An environmental perception-based underwater snake robot path planning system

Abstract

The application relates to the technical field of autonomous navigation of underwater robots, and discloses an underwater snake-shaped robot path planning system based on environment perception. The system environment perception module constructs a current environment state model; the environment evolution prediction module jointly predicts obstacles, ocean currents and risk areas; the dynamic passable domain generation module evaluates the passability of each grid; the dynamic safety domain construction module introduces a safety buffer zone and constructs a four-dimensional space-time safety domain; and the adaptive rolling horizon planning module dynamically adjusts the prediction time domain according to the environment complexity and solves the optimal path. The autonomous navigation and safety obstacle avoidance of the robot in a complex dynamic underwater environment are realized.

Description

Technical Field

[0001] This invention relates to the field of autonomous navigation technology for underwater robots, specifically to a path planning system for an underwater snake-like robot based on environmental perception. Background Technology

[0002] Current underwater robot path planning technologies are mainly developing along four lines: Classical methods based on static environment maps assume the environment is static and cannot cope with dynamic obstacles; Reactive obstacle avoidance methods based on real-time sensor data only respond to the current environment, lacking foresight and prone to getting trapped in local minima; Rolling time-domain planning methods based on model predictive control have a certain foresight capability, but the prediction time domain length is fixed and cannot be dynamically adjusted according to environmental complexity, and the environmental prediction is limited to the extrapolation of obstacle positions, without systematically incorporating various factors such as ocean current evolution and risk propagation; Robust planning methods that consider uncertainty improve safety margins, but treat prediction errors, ocean current fluctuations and positioning errors in isolation, lacking a unified spatiotemporal coupled safety measurement framework, resulting in overly conservative paths and high energy consumption.

[0003] In summary, existing technologies generally suffer from three major common deficiencies: First, they lack the ability to jointly predict and make decisions regarding obstacle movement, ocean current evolution, and risk zone expansion; second, there is a disconnect between drivability assessment and safety margin modeling, insufficient comprehensive consideration of multi-source uncertainties, and an imperfect path safety assurance mechanism; third, the rolling time-domain planning window is fixed and rigid, lacking the ability to adaptively adjust to environmental complexity. These deficiencies are interconnected and mutually restrictive, and local improvements in a single aspect cannot fundamentally solve the problems. There is an urgent need to construct a systematic solution that deeply couples environmental perception, evolutionary prediction, drivability assessment, safety domain construction, and adaptive rolling planning to meet the practical needs of snake-like robots for autonomous navigation and safe obstacle avoidance in complex and dynamic underwater environments. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an underwater snake robot path planning system based on environmental perception. This system has advantages such as strong predictability, sufficient safety margin, time-domain adaptability, and consideration of both energy consumption and smoothness. It enables autonomous navigation and safe obstacle avoidance in complex underwater environments, thus solving the aforementioned technical problems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: including an environment perception module, an environment evolution prediction module, a dynamic traversable domain generation module, a dynamic safety domain construction module, and an adaptive rolling domain planning module; The environmental perception module is used to acquire real-time information about the environment around the underwater snake robot through multi-source sensors, synchronize the data of each sensor in time and unify the coordinates, extract obstacle information in the environment and estimate the local ocean current speed, direction and intensity of environmental disturbance, and construct an environmental state model at the current moment. The environmental evolution prediction module is used to predict the future location of dynamic obstacles based on the environmental state model, establish a local flow field model by combining sonar observation results and predict the future trend of ocean current direction and velocity changes, construct an environmental risk propagation model and predict the future expansion trend of risk areas, thus forming the future environmental state. The dynamic traversable domain generation module is used to divide the environmental space into multiple discrete grid cells based on the future environmental state, and calculate a traversability index for each grid cell. Generate dynamic passable domains ; The dynamic security domain construction module is used in the dynamic accessibility domain Based on this, a safety buffer zone is introduced around each passable grid. The buffer zone is dynamically adjusted according to the obstacle prediction error, the degree of ocean current fluctuation, and the robot positioning error. The safety domains at different prediction times are time-series superimposed to construct a four-dimensional dynamic safety domain. The adaptive rolling domain planning module is used to calculate the environmental complexity index based on the environmental state. The prediction time domain length is dynamically adjusted according to the environmental complexity, the optimal path is solved under the constraints of the four-dimensional dynamic safety domain, and the future planned path is continuously updated through a rolling optimization mechanism.

[0006] Preferably, the environmental perception module is equipped with sensors including a forward-looking sonar, a multibeam sonar, an inertial measurement unit, a Doppler velocimeter, and a depth sensor; the forward-looking sonar is used to detect the position and contour information of obstacles in front of the robot, the multibeam sonar is used to acquire three-dimensional point cloud data of the surrounding environment, the Doppler velocimeter is used to measure the robot's motion speed relative to the seabed or water body, the inertial measurement unit is used to acquire attitude angle and angular velocity information, and the depth sensor is used to acquire the water depth where the robot is located in real time.

[0007] Preferably, in the environmental evolution prediction module, different motion models are used for prediction of different types of dynamic obstacles, including schools of fish, floating objects, and other autonomous underwater vehicles.

[0008] Preferably, the environmental evolution prediction module incorporates flow field factors as environmental constraints into the prediction model.

[0009] Preferably, the accessibility index The expression is as follows: ; in, Represents grid cells exist The comprehensive accessibility index value at any given time; This represents the normalized obstacle distance factor; Indicates the ocean current direction fit factor; This represents the normalized risk level factor; Indicates the perceived credibility factor; , , and Indicates the weighting coefficient; This represents the row index of the discrete mesh element in the environment within a two-dimensional horizontal plane; This represents the column index of the discrete mesh element in the environment within a two-dimensional horizontal plane. Indicates the current planning time.

[0010] Preferably, the dynamic traversable domain The expression is as follows: ; in, Indicates time A dynamic set of passable domains; The planar spatial index representing the discrete mesh cell of the environment; Represents grid cells exist The comprehensive accessibility index value at any given time; This indicates the preset passability threshold.

[0011] Preferably, the passable area The expression for the safety buffer radius is as follows: ; in, Indicates at time Grid cells The radius of the safety space that needs to be reserved around it; The standard deviation of the uncertainty in the predicted location of the obstacle; Indicates the degree of ocean current fluctuation; This represents the standard deviation of the robot's positioning error; , and These are the weighting coefficients; This represents the row index of the discrete mesh element in the environment within a two-dimensional horizontal plane; This represents the column index of the discrete mesh element in the environment within a two-dimensional horizontal plane. Indicates the current planning time.

[0012] Preferably, the environmental complexity index The expression is as follows: ; in, This represents the normalized dynamic obstacle quantity factor; This represents the normalized obstacle velocity variance factor; This represents the normalized flow field rate of change factor. This represents the normalized risk area proportion factor; , , and This represents the weighting coefficient.

[0013] Preferably, the step of dynamically adjusting the prediction time domