Navigation and positioning method for unmanned offshore work platform vehicle in deep sea and weak signal

Abstract

The present invention discloses a navigation and positioning method for an unmanned working vehicle on an offshore work platform in deep sea and weak signal conditions. First, an ArUco marking and positioning layer is constructed to obtain the position information and posture information of the marked unmanned working vehicle; a YOLOv10 obstacle detection layer is constructed to detect obstacles in an image in real time and obtain obstacle boundary box information; a K-means clustering layer is constructed to cluster the detected obstacles and classify adjacent or similar obstacles into one category; a path planning layer is constructed to calculate the optimal path from the starting point to the target point based on the obtained position information and posture information of the unmanned working vehicle, the clustering results of the obstacles, and the static map information GIS, and dynamically update the path and obstacle avoidance strategy to ensure the safe and efficient operation of the unmanned working vehicle in a complex environment.

Description

Technical Field

[0001] The present invention belongs to the technical field of navigation methods, and in particular relates to a navigation and positioning method for an unmanned offshore work platform vehicle under deep sea and weak signal conditions. Background Art

[0002] With the increase in marine resource exploration and deep-sea operations, unmanned operating vehicles (UGVs) have gradually become an important tool for performing complex tasks in offshore and deep-sea environments. However, in these extreme environments, UGVs face many challenges, especially when GPS signals are weak or completely lost. Traditional navigation and positioning methods often perform poorly under these conditions and have difficulty providing reliable path planning and obstacle avoidance capabilities. Therefore, a new navigation and positioning method is needed to solve the problems in these environments in order to improve the autonomy and reliability of UGVs in offshore and deep-sea operations. Summary of the invention

[0003] The present invention is mainly to overcome the deficiencies of the prior art and to provide a navigation and positioning method for an unmanned offshore work platform vehicle in deep sea and weak signal conditions.

[0004] The present invention is achieved through the following technical solutions:

[0005] A navigation and positioning method for an unmanned working vehicle on an offshore work platform in deep sea and weak signal conditions comprises the following steps:

[0006] Step 1: construct an ArUco marker positioning layer to detect the ArUco markers on the unmanned vehicle in the captured image and calculate the position information R and posture information T of the marked unmanned vehicle;

[0007] Step 2: Build the YOLOv10 obstacle detection layer to detect obstacles in the image in real time and obtain obstacle bounding box information. ,in( x 1, y 1) and ( x 2, y 2) are the coordinates of the upper left and lower right corners of the bounding box, c is the confidence score;

[0008] Step 3: Construct a K-means clustering layer to cluster the detected obstacles and classify adjacent or similar obstacles into one category. Its input is the obstacle bounding box information D obtained in step 2, and its output is the two-dimensional coordinates of the cluster center of the obstacle. ;

[0009] Step 4: Construct the path planning layer based on the position information R and posture information T of the unmanned vehicle obtained in step 1 and the two-dimensional coordinates of the cluster center of the obstacles obtained in step 3. , and static map information GIS, calculate the optimal path from the starting point to the target point, and dynamically update the path and obstacle avoidance strategy.

[0010] In the above technical solution, in step 1, ArUco markers are placed on the unmanned vehicle, and the camera captures and analyzes the marker image to achieve accurate positioning of the unmanned vehicle; the input is the image captured by the camera , the output is the position information R and posture information T of the marked unmanned vehicle;

[0011] The position information R and posture information T are solved using the solvePnP function in OpenCV. Let the detected ArUco marker corner point be , the coordinates of these corner points in the world coordinate system are , use the PnP algorithm to solve the rotation vector and translation vector, including the following formula:

[0012] ;

[0013] ;

[0014] ;

[0015] ;

[0016] Where: distCoeffs is the distortion coefficient, rvec is the rotation vector, tvec is the translation vector, M is the camera's intrinsic matrix, and H is the homography matrix obtained through ArUco marker detection. f x and f y is the focal length, c x and c y is the optical center coordinate, rvec= i u, u is the unit rotation axis, i is the angle of rotation about this axis.

[0017] In the above technical solution, in step 2, the objective function of YOLOv10 obstacle detection is:

[0018] ;

[0019] The loss function of YOLOv10 obstacle detection is:

[0020] ;

[0021] L coordis the positioning loss:

[0022] ;

[0023] L conf is the confidence loss:

[0024] ;

[0025] L class For the category loss:

[0026] ;

[0027] in: S : The size of the grid; B : The number of bounding boxes predicted for each grid cell; If i The first grid cell j If a bounding box contains an object, it is 1, otherwise it is 0; If i The first grid cell j If a bounding box does not contain an object, it is 1, otherwise it is 0; Predicted bounding box parameters; The actual bounding box parameters; Predicted class probabilities; The actual class label; It is a weight coefficient used to adjust the weight associated with the loss term of undetected targets in the loss function, and its value is between 0 and 1; classes: The number of target categories that the model needs to detect and classify, and its value is equal to the total number of categories defined in the dataset.

[0028] In the above technical solution, in step 3, the objective function of K-means clustering is to minimize the sum of the squares of the distances from each sample in the cluster to the cluster center:

[0029] ;

[0030] ;

[0031] in: K : the number of clusters; C k : No. k clusters; No. k The center of the cluster; k i : No. i The feature vector of an obstacle contains the location, size, shape and other feature information of the obstacle.

[0032] In the above technical solution, in step 4, the A* algorithm is used to find the shortest path from the starting point to the end point in a static environment;

[0033] The A* algorithm evaluates the total cost of each node, which consists of two parts: the actual cost from the starting point to the current node n g (n) and the estimated cost from the current node to the end point h (n), total cost f The formula for (n) is:

[0034] ;

[0035] ;

[0036] ;

[0037] ;

[0038] in:( x r , y r ): The current node position of the unmanned vehicle is obtained from the position information R of the unmanned vehicle; ( x t , y t ): the target node position of the unmanned vehicle; r is the influence radius of the obstacle, which is measured by the binocular camera on the unmanned vehicle.

[0039] In the above technical solution, in step 4, a DWA algorithm is used to adjust the speed of the unmanned vehicle in real time in a dynamic environment to avoid obstacles;

[0040] The DWA algorithm samples multiple groups of speeds in the speed space, simulates the trajectories of the unmanned vehicle at these speeds for a certain period of time, and then evaluates these trajectories and selects the optimal trajectory to drive the unmanned vehicle; its steps include speed sampling, trajectory prediction and trajectory evaluation; speed sampling space is the unmanned vehicle linear velocity v and angular velocity The set of constraints on which the set is subject includes: velocity boundary constraints, acceleration constraints, and obstacle avoidance constraints;

[0041] Speed ​​Boundary Limits:

[0042] ;

[0043] in, v min: The minimum linear speed allowed by the unmanned vehicle, which indicates the lowest speed that the unmanned vehicle can reach during its travel; v max : The maximum linear speed allowed by the unmanned vehicle, which indicates the maximum speed that the unmanned vehicle can reach during its travel; oh min : The minimum angular velocity allowed by the unmanned vehicle, which indicates the lowest rotation speed that the unmanned vehicle can reach during the turning process; oh max : The maximum angular velocity allowed by the unmanned vehicle, which indicates the maximum rotation speed that the unmanned vehicle can reach during the turning process; V m : represents the sampling space under the speed boundary limit;

[0044] Acceleration Limitation:

[0045] ;

[0046] Among them, α vmax : The linear acceleration limit of the unmanned vehicle, which indicates the maximum value of the linear velocity change of the unmanned vehicle in unit time; α ωmax : angular acceleration limit of the unmanned vehicle, indicating the maximum value of the angular velocity change of the unmanned vehicle per unit time; Δt: time step, indicating the time interval of the unmanned vehicle performing uniform linear motion or circular motion within a period of time; V d : represents the sampling space under acceleration limit;

[0047] Obstacle avoidance constraints:

[0048] ;

[0049] in, Indicates that at a given speed Next, the distance between the simulated trajectory and the nearest obstacle; : represents the sampling space under obstacle avoidance constraints;

[0050] The final velocity sampling space is the intersection of these three sets , represents the set of all speeds that meet the conditions under speed boundary restrictions, acceleration restrictions and obstacle avoidance constraints; the unmanned vehicle will select the appropriate linear speed from this set and angular velocity Perform trajectory planning;

[0051] Assume that the unmanned vehicle has If the unmanned vehicle performs uniform linear motion or circular motion, the position update formula of the unmanned vehicle is:

[0052] ;

[0053] in, The horizontal coordinate of the current position of the unmanned vehicle in the global coordinate system; The ordinate of the current position of the unmanned vehicle in the global coordinate system; The current heading angle of the unmanned vehicle, which indicates the direction the unmanned vehicle is pointing in the global coordinate system; The current linear speed of the unmanned vehicle, which indicates the speed at which the unmanned vehicle moves in the current direction; The current angular velocity of the unmanned operating vehicle indicates the speed at which the unmanned operating vehicle rotates.

