An intelligent underwater path planning system and method based on machine vision

Abstract

The present invention relates to the field of robots, and in particular to an intelligent underwater path planning system and method based on machine vision, wherein the system comprises an underwater environment monitoring module, a fluid simulation unit, and a path planning module; the underwater environment monitoring module is used to monitor the velocity and direction of water flow around entities and bodies in the underwater environment; the fluid simulation unit performs fluid mechanics simulation analysis on the influence of water flow on the motion characteristics of the body according to the water flow monitoring data around the body; the path planning module dynamically updates the travel cost of the body moving in each direction at each spatial node according to the body mass, movement speed, and the combined external force of the environment, and then plans the minimum travel cost path of the body, and also sets a spatial obstacle avoidance threshold according to the error value between the predicted position and the actual position of the body when moving, and corrects the travel cost of the spatial nodes around the environmental entities, thereby realizing early obstacle avoidance of the obstacle entities.

Description

Technical Field

[0001] The present invention relates to the field of robotics technology, and in particular to an intelligent underwater path planning system and method based on machine vision. Background Art

[0002] With the rapid development of robot technology, the path planning and obstacle avoidance technology of related land robots has been continuously updated, but there are some technical barriers to the path planning technology of underwater robots. On the one hand, due to the greater resistance of robots in underwater environments, and the influence of fluid buoyancy is more obvious than that in land environments, the robot needs to comprehensively consider obstacles in three-dimensional space when moving and plan a three-dimensional route. On the other hand, in underwater path planning, the force of the surrounding fluid on the robot cannot be comprehensively considered, and the influence of the resistance and thrust of the water flow on the endurance of the robot during the movement process cannot be comprehensively considered. It is impossible to avoid the energy loss of resisting the water flow, and it is impossible to comprehensively utilize the thrust of the water flow. In addition, it is also necessary to comprehensively consider the deviation of the actual position of the body caused by the changing water flow, which in turn creates the risk of obstacle avoidance of the body.

[0003] Therefore, it is necessary to use an intelligent underwater path planning system and method based on machine vision to solve the above technical obstacles. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent underwater path planning system and method based on machine vision to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] An intelligent underwater path planning method based on machine vision, the method comprising the following analysis steps:

[0007] Step S100: When the aircraft starts to move, a three-dimensional coordinate system is constructed with the center point of the aircraft structure as the coordinate origin, the horizontal component of the aircraft facing direction as the positive semi-axis of the x-axis, the right side of the horizontal plane perpendicular to the x-axis direction as the positive semi-axis of the y-axis, and the vertical upward direction as the positive semi-axis of the z-axis. The map is divided into unit space nodes by unit coordinates; the underwater environment is divided, and each area of ​​the environment is analyzed in detail to plan the path more accurately;

[0008] Step S200: using sonar to monitor the surrounding environment of the body in real time, and obtaining the distance between each position of the underwater environment and the entity and the coordinate origin;

[0009] Step S300: Setting a visual detection threshold and monitoring the position coordinates of entities around the body in the underwater environment in real time. When the distance between the entity part and the coordinate origin is less than the visual detection threshold, using a binocular vision sensor for three-dimensional reconstruction;

[0010] Step S400: monitoring the direction and speed of the water flow around the body, and performing fluid-solid coupling simulation on the fluid in the surrounding environment of the body to analyze the influence of the water flow on the motion characteristics of the body;

[0011] Step S500: calculating the travel cost of the body in a dynamic environment in real time according to the influence of the water flow on the movement characteristics of the body, and planning a minimum travel cost path;

[0012] Step S600: According to the position of the aircraft and the orientation of the three-dimensional reconstructed model of the environment entity, the travel cost of the obstacle position is corrected, and obstacle avoidance operations are performed on the aircraft.

