A deep-sea mining vehicle three-dimensional ocean current constraint path planning cooperation method and system

Abstract

The application belongs to the field of path planning cooperative control, and particularly relates to a three-dimensional ocean current constraint path planning cooperative method and system for a deep-sea mining vehicle. The method collects multi-source data such as seabed topography, obstacles, collection tasks and ocean current observation, establishes a double-coordinate system conversion relationship to generate a layered tetrahedral mesh, constructs a divergence free vector Gaussian process model, and outputs key indicators such as an ocean current mean vector and a covariance matrix. Then, a vehicle-terrain contact constraint model is constructed to output a vehicle feasible motion domain, a cable safety margin and a space envelope point cloud. Subsequently, a Finsler navigation metric is constructed, a multi-dimensional cost is fused to generate an anisotropic cost tensor, and path planning is converted into a geodesic line solving problem. A three-dimensional anisotropic fast marching algorithm is used to solve a third-order B-spline main path that satisfies the constraints, an ellipsoid safety pipeline is generated, interference checking is completed, and a compliant path is output. The application improves the environmental adaptability and operation safety of deep-sea mining vehicle path planning.

Description

Technical Field

[0001] This invention pertains to the field of collaborative control, specifically a collaborative method and system for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles. Background Technology

[0002] Solid mineral resources such as deep-sea polymetallic nodules are a core area of ​​global marine resource development. Tracked deep-sea mining vehicles, as the core equipment for seabed mining operations, directly determine mining efficiency, operational safety, and marine environmental protection effectiveness through the rationality of their path planning.

[0003] Existing solutions mostly use two-dimensional planar grid models, which are not adapted to the mechanical properties of complex three-dimensional seabed topography and non-uniform geological stratification, and are prone to safety accidents such as excessive ground pressure and vehicle rollover. The impact of three-dimensional ocean currents is not adequately handled. Most of them simply simplify the ocean currents as fixed resistance terms without considering the spatiotemporal uncertainty and anisotropy of ocean currents. They cannot avoid the problem of sediment plume backflow contaminating the mined area, and also lack the ability to model high-precision flow fields that meet the incompressibility of fluids. The motion of mining vehicles and the mechanical constraints of umbilical cables are not fully coupled, which can easily lead to the risk of cable bending overload, entanglement and breakage.

[0004] Therefore, a collaborative method and system for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles is needed to solve the above problems. Summary of the Invention

[0005] To address the technical problems mentioned in the background, this invention provides a collaborative method and system for three-dimensional ocean current-constrained path planning for deep-sea mining vehicles.

[0006] The objective of this invention can be achieved through the following technical solutions: The first aspect of this invention provides a collaborative method for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles, comprising the following steps: Step 1: Collect data on seabed topography, obstacles, collection tasks, mother ship location, and mining vehicle status; establish a dual coordinate system transformation relationship to generate a layered tetrahedral mesh; and output the environmental model and parameter set. Based on the layered tetrahedral mesh, acquire multi-source ocean current observation data; construct a divergence free vector Gaussian process model; and output the ocean current mean vector, common covariance matrix, and vertical shear strength index of each mesh. Step 2: Construct a vehicle-terrain contact constraint model and an umbilical cable quasi-static Cosserat model based on a layered tetrahedral mesh and mining vehicle parameters, and output the vehicle's feasible motion domain, cable safety margin function, and cable spatial envelope point cloud. Step 3: Based on the hierarchical tetrahedral mesh and ocean current mean vector, construct the Finsler navigation metric, integrate energy consumption, risk, plume recirculation and cable cost to generate a cost tensor, output the anisotropic cost tensor, and then transform the path planning into a geodesic problem. Step 4: Starting from the initial pose of the mining vehicle and ending at the target point of the collection block, embed the cost tensor into the three-dimensional anisotropic fast travel algorithm to solve for the optimal arrival value function. Then, through gradient descent backtracking, obtain the cubic B-spline main path that satisfies the curvature and slope constraints. Step 5: Generate an ellipsoidal safety pipeline based on the ocean current covariance matrix and tracking error, incorporate it into the cable spatial envelope point cloud for interference verification, and output the compliant path of the safety pipeline.

[0007] In this application, based on the data collected in step one, including seabed topography, obstacles, data collection tasks, mother ship location, and mining vehicle status, a dual coordinate system transformation relationship is established to generate a layered tetrahedral mesh, and an environmental model and parameter set are output. Multi-source ocean current observation data are obtained based on the layered tetrahedral mesh, and a divergence free vector Gaussian process model is constructed. The mean vector of ocean currents, common covariance matrix, and vertical shear strength index of each mesh are output. The specific steps are as follows: Raw point cloud data of seabed topography was collected using a multibeam echo sounder to obtain topographic point cloud data. The exploration grade data of polymetallic nodules on the seabed was used to delineate areas to be collected. The horizontal boundaries of each area were represented by an ordered vertex sequence to obtain area boundary data. Location records of known reefs and abandoned mining equipment were extracted from the historical operation database. Obstacles were verified on-site using side-scan sonar to obtain obstacle data. The positions of the mother ship and relay station were collected in real time using GPS sensors. The initial pose of the mining vehicle was obtained through vehicle-mounted sensors and battery management. Using the horizontal projection of the initial pose of the mining vehicle as the origin, with the X-axis pointing due east, the Y-axis pointing due north, and the Z-axis pointing vertically upward, a right-handed Cartesian coordinate system was formed. This right-handed Cartesian coordinate system was transformed into a local seabed coordinate system using a homogeneous transformation matrix. An initial point set was generated based on the topographic point cloud data and a preset vertical layer height of the sea surface. A constrained tetrahedral mesh was obtained by embedding the seabed topographic surface, obstacles, and preset restricted area boundaries as hard constraints into the initial point set. All environmental and mission parameters, including terrain elevation, obstacle markers, mothership and relay station locations, initial pose of the mining vehicle, remaining energy, and block boundary data, are mapped to a tetrahedral mesh using bilinear interpolation. Three-dimensional flow velocity data retrieved from the mining vehicle's onboard Doppler log is collected. The influence of ship motion on flow velocity measurement is corrected to obtain a preprocessed set of observation points. This set is then mapped to a hierarchical tetrahedral mesh using inverse distance weighted interpolation to obtain the observation values ​​on the mesh. A divergence-free kernel function is set based on any two observation points within the mesh. Several parts of a tetrahedral mesh are randomly selected as induction points. The common covariance matrix of the induction points and the mesh is obtained based on the covariance matrix of the mesh and the covariance matrix of the induction points. The posterior distribution of the current velocity is predicted for each mesh, and the mean vector of the ocean current is output. The shear strength is then calculated based on the vertical velocity profile. The vertical gradient of the horizontal velocity is calculated for each mesh using the central difference method. The vertical shear strength index is obtained based on the vertical gradient and the shear strength.

