A Collision Avoidance Method for Unmanned Surface Vessel Swarms in Multi-Obstacle Environments Based on Safety Obstacle Backstepping

Abstract

This invention proposes a collision avoidance method for unmanned surface vessel (USV) swarms in multi-obstacle environments based on safety obstacle backstepping, belonging to the field of cooperative control of unmanned aquatic equipment. It solves the problems of high system complexity and computational burden in existing collision avoidance control methods for USV swarms in complex water environments with multiple obstacles, due to multi-constraint coupling. It includes establishing a fully feedback two-layer affine system for USVs; constructing multiple control obstacle functions based on the collision avoidance requirements between multiple obstacles and USVs; introducing a smooth top-level control obstacle function to approximate the lower bound of the multiple control obstacle functions, transforming the multi-constraint collision avoidance problem into a single-layer safety constraint; constructing a global control obstacle function for the fully feedback two-layer affine system of USVs based on the control obstacle function backstepping method; and constructing an explicit safety collision avoidance cooperative control algorithm by modifying the nominal control law based on a smooth Lagrange operator. It is mainly used for collision avoidance in USV swarms in multi-obstacle environments.

Description

Technical Field

[0001] This invention belongs to the field of collaborative control of unmanned watercraft, and in particular relates to a collision avoidance method for unmanned surface vessels in multi-obstacle environments based on safety obstacle backstepping. Background Technology

[0002] In confined waters such as ports, islands, and narrow waterways, unmanned surface vessel (USV) swarms often face numerous obstacles when performing collaborative detection, patrol, or search and rescue missions. Faced with complex and ever-changing multi-obstacle environments, USV swarms must ensure safe collision avoidance between themselves and with obstacles to guarantee the overall safety and reliability of the swarm system. However, existing USV collision avoidance control methods still have significant shortcomings in dealing with multi-obstacle scenarios. For example, while behavior- or potential field-based methods are computationally simple, they are prone to getting trapped in local minima under multiple constraint couplings, leading to collision avoidance failure. Furthermore, existing swarm control schemes often prioritize maintaining a fixed formation, failing to fully consider the unstructured and time-varying nature of obstacle distribution, making it difficult to achieve flexible and efficient collision avoidance maneuvers in congested waters.

[0003] Unmanned surface vessels (USVs) possess complex dynamic characteristics such as underactuation, strong nonlinearity, and hydrodynamic coupling, which further increases the design difficulty of swarm safety collision avoidance controllers. On the one hand, there is an inherent conflict between collision avoidance requirements and mission objectives such as trajectory tracking or formation keeping: overemphasizing collision avoidance may lead to mission deviation, while overemphasizing mission accuracy may introduce collision risks. Existing methods often employ strategies such as weighted summation or priority hierarchies to handle this multi-objective conflict, but these strategies cannot guarantee that safety constraints and mission accuracy can be simultaneously satisfied under all operating conditions. On the other hand, when multiple USVs in a swarm approach multiple obstacles simultaneously, the number of constraints increases dramatically. Traditional control frameworks require solving high-dimensional optimization problems, and the computational burden increases exponentially with the number of obstacles, severely limiting the real-time performance of control methods. Therefore, how to achieve low-complexity, high-reliability collision avoidance control under multi-constraint, multi-objective conditions has become a key issue that needs to be considered in the design of collision avoidance algorithms.

[0004] Existing collision avoidance control methods for unmanned surface vessel (USV) swarms still have significant shortcomings in engineering practicality. While online optimization-based collision avoidance control algorithms can theoretically achieve collision avoidance control under multiple constraints, they suffer from complex parameter tuning processes and excessive online computation, making them difficult to deploy on USV embedded platforms with limited computing resources and unsuitable for practical engineering applications. Furthermore, in navigation scenarios with densely distributed obstacles, existing algorithms are prone to frequent control law switching and drastic fluctuations in control output, which not only increases the wear and tear on USV actuators and navigation energy consumption but also affects the long-term operational stability of the swarm control system. Currently, there is no mature technical solution that can simultaneously achieve low complexity, high real-time performance, and strong engineering adaptability in the control algorithm while ensuring safe collision avoidance performance of USV swarms, making it difficult to meet the actual operational needs of USV swarms in multi-obstacle environments. Summary of the Invention