[0013] According to the above technical solution, in step S300, the propagation speed of sound waves in the water environment around the body is measured, and the movement distance of the body and the transmission distance of the sound waves are calculated according to the sending and collecting time of the sound waves, and then the orientation of the entity position and its coordinates in the three-dimensional coordinate system are calculated according to the trigonometric function of the position of the body sending the sound waves, the position of the sound waves collected, and the position of the entity reflecting the sound waves; the sonar is used to measure the position and distance of the entity in the underwater environment, so that the environmental characteristics and the entity information therein can be preliminarily grasped in the underwater environment with low visibility;

[0014] Set a visual detection threshold. When the distance between the coordinates of an entity position point and the origin coordinates is less than the visual detection threshold, turn on active lighting and binocular vision sensors to capture images of the entity position points, and calculate the coordinate dot matrix of each point in the visible part of the origin in the entity, and then reconstruct the outline of the entity in three dimensions. When it is detected that the entity in the environment is close to the body, turn on the binocular vision sensor, and use active lighting to monitor the entity, and adaptively turn on the visual monitoring system to reduce the endurance pressure.

[0015] According to the above technical solution, in step S400, a model is established in a three-dimensional coordinate system according to the body contour data. When the body moves, the water flow direction and speed in each unit coordinate space around the body contour of the current space node are collected, and the fluid-solid coupling simulation is performed on the fluid unit around the body contour to obtain the motion velocity component v affected by the water flow when the body moves. 流 When the body moves from one spatial node to the next spatial node, the surrounding fluid information is collected in real time, and through fluid mechanics simulation, the water flow effects on the body in the underwater environment can be more accurately analyzed.

[0016] According to the above technical solution, in step S500, the power data required for the center of the body structure to move one unit coordinate space in each direction in still water of the current water density is collected; when the body starts to move, the travel cost of each unit coordinate space in each direction in the map is initialized to the power required for the center of the body structure to move one unit coordinate space in each direction in the still water state, and the minimum travel cost path is pre-planned using the D*Lite algorithm based on the initial travel cost of each unit coordinate space in each direction in the underwater environment;

[0017] When the body is moving, the gravity G and the inherent buoyancy F of the water body at the current spatial node are obtained and calculated. 浮 , and according to the velocity component v of the water flow when the body moves 流 and the displacement from the current spatial node to the next spatial node to calculate the power F of the body moving one spatial node 动 , and set the travel cost from the current spatial node to the next spatial node to F 动 The value of; According to the equation: Calculate the power required for the body to move to the next spatial node; where t is the time taken to move forward a spatial node, m is the body's own mass, v0 is the body's speed at the current spatial node, x is the displacement of the body to the next spatial node, and F 动 、F 浮 ,G,v0,v 流 , x participates in the calculation as a vector;

[0018] Analyze the relevant factors of the underwater environment that affect the movement state of the aircraft, calculate the required power and direction according to the direction of the aircraft to the next spatial node after vector addition and subtraction, and dynamically update the cost of the aircraft moving in each direction at each spatial node based on the thrust of the water flow on the moving aircraft, so as to more accurately optimize the aircraft's endurance by utilizing the characteristics of the underwater environment.

[0019] According to the above technical solution, in step S600, the predicted position of the center of the body structure before traveling and the actual position after traveling are monitored in real time, and the number of traveling steps n between the two positions is calculated. 误差 , set the spatial obstacle avoidance threshold n 避障 =α*v*n 误差 +C; where α is the obstacle avoidance risk coefficient, v is the speed of the aircraft, and C is the basic obstacle avoidance threshold;

[0020] According to the spatial obstacle avoidance threshold n 避障 , determine the position of the spatial node outside the boundary of the entity model, if there is a travel number less than n 避障If the marked spatial node is reachable, the current spatial node is recorded as the obstacle avoidance spatial node; when the body structure center moves from the current spatial node to the next spatial node, if the body outline and the obstacle avoidance spatial node coincide with each other, the cost of the current spatial node moving in the direction where the collision will occur is set to infinity; when the body structure center moves from the current spatial node to the next spatial node, if the body outline and the obstacle avoidance spatial node do not coincide with each other, the aircraft moves along the original path;

[0021] Dynamically set the spatial obstacle avoidance threshold of the aircraft. When the aircraft travels at different speeds and with different water flow fluents, different values ​​are assigned to the spatial obstacle avoidance threshold, realizing adaptive obstacle avoidance operation of the aircraft in different underwater situations. Adding the basic obstacle avoidance threshold during calculation prevents the obstacle avoidance operation from failing when the aircraft is temporarily suspended, resulting in the inability to avoid moving obstacles.