[0008] In this application, based on step two, a vehicle-terrain contact constraint and umbilical cable quasi-static Cosserat model is constructed based on the layered tetrahedral mesh and mining vehicle parameters. The model outputs the vehicle's feasible motion domain, the cable safety margin function, and the cable spatial envelope point cloud. The specific steps are as follows: Geological stratification information and seabed geological survey data from a layered tetrahedral mesh are loaded and integrated into the bottom-layer mechanical parameters. Corresponding strata mechanical parameters, including internal friction angle, cohesion, elastic modulus, and foundation bearing capacity characteristic values, are embedded into the layered tetrahedral mesh. Linear interpolation maps the element mechanical parameters to the corresponding mesh, generating a topographic mechanical property field. Mining vehicle parameters are acquired, including vehicle mass, maximum drive wheel torque, maximum design climbing angle, minimum turning radius, and maximum allowable ground pressure. Collision detection is performed on the vehicle chassis, tracks, and layered tetrahedral mesh to determine... The system calculates the average grounding specific pressure based on the contact point set, setting an average grounding specific pressure ≤ preset grounding specific pressure as a grounding specific pressure constraint. Based on the terrain's internal friction angle and cohesion, it calculates the maximum track adhesion force, setting a real-time traction force ≤ maximum track adhesion force as a track slip constraint. Key parameters of the umbilical cable are input, including total length, linear density, torsional stiffness, outer diameter, minimum allowable bending radius, and maximum allowable tensile force. The cable is discretized according to a preset length to obtain several nodes. A quasi-static equilibrium equation for the cable is established based on Cosserat rod theory, including internal force equilibrium and moment equilibrium. Centered on the current mining vehicle position, a candidate region is defined on the seabed surface of a layered tetrahedral mesh. The seabed mesh nodes within this region are used as candidate locations. Track slip constraints and ground pressure constraints are verified sequentially for each candidate location. If all constraints are satisfied, the location is marked as feasible; otherwise, it is marked as infeasible. All feasible locations are fitted using a convex hull algorithm to form a feasible motion domain for the vehicle, which is stored as a set of mesh nodes. For each vehicle position within the feasible motion domain, it is set as the boundary condition of the lower end of the umbilical cable. The quasi-static equilibrium equation is solved using the Newton-Raphson method to obtain the tensile force and curvature of each node of the cable. Based on the tensile force and curvature, the tensile safety margin and bending safety margin are calculated. A comprehensive safety margin function is set based on the tensile safety margin and bending safety margin. For all vehicle positions that meet the safety margin threshold, the coordinates of all nodes of the corresponding cable are extracted to generate a cable spatial envelope point cloud.

[0009] In this application, based on step three, a Finsler navigation metric is constructed using a layered tetrahedral mesh and ocean current mean vector. Energy consumption, risk, and cable costs are fused to generate a cost tensor, which is then output as an anisotropic cost tensor. The path planning is then transformed into a geodesic problem. The specific steps are as follows: Based on the geological properties of the tetrahedral mesh, the equivalent friction coefficient is obtained through linear interpolation. An angle tensor is constructed based on the equivalent friction coefficient. The horizontal component of the ocean current mean vector is then extracted from each mesh node. A tangent vector field is constructed using the ocean current horizontal component. The angle tensor and tangent vector field are integrated to obtain the Finsler navigation metric. Energy consumption cost is calculated based on the mining vehicle's power consumption. Slope risk is obtained by dividing the real-time slope by the maximum climbing angle, and ground pressure risk is obtained by dividing the real-time ground pressure ratio by the maximum ground pressure ratio. The slope risk and ground pressure ratio risk are fused to obtain the risk cost. The cable cost is obtained by subtracting 1 from the comprehensive safety margin function. A comprehensive cost factor is obtained by fusing energy consumption cost, risk cost, and cable cost through a linear weighting method. This comprehensive cost factor is then embedded into the Finsler navigation metric to generate an anisotropic cost tensor field. Based on the anisotropic cost tensor field, finding the optimal mining vehicle path is transformed into finding the geodesic with the shortest cost length in the Finsler space. Specifically, the starting point of the mining vehicle is set as a layered tetrahedral grid node, and the ending point is the target mining area node. The path length is obtained by arbitrarily connecting the grid node and the target mining area node. Then, the tangent vector of the path is obtained. Based on the path length and the tangent vector, the cost length of the path is obtained. The geodesic with the minimum cost length is set as the optimal path.

