Flexible unmanned ship collision-free following control method based on potential field prediction
Through the hybrid line of sight and improved artificial potential field method, combined with the nonlinear model prediction controller, the problem of stagnation of unmanned boats caused by traditional potential field method is solved, and the efficient collision avoidance of unmanned boats in complex waters is achieved, and the improvement of the efficient collision avoidance of unmanned boats in complex waters is improved. The real-time tracking performance of the follower boat and the collision-free navigation capability of the system.
Patent Information
- Application Number
- CN202411249468.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-06
- Publication Date
- 2024-11-12
AI Technical Summary
The traditional artificial potential field method has saddle points where gravity equals repulsive force in unmanned boat collision avoidance, causing the movement of unmanned boats to stagnate, threatening safe navigation, and the prior art is difficult to avoid obstacles efficiently in complex waters in real time.
Using a flexible guidance method of unmanned boats based on hybrid line of sight, the improved artificial potential field method generates collision-free trajectories, and through a nonlinear model prediction controller scheme, the bias proportional guidance law when there is no collision risk and when there is collision risk The potential field prediction guide law is combined with the virtual pilot boat and the tangential repulsive force potential field to design a collision-free trajectory.
It effectively improves the real-time tracking efficiency of the follower boat on the pilot boat, reduces the stagnation of motion, increases the probability of the system obtaining the best collision-free route, and provides theoretical support for the new generation of water unmanned boat collision avoidance system.
Smart Images

Figure CN118938938A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of unmanned boats, and in particular relates to a collision-free following control method of a flexible unmanned boat based on potential field prediction. Background Art
[0002] Exploring advanced unmanned boat formation collision avoidance intelligent autonomous technology and building an efficient and highly reliable unmanned boat intelligent collision avoidance system are of great significance to improving the safe operation capability of unmanned boats in complex waters. As an important part of the unmanned system on water, a real-time and accurate collision avoidance system is the key to ensuring the normal operation of unmanned boats. Researching unmanned boat collision avoidance methods can not only provide theoretical support for the intelligent and autonomous collaborative operation of the new generation of unmanned boats, but also provide solid technical support for the safe navigation of unmanned boats at sea, thereby improving autonomous shipping capabilities.
[0003] An unmanned boat is an intelligent ship that can autonomously perform multiple tasks in complex water environments without human intervention. Compared with a single unmanned boat, a fleet of unmanned boats has a stronger transport capacity, and they play an increasingly important role in water operations (such as freight, passenger transport, sightseeing, maintenance, etc.). Faced with a complex water environment, how to avoid obstacles in real time and efficiently has not yet been well solved. The artificial potential field method is a reliable method for collision avoidance technology at this stage. However, the traditional artificial potential field method has a saddle point where gravity is equal to repulsion, which will cause the unmanned boat to have a local motion stagnation phenomenon, which greatly threatens the safe navigation of the unmanned boat. Summary of the invention
[0004] In order to overcome the shortcomings of the prior art, the present invention provides a method for flexible unmanned boat collision-free following control based on potential field prediction, which includes the following steps: (1) In view of the formation control problem during the journey, a flexible unmanned boat guidance method based on mixed line of sight is proposed, so that the follower boat can automatically track the pilot boat; (2) In order to avoid potential obstacles, an improved artificial potential field method is used to generate a collision-free trajectory, so that the unmanned boat can directly bypass the obstacles; (3) In order to achieve the stable and accurate tracking of the pilot boat by the follower boat, a nonlinear model predictive controller scheme is proposed. Compared with the prior art, the present invention effectively improves the real-time tracking efficiency of the follower boat to the pilot boat, greatly avoids the situation of motion stagnation of the ship, increases the probability of the system obtaining the best collision-free route, and provides theoretical support for the development of a new generation of collision avoidance systems for unmanned boats on water.