[0005] In view of this, the present invention aims to propose a collision avoidance method for unmanned surface vessel (USV) swarms in multi-obstacle environments based on safety obstacle backstepping, in order to solve the problems of existing USV swarm collision avoidance control methods in complex water environments with multiple obstacles, such as high system complexity under multiple constraint coupling, large computational burden, difficulty in adapting to the underactuated and strongly nonlinear dynamic characteristics of USVs, inability to take into account both safety collision avoidance hard constraints and collaborative operation task requirements, and insufficient real-time performance and engineering practicality.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a collision avoidance method for unmanned surface vessels (USVs) swarms in multi-obstacle environments based on safety obstacle backstepping, comprising the following steps: Step 1: Establish a dual-layer affine system with full feedback for the unmanned surface vessel; Step 2: Based on the collision avoidance requirements between multiple obstacles and unmanned surface vessels, construct multiple control obstacle functions, introduce a smooth top-level control obstacle function to approximate the lower bound of the multiple control obstacle functions, and transform the multi-constraint collision avoidance problem into a single-layer safety constraint. Step 3: Based on the backstepping method of the control obstacle function, construct a global control obstacle function for the unmanned surface vessel's full feedback two-layer affine system; Step 4: Modify the nominal control law based on the smooth Lagrange operator to construct an explicit safe collision avoidance cooperative control algorithm.

[0007] Furthermore, the process of establishing the unmanned surface vessel's fully feedback dual-layer affine system in step 1 is as follows: a kinematic and dynamic model of the unmanned surface vessel is established based on the geodetic coordinate system and the body coordinate system, a position state transformation is introduced to handle the underactuated characteristics of the unmanned surface vessel, and a fully feedback dual-layer affine system of the unmanned surface vessel is established.

[0008] Furthermore, unmanned surface vessels The kinematic and dynamic model is represented as follows: in Indicates position and heading angle , Indicates longitudinal velocity lateral velocity and bow angular velocity The velocity vector in the body coordinate system. For control input, where and Representing the longitudinal thrust and bow moment respectively, the rotation matrix is ​​expressed as: It is the inertia matrix. Represents the Coriolis-centripetal force matrix. For hydrodynamic damping matrix, , and This represents the hydrodynamic damping coefficient.

[0009] Furthermore, the position state transformation is represented as ,in If the parameters are transformed, then the unmanned surface vessel model can be represented as: , in and This represents a nonlinear term, while It is to satisfy Input mapping function, The control input of the transformed system is expressed as follows: The unmanned surface vessel's fully feedback dual-layer affine system is: .

[0010] Furthermore, in step 2, the multiple control obstacle function is used to modify the existing nominal cooperative controller of the unmanned surface vessel. Make corrections, saying the cooperative controller adjusts according to the desired speed. The design includes multiple obstacle avoidance control functions, including obstacle avoidance control functions between unmanned surface vessels and obstacle avoidance control functions for individual obstacles. Unmanned surface vessel obstacle avoidance control function: , , Maintain a safe distance between unmanned surface vessels; Obstacle avoidance control function: , , Obstacles The collision avoidance radius.

[0011] Furthermore, in step 2, the process of introducing a smooth top-level control obstacle function to approximate the lower bound of the multi-control obstacle function is as follows: based on Boolean logic operations, the top-level position of the unmanned surface vessel is always within the intersection range of the safety set, and the lower bound of the multi-control obstacle function is uniformly represented by a minimum value function. Represented as: Smooth the top-level control barrier function: in It is a positive constant. A smooth top-level control barrier function is introduced to approximate the lower bound of the multi-control barrier function.