[0054] The advantages and beneficial effects of the present invention are:

[0055] This paper combines ArUco tagging technology for high-precision visual positioning, and uses the YOLOv10 network for real-time obstacle detection and recognition. In addition, the K-means clustering algorithm is used to effectively cluster the detected obstacles to improve the robustness of obstacle avoidance. Finally, the ArUco tagging and clustering results are combined to generate an optimized navigation route to ensure the safe and efficient operation of the unmanned vehicle in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a flow chart of the navigation and positioning method of the offshore work platform unmanned operation vehicle under deep sea and weak signal of the present invention.

[0057] For ordinary technicians in this field, other relevant drawings can be obtained based on the above drawings without any creative work. DETAILED DESCRIPTION

[0058] In order to enable those skilled in the art to better understand the technical solution of the present invention, the technical solution of the present invention is further described below in conjunction with specific embodiments.

[0059] The present invention designs a navigation and positioning method for an unmanned working vehicle on an offshore work platform in deep sea and weak signal conditions, and the specific steps are as follows:

[0060] Step 1: Construct an ArUco marker positioning layer to detect the ArUco markers on the unmanned operating vehicle in the image and calculate the position information R and posture information T of the marked unmanned operating vehicle.

[0061] Specifically, ArUco tagging is a positioning method based on computer vision. By placing ArUco tags on the unmanned vehicle, the camera captures the tag image and analyzes it, thereby achieving accurate positioning of the unmanned vehicle. The input is the image captured by the camera , the output is the position information R and posture information T of the marked unmanned operating vehicle.

[0062] The position and attitude calculation formula is based on the PnP problem, which is solved by the solvePnP function in OpenCV. Suppose the detected ArUco marker corner point is , the coordinates of these corner points in the world coordinate system are , use the PnP algorithm to solve the rotation vector and translation vector:

[0063] ;

[0064] ;

[0065] ;

[0066] ;

[0067] Where: distCoeffs is the distortion coefficient, rvec is the rotation vector, tvec is the translation vector, M is the camera's intrinsic matrix, and H is the homography matrix obtained through ArUco marker detection. f x and f y is the focal length, c x and c y is the optical center coordinate, rvec= i u, u is the unit rotation axis, i is the angle of rotation about this axis.

[0068] Step 2: Build the YOLOv10 obstacle detection layer to detect obstacles in the image in real time. Its input is image , the output is the detected obstacle bounding box information ,in( x 1, y 1) and ( x 2, y 2) are the coordinates of the upper left and lower right corners of the bounding box, c is the confidence score.

[0069] The objective function of YOLOv10 obstacle detection is:

[0070] ;

[0071] The loss function of YOLOv10 obstacle detection is:

[0072] ;

[0073] L coord is the positioning loss:

[0074] ;

[0075] L conf is the confidence loss:

[0076] ;

[0077] L class For the category loss:

[0078] ;

[0079] in: S : The size of the grid; B : The number of bounding boxes predicted for each grid cell; If i The first grid cell j If a bounding box contains an object, it is 1, otherwise it is 0; If i The first grid cell j If a bounding box does not contain an object, it is 1, otherwise it is 0; Predicted bounding box parameters; The actual bounding box parameters; Predicted class probabilities; The actual class label; It is a weight coefficient used to adjust the weight associated with the loss term of the undetected target (ie, the background area) in the loss function, and its value is between 0 and 1; classes: The number of target categories that the model needs to detect and classify, and its value is equal to the total number of categories defined in the dataset.

[0080] Step 3: Construct a K-means clustering layer to cluster the detected obstacles and classify adjacent obstacles into one category. Its input is the obstacle bounding box information D obtained in step 2, and its output is the two-dimensional coordinates of the cluster center of the obstacle. .

[0081] The objective function of K-means clustering is to minimize the sum of the squared distances from each sample in each cluster to the cluster center:

[0082] ;

[0083] ;

[0084] in: K : the number of clusters;C k : No. k clusters; No. k The center of the cluster; k i : No. i The feature vector of an obstacle contains the location, size, shape and other feature information of the obstacle.

[0085] Step 4: Construct the path planning layer based on the position information R and posture information T of the unmanned vehicle obtained in step 1 and the two-dimensional coordinates of the cluster center of the obstacles obtained in step 3. , and static map information GIS, calculate the optimal path from the starting point to the target point, and dynamically update the path and obstacle avoidance strategy.

[0086] The input is: the position information R and posture information T of the unmanned vehicle obtained in step 1, and the two-dimensional coordinates of the cluster center of the obstacle obtained in step 3 , as well as static map information GIS and the two-dimensional coordinates of the destination that the unmanned vehicle needs to reach ; Output: updated safe obstacle avoidance navigation path.

[0087] The A* algorithm is used to find the shortest path from the starting point to the end point in a static environment;

[0088] The A* algorithm evaluates the total cost of each node, which consists of two parts: the actual cost from the starting point to the current node n g (n) and the estimated cost from the current node to the end point h (n), total cost f The formula for (n) is:

[0089] ;

[0090] ;

[0091] ;

[0092] ;

[0093] in:( x r , y r ): The current node position of the unmanned vehicle is obtained from the position information R of the unmanned vehicle; ( x t , y t ): the target node position of the unmanned vehicle; r is the influence radius of the obstacle, which is measured by the binocular camera on the unmanned vehicle.

[0094] The DWA algorithm is used to adjust the speed of the unmanned vehicle in real time in a dynamic environment to avoid obstacles.

[0095] The DWA algorithm samples multiple groups of speeds in the speed space, simulates the trajectories of the unmanned vehicle at these speeds for a certain period of time, and then evaluates these trajectories and selects the optimal trajectory to drive the unmanned vehicle; its steps include speed sampling, trajectory prediction and trajectory evaluation; speed sampling space is the unmanned vehicle linear velocity v and angular velocity The set of constraints on which the set is subject includes: velocity boundary constraints, acceleration constraints, and obstacle avoidance constraints;

[0096] Speed ​​Boundary Limits:

[0097] ;

[0098] in, v min : The minimum linear speed allowed by the unmanned vehicle, which indicates the lowest speed that the unmanned vehicle can reach during its travel; v max : The maximum linear speed allowed by the unmanned vehicle, which indicates the maximum speed that the unmanned vehicle can reach during its travel; oh min : The minimum angular velocity allowed by the unmanned vehicle, which indicates the lowest rotation speed that the unmanned vehicle can reach during the turning process; oh max : The maximum angular velocity allowed by the unmanned vehicle, which indicates the maximum rotation speed that the unmanned vehicle can reach during the turning process; V m : represents the sampling space under the speed boundary limit;

[0099] Acceleration Limitation:

[0100] ;

[0101] Among them, α vmax : The linear acceleration limit of the unmanned vehicle, which indicates the maximum value of the linear velocity change of the unmanned vehicle in unit time; α ωmax : angular acceleration limit of the unmanned vehicle, indicating the maximum value of the angular velocity change of the unmanned vehicle per unit time; Δt: time step, indicating the time interval of the unmanned vehicle performing uniform linear motion or circular motion within a period of time; V d : represents the sampling space under acceleration limit;

[0102] Obstacle avoidance constraints:

[0103] ;

[0104] in, Indicates that at a given speed Next, the distance between the simulated trajectory and the nearest obstacle; : represents the sampling space under obstacle avoidance constraints;

[0105] The final velocity sampling space is the intersection of these three sets , represents the set of all speeds that meet the conditions under speed boundary restrictions, acceleration restrictions and obstacle avoidance constraints; the unmanned vehicle will select the appropriate linear speed from this set and angular velocity Perform trajectory planning;

[0106] Assume that the unmanned vehicle has If the unmanned vehicle performs uniform linear motion or circular motion, the position update formula of the unmanned vehicle is:

[0107] ;

[0108] in, The horizontal coordinate of the current position of the unmanned vehicle in the global coordinate system; The ordinate of the current position of the unmanned vehicle in the global coordinate system; The current heading angle of the unmanned vehicle, which indicates the direction the unmanned vehicle is pointing in the global coordinate system; The current linear speed of the unmanned vehicle, which indicates the speed at which the unmanned vehicle moves in the current direction; The current angular velocity of the unmanned operating vehicle indicates the speed at which the unmanned operating vehicle rotates.