[0022] An intelligent underwater path planning system based on machine vision, the system comprises an underwater environment monitoring module, a fluid simulation unit, and a path planning module; the output end of the underwater environment monitoring module is connected to the fluid simulation unit and the input end of the path planning module; the output end of the fluid simulation unit is connected to the input end of the path planning module;

[0023] The underwater environment monitoring module is used to monitor the flow rate and direction of water around entities and bodies in the underwater environment; the fluid simulation unit performs fluid mechanics simulation analysis on the impact of water flow on the movement characteristics of the body based on the water flow monitoring data around the body; the path planning module dynamically updates the travel cost of the body in each direction at each spatial node according to the body mass, movement speed, and the combined external force of the environment, and then plans the minimum travel cost path for the body. It also sets the spatial obstacle avoidance threshold according to the error value between the predicted position and the actual position of the body when moving, and corrects the travel cost of the spatial nodes around the environmental entities, thereby achieving early obstacle avoidance of obstacle entities.

[0024] The underwater environment monitoring module includes an environment entity monitoring unit and a water flow monitoring unit; the output end of the environment entity monitoring unit is connected to the input end of the path planning module; the water flow monitoring unit is connected to the input end of the fluid simulation unit;

[0025] The environmental entity monitoring unit uses sonar to perform initial monitoring operations on the environment. When the distance between the environmental entity and the body is less than the visual detection threshold, the binocular vision sensor is used to reconstruct the environmental entity in three dimensions; the water flow monitoring unit is used to monitor the flow velocity and flow direction information of the water flow around the body.

[0026] The path planning module includes a travel cost analysis unit, a path analysis unit, and an obstacle avoidance correction unit; the input end of the travel cost analysis unit is connected to the output end of the underwater environment monitoring module and the fluid simulation unit; the input end of the obstacle avoidance correction unit is connected to the output end of the underwater environment monitoring module; the output ends of the travel cost analysis unit and the obstacle avoidance correction unit are connected to the input end of the path analysis unit;

[0027] The travel cost analysis unit analyzes the travel cost of the aircraft moving from the current spatial node to all directions according to the gravity of the aircraft, the inherent buoyancy of the current water body on the aircraft, and the cooperative force of the water body when the aircraft moves; the path analysis unit uses the D*Lite algorithm to perform minimum travel cost path planning according to the dynamically updated travel cost; the obstacle avoidance correction unit dynamically sets the spatial obstacle avoidance threshold according to the detected environmental entity position and the error value between the actual position and the predicted position of the aircraft, and then corrects the travel cost of moving to the spatial nodes around the environmental entity.

[0028] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0029] The present invention is an environmental monitoring method that combines sonar and binocular vision when the robot is traveling in an unknown underwater environment, so as to ensure that the robot can grasp map information more accurately when traveling in the environment; when planning the underwater path for the robot, the force of the underwater fluid on the robot is comprehensively considered, and the fluid thrust and buoyancy are analyzed by fluid simulation, and the fluid thrust and buoyancy are utilized to further improve the energy endurance of the underwater robot; when the robot is traveling in the underwater environment, the travel cost of the robot at each spatial node is dynamically updated according to the change of the water flow around the robot and the collected environmental entity position, so that the robot can adjust the travel route in real time according to the change of the surrounding environment in the unknown changing underwater environment; a spatial obstacle avoidance threshold is set, in which the actual position error of the robot when traveling and the travel speed of the robot are introduced as calculation parameters, the position deviation of the robot in the underwater environment of different states is fully considered, and the obstacle avoidance distance of the robot to the environmental entity is appropriately adjusted to ensure that the robot can safely avoid obstacles when traveling at different speeds. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0031] Figure 1 It is a structural connection diagram of an intelligent underwater path planning system based on machine vision of the present invention;

