned vessel docking and undocking control methods: Vector motion control for unmanned surface vessels (USVs), namely, achieving independent or coordinated control of the longitudinal, lateral, and yaw degrees of freedom, is a core technology for realizing highly maneuverable operations (such as autonomous berthing and unberthing). However, this process faces multiple severe challenges. First, there is the complexity and uncertainty of the model. The dynamic model of a USV exhibits highly nonlinear and strongly coupled characteristics. Its hydrodynamic coefficients change significantly due to dynamic variations in speed, load, draft, and sea state (wind, waves, and currents), making it difficult to establish an accurate mathematical model and resulting in serious model mismatch problems. Second, there is the computational complexity and real-time requirement: Vector control requires real-time solving of complex kinematic equations and distributing control commands to multiple thrusters, which places extremely high demands on the computational power and response speed of the control system. Simultaneously, there is the strong interference from environmental disturbances: External disturbances in the marine environment, such as wind, waves, and currents, are time-varying and unpredictable. They directly affect the hull, posing a continuous challenge to the stability and accuracy of the control system. Finally, the dynamic response of the actuator needs to be considered: The thruster itself has dynamic response lag and nonlinear characteristics (such as dead zone and saturation), which makes it difficult to accurately model the transition from control commands to actual thrust output.

[0028] To address the above issues, an incremental proportional-integral-derivative (PID) control algorithm can be used in the initial control design. Incremental PID avoids integral saturation and achieves smoother control switching by calculating the increment of the control variable rather than its absolute value. Its core idea is to dynamically adjust the control output based on the system deviation.

[0029] After obtaining the expressions for the proportional, integral, and derivative elements, we can obtain the expression for the PID controller output u(t):

[0030] Where K p Ki K d These are the three parameters of PID control, which adjust the error in practice. These three parameters directly affect the stability, response speed, overshoot, and steady-state accuracy of the control system.

[0031] In the formula, e(t) is the system error (the difference between the actual value and the set value).

[0032]

[0033] Where u d v d r d , respectively, the expected longitudinal velocity, lateral velocity, and vector motion of the unmanned surface vessel. Where u, v, and r are the actual longitudinal velocity, lateral velocity, and velocity of the unmanned surface vessel's vector motion, respectively. By designing longitudinal and lateral speed controllers and a bow speed controller, vector motion control can be achieved where the lateral and longitudinal speeds of the unmanned surface vessel are not zero, but the bow angular velocity is zero. This enables the unmanned surface vessel to perform berthing and unberthing operations in an attitude parallel to the unmanned dock on the shore.

[0034] Although PID controllers are widely used due to their simple structure and robustness, their limitations in fixed parameters are becoming increasingly apparent in complex tasks such as autonomous berthing and unberthing of USVs. First, there is the issue of adaptability to nonlinear systems: traditional PID controllers are linear, while the dynamic model of a USV is a typical nonlinear system. When the operating point changes significantly, a fixed set of PID parameters struggles to maintain optimal control performance across the entire operating range. Second, there are parameter perturbations and external disturbances: faced with the aforementioned changes in hydrodynamic coefficients and complex environmental disturbances, fixed-parameter PID controllers may fail to adapt promptly, easily leading to overshoot, oscillations, or even slow response, severely impacting the safety and efficiency of berthing and unberthing. Therefore, to improve the control performance of USVs in complex and variable marine environments, more advanced, adaptive control strategies must be introduced.

[0035] To overcome the shortcomings of traditional PID control, this paper introduces a fuzzy adaptive PID control strategy. This method combines the reasoning and decision-making capabilities of fuzzy logic with the robustness and reliability of PID control, achieving complementary advantages. Its core idea is to use a fuzzy controller to automatically adjust the three key parameters (Kp, Ki, Kd) of the PID controller online based on the real-time error e and the rate of change of error ec. This dynamic adjustment mechanism allows the controller to "sense" the current state of the system and, like an experienced operator, optimize the control strategy in real time, thereby effectively coping with the uncertainty of model parameters and disturbances in the external environment, achieving better dynamic performance and stronger robustness.

[0036] The fuzzy PID algorithm combines PID and fuzzy control to achieve complementary advantages. By dynamically adjusting PID parameters through fuzzy control, it adapts to changes in model parameters, thereby achieving better control performance. This paper applies the fuzzy PID algorithm to heading control in a study of autonomous berthing and unberthing of ships to address the problem of uncertain ship mathematical model parameters. The specific control structure of the fuzzy PID algorithm is shown in Figure 1. The specific workflow of the fuzzy adaptive PID controller is as follows, and its structure is as follows: Figure 1 As shown.

[0037] First, input extraction: In each control cycle, the system error e is calculated using the formula, and its rate of change ec(t) = de(t) / dt is further calculated. These two precise quantities serve as the inputs to the fuzzy controller. Next, the inputs are fuzzified: The precise input quantities e(t) and ec(t) are converted into fuzzy linguistic variables, and their corresponding membership degrees are determined. This process is called fuzzification, resulting in fuzzy sets E and Ec. Based on a pre-defined fuzzy rule base (usually built from expert knowledge or empirical data), logical reasoning is performed on the fuzzified inputs. The rule form in the rule base is typically: "The fuzzy rule base includes multiple rules, each in the form: if E is A and Ec is B, then ΔKp is C, ΔKi is D, and ΔKd is E, where A, B, C, D, and E are fuzzy sets." By activating all relevant rules, a new fuzzy output set is obtained. The output set obtained from fuzzy reasoning (the fuzzy values ​​of ΔKp, ΔKi, and ΔKd) is converted into precise numerical values. This process is called defuzzification, and commonly used methods include the centroid method and the maximum membership method.