[0012] Furthermore, step 3 integrates multiple security requirement constraints into a continuously differentiable top-level control barrier function, the global control barrier function being: in It is a positive constant.

[0013] Furthermore, in step 4, the smoothing Lagrange operator... for: in It is a positive constant coefficient.

[0014] Furthermore, the explicit safety collision avoidance cooperative control algorithm modifies the nominal control law through a safety filter optimizer, and the modified safety control algorithm is as follows: Among them It is continuously differentiable function.

[0015] Furthermore, the effectiveness of the unmanned surface vessel swarm collision avoidance method in multi-obstacle environments was verified through simulation experiments.

[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: The proposed collision avoidance method for unmanned surface vessel (USV) swarms in multi-obstacle environments based on safety obstacle backstepping achieves effective approximation of the lower bound of multiple control obstacle functions by designing a smooth top-level control safety function. This transforms the multi-constraint collision avoidance problem into a single-layer safety constraint form, significantly reducing the complexity of the collision avoidance control system in multi-obstacle scenarios. It effectively solves the pain point of traditional methods where the number of constraints increases sharply when multiple USVs approach multiple obstacles simultaneously in a swarm, requiring the solution of high-dimensional optimization problems and the computational burden increases exponentially with the number of obstacles. It avoids the problem of traditional behavior- or potential field-based methods easily getting trapped in local minima under multi-constraint coupling, leading to collision avoidance failure. This greatly improves the simplicity, real-time performance, and adaptability to complex obstacle environments. It can adapt to unstructured, time-varying obstacle distribution scenarios in restricted waters such as ports, islands, and narrow waterways, enabling USV swarms to achieve flexible and efficient collision avoidance maneuvers in congested waters, fully meeting the core requirements of collaborative and flexible collision avoidance in USV swarm systems.

[0017] This invention proposes a global control obstacle function suitable for a fully feedback two-layer affine model of unmanned surface vessels (USVs) by combining the backstepping method of the control obstacle function. This function can simultaneously encode multiple obstacle safety constraints and the underactuated dynamic characteristics of USVs, effectively adapting to the complex dynamic characteristics of USVs, such as underactuation, strong nonlinearity, and hydrodynamic coupling. Under the premise of ensuring collision avoidance safety, it effectively coordinates the conflict between collision avoidance behavior and mission objectives such as trajectory tracking or formation keeping. It solves the problem that existing strategies such as weighted summation or priority hierarchical are difficult to satisfy both safety constraints and mission accuracy under all operating conditions. It avoids the drawbacks of overemphasizing collision avoidance leading to mission deviation and over-pursuing mission accuracy leading to collision risks, effectively improving the safety and reliability of USV cooperative control.

[0018] This invention constructs an explicit safety collision avoidance cooperative control strategy by introducing a smooth Lagrange operator to modify the nominal control law. This strategy can achieve a smooth transition of the control law in dense obstacle environments, effectively avoiding actuator wear and increased energy consumption caused by frequent control switching or drastic output, thus improving the long-term operational stability and engineering reliability of the unmanned surface vessel (USV) swarm system. Furthermore, the proposed method does not rely on complex online optimization solutions, resulting in low computational burden. It solves the problems of existing online optimization-based algorithms, such as complex parameter tuning, large online computation volume, and difficulty in deployment on USV embedded platforms with limited computing resources. This significantly improves the engineering practicality of the method and facilitates the development and application of practical systems. Attached Figure Description

[0019] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram illustrating the collision avoidance of unmanned surface vessels (USVs) in multi-obstacle environments as described in this invention. Figure 2 This is a flowchart of a multi-obstacle environment unmanned surface vessel swarm collision avoidance method based on safety obstacle backstepping, as described in this invention. Figure 3 This is a trajectory diagram of the unmanned surface vessel swarm's obstacle avoidance maneuver in complex environments as described in this invention; Figure 4 This is a graph showing the variation of the smooth top-level control barrier function described in this invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present invention can be combined with each other, and the described embodiments are only some embodiments of the present invention, not all embodiments.