[0032] Figure 2 It is an organizational structure diagram of an intelligent underwater path planning system based on machine vision of the present invention;

[0033] Figure 3 It is a flow chart of an intelligent underwater path planning method based on machine vision of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] See also Figure 1-Figure 3 , the present invention provides a technical solution:

[0036] An intelligent underwater path planning method based on machine vision includes the following analysis steps:

[0037] Step S100: When the aircraft starts to move, a three-dimensional coordinate system is constructed with the center point of the aircraft structure as the coordinate origin, the horizontal component of the aircraft facing direction as the positive semi-axis of the x-axis, the right side of the horizontal plane perpendicular to the x-axis direction as the positive semi-axis of the y-axis, and the vertical upward direction as the positive semi-axis of the z-axis. The map is divided into unit space nodes by unit coordinates; the underwater environment is divided, and each area of ​​the environment is analyzed in detail to plan the path more accurately;

[0038] In actual use, a float can be placed synchronously when the robot enters the water as the origin of the three-dimensional coordinate system. At the same time, the float can send a sonar signal of a specific frequency to locate the body, and then the actual position of the body underwater is located according to the actual position of the float, avoiding the high absorption rate of electromagnetic waves by the water body, which leads to the inability to realize the positioning function when the communication environment is poor and there is a lack of reference objects.

[0039] Step S200: using sonar to monitor the surrounding environment of the body in real time, and obtaining the distance between each position of the underwater environment and the entity and the coordinate origin;

[0040] Step S300: Setting a visual detection threshold and monitoring the position coordinates of entities around the body in the underwater environment in real time. When the distance between the entity part and the coordinate origin is less than the visual detection threshold, using a binocular vision sensor for three-dimensional reconstruction;

[0041] Step S400: monitoring the direction and speed of the water flow around the body, fitting the water flow around the body into a fluid unit, and analyzing the influence of the water flow on the movement characteristics of the body;

[0042] Step S500: calculating the travel cost of the body in a dynamic environment in real time according to the influence of the water flow on the movement characteristics of the body, and planning a minimum travel cost path;

[0043] Step S600: According to the position of the aircraft and the orientation of the three-dimensional reconstructed model of the environment entity, the travel cost of the obstacle position is corrected, and obstacle avoidance operations are performed on the aircraft.

[0044] In step S300, the propagation speed of sound waves in the water environment around the body is measured, and the movement distance of the body and the transmission distance of the sound waves are calculated according to the sending and collecting time of the sound waves, and then the orientation of the entity position and its coordinates in the three-dimensional coordinate system are calculated according to the trigonometric function of the body sending sound wave position, the collecting sound wave position, and the entity position point reflecting the sound wave; the sonar is used to measure the position and distance of the entity in the underwater environment, and the environmental characteristics and the entity information therein can be preliminarily grasped in the underwater environment with low visibility; in actual operation, sound wave signals of different frequencies can be periodically sent in sequence to avoid the simultaneous collection of distant reflections and near reflections of the same-frequency signals of different periods, resulting in signal processing errors and misjudgment of environmental data;

[0045] Set a visual detection threshold. When the distance between the coordinates of the entity position point and the origin coordinates is less than the visual detection threshold, turn on active lighting and binocular vision sensors to collect images of the entity position points, and calculate the coordinate dot matrix of each point in the visible part of the origin of the entity, and then reconstruct the outline of the entity in three dimensions; when it is detected that the entity in the environment is close to the body, turn on the binocular vision sensor, and use active lighting to monitor the entity, and adaptively turn on the visual monitoring system to reduce the endurance pressure;

[0046] When setting the visual detection threshold, the value can be set according to the visibility of the actual underwater environment; when the water is turbid, the light is poor, and the dive is deep, the effective collection distance of the visual sensor is short, and the visual detection threshold can be reduced to reduce the energy loss of the visual collection system when it is unable to collect effective images; and in some known smaller waters, when the underwater visibility is high, the visual detection threshold can be increased. If the maximum size of the known water area is smaller than the visual detection threshold, sonar detection can be omitted and the visual collection system can be used directly to collect images and reconstruct three-dimensional reconstruction of the underwater environment.

[0047] In step S400, a model is established in a three-dimensional coordinate system according to the body contour data. When the body moves, the flow direction and velocity of the water flow in each unit coordinate space around the body contour of the current space node are collected, and the fluid-solid coupling simulation is performed on the fluid unit around the body contour to obtain the motion velocity component v affected by the water flow when the body moves. 流 By collecting the fluid information around the body in real time and performing fluid mechanics simulation, the effects of water flow on the body while it is moving in an underwater environment can be analyzed more accurately.

[0048] In actual use, when the body is moving underwater, the velocity and direction of the surrounding water flow are collected in real time. The body model and water flow motion data are used as simulation inputs for fluid-solid coupling simulation in COMSOL software to obtain the motion speed curve of the body simulated in the underwater environment. The obtained curve data represents the expected data of the body's travel speed when the body is at the current movement speed and no additional power is applied in the underwater environment. In a shorter environmental data monitoring period, the result error caused by environmental mutations can be ignored, and the simulation prediction data is used as the body speed prediction data under the influence of environmental water flow; starting from the starting point, partial data of the movement time length of a spatial node is selected in the curve, and the average value is taken after discrete sampling at equal intervals as the velocity component generated by the water flow effect on the body when it is moving.

[0049] In step S500, the power data required for the center of the body structure to move one unit coordinate space in each direction in still water of the current water density is collected; when the body starts to move, the travel cost of each unit coordinate space in each direction in the map is initialized to the power required for the center of the body structure to move one unit coordinate space in each direction in the still water state, and the minimum travel cost path is pre-planned using the D*Lite algorithm based on the initial travel cost of each unit coordinate space in each direction in the underwater environment;

[0050] When the body is moving, the gravity G and the inherent buoyancy F of the water body at the current spatial node are obtained and calculated. 浮 , and according to the velocity component v of the water flow when the body moves 流 and the displacement from the current spatial node to the next spatial node to calculate the power F of the body moving one spatial node 动 , and set the travel cost from the current spatial node to the next spatial node to F 动 The value of; According to the equation: Calculate the power required for the body to move to the next spatial node; where t is the time taken to move forward a spatial node, m is the body's own mass, v0 is the body's speed at the current spatial node, x is the displacement of the body to the next spatial node, and F 动 、F 浮 ,G,v0,v 流 , x participates in the calculation as a vector;

[0051] When the body is actually moving, if it encounters a water flow in the opposite direction of movement, and exerts resistance on the body, assuming that the body's gravity and buoyancy cancel each other out, v 流 The direction of the component in the direction of motion is opposite to v0, as shown in the above equation, v0+v 流 Some vector sum values ​​are smaller than those in still water, so it is necessary to increase F 动 Value, the body needs to provide greater power to overcome the water resistance, and the cost of the current direction is also updated to the increased F动 Value; if the body encounters a water flow in the same direction as the movement, it will exert a forward thrust on the body. Assuming that the body's gravity and buoyancy cancel each other out, v 流 The direction of the motion component of v0 is the same as that of v0, as shown in the above equation, v0+v 流 The sum of some vectors is larger than that in still water, which can reduce F 动 Value, the body can use the water thrust to reduce power output and improve endurance, and the cost of the current direction is also corrected to the reduced power size; when the body's gravity and the resultant force are downward in the vertical direction, assuming v0+v 流 The vertical component is 0, and the x vertical component is upward, as shown in the above equation, F 动 A larger power component is required in the vertical direction to overcome the combined force of gravity and buoyancy, so a larger power is required;

[0052] Analyze the relevant factors of the underwater environment that affect the movement state of the aircraft, calculate the required power and direction according to the direction of the aircraft to the next spatial node after vector addition and subtraction, and dynamically update the cost of the aircraft moving in each direction at each spatial node based on the thrust of the water flow on the moving aircraft, so as to more accurately optimize the aircraft's endurance by utilizing the characteristics of the underwater environment.

[0053] In step S600, the predicted position of the center of the body structure before traveling and the actual position after traveling are monitored in real time, and the number of travel steps n between the two positions is calculated. 误差 , set the spatial obstacle avoidance threshold n 避障 =α*v*n 误差 +C; where α is the obstacle avoidance risk coefficient, v is the speed of the aircraft, and C is the basic obstacle avoidance threshold;

[0054] According to the spatial obstacle avoidance threshold n 避障 , determine the position of the spatial node outside the boundary of the entity model, if there is a travel number less than n 避障 If the marked spatial node is reachable, the current spatial node is recorded as the obstacle avoidance spatial node; when the body structure center moves from the current spatial node to the next spatial node, if the body outline and the obstacle avoidance spatial node coincide with each other, the cost of the current spatial node moving in the direction where the collision will occur is set to infinity; when the body structure center moves from the current spatial node to the next spatial node, if the body outline and the obstacle avoidance spatial node do not coincide with each other, the aircraft moves along the original path;

[0055] When actually setting the obstacle avoidance risk coefficient, it can be set according to the visual detection threshold and the water flow rate. When the underwater visibility is good, the visual detection threshold value of the aircraft increases, and the obstacle avoidance risk coefficient can be reduced accordingly; when the water flow rate is fast, the azimuth deviation of the aircraft due to underwater inertia and buoyancy will increase accordingly, and the obstacle avoidance risk coefficient can be increased at this time; when setting the basic obstacle avoidance threshold, it can be set according to the size and mass of the aircraft itself. The minimum value of the basic obstacle avoidance threshold of the aircraft in actual use should be greater than the maximum distance from the center of the aircraft structure to the outer contour, to avoid the situation where the distance between the center of the aircraft structure and the obstacle is greater than the spatial obstacle avoidance threshold and the outer contour has collided with the obstacle during obstacle avoidance monitoring.

[0056] Dynamically set the spatial obstacle avoidance threshold of the aircraft. When the aircraft travels at different speeds and with different water flow fluents, different values ​​are assigned to the spatial obstacle avoidance threshold, realizing adaptive obstacle avoidance operation of the aircraft in different underwater situations. Adding the basic obstacle avoidance threshold during calculation prevents the obstacle avoidance operation from failing when the aircraft is temporarily suspended, resulting in the inability to avoid moving obstacles.

[0057] An intelligent underwater path planning system based on machine vision, comprising an underwater environment monitoring module, a fluid simulation unit, and a path planning module;

[0058] The underwater environment monitoring module is used to monitor the flow rate and direction of water around entities and bodies in the underwater environment; the fluid simulation unit performs fluid mechanics simulation based on the water flow monitoring data around the body to analyze the impact of water flow on the body's motion characteristics; the path planning module dynamically updates the cost of the body's movement in each direction at each spatial node based on the body's mass, movement speed, and the combined external force of the environment, and then plans the body's minimum travel cost path. It also sets the spatial obstacle avoidance threshold based on the error value between the predicted position and the actual position of the body when it is moving, and corrects the travel cost of the spatial nodes around the environmental entities, thereby achieving early obstacle avoidance of obstacle entities.

[0059] The underwater environment monitoring module includes an environmental entity monitoring unit and a water flow monitoring unit;

[0060] The environmental entity monitoring unit uses sonar to perform initial monitoring operations on the environment. When the distance between the environmental entity and the aircraft body is less than the visual detection threshold, the binocular vision sensor is used to reconstruct the environmental entity in three dimensions; the water flow monitoring unit is used to monitor the flow velocity and direction information of the water flow around the aircraft body.