[0109] The present invention is described above by way of example. It should be noted that, without departing from the core of the present invention, any simple deformation, modification or other equivalent replacement that can be made by those skilled in the art without inventive effort falls within the protection scope of the present invention.

Claims

1. A navigation and positioning method for an unmanned offshore work platform vehicle in deep sea and weak signal conditions, characterized in that: The following steps are involved: Step 1: construct an ArUco marker positioning layer to detect the ArUco markers on the unmanned vehicle in the captured image and calculate the position information R and posture information T of the marked unmanned vehicle; In step 1, ArUco markers are placed on the unmanned vehicle, and the camera captures and analyzes the marker image to achieve accurate positioning of the unmanned vehicle; the input is the image I captured by the camera, and the output is the position information R and posture information T of the marked unmanned vehicle; The position information R and attitude information T are solved using the solvePnP function in OpenCV. Assume that the detected ArUco marker corner points are p1, p2, p3, and p4, and the coordinates of these corner points in the world coordinate system are P1, P2, P3, and P4. The PnP algorithm is used to solve the rotation vector and the translation vector, including the following formulas: solvePnP([P1,P2,P3,P4],[P1,P2,p3,p4],M,distCoeffs,rvec,tvec) R=I+sin(θ)M+(1-cos(θ))M 2 T=M -1 H Where: distCoeffs is the distortion coefficient, rvec is the rotation vector, tvec is the translation vector, M is the camera's intrinsic matrix, H is the homography matrix obtained by ArUco marker detection, and f x and f y is the focal length, c x and c y is the coordinate of the optical center, rvec = θu, u is the unit rotation axis, θ is the angle of rotation around the axis; Step 2: Build a YOLOv10 obstacle detection layer to detect obstacles in the image in real time and obtain the obstacle bounding box information D = (x1, y1, x2, y2, c), where (x1, y1) and (x2, y2) are the coordinates of the upper left corner and lower right corner of the bounding box, and c is the confidence score; In step 2, the objective function of YOLOv10 obstacle detection is: The loss function of YOLOv10 obstacle detection is: L L =L coord +L conf +L class L coord is the positioning loss: L conf is the confidence loss: L class The category loss is: Where: S: the size of the grid; B: the number of bounding boxes predicted for each grid cell; 1 if the jth bounding box in the i-th grid cell contains an object, 0 otherwise; If the jth bounding box in the i-th grid cell does not contain an object, it is 1, otherwise it is 0; (x i ,y i ,w i ,h i ): predicted bounding box parameters; The actual bounding box parameters; p i (c): predicted class probability; The actual class label; λ noobj : is a weight coefficient used to adjust the weight associated with the loss term of undetected targets in the loss function, with a value between 0 and 1; classes: the number of target categories that the model needs to detect and classify, and the value is equal to the total number of categories defined in the dataset; Step 3: Construct a K-means clustering layer to cluster the detected obstacles and classify adjacent or similar obstacles into one category. Its input is the obstacle bounding box information D obtained in step 2, and its output is the two-dimensional coordinates of the cluster center C = (x c ,y c ); Step 4: Construct the path planning layer. According to the position information R and posture information T of the unmanned vehicle obtained in step 1, the two-dimensional coordinates C of the cluster center of the obstacle obtained in step 3 = (x c ,y c ), as well as static map information GIS, calculate the optimal path from the starting point to the target point, and dynamically update the path and obstacle avoidance strategy.

2. The navigation and positioning method for an unmanned offshore work platform vehicle in deep sea and weak signal conditions according to claim 1 is characterized by: In step 3, the objective function of K-means clustering is to minimize the sum of squared distances from samples in each cluster to the cluster center: Where: K: the number of clusters; C k : kth cluster; μ k : The center of the kth cluster; k i : The feature vector of the ith obstacle, including the location, size, shape and other feature information of the obstacle.