[0021] See Figure 1-4 This embodiment describes a collision avoidance method for unmanned surface vessel (USV) swarms in multi-obstacle environments based on safety obstacle backstepping. This method is applicable to collision avoidance control scenarios in restricted waterways such as ports, islands, and narrow waterways, where USV swarms perform collaborative detection, patrol, or search and rescue missions. The overall collision avoidance scenario is illustrated in the schematic diagram of USV swarm collision avoidance in multi-obstacle environments. Figure 1 As shown, the overall process of the method is illustrated in the flowchart of a multi-obstacle environment unmanned surface vessel swarm collision avoidance method based on safety obstacle backstepping. Figure 2 As shown.

[0022] This implementation first establishes a fully feedback, dual-layer affine system for unmanned surface vessels. (The last sentence appears to be incomplete and possibly refers to a different implementation.) An unmanned surface vessel (USV) swarm is first constructed using a geodetic coordinate system and a body coordinate system to establish the kinematic and dynamic models of the USVs. The kinematic and dynamic model is represented as follows: in Indicates position and heading angle , Indicates longitudinal velocity lateral velocity and bow angular velocity The velocity vector in the body coordinate system. For control input, where and Representing the longitudinal thrust and bow moment respectively, the rotation matrix is ​​expressed as: It is the inertia matrix. Represents the Coriolis-centripetal force matrix. For hydrodynamic damping matrix, , and This represents the hydrodynamic damping coefficient.

[0023] To address the underactuation problem of unmanned surface vessels (USVs), this implementation introduces a position-state transformation with transformation parameters to convert the original USV model. The position-state transformation is expressed as follows: ,in If the parameters are transformed, then the unmanned surface vessel model can be represented as: , in and This represents a nonlinear term, while It is to satisfy Input mapping function, The control input of the transformed system is expressed as follows: The unmanned surface vessel's fully feedback dual-layer affine system is: .

[0024] The full feedback dual-layer affine system of unmanned surface vessels (USVs) provides a system foundation adapted to the dynamic characteristics of USVs for the design of subsequent collision avoidance control algorithms.

[0025] Based on the completed unmanned surface vessel (USV) system model, and considering the collision avoidance requirements between multiple obstacles and the USV, a multi-control obstacle function is constructed. A smooth top-level control obstacle function is introduced to approximate the lower bound of the multi-control obstacle function, transforming the multi-constraint collision avoidance problem into a single-layer safety constraint problem. Based on the safety collision avoidance requirements between multiple obstacles and the USV, a smooth top-level control safety function is designed to approximate the lower bound of the multi-control obstacle function and reduce the complexity of the collision avoidance system.

[0026] This implementation uses a control obstacle function based on the desired velocity. Designed Nominal Cooperative Controller for Unmanned Surface Vessels To correct this, the obstacle avoidance task essentially involves setting constraints on the relative positions of each unmanned surface vessel (USV) with multiple obstacles and other USVs. Therefore, considering a rigorous feedback model for USVs, two different models are constructed to address the issue of USVs' relative positions to obstacles and other USVs. Position-dependent top-level control barrier function, 1) Obstacle avoidance control function between unmanned surface vessels: , , Maintain a safe distance between unmanned surface vessels; 2) Obstacle avoidance control function: , , Obstacles The collision avoidance radius.

[0027] Based on Boolean logic operations, the top position of the unmanned surface vessel must always be within the intersection of the following safety sets to satisfy all obstacle avoidance requirements.

[0028] This security constraint can be expressed by a minimum value function as follows: The following smooth top-level control barrier function is introduced to approximate the lower bound of the above multi-control barrier function.

[0029] Smooth the top-level control barrier function: Among them A positive constant, whose Lie derivative along the top layer of the unmanned surface vessel system is expressed as: gradient Represented as: satisfy .