[0010] In this application, based on step four, starting from the initial pose of the mining vehicle and ending at the target point of the collected block, the cost tensor is embedded into a three-dimensional anisotropic fast traversal algorithm for solution to obtain the optimal arrival value function. The output cubic B-spline principal path satisfying curvature and slope constraints is obtained through gradient descent backtracking. The specific steps are as follows: Input the initial pose of the mining vehicle and map it to the grid using the nearest node method. Traverse the grid nodes and find the node with the smallest Euclidean distance from the initial pose. Set it as the grid start point. Input the target point set of the mining block and repeat the above mapping steps for each target point to obtain the grid target node set. Then select the target node closest to the initial position as the main path end point. The node with the smallest arrival value is selected from the target node set of the grid, marked as frozen and moved out of the narrow-band grid target node set. The cost tensors of the smallest node and the unfrozen nodes are obtained by centroid coordinate interpolation. The anisotropic cost length increment of the smallest node and the unfrozen nodes is calculated based on the cost tensor. The arrival value of the smallest node and the corresponding anisotropic cost length increment are added to obtain the optimal arrival value of the grid node. The dual value of its velocity vector is calculated based on the cost tensor to obtain the optimal backtracking direction. A fixed step size is moved based on the optimal backtracking direction to obtain a new path point. If the new path point is within the preset cell, it is added to the initial path point set. If the new path point is outside the preset cell, its intersection with the cell surface is calculated, the intersection is added to the initial path point set, and the backtracking continues in the adjacent tetrahedral grid. When the distance between the backtracking point and the grid starting point is less than the fixed step size, the grid starting point is added to the initial path point set, and a discrete initial path is obtained. Assign parameters to the points of the discrete initial path according to the cumulative chord length, obtain the curvature of each parameter, set curvature < preset curvature threshold as curvature constraint, set slope < preset slope as slope constraint, embed the curvature constraint and slope constraint into the discrete initial path, and perform cubic B-spline curve transformation to obtain the main path point set.

[0011] In this application, based on step five, where an ellipsoidal safety pipeline is generated based on the common covariance matrix and tracking error, and the cable spatial envelope point cloud is incorporated for interference verification to output a compliant path for the safety pipeline, the specific steps are as follows: Based on the calibration accuracy of the mining vehicle motion control, the tracking errors of the tangential and normal directions of the path are decomposed to generate a tracking error covariance matrix. The common covariance matrix and the tracking error covariance matrix are superimposed to obtain the covariance matrix of the path point. The tangential vector is obtained by normalizing the first derivative of the path point. The direction of the curvature center of the path is taken as the principal normal vector. The tangential vector and the principal normal vector are multiplied to obtain the secondary normal vector. An orthogonal transformation matrix is ​​generated through the secondary normal vector. The set of spatial points that satisfy the quadratic form of the path point covariance matrix and the orthogonal transformation matrix ≤ the confidence threshold value is obtained to obtain the safety ellipsoid of the point. The safety ellipsoids of all points on the main path point set are connected in series along the path to generate the ellipsoidal safety pipeline of the main path. Collision detection is performed on the voxel point cloud of the ellipsoidal safety pipeline and the spatial envelope point cloud of the mining vehicle cable. If a collision occurs, a new path is regenerated; otherwise, a compliant path for the safety pipeline is output.

[0012] A second aspect of this invention provides a collaborative system for three-dimensional ocean current-constrained path planning for deep-sea mining vehicles, employing the method described above, including a data acquisition module, a vehicle analysis module, a baseline analysis module, a path analysis module, and a compliant path output module. The acquisition module collects data on seabed topography, obstacles, acquisition tasks, mother ship location, and mining vehicle status. It establishes a dual coordinate system transformation relationship to generate a layered tetrahedral mesh and outputs an environmental model and parameter set. Based on the layered tetrahedral mesh, it acquires multi-source ocean current observation data, constructs a divergence free vector Gaussian process model, and outputs the ocean current mean vector, ocean current covariance matrix, and vertical shear strength index of each mesh. The vehicle analysis module constructs a vehicle-terrain contact constraint model and an umbilical cable quasi-static Cosserat model based on a hierarchical tetrahedral mesh and mining vehicle parameters, and outputs the vehicle's feasible motion domain, cable safety margin function, and cable spatial envelope point cloud. The geodesic analysis module constructs a Finsler navigation metric based on a layered tetrahedral mesh and ocean current mean vector. It integrates energy consumption, risk, plume recirculation, and cable cost to generate a cost tensor, outputs an anisotropic cost tensor, and then transforms the path planning into a geodesic problem. The path analysis module takes the initial pose of the mining vehicle as the starting point and the target point of the collection block as the ending point. It embeds the cost tensor into the three-dimensional anisotropic fast travel algorithm for solution to obtain the optimal arrival value function. Through gradient descent backtracking, it obtains the cubic B-spline main path that satisfies the curvature and slope constraints. The compliance path output module generates an ellipsoidal safety pipeline based on the common covariance matrix and tracking error, incorporates the cable spatial envelope point cloud for interference verification, and outputs the compliance path of the safety pipeline.

[0013] Compared with the prior art, the beneficial effects of the present invention are: This invention integrates vehicle-terrain contact constraints and static equations into the path planning process beforehand, rather than relying on post-process verification as in traditional solutions. The vehicle-terrain contact constraints fully incorporate real mechanical parameters such as the internal friction angle, cohesion, and bearing capacity of the seabed strata, defining the feasible motion domain for vehicles from multiple dimensions including ground pressure, track slippage, and climbing stability. This reduces the operational risks of vehicles getting stuck, slipping, or overturning in soft seabed environments. The static equations output cable safety margin functions and spatial envelope point clouds, mitigating the risks of cable bending overload, tensile breakage, and spatial entanglement from the path planning source, reducing unnecessary replanning and improving overall planning efficiency. Based on the ocean current covariance matrix and vehicle tracking error, the ellipsoidal safety pipeline adaptively adjusts its boundary according to the uncertainty of the flow field and the control deviation. Compared with the fixed-width safety channel, it avoids overly conservative path planning while ensuring operational safety. Based on multi-vehicle collaborative operation scenarios, it is adapted to the weak communication environment of deep-sea acoustic communication with low bandwidth, high latency and easy interruption. It realizes the elimination of path conflicts under sparse information exchange through a distributed collaborative mechanism, without the need for a centralized control node. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. The following drawings are not drawn to scale according to the actual size, but are intended to illustrate the main idea of ​​the present invention.