3. The navigation and positioning method for an unmanned offshore work platform vehicle in deep sea and weak signal conditions according to claim 1 is characterized by: In step 4, the A* algorithm is used to find the shortest path from the starting point to the end point in a static environment; The A* algorithm evaluates the total cost of each node, which consists of two parts: the actual cost g(n) from the starting point to the current node n and the estimated cost h(n) from the current node to the end point. The formula for the total cost f(n) is: f(n)=g(n)+h(n) h(n)=|x t -x r |+|y t -y r | g(n)=0, when Where: (x r ,y r ): The current node position of the unmanned vehicle, obtained from the position information R of the unmanned vehicle; (x t ,y t ): the target node position of the unmanned vehicle; r is the influence radius of the obstacle, which is measured by the binocular camera on the unmanned vehicle.

4. The method for navigation and positioning of an unmanned offshore work platform vehicle in deep sea and weak signal conditions according to claim 3 is characterized by: In step 4, the DWA algorithm is used to adjust the speed of the unmanned vehicle in real time in a dynamic environment to avoid obstacles; The DWA algorithm samples multiple sets of speeds in the speed space and simulates the trajectories of the unmanned vehicle at these speeds for a certain period of time. It then evaluates these trajectories and selects the optimal trajectory to drive the unmanned vehicle. The steps include velocity sampling, trajectory prediction and trajectory evaluation; the velocity sampling space V s is the set of the unmanned vehicle’s linear velocity v and angular velocity ω, which is subject to constraints including velocity boundary constraints, acceleration constraints, and obstacle avoidance constraints; Speed ​​Boundary Limits: V m ={(v,ω)|v∈[v min ,v max ],ω∈[ω min ,ω max ]} Among them, v min : The minimum linear speed allowed by the unmanned vehicle, which means the lowest speed that the unmanned vehicle can reach during its travel; v max : The maximum linear speed allowed by the unmanned vehicle, which indicates the maximum speed that the unmanned vehicle can reach during its travel; ω min : The minimum angular velocity allowed by the unmanned vehicle, indicating the lowest rotation speed that the unmanned vehicle can reach during the turning process; ω max V: The maximum angular velocity allowed by the unmanned vehicle, which indicates the maximum rotation speed that the unmanned vehicle can reach during the turning process; m : represents the sampling space under the speed boundary limit; Acceleration Limitation: V d ={(v,ω)|v∈[va vmax ·Δt,v+a vmax ·Δt],ω∈[ω-a ωmax ·Δt,ω+a ωmax ·Δt]} Among them, α vmax : The linear acceleration limit of the unmanned vehicle, which indicates the maximum value of the linear velocity change of the unmanned vehicle in unit time; α ωmax : angular acceleration limit of the unmanned vehicle, indicating the maximum value of the angular velocity change of the unmanned vehicle per unit time; Δt: time step, indicating the time interval of the unmanned vehicle performing uniform linear motion or circular motion within a period of time; V d : represents the sampling space under acceleration limit; Obstacle avoidance constraints: Where dist(v,ω) represents the distance between the simulated trajectory and the nearest obstacle at a given speed (v,ω); V a : represents the sampling space under obstacle avoidance constraints; The final velocity sampling space is the intersection of these three sets V s =V m ∩V d ∩V a , represents the set of all speeds that meet the conditions under the speed boundary limit, acceleration limit and obstacle avoidance constraint; the unmanned vehicle will select the appropriate linear speed v and angular speed ω from this set for trajectory planning; Assuming that the unmanned vehicle performs uniform linear motion or circular motion in each time step Δt, the position update formula of the unmanned vehicle is: x(t)=x(t-1)+v(t)·Δt·cos(θ(t-1)) y(t)=y(t-1)+v(t)·Δt·sin(θ(t-1)) θ(t)=θ(t-1)+ω(t)·Δt Among them, x(t): the horizontal coordinate of the current position of the unmanned operating vehicle in the global coordinate system; y(t): the vertical coordinate of the current position of the unmanned operating vehicle in the global coordinate system; θ(t): the current heading angle of the unmanned operating vehicle, indicating the direction pointed by the unmanned operating vehicle in the global coordinate system; v(t): the current linear velocity of the unmanned operating vehicle, indicating the moving speed of the unmanned operating vehicle in the current direction; ω(t): the current angular velocity of the unmanned operating vehicle, indicating the rotation speed of the unmanned operating vehicle.

Images