[0030] Define the minimum value of the multi-control barrier function: Then the smooth top-level control barrier function satisfies: therefore, If it is established, then Approximates the lower bound of a multi-control barrier function ,Right now: Subsequently, based on the backstepping method of the control obstacle function, a global control obstacle function suitable for the fully feedback two-layer affine model of unmanned surface vessels (USVs) is constructed. Multiple safety requirements and constraints are integrated into a continuously differentiable top-level control obstacle function. For the aforementioned fully feedback two-layer affine system of USVs, a global control obstacle function with a positive constant is designed so that the function can simultaneously encode multiple obstacle safety constraints and the underactuated dynamic characteristics of USVs. Under the premise of ensuring collision avoidance safety, it can achieve coordinated coordination between collision avoidance behavior and mission objectives such as trajectory tracking and formation keeping, thus solving the control design problem under multi-objective conflict.

[0031] The global control barrier function is designed as follows: in It is a positive constant.

[0032] Based on the global control obstacle function, a nominal control law correction problem is constructed using the smooth Lagrange operator, and an explicit safe collision avoidance cooperative control algorithm is designed.

[0033] Considering the top-level kinematic model of the unmanned surface vessel, the safe virtual speed It was determined based on the following optimization problem: Among them It is continuously differentiable. Function. Consider the top-level affine model of the unmanned surface vessel, and the safe virtual velocity. The differentiable explicit solution can be obtained as follows: in The smoothing Lagrange operator is defined as follows: in It is a positive constant coefficient.

[0034] To achieve the goals of cooperative control and collision avoidance, the following safety filter optimizer is designed to modify the nominal cooperative controller. .

[0035] The revised security control algorithm is as follows: Among them It is continuously differentiable function.

[0036] To further illustrate the structure of the safety controller, its explicit expression is given, along with the following Lie derivatives: The gradient is expressed as: in The expression for the gradient has been given, where the gradient is... It is represented as: 1) When , 2) When , in: and The expression is as follows: because and The Lie derivative of the top-level control barrier function satisfies: This implementation also verifies the effectiveness of the proposed collision avoidance method through simulation experiments. During the simulation, the obstacle avoidance maneuvers of multiple unmanned surface vessels (USVs) in a complex multi-obstacle environment are simulated. The resulting obstacle avoidance maneuver trajectories of the USV swarm in complex environments are shown in the USV swarm obstacle avoidance maneuver trajectory diagram. Figure 3 As shown, the change data of the smooth top-level control barrier function during the simulation were collected simultaneously, and the change curve of the smooth top-level control barrier function was plotted. Figure 4 Through simulation data and trajectory results verification, the method proposed in this embodiment can effectively meet the collaborative and flexible collision avoidance requirements of unmanned surface vessel (USV) swarm systems, maintain safe collision avoidance constraints in multi-obstacle environments, and take into account the accuracy of collaborative task execution, thereby effectively improving the safety and reliability of USV collaborative control.

[0037] The specific embodiments of the present invention disclosed above are merely illustrative of the invention. These embodiments do not exhaustively describe all details, nor do they limit the invention to the specific embodiments described. Many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A collision avoidance method for unmanned surface vessels (USVs) swarms in multi-obstacle environments based on safety obstacle backstepping, characterized in that: Includes the following steps: Step 1: Establish a dual-layer affine system with full feedback for the unmanned surface vessel; Step 2: Based on the collision avoidance requirements between multiple obstacles and unmanned surface vessels, construct multiple control obstacle functions, introduce a smooth top-level control obstacle function to approximate the lower bound of the multiple control obstacle functions, and transform the multi-constraint collision avoidance problem into a single-layer safety constraint. Step 3: Based on the backstepping method of the control obstacle function, construct a global control obstacle function for the unmanned surface vessel's full feedback two-layer affine system; Step 4: Modify the nominal control law based on the smooth Lagrange operator to construct an explicit safe collision avoidance cooperative control algorithm.