[0061] The path planning module includes a travel cost analysis unit, a path analysis unit, and an obstacle avoidance correction unit;

[0062] The travel cost analysis unit analyzes the travel cost of the aircraft moving from the current spatial node to all directions based on the gravity of the aircraft, the inherent buoyancy of the current water body on the aircraft, and the cooperative force of the water body when the aircraft moves; the path analysis unit uses the D*Lite algorithm to perform minimum travel cost path planning based on the dynamically updated travel cost; the obstacle avoidance correction unit dynamically sets the spatial obstacle avoidance threshold based on the detected environmental entity position and the error value between the actual position and the predicted position of the aircraft, and then corrects the travel cost of moving to the spatial nodes around the environmental entity.

[0063] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0064] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent underwater path planning method based on machine vision, characterized in that: The method comprises the following analysis steps: Step S100: When the aircraft starts to move, a three-dimensional coordinate system is constructed with the center point of the aircraft structure as the coordinate origin, the horizontal component of the aircraft facing direction as the positive semi-axis of the x-axis, the right side of the horizontal plane perpendicular to the x-axis direction as the positive semi-axis of the y-axis, and the vertical upward direction as the positive semi-axis of the z-axis, and the map is divided into unit space nodes with unit coordinates; Step S200: using sonar to monitor the surrounding environment of the body in real time, and obtaining the distance between each position of the underwater environment and the entity and the coordinate origin; Step S300: Setting a visual detection threshold and monitoring the position coordinates of entities around the body in the underwater environment in real time. When the distance between the entity part and the coordinate origin is less than the visual detection threshold, using a binocular vision sensor for three-dimensional reconstruction; Step S400: monitoring the direction and speed of the water flow around the body, and performing fluid-solid coupling simulation on the fluid in the surrounding environment of the body to analyze the influence of the water flow on the motion characteristics of the body; Step S500: calculating the travel cost of the body in a dynamic environment in real time according to the influence of the water flow on the movement characteristics of the body, and planning a minimum travel cost path; Step S600: According to the position of the aircraft and the orientation of the three-dimensional reconstructed model of the environment entity, the travel cost of the obstacle position is corrected, and the obstacle avoidance operation is performed on the aircraft; Among them, in step S400, a model is established in a three-dimensional coordinate system according to the body contour data. When the body moves, the flow direction and speed of the water flow in each unit coordinate space around the body contour of the current space node are collected, and the fluid-solid coupling simulation of the environmental fluid around the body contour is performed to obtain the motion velocity component v affected by the water flow when the body moves. 流 ; In step S500, the power data required for the center of the body structure to move one unit coordinate space in each direction in still water of the current water density is collected; when the body starts to move, the travel cost of each unit coordinate space in each direction in the map is initialized to the power required for the center of the body structure to move one unit coordinate space in each direction in the still water state, and D is used according to the initial travel cost of each unit coordinate space in each direction in the underwater environment. * Lite algorithm performs minimum travel cost path pre-planning; When the body is moving, the gravity G and the inherent buoyancy F of the water body at the current spatial node are obtained and calculated. 浮 , and according to the velocity component v of the water flow when the body moves 流 and the displacement from the current spatial node to the next spatial node to calculate the power F of the body moving one spatial node 动 , and set the travel cost from the current spatial node to the next spatial node to F 动 The value of; According to the equation: Calculate the power required for the body to move to the next spatial node; where t is the time taken to move forward a spatial node, m is the body's own mass, v0 is the body's speed at the current spatial node, x is the displacement of the body to the next spatial node, and F 动 、F 浮 ,G,v0,v 流 , x participates in the calculation as a vector; In step S600, the predicted position of the center of the aircraft structure before traveling and the actual position after traveling are monitored in real time, and the number of travel steps n between the two positions is calculated. 误差 , set the spatial obstacle avoidance threshold n 避障 =α*v*n 误差 +C; where α is the obstacle avoidance risk coefficient, v is the speed of the aircraft, and C is the basic obstacle avoidance threshold; According to the spatial obstacle avoidance threshold n 避障 , determine the position of the spatial node outside the boundary of the entity model, if there is a travel number less than n 避障 If the marked spatial node is reachable, the current spatial node is recorded as the obstacle avoidance spatial node; when the body outline and the obstacle avoidance spatial node coincide with each other when the body structure center moves from the current spatial node to the next spatial node, the cost of the current spatial node moving in the direction where the collision will occur is set to infinity; when the body outline and the obstacle avoidance spatial node do not coincide with each other when the body structure center moves from the current spatial node to the next spatial node, the original path is followed.