[0005] The technical solution adopted by the present invention to solve the technical problem is as follows:
[0006] Step 1: Construct a tracking guidance law based on bias proportional guidance when there is no collision risk;
[0007] Step 2: Construct a collision avoidance guidance law based on potential field prediction when there is a collision risk;
[0008] Step 3: Construct a nonlinear prediction model method based on full state integration;
[0009] Preferably, the step 1 is specifically:
[0010] The bow angles of the lead boat and the following boat are defined as is the line of sight angle between the following boat and the pilot boat; the relative position and relative velocity vector between the pilot boat and the following boat are defined as p = p t -p s and V w =V t -V s , define p s and p t is the position scalar of the lead boat and the following boat, V s and V t is the speed scalar of the lead boat and the following boat;
[0011] Establish an inertial relative coordinate system {W} with the pilot boat as the origin, and the relative lateral acceleration a w Perpendicular to the relative velocity V w , the relative viewing angle Θ is defined as Relative pitch angle ψ w It is expressed as:
[0012]
[0013] Among them, the coefficient
[0014] In the inertial relative coordinate system {W}, there exists and Relative speed V w The scalar V w and the relative lateral acceleration a w The scalar a w for:
[0015]
[0016] a w =a s cos(ψ-ψ w )-a t cos(ψ t -ψ w )
[0017] Among them, a t 、a s are the lateral accelerations of the lead boat and the following boat respectively;
[0018] In the coordinate system {W}, because and variable and the partial derivative of Θ with respect to p is written as and
[0019] After the angle constraint is introduced, the relative lateral acceleration a in the inertial relative coordinate system is w for:
[0020]
[0021] The lateral acceleration of the unmanned boat is a s It is expressed as:
[0022]
[0023] Where K is a constant coefficient.
[0024] Preferably, the step 2 is specifically:
[0025] The obstacle avoidance process is divided into two sub-steps: (i) Determine the reference pitch angle ψ * ; (ii) Design an obstacle avoidance guidance law to ensure that the position vector When ψ(t′ f )=ψ * ; The vector representing the distance between the unmanned boat and the obstacle, t′ f It is the moment when the following boat approaches the obstacle tangent point;
[0026] The yaw angle ψ is calculated based on the predicted artificial potential field method. * , gravitational potential field Based on the tangent point of the obstacle contour, we have:
[0027]
[0028] Where i is the number of the obstacle, and are the minimum and maximum contour angles of the obstacle respectively; the coefficient κ1 is a positive constant; r D The radius of the obstacle;
[0029] Tangential force of the ith obstacle Using gravitational potential field The negative gradient of
[0030]
[0031] in, is a unit vector, which represents the sight angle from the following boat to the tangent point of the i-th obstacle, and can be rewritten as is a vector scalar value of ; represents the distance angle between the unmanned boat and the obstacle mass center, and the lateral potential field Located on the following boat, as the distance between the following boat and the obstacle increases, the potential energy decreases. Written as:
[0032]
[0033] Among them, the coefficient κ2 is a positive constant;
[0034] The lateral force of the ith obstacle It is expressed as:
[0035]
[0036] in, is a unit vector perpendicular to the line connecting the following boat and the i-th obstacle; Can be rewritten as in is the rotation matrix, is a vector The scalar magnitude of the force F Σ (p) is written as:
[0037]
[0038] The reference yaw angle is equal to the direction angle F of the resultant force Σ (p), and obtain the reference yaw angle ψ * =∠F Σ (p);
[0039] Lateral reference acceleration of the following boat In the inertial system, it is expressed as:
[0040]
[0041] in, and is a positive constant;
[0042] Preferably, the step 3 is specifically:
[0043] Set the following prediction model:
[0044]
[0045] The path information of the follower boat is converted into control input to form the control space X;
[0046] Substituting the control input of the input space into the prediction model for iteration, predicting the future position and posture state of the boat, and forming a state space U;
[0047] Substitute the posture state in the state space into the following formula to optimize the objective function:
[0048]
[0049] Where h(p(t),u(t)) represents the objective function, l(p(t),u(t)) is the phase cost function, and L(p(t)) is the terminal cost function.
[0050] The sequential quadratic programming algorithm is used to solve the optimal problem and find the sequence that minimizes the objective function, that is, the set closest to the reference path, which is then transferred to the controlled boat.
[0051] The beneficial effects of the present invention are as follows:
[0052] (1) The present invention designs a tracking guidance law based on biased proportional guidance when there is no collision risk, which can flexibly select the optimal motion trajectory according to the environment, thereby providing an effective method for the unmanned boat formation guidance system.
[0053] (2) The present invention improves the gravitational potential field function and the repulsive potential field function by introducing a virtual pilot boat and a tangential repulsive potential field, thereby reducing the local stagnation of the following boat in the traditional artificial potential field method and achieving comprehensive collision avoidance.
[0054] (3) The present invention realizes optimal control trajectory tracking in real time and accurately by setting a nonlinear prediction model. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] Figure 1 It is a flow chart of the following control method of the present invention;
[0056] Figure 2 This is a schematic diagram of the path of the unmanned boat fleet of the present invention when encountering obstacles;
[0057] Figure 3 It is a schematic diagram of the trajectory of the hybrid track line algorithm in the experimental example of the present invention;
[0058] Figure 4 It is a schematic diagram of the speed of the hybrid track line algorithm in the experimental example of the present invention;
[0059] Figure 5 A schematic diagram of collision avoidance results in an embodiment of the present invention;
[0060] Figure 6 Schematic diagram of the motion relationship between the pilot boat and the follower boat in an embodiment of the present invention. DETAILED DESCRIPTION
[0061] The present invention is further described below in conjunction with the accompanying drawings and embodiments.
[0062] The purpose of the present invention is to provide a flexible unmanned boat collision-free following control method based on potential field prediction in order to solve the tracking and collision avoidance control problems in multi-boat target following mission scenarios. The guidance and control system established by the present invention takes into account the obstacle detection, collision avoidance and trajectory tracking of the unmanned boat, and realizes the accurate tracking and interception of the target by the unmanned boat in a complex environment. The system flow is as follows Figure 1 As shown in the figure, the unmanned boat obtains the information of surface obstacles through millimeter wave radar, including the location, shape, size, etc. of the obstacles, analyzes and evaluates the obstacles, determines the danger level and impact range of the obstacles, and plans a safe and reasonable collision avoidance trajectory. By using the onboard sensors, the information of the waters around the unmanned boat is obtained in real time and used in guidance and control. During the target tracking and interception process, the reference heading angle is calculated to determine the correct heading direction. In addition, when encountering obstacles, effective collision avoidance strategies are adopted to ensure the safety of the mission. The obstacle detection and collision avoidance technology of the unmanned boat is used to avoid collision while ensuring that the trajectory tracking and target tracking of the unmanned boat are not affected. According to the reference heading angle, a target tracking controller based on model prediction is designed to control the propeller and steering gear of the unmanned boat so that the unmanned boat can track the reference heading angle.
[0063] A collision-free following control method for a flexible unmanned boat based on potential field prediction comprises the following steps:
[0064] (1) Based on the environmental detection information processing, the hybrid track method is used to guide the following ship and plan the best tracking path;
[0065] According to the nonlinear tracking model of the unmanned boat, a predictive control method is designed to set the input state space of the position, speed, angle, etc. of the formation object. The control quantity of the input space is input into the predictive control model and iterated to predict the posture state of the formation boat to form a state space. The posture state in the state space is substituted into the optimization objective function, and the sequential quadratic programming algorithm is used to solve the optimal problem, and the controlled variable that minimizes the objective function is found, and it is passed to the controlled unmanned boat to complete the trajectory tracking boat.
[0066] The following motion model of the two boats is established, and the position, speed and heading angle deviation of the following boat and the pilot boat are used as system inputs. The yaw angle constraint in the tracking process is considered, and the bias proportional guidance law is used to calculate the speed and heading reference heading angle, angular velocity and speed that the following boat needs to adjust at the next moment, so as to achieve the tracking of the pilot boat by the following boat. The trajectory calculation results are shown in Figure 2. Figure 3 As shown, the speed is Figure 4 shown.
[0067] (2) Generate collision-free trajectories based on the obstacle avoidance algorithm based on the improved artificial potential field method;
[0068] The reference obstacle avoidance bow angle is calculated according to the virtual resultant force, and the lateral reference acceleration of the following boat is calculated through the offset proportional guidance law. The following unmanned boat is guided by this acceleration to avoid obstacles and move toward the target at the same time.
[0069] (3) Adopt nonlinear model predictive control scheme to accurately track the optimal control trajectory in real time.
[0070] The dynamic prediction mechanism of the pilot boat's motion is introduced to predict the position of the pilot boat after a period of time, and a virtual boat potential field is set here. The origin of the potential field is located at the position of the virtual pilot boat. The potential field contains the gravitational potential field function, which generates an assumed gravitational force to guide the follower boat to move towards the pilot boat. At the same time, an obstacle avoidance potential field function is set. The collision avoidance potential field function contains a repulsive potential field function, which generates an assumed repulsive force to guide the follower boat to avoid obstacles. According to the gravitational potential field function and the repulsive potential field function, the virtual gravitational force and virtual repulsive force on the follower boat are calculated, and then the size and direction of the virtual resultant force are obtained. According to the virtual resultant force, the reference trajectory point at the next moment is calculated, and multiple iterations are performed to obtain a collision-free trajectory. Figure 2 This is the simulated trajectory of the two boats when they bypass an obstacle. The blue line and red line represent the collision-free trajectories of the follower boat and the lead boat, respectively.
[0071] Example:
[0072] like Figure 3 As shown, taking two unmanned boats in a circular formation as an example, the specific implementation steps of the present invention are unfolded, and the implementation method of the scheme of the present invention is described in detail in conjunction with the accompanying drawings.
[0073] The present invention proposes a collision-free following control method for a flexible unmanned boat based on potential field prediction, which includes the following two steps.
[0074] Step 1: Collect the speed, heading angle and position information of the following boat and the pilot boat, and calculate the reference speed, heading angle and angular velocity.
[0075] Step 2: If an obstacle is detected nearby, the follower boat will temporarily leave the formation, bypass the obstacle and then return to the formation.
[0076] 1) Under the control of the nonlinear prediction model method, the follower boat moves along the reference trajectory.
[0077] 2) When an obstacle is detected nearby, the position of the obstacle is recorded. The following boat leaves the formation and moves according to the collision-free trajectory obtained by the improved artificial potential field method.
[0078] 3) After bypassing the obstacle, the follower boat replans its route and continues to follow the lead boat. Figure 6 shown.
[0079] The final solution of the collision-free path in this embodiment is as follows Figure 5 shown.
[0080] The present invention designs a hybrid track line algorithm, so that the unmanned boat can flexibly select the optimal motion trajectory according to the environment, and provides an effective method for the unmanned boat formation guidance system; by introducing a virtual pilot ship and a tangential repulsive potential field, the gravitational potential field function and the repulsive potential field function are improved, and the local stagnation of the following boat in the traditional artificial potential field method is reduced. Comprehensive collision avoidance is achieved; by setting a nonlinear prediction model, the optimal control trajectory tracking is achieved in real time and accurately.
Claims
1. A collision-free following control method for a flexible unmanned boat based on potential field prediction, characterized in that: The steps include: Step 1: Construct a tracking guidance law based on bias proportional guidance when there is no collision risk; Step 2: Construct a collision avoidance guidance law based on potential field prediction when there is a collision risk; Step 3: Construct a nonlinear prediction model method based on full-state integration.
2. The method for controlling a flexible unmanned boat without collision based on potential field prediction according to claim 1, characterized in that: The step 1 is specifically as follows: The bow angles of the lead boat and the following boat are defined as is the line of sight angle between the following boat and the pilot boat; the relative position and relative velocity vector between the pilot boat and the following boat are defined as p = p t -p s and V w =V t -V s , define p s and p t is the position scalar of the lead boat and the following boat, V s and V t is the speed scalar of the lead boat and the following boat; Establish an inertial relative coordinate system {W} with the pilot boat as the origin, and the relative lateral acceleration a w Perpendicular to the relative velocity V w , the relative viewing angle Θ is defined as Relative pitch angle ψ w It is expressed as: Among them, the coefficient In the inertial relative coordinate system {W}, there exists and Relative speed V w The scalar V w and the relative lateral acceleration a w The scalar a w for: a w =a s cos(ψ-ψ w )-a t cos(ψ t -ψ w ) Among them, a t 、a s are the lateral accelerations of the lead boat and the following boat respectively; In the coordinate system {W}, because and variable and the partial derivative of Θ with respect to p is written as and After the angle constraint is introduced, the relative lateral acceleration a in the inertial relative coordinate system is w for: The lateral acceleration of the unmanned boat is a s It is expressed as: Where K is a constant coefficient.
3. The method for collision-free following control of a flexible unmanned boat based on potential field prediction according to claim 2 is characterized in that: The step 2 is specifically as follows: The obstacle avoidance process is divided into two sub-steps: (i) Determine the reference pitch angle ψ * ; (ii) Design an obstacle avoidance guidance law to ensure that the position vector When ψ(t′ f )=ψ * ; The vector representing the distance between the unmanned boat and the obstacle, t′ f It is the moment when the following boat approaches the obstacle tangent point; The yaw angle ψ is calculated based on the predicted artificial potential field method. * , gravitational potential field Based on the tangent point of the obstacle contour, we have: Where i is the number of the obstacle, and are the minimum and maximum contour angles of the obstacle respectively; the coefficient κ1 is a positive constant; r D The radius of the obstacle; Tangential force of the ith obstacle Using gravitational potential field The negative gradient of in, is a unit vector, which represents the sight angle from the following boat to the tangent point of the i-th obstacle, and can be rewritten as is a vector scalar value of ; represents the distance angle between the unmanned boat and the obstacle mass center, and the lateral potential field Located on the following boat, as the distance between the following boat and the obstacle increases, the potential energy decreases. Written as: Among them, the coefficient κ2 is a positive constant; The lateral force of the ith obstacle It is expressed as: in, is a unit vector perpendicular to the line connecting the following boat and the i-th obstacle; Can be rewritten as in is the rotation matrix, is a vector The scalar magnitude of the force F Σ (p) is written as: The reference yaw angle is equal to the direction angle F of the resultant force Σ (p), and obtain the reference yaw angle ψ * =∠F Σ (p); Lateral reference acceleration of the following boat In the inertial system, it is expressed as: in, and Is a normal number.
4. The method for controlling a flexible unmanned boat without collision based on potential field prediction according to claim 3 is characterized in that: The step 3 is specifically as follows: Set the following prediction model: The path information of the follower boat is converted into control input to form the control space X; Substituting the control input of the input space into the prediction model for iteration, predicting the future position and posture state of the boat, and forming a state space U; Substitute the posture state in the state space into the following formula to optimize the objective function: Where h(p(t),u(t)) represents the objective function, l(p(t),u(t)) is the phase cost function, and L(p(t)) is the terminal cost function; The sequential quadratic programming algorithm is used to solve the optimal problem and find the sequence that minimizes the objective function, that is, the set closest to the reference path, which is then transferred to the controlled boat.
Citation Information
Cited By
Unmanned ship cluster safety formation reconstruction control system based on local detection information
CN120085577A
Unmanned ship cluster obstacle avoidance method integrating artificial potential field method and dynamic window method
CN122239724A