[0015] Figure 1 This is a diagram illustrating the method steps of the present invention.

[0016] Figure 2 This is a module connection diagram of the present invention. Detailed Implementation

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are also within the scope of protection of the present invention.

[0018] Please refer to Figure 1 As shown, the first aspect of this invention provides a collaborative method for three-dimensional ocean current-constrained path planning for deep-sea mining vehicles, comprising the following steps: Step 1: Collect data on seabed topography, obstacles, collection tasks, mother ship location, and mining vehicle status; establish a dual coordinate system transformation relationship to generate a layered tetrahedral mesh; and output the environmental model and parameter set. Based on the layered tetrahedral mesh, acquire multi-source ocean current observation data; construct a divergence free vector Gaussian process model; and output the ocean current mean vector, common covariance matrix, and vertical shear strength index of each mesh. Step 2: Construct a vehicle-terrain contact constraint model and an umbilical cable quasi-static Cosserat model based on a layered tetrahedral mesh and mining vehicle parameters, and output the vehicle's feasible motion domain, cable safety margin function, and cable spatial envelope point cloud. Step 3: Based on the hierarchical tetrahedral mesh and ocean current mean vector, construct the Finsler navigation metric, integrate energy consumption, risk, plume recirculation and cable cost to generate a cost tensor, output the anisotropic cost tensor, and then transform the path planning into a geodesic problem. Step 4: Starting from the initial pose of the mining vehicle and ending at the target point of the collection block, embed the cost tensor into the three-dimensional anisotropic fast travel algorithm to solve for the optimal arrival value function. Then, through gradient descent backtracking, obtain the cubic B-spline main path that satisfies the curvature and slope constraints. Step 5: Generate an ellipsoidal safety pipeline based on the ocean current covariance matrix and tracking error, incorporate it into the cable spatial envelope point cloud for interference verification, and output the compliant path of the safety pipeline.

[0019] In this application, based on the data collected in step one, including seabed topography, obstacles, data collection tasks, mother ship location, and mining vehicle status, a dual coordinate system transformation relationship is established to generate a layered tetrahedral mesh, and an environmental model and parameter set are output. Multi-source ocean current observation data are obtained based on the layered tetrahedral mesh, and a divergence free vector Gaussian process model is constructed. The mean vector of ocean currents, common covariance matrix, and vertical shear strength index of each mesh are output. The specific steps are as follows: Raw point cloud data of seabed topography was collected using a multibeam echo sounder to obtain topographic point cloud data. The exploration grade data of polymetallic nodules on the seabed was used to delineate areas to be collected. The horizontal boundaries of each area were represented by an ordered vertex sequence to obtain area boundary data. Location records of known reefs and abandoned mining equipment were extracted from the historical operation database. Obstacles were verified on-site using side-scan sonar to obtain obstacle data. The positions of the mother ship and relay station were collected in real time using GPS sensors. The initial pose of the mining vehicle and the voltage, current, and temperature parameters of the power battery pack were obtained through vehicle-mounted sensors and battery management. Using the horizontal projection of the initial pose of the mining vehicle as the origin, with the X-axis pointing due east, the Y-axis pointing due north, and the Z-axis pointing vertically upward, a right-handed Cartesian coordinate system was formed. This right-handed Cartesian coordinate system was transformed into a local seabed coordinate system using a homogeneous transformation matrix. An initial point set was generated based on the topographic point cloud data and a preset vertical layer height of the sea surface. A constrained tetrahedral mesh was obtained by embedding the seabed topographic surface, obstacles, and preset restricted area boundaries as hard constraints into the initial point set. All environmental and mission parameters, including terrain elevation, obstacle markers, mothership and relay station locations, initial pose of the mining vehicle, remaining energy, and block boundary data, are mapped to a tetrahedral mesh using bilinear interpolation. Three-dimensional flow velocity data retrieved from the mining vehicle's onboard Doppler log is collected. The influence of ship motion on flow velocity measurement is corrected to obtain a preprocessed set of observation points. This set is then mapped to a hierarchical tetrahedral mesh using inverse distance weighted interpolation to obtain the observation values ​​on the mesh. A divergence-free kernel function is set based on any two observation points within the mesh. Several parts of a tetrahedral mesh are randomly selected as induction points. The common covariance matrix of the induction points and the mesh is obtained based on the covariance matrix of the mesh and the covariance matrix of the induction points. The posterior distribution of the current velocity is predicted for each mesh, and the mean vector of the ocean current is output. The shear strength is then calculated based on the vertical velocity profile. The vertical gradient of the horizontal velocity is calculated for each mesh using the central difference method. The vertical shear strength index is obtained based on the vertical gradient and the shear strength.