[0038] The overall idea of ​​the fuzzy PID algorithm is based on the PID principle, with the addition of a fuzzy controller to adjust K. p K i K d The three parameters are tuned to obtain the optimal three parameters. The specific workflow is as follows: First, based on the principle of PID, the system deviation e and the rate of change of deviation ec are obtained. Then, these are used as inputs to the fuzzy controller for fuzzification processing, thereby obtaining E and Ec. c Then, a new fuzzy output set is obtained according to the fuzzy rules. Defuzzification is then performed to obtain the setpoint values ​​for the three parameters: ΔK. p ΔK i ΔK d Through the formula:

[0039] The new PID parameters are obtained and used as inputs to the PID controller to obtain the new output parameters U, where... , , The baseline values ​​for adjusting the three parameters.

[0040] The nominal parameters obtained by defuzzification are used as input to obtain new output values. These new output values ​​are then used as the power output for the unmanned vessel, enabling it to perform berthing and unberthing operations with the unmanned dock in an attitude parallel to the shore.

[0041] Through this closed-loop adaptive adjustment process, the fuzzy PID controller ensures that the USV can still accurately track the desired trajectory (such as a parallel attitude to the dock) even when subjected to strong external disturbances or changes in its own model parameters, ultimately achieving safe, smooth, and efficient autonomous berthing and unberthing operations.

[0042] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A fuzzy control method for the safe and autonomous berthing and unberthing of unmanned vessels, characterized in that, Includes the following steps: (1) The surrounding environment is scanned in real time by the three-dimensional lidar carried by the unmanned ship to generate point cloud data. The point cloud processing algorithm is used to identify the structured outline of the unmanned dock, construct a three-dimensional geometric model, and calculate the relative pose between the unmanned ship and the dock, including position and attitude. (2) Based on the relative pose, determine the desired longitudinal velocity u of the unmanned surface vessel's vector motion. d Desired lateral velocity v d and desired turning angular velocity r d The system obtains the actual longitudinal velocity u, actual lateral velocity v, and actual turning angular velocity r of the unmanned surface vessel's vector motion, and calculates the system error e(t) and the error rate of change ec(t), where: The systematic error e(t) is defined as: The error rate of change ec(t) is defined as ec(t) = de(t) / dt; (3) Input the system error e(t) and the error change rate ec(t) into the fuzzy controller and perform fuzzification processing to obtain fuzzy linguistic variables E and Ec; based on the preset fuzzy rule base, perform fuzzy inference on E and Ec to output the adjustment amounts ΔKp, ΔKi, and ΔKd of the PID parameters; through defuzzification processing, convert the adjustment amounts into precise values, and calculate the adjusted PID parameters according to the following formula: Among them, K p0 K i0 K d0 This is the baseline value for the PID parameters; Using the adjusted PID parameters ΔKp, ΔKi, and ΔKd, calculate the control output u(t) according to the PID controller output formula: (4) Decompose the control output u(t) into thrust commands for each propeller of the unmanned vessel, and control the unmanned vessel to achieve coordinated operation of longitudinal, lateral and bow movements, so that the unmanned vessel can complete autonomous berthing and unberthing in an attitude parallel to the shore.

2. The method according to claim 1, characterized in that, In the environmental perception step, the point cloud processing algorithm includes point cloud segmentation, feature extraction, and model matching to accurately identify the dock outline and calculate the relative pose.

3. The method according to claim 1, characterized in that, The fuzzy rule base includes multiple rules, each rule having the following form: if E is A and Ec is B, then ΔKp is C, ΔKi is D, and ΔKd is E, where A, B, C, D, and E are fuzzy sets.

4. The method according to claim 1, characterized in that, The defuzzing process employs either the centroid method or the maximum membership method.

5. The method according to claim 1, characterized in that, The PID controller is an incremental PID controller, which avoids integral saturation by calculating the increment of the control quantity.

6. The method according to claim 1, characterized in that, The thruster control steps include intelligently decomposing and distributing the total thrust and torque requirements to multiple independent thrusters, including the main thruster and the side thrusters, to achieve six degrees of freedom maneuverability of the unmanned vessel.

7. A system for implementing the method of any one of claims 1-6, characterized in that, include: 3D LiDAR is used to scan the environment and generate point cloud data; The processing unit is used to run point cloud processing algorithms and fuzzy adaptive PID control algorithms. Control unit, used to generate thruster commands; Multiple thrusters, controlled by the control unit, perform berthing and unberthing operations.

8. The system according to claim 7, characterized in that, The processing unit includes a fuzzy controller and a PID controller, wherein the fuzzy controller dynamically adjusts the parameters of the PID controller.

Images