2. The method for collision avoidance of unmanned surface vessels (USVs) swarms in multi-obstacle environments based on safety obstacle backstepping as described in claim 1, characterized in that: The process of establishing the unmanned surface vessel's fully feedback dual-layer affine system in step 1 is as follows: establish the kinematic and dynamic model of the unmanned surface vessel based on the geodetic coordinate system and the body coordinate system, introduce position state transformation to handle the underactuated characteristics of the unmanned surface vessel, and establish the unmanned surface vessel's fully feedback dual-layer affine system.

3. The method for collision avoidance of unmanned surface vessels (USVs) in multi-obstacle environments based on safety obstacle backstepping as described in claim 2, characterized in that: Unmanned Surface Vessel The kinematic and dynamic model is represented as follows: in Indicates position and heading angle , Indicates longitudinal velocity lateral velocity and bow angular velocity The velocity vector in the body coordinate system. For control input, where and Representing the longitudinal thrust and bow moment respectively, the rotation matrix is ​​expressed as: It is the inertia matrix. Represents the Coriolis-centripetal force matrix. For hydrodynamic damping matrix, , and This represents the hydrodynamic damping coefficient.

4. The method for collision avoidance of unmanned surface vessels (USVs) in multi-obstacle environments based on safety obstacle backstepping as described in claim 3, characterized in that: Position state transformation is represented as ,in If the parameters are transformed, then the unmanned surface vessel model can be represented as: ， in and This represents a nonlinear term, while It is to satisfy Input mapping function, The control input of the transformed system is expressed as follows: The unmanned surface vessel's fully feedback dual-layer affine system is: 。 5. The method for collision avoidance of unmanned surface vessels (USVs) swarms in multi-obstacle environments based on safety obstacle backstepping as described in claim 1, characterized in that: In step 2, the multiple control obstacle function is used to modify the existing nominal cooperative controller of the unmanned surface vessel. Make corrections, saying the cooperative controller adjusts according to the desired speed. The design includes multiple obstacle avoidance control functions, including obstacle avoidance control functions between unmanned surface vessels and obstacle avoidance control functions for individual obstacles. Unmanned surface vessel obstacle avoidance control function: , , Maintain a safe distance between unmanned surface vessels; Obstacle avoidance control function: , , For obstacles The collision avoidance radius.

6. A collision avoidance method for unmanned surface vessels (USVs) swarms in multi-obstacle environments based on safety obstacle backstepping, as described in claim 5, is characterized in that: Step 2, which introduces a smooth top-level control obstacle function to approximate the lower bound of the multi-control obstacle function, involves: based on Boolean logic operations, ensuring that the top-level position of the unmanned surface vessel (USV) always falls within the intersection of the safety sets, and using a minimum value function to uniformly represent the lower bound of the multi-control obstacle function. Represented as: Smooth the top-level control barrier function: in It is a positive constant. A smooth top-level control barrier function is introduced to approximate the lower bound of the multi-control barrier function.

7. The method for collision avoidance of unmanned surface vessels (USVs) in multi-obstacle environments based on safety obstacle backstepping as described in claim 1, characterized in that: Step 3 integrates multiple security requirement constraints into a continuously differentiable top-level control barrier function, the global control barrier function being: in It is a positive constant.

8. A collision avoidance method for unmanned surface vessels (USVs) swarms in multi-obstacle environments based on safety obstacle backstepping, as described in claim 1, is characterized in that: The smoothing Lagrange operator in step 4 for: in It is a positive constant coefficient.

9. A collision avoidance method for unmanned surface vessels (USVs) swarms in multi-obstacle environments based on safety obstacle backstepping, as described in claim 8, is characterized in that: The explicit safety collision avoidance cooperative control algorithm modifies the nominal control law through a safety filter optimizer. The modified safety control algorithm is as follows: Among them It is continuously differentiable function.

10. A collision avoidance method for unmanned surface vessels (USVs) swarms in multi-obstacle environments based on safety obstacle backstepping, as described in claim 1, is characterized in that: The effectiveness of the unmanned surface vessel swarm collision avoidance method in multi-obstacle environments was verified through simulation experiments.

Images