[0020] In this application, based on step two, a vehicle-terrain contact constraint and umbilical cable quasi-static Cosserat model is constructed based on the layered tetrahedral mesh and mining vehicle parameters. The model outputs the vehicle's feasible motion domain, the cable safety margin function, and the cable spatial envelope point cloud. The specific steps are as follows: Geological stratification information and seabed geological survey data from a layered tetrahedral mesh are loaded and integrated into the bottom-layer mechanical parameters. Corresponding strata mechanical parameters, including internal friction angle, cohesion, elastic modulus, and foundation bearing capacity characteristic values, are embedded into the layered tetrahedral mesh. Linear interpolation maps the element mechanical parameters to the corresponding mesh, generating a topographic mechanical property field. Mining vehicle parameters are acquired, including vehicle mass, maximum drive wheel torque, maximum design climbing angle, minimum turning radius, and maximum allowable ground pressure. Collision detection is performed on the vehicle chassis, tracks, and layered tetrahedral mesh to determine... The system calculates the average grounding specific pressure based on the contact point set, setting an average grounding specific pressure ≤ preset grounding specific pressure as a grounding specific pressure constraint. Based on the terrain's internal friction angle and cohesion, it calculates the maximum track adhesion force, setting a real-time traction force ≤ maximum track adhesion force as a track slip constraint. Key parameters of the umbilical cable are input, including total length, linear density, torsional stiffness, outer diameter, minimum allowable bending radius, and maximum allowable tensile force. The cable is discretized according to a preset length to obtain several nodes. A quasi-static equilibrium equation for the cable is established based on Cosserat rod theory, including internal force equilibrium and moment equilibrium. Centered on the current mining vehicle position, a candidate region is defined on the seabed surface of a layered tetrahedral mesh. The seabed mesh nodes within this region are used as candidate locations. Track slip constraints and ground pressure constraints are verified sequentially for each candidate location. If all constraints are satisfied, the location is marked as feasible; otherwise, it is marked as infeasible. All feasible locations are fitted using a convex hull algorithm to form a feasible motion domain for the vehicle, which is stored as a set of mesh nodes. For each vehicle position within the feasible motion domain, it is set as the boundary condition of the lower end of the umbilical cable. The quasi-static equilibrium equation is solved using the Newton-Raphson method to obtain the tensile force and curvature of each node of the cable. Based on the tensile force and curvature, the tensile safety margin and bending safety margin are calculated. A comprehensive safety margin function is set based on the tensile safety margin and bending safety margin. For all vehicle positions that meet the safety margin threshold, the coordinates of all nodes of the corresponding cable are extracted to generate a cable spatial envelope point cloud.

[0021] In this application, based on step three, a Finsler navigation metric is constructed using a layered tetrahedral mesh and ocean current mean vector. Energy consumption, risk, and cable costs are fused to generate a cost tensor, which is then output as an anisotropic cost tensor. The path planning is then transformed into a geodesic problem. The specific steps are as follows: Based on the geological properties of the tetrahedral mesh, the equivalent friction coefficient is obtained through linear interpolation. An angle tensor is constructed based on the equivalent friction coefficient. The horizontal component of the ocean current mean vector is then extracted from each mesh node. A tangent vector field is constructed using the ocean current horizontal component. The angle tensor and tangent vector field are integrated to obtain the Finsler navigation metric. Energy consumption cost is calculated based on the mining vehicle's power consumption. Slope risk is obtained by dividing the real-time slope by the maximum climbing angle, and ground pressure risk is obtained by dividing the real-time ground pressure ratio by the maximum ground pressure ratio. The slope risk and ground pressure ratio risk are fused to obtain the risk cost. The cable cost is obtained by subtracting 1 from the comprehensive safety margin function. A comprehensive cost factor is obtained by fusing energy consumption cost, risk cost, and cable cost through a linear weighting method. This comprehensive cost factor is then embedded into the Finsler navigation metric to generate an anisotropic cost tensor field. Based on the anisotropic cost tensor field, finding the optimal mining vehicle path is transformed into finding the geodesic with the shortest cost length in the Finsler space. Specifically, the starting point of the mining vehicle is set as a layered tetrahedral grid node, and the ending point is the target mining area node. The path length is obtained by arbitrarily connecting the grid node and the target mining area node. Then, the tangent vector of the path is obtained. Based on the path length and the tangent vector, the cost length of the path is obtained. The geodesic with the minimum cost length is set as the optimal path.

[0022] In this application, based on step four, starting from the initial pose of the mining vehicle and ending at the target point of the collected block, the cost tensor is embedded into a three-dimensional anisotropic fast traversal algorithm for solution to obtain the optimal arrival value function. The output cubic B-spline principal path satisfying curvature and slope constraints is obtained through gradient descent backtracking. The specific steps are as follows: Input the initial pose of the mining vehicle and map it to the grid using the nearest node method. Traverse the grid nodes and find the node with the smallest Euclidean distance from the initial pose. Set it as the grid start point. Input the target point set of the mining block and repeat the above mapping steps for each target point to obtain the grid target node set. Then select the target node closest to the initial position as the main path end point. The node with the smallest arrival value is selected from the target node set of the grid, marked as frozen and moved out of the narrow-band grid target node set. The cost tensors of the smallest node and the unfrozen nodes are obtained by centroid coordinate interpolation. The anisotropic cost length increment of the smallest node and the unfrozen nodes is calculated based on the cost tensor. The arrival value of the smallest node and the corresponding anisotropic cost length increment are added to obtain the optimal arrival value of the grid node. The dual value of its velocity vector is calculated based on the cost tensor to obtain the optimal backtracking direction. A fixed step size is moved based on the optimal backtracking direction to obtain a new path point. If the new path point is within the preset cell, it is added to the initial path point set. If the new path point is outside the preset cell, its intersection with the cell surface is calculated, the intersection is added to the initial path point set, and the backtracking continues in the adjacent tetrahedral grid. When the distance between the backtracking point and the grid starting point is less than the fixed step size, the grid starting point is added to the initial path point set, and a discrete initial path is obtained. Assign parameters to the points of the discrete initial path according to the cumulative chord length, obtain the curvature of each parameter, set curvature < preset curvature threshold as curvature constraint, set slope < preset slope as slope constraint, embed the curvature constraint and slope constraint into the discrete initial path, and perform cubic B-spline curve transformation to obtain the main path point set.