2. The method for intelligent underwater path planning based on machine vision according to claim 1, characterized in that: In step S300, the propagation speed of sound waves in the water environment around the body is measured, and the movement distance of the body and the transmission distance of the sound waves are calculated according to the time of sending and collecting the sound waves, and then the orientation of the entity position and its coordinates in the three-dimensional coordinate system are calculated according to the trigonometric functions of the position of the body sending the sound waves, the position of the sound waves collected, and the entity position point reflecting the sound waves; A visual detection threshold is set. When the distance between the coordinates of an entity position point and the coordinates of the origin is less than the visual detection threshold, active lighting and binocular vision sensors are turned on to collect images of the entity position points, and the coordinate lattice of each point in the visible part of the origin in the entity is calculated, and then the outline of the entity is reconstructed in three dimensions, and all spatial nodes touched by the entity three-dimensional model in the three-dimensional coordinate system are marked as obstacle spatial nodes.

3. An intelligent underwater path planning system based on machine vision using the intelligent underwater path planning method based on machine vision according to any one of claims 1 to 2, characterized in that: The system includes an underwater environment monitoring module, a fluid simulation unit, and a path planning module; The underwater environment monitoring module is used to monitor the flow rate and direction of water around entities and bodies in the underwater environment; the fluid simulation unit performs fluid mechanics simulation analysis on the influence of water flow on the motion characteristics of the body based on the water flow monitoring data around the body; the path planning module dynamically updates the travel cost of the body in each direction at each spatial node according to the body mass, movement speed, and the combined external force of the environment, and then plans the minimum travel cost path of the body, and also sets the spatial obstacle avoidance threshold according to the error value between the predicted position and the actual position of the body when it is moving, and corrects the travel cost of the spatial nodes around the environmental entities, thereby achieving early obstacle avoidance of obstacle entities; The output end of the underwater environment monitoring module is connected to the fluid simulation unit and the input end of the path planning module; the output end of the fluid simulation unit is connected to the input end of the path planning module.

4. The intelligent underwater path planning system based on machine vision according to claim 3, characterized in that: The underwater environment monitoring module includes an environment entity monitoring unit and a water flow monitoring unit; The environmental entity monitoring unit uses sonar to perform initial monitoring operations on the environment. When the distance between the environmental entity and the aircraft body is less than the visual detection threshold, the binocular visual sensor is used to perform three-dimensional reconstruction of the environmental entity. The water flow monitoring unit is used to monitor the flow velocity and flow direction information of the water flow around the aircraft body. The output end of the environmental entity monitoring unit is connected to the input end of the path planning module; the water flow monitoring unit is connected to the input end of the fluid simulation unit.

5. The intelligent underwater path planning system based on machine vision according to claim 3, characterized in that: The path planning module includes a travel cost analysis unit, a path analysis unit, and an obstacle avoidance correction unit; The travel cost analysis unit analyzes the travel cost of the body moving from the current spatial node to each direction according to the body's gravity, the inherent buoyancy of the current water body on the body, and the cooperative force of the water body when the body moves; the path analysis unit uses D according to the dynamically updated travel cost * Lite algorithm is used to plan the minimum travel cost path; the obstacle avoidance correction unit dynamically sets the spatial obstacle avoidance threshold according to the detected environmental entity position and the error value between the actual position of the body and the predicted position, and then corrects the travel cost of traveling to the spatial nodes around the environmental entity; The input end of the traveling cost analysis unit is connected to the output end of the underwater environment monitoring module and the fluid simulation unit; the input end of the obstacle avoidance correction unit is connected to the output end of the underwater environment monitoring module; The output ends of the travel cost analysis unit and the obstacle avoidance correction unit are connected to the input end of the path analysis unit.

Images