[0023] In this application, based on step five, where an ellipsoidal safety pipeline is generated based on the common covariance matrix and tracking error, and the cable spatial envelope point cloud is incorporated for interference verification to output a compliant path for the safety pipeline, the specific steps are as follows: Based on the calibration accuracy of the mining vehicle motion control, the tracking errors of the tangential and normal directions of the path are decomposed to generate a tracking error covariance matrix. The common covariance matrix and the tracking error covariance matrix are superimposed to obtain the covariance matrix of the path point. The tangential vector is obtained by normalizing the first derivative of the path point. The direction of the curvature center of the path is taken as the principal normal vector. The tangential vector and the principal normal vector are multiplied to obtain the secondary normal vector. An orthogonal transformation matrix is ​​generated through the secondary normal vector. The set of spatial points that satisfy the quadratic form of the path point covariance matrix and the orthogonal transformation matrix ≤ the confidence threshold value is obtained to obtain the safety ellipsoid of the point. The safety ellipsoids of all points on the main path point set are connected in series along the path to generate the ellipsoidal safety pipeline of the main path. Collision detection is performed on the voxel point cloud of the ellipsoidal safety pipeline and the spatial envelope point cloud of the mining vehicle cable. If a collision occurs, a new path is regenerated; otherwise, a compliant path for the safety pipeline is output.

[0024] Please refer to Figure 2 As shown, a second aspect of the present invention provides a collaborative system for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles, employing the method described above, including a data acquisition module, a vehicle analysis module, a baseline analysis module, a path analysis module, and a compliant path output module. The acquisition module collects data on seabed topography, obstacles, acquisition tasks, mother ship location, and mining vehicle status. It establishes a dual coordinate system transformation relationship to generate a layered tetrahedral mesh and outputs an environmental model and parameter set. Based on the layered tetrahedral mesh, it acquires multi-source ocean current observation data, constructs a divergence free vector Gaussian process model, and outputs the ocean current mean vector, ocean current covariance matrix, and vertical shear strength index of each mesh. The vehicle analysis module constructs a vehicle-terrain contact constraint model and an umbilical cable quasi-static Cosserat model based on a hierarchical tetrahedral mesh and mining vehicle parameters, and outputs the vehicle's feasible motion domain, cable safety margin function, and cable spatial envelope point cloud. The geodesic analysis module constructs a Finsler navigation metric based on a layered tetrahedral mesh and ocean current mean vector. It integrates energy consumption, risk, plume recirculation, and cable cost to generate a cost tensor, outputs an anisotropic cost tensor, and then transforms the path planning into a geodesic problem. The path analysis module takes the initial pose of the mining vehicle as the starting point and the target point of the collection block as the ending point. It embeds the cost tensor into the three-dimensional anisotropic fast travel algorithm for solution to obtain the optimal arrival value function. Through gradient descent backtracking, it obtains the cubic B-spline main path that satisfies the curvature and slope constraints. The compliance path output module generates an ellipsoidal safety pipeline based on the common covariance matrix and tracking error, incorporates the cable spatial envelope point cloud for interference verification, and outputs the compliance path of the safety pipeline.

[0025] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, ATA hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state ATA hard disk.

[0026] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0027] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0028] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0029] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0030] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0031] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable ATA hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0032] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A collaborative method for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles, characterized in that, Includes the following steps: Step 1: Collect seabed topography, obstacles, collection tasks, mother ship location and mining vehicle status data, establish dual coordinate system transformation relationship to generate layered tetrahedral mesh, and output environmental model and parameter set. Based on the layered tetrahedral mesh, obtain multi-source ocean current observation data, construct divergence free vector Gaussian process model, and output ocean current mean vector, common covariance matrix and vertical shear strength index of each mesh. Step 2: Construct a vehicle-terrain contact constraint model and quasi-static equilibrium equations based on hierarchical tetrahedral mesh and mining vehicle parameters, and output the vehicle feasible motion domain, cable safety margin function and cable spatial envelope point cloud; Step 3: Based on the hierarchical tetrahedral mesh and ocean current mean vector, construct the Finsler navigation metric, integrate energy consumption, risk, plume recirculation and cable cost to generate a cost tensor, output the anisotropic cost tensor, and then transform the path planning into a geodesic problem. Step 4: Starting from the initial pose of the mining vehicle and ending at the target point of the collection block, embed the cost tensor into the three-dimensional anisotropic fast travel algorithm to solve for the optimal arrival value function. Then, through gradient descent backtracking, obtain the cubic B-spline main path that satisfies the curvature and slope constraints. Step 5: Generate an ellipsoidal safety pipeline based on the ocean current covariance matrix and tracking error, incorporate it into the cable spatial envelope point cloud for interference verification, and output the compliant path of the safety pipeline.

2. The collaborative method for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles according to claim 1, characterized in that, The specific steps are as follows: Data on seabed topography, obstacles, data collection tasks, mother ship location, and mining vehicle status are collected; a dual coordinate system transformation relationship is established to generate a layered tetrahedral mesh; and an environmental model and parameter set are output. Raw point cloud data of seabed topography is collected using a multibeam echo sounder to obtain topographic point cloud data. The exploration grade data of polymetallic nodules on the seabed is used to delineate the blocks to be collected. The horizontal boundaries of each block to be collected are represented by an ordered vertex sequence to obtain block boundary data. The location records of known reefs and abandoned mining equipment are extracted from the historical operation database. Obstacles are verified on-site using side-scan sonar to obtain obstacle data. The location of the mother ship and the relay station are collected in real time using GPS sensors. The initial pose of the mining vehicle is obtained through vehicle-mounted sensors and vehicle-mounted battery management. Using the horizontal projection of the initial pose of the mining vehicle as the origin, with the X-axis pointing due east, the Y-axis pointing due north, and the Z-axis pointing vertically upward, a right-handed Cartesian coordinate system is formed. This right-handed Cartesian coordinate system is then transformed into a local seabed coordinate system using a homogeneous transformation matrix. An initial point set is generated based on the terrain point cloud data and the preset vertical layering height of the sea surface. The seabed terrain surface, obstacles, and the preset restricted area boundary are then embedded into the initial point set as hard constraints using a constrained tetrahedral mesh, resulting in a layered tetrahedral mesh.

3. The collaborative method for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles according to claim 2, characterized in that, Based on multi-source ocean current observation data acquired using a layered tetrahedral grid, a divergence free vector Gaussian process model is constructed, outputting the ocean current mean vector, common covariance matrix, and vertical shear intensity index for each grid. The specific steps are as follows: All environmental and mission parameters, including terrain elevation, obstacle markers, mothership and relay station locations, initial pose of the mining vehicle, remaining energy, and block boundary data, are mapped to a tetrahedral mesh using bilinear interpolation. Three-dimensional flow velocity data retrieved from the mining vehicle's onboard Doppler log is collected. The influence of ship motion on flow velocity measurement is corrected to obtain a preprocessed set of observation points. This set is then mapped to a hierarchical tetrahedral mesh using inverse distance-weighted interpolation to obtain the observation values ​​on the mesh. A divergence-free kernel function is set based on any two observation points within the mesh. Several parts of a tetrahedral mesh are randomly selected as induction points. The common covariance matrix of the induction points and the mesh is obtained based on the covariance matrix of the mesh and the covariance matrix of the induction points. The posterior distribution of the current velocity is predicted for each mesh, and the mean vector of the ocean current is output. The shear strength is then calculated based on the vertical velocity profile. The vertical gradient of the horizontal velocity is calculated for each mesh using the central difference method. The vertical shear strength index is obtained based on the vertical gradient and the shear strength.

4. The collaborative method for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles according to claim 1, characterized in that, The specific steps for constructing vehicle-terrain contact constraints and quasi-static equilibrium equations based on layered tetrahedral meshes and mining vehicle parameters are as follows: Geological stratification information and seabed geological survey data from a layered tetrahedral mesh are loaded and integrated into the bottom-layer mechanical parameters. Corresponding strata mechanical parameters, including internal friction angle, cohesion, elastic modulus, and foundation bearing capacity characteristic values, are embedded into the layered tetrahedral mesh. Linear interpolation maps the element mechanical parameters to the corresponding mesh, generating a topographic mechanical property field. Mining vehicle parameters are acquired, including vehicle mass, maximum drive wheel torque, maximum design climbing angle, minimum turning radius, and maximum allowable ground pressure. Collision detection is performed on the vehicle chassis, tracks, and layered tetrahedral mesh to determine... The system calculates the average grounding specific pressure based on the contact point set, setting an average grounding specific pressure ≤ preset grounding specific pressure as a grounding specific pressure constraint. Based on the terrain's internal friction angle and cohesion, it calculates the maximum track adhesion force, setting a real-time traction force ≤ maximum track adhesion force as a track slip constraint. Key parameters of the umbilical cable are input, including total length, linear density, torsional stiffness, outer diameter, minimum allowable bending radius, and maximum allowable tensile force. The cable is discretized according to a preset length to obtain several nodes. A quasi-static equilibrium equation for the cable is established based on Cosserat rod theory, including internal force equilibrium and moment equilibrium.

5. The collaborative method for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles according to claim 1, characterized in that, The specific steps for outputting the vehicle's feasible motion domain, cable safety margin function, and cable spatial envelope point cloud are as follows: Centered on the current mining vehicle position, a candidate region is delineated on the seabed surface of a layered tetrahedral mesh. The seabed mesh nodes within this region are used as candidate locations. Track slip constraints and ground pressure constraints are verified sequentially for each candidate location. If all constraints are satisfied, the location is marked as feasible; otherwise, it is marked as infeasible. All feasible locations are fitted using a convex hull algorithm to form a feasible motion domain for the vehicle, which is stored as a set of mesh nodes. For each vehicle position within the feasible motion domain, it is set as the boundary condition of the lower end of the umbilical cable. The quasi-static equilibrium equation is solved using the Newton-Raphson method to obtain the tensile force and curvature of each node of the cable. Based on the tensile force and curvature, the tensile safety margin and bending safety margin are calculated. A comprehensive safety margin function is set based on the tensile safety margin and bending safety margin. For all vehicle locations that meet the safety margin threshold, extract all node coordinates of their corresponding cables to generate a cable spatial envelope point cloud.

6. The collaborative method for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles according to claim 1, characterized in that, Based on a layered tetrahedral mesh and ocean current mean vector, a Finsler navigation metric is constructed. Energy consumption, risk, and cable costs are fused to generate a cost tensor, which outputs an anisotropic cost tensor. The path planning is then transformed into a geodesic problem. The specific steps are as follows: Based on the geological properties of the tetrahedral mesh, the equivalent friction coefficient is obtained through linear interpolation. An angle tensor is constructed based on the equivalent friction coefficient. The horizontal component of the ocean current mean vector is then extracted from each mesh node. A tangent vector field is constructed using the ocean current horizontal component. The angle tensor and tangent vector field are integrated to obtain the Finsler navigation metric. Energy consumption cost is calculated based on the mining vehicle's power consumption. Slope risk is obtained by dividing the real-time slope by the maximum climbing angle, and ground pressure risk is obtained by dividing the real-time ground pressure ratio by the maximum ground pressure ratio. The slope risk and ground pressure ratio risk are fused to obtain the risk cost. The cable cost is obtained by subtracting 1 from the comprehensive safety margin function. A comprehensive cost factor is obtained by fusing energy consumption cost, risk cost, and cable cost through a linear weighting method. This comprehensive cost factor is then embedded into the Finsler navigation metric to generate an anisotropic cost tensor field. Based on the anisotropic cost tensor field, finding the optimal mining vehicle path is transformed into finding the geodesic with the shortest cost length in the Finsler space. Specifically, the starting point of the mining vehicle is set as a layered tetrahedral grid node, and the ending point is the target mining area node. The path length is obtained by arbitrarily connecting the grid node and the target mining area node. Then, the tangent vector of the path is obtained. Based on the path length and the tangent vector, the cost length of the path is obtained. The geodesic with the minimum cost length is set as the optimal path.

7. The collaborative method for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles according to claim 1, characterized in that, Starting from the initial pose of the mining vehicle and ending at the target point of the mining block, the cost tensor is embedded into a 3D anisotropic fast traversal algorithm for solution, yielding the optimal arrival function. Gradient descent backtracking is then used to obtain the cubic B-spline principal path that satisfies curvature and slope constraints. The specific steps are as follows: Input the initial pose of the mining vehicle and map it to the grid using the nearest node method. Traverse the grid nodes and find the node with the smallest Euclidean distance from the initial pose. Set it as the grid start point. Input the target point set of the mining block and repeat the above mapping steps for each target point to obtain the grid target node set. Then select the target node closest to the initial position as the main path end point. The node with the smallest arrival value is selected from the target node set of the grid, marked as frozen and moved out of the narrow-band grid target node set. The cost tensors of the smallest node and the unfrozen nodes are obtained by centroid coordinate interpolation. The anisotropic cost length increment of the smallest node and the unfrozen nodes is calculated based on the cost tensor. The arrival value of the smallest node and the corresponding anisotropic cost length increment are added to obtain the optimal arrival value of the grid node. The dual value of its velocity vector is calculated based on the cost tensor to obtain the optimal backtracking direction. A fixed step size is moved based on the optimal backtracking direction to obtain a new path point. If the new path point is within the preset cell, it is added to the initial path point set. If the new path point is outside the preset cell, its intersection with the cell surface is calculated, the intersection is added to the initial path point set, and the backtracking continues in the adjacent tetrahedral grid. When the distance between the backtracking point and the grid starting point is less than the fixed step size, the grid starting point is added to the initial path point set, and a discrete initial path is obtained. Assign parameters to the points of the discrete initial path according to the cumulative chord length, obtain the curvature of each parameter, set curvature < preset curvature threshold as curvature constraint, set slope < preset slope as slope constraint, embed the curvature constraint and slope constraint into the discrete initial path, and perform cubic B-spline curve transformation to obtain the main path point set.

8. The collaborative method for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles according to claim 1, characterized in that, An ellipsoidal safety pipeline is generated based on the common covariance matrix and tracking error. The compliance path of the safety pipeline is then output through interference verification by incorporating the cable spatial envelope point cloud. The specific steps are as follows: Based on the calibration accuracy of the mining vehicle motion control, the tracking errors of the tangential and normal directions of the path are decomposed to generate a tracking error covariance matrix. The common covariance matrix and the tracking error covariance matrix are superimposed to obtain the covariance matrix of the path point. The tangential vector is obtained by normalizing the first derivative of the path point. The direction of the curvature center of the path is taken as the principal normal vector. The tangential vector and the principal normal vector are multiplied to obtain the secondary normal vector. An orthogonal transformation matrix is ​​generated through the secondary normal vector. The set of spatial points that satisfy the quadratic form of the path point covariance matrix and the orthogonal transformation matrix ≤ the confidence threshold value is obtained to obtain the safety ellipsoid of the point. The safety ellipsoids of all points on the main path point set are connected in series along the path to generate the ellipsoidal safety pipeline of the main path. Collision detection is performed on the voxel point cloud of the ellipsoidal safety pipeline and the spatial envelope point cloud of the mining vehicle cable. If a collision occurs, a new path is regenerated; otherwise, a compliant path for the safety pipeline is output.

9. A collaborative system for three-dimensional ocean current-constrained path planning of deep-sea mining vehicles, implementing the method of any one of claims 1-8, characterized in that, It includes a data acquisition module, a vehicle analysis module, a baseline analysis module, a route analysis module, and a compliance route output module. The acquisition module collects data on seabed topography, obstacles, acquisition tasks, mother ship location, and mining vehicle status. It establishes a dual coordinate system transformation relationship to generate a layered tetrahedral mesh and outputs an environmental model and parameter set. Based on the layered tetrahedral mesh, it acquires multi-source ocean current observation data, constructs a divergence free vector Gaussian process model, and outputs the ocean current mean vector, ocean current covariance matrix, and vertical shear strength index for each mesh. The vehicle analysis module constructs a vehicle-terrain contact constraint model and an umbilical cable quasi-static Cosserat model based on a hierarchical tetrahedral mesh and mining vehicle parameters, and outputs the vehicle's feasible motion domain, cable safety margin function, and cable spatial envelope point cloud. The geodesic analysis module constructs a Finsler navigation metric based on a layered tetrahedral mesh and ocean current mean vector. It integrates energy consumption, risk, plume recirculation, and cable cost to generate a cost tensor, outputs an anisotropic cost tensor, and then transforms the path planning into a geodesic problem. The path analysis module takes the initial pose of the mining vehicle as the starting point and the target point of the collection block as the ending point. It embeds the cost tensor into the three-dimensional anisotropic fast travel algorithm for solution to obtain the optimal arrival value function. Through gradient descent backtracking, it obtains the cubic B-spline main path that satisfies the curvature and slope constraints. The compliance path output module generates an ellipsoidal safety pipeline based on the common covariance matrix and tracking error, incorporates the cable spatial envelope point cloud for interference verification, and outputs the compliance path of the safety pipeline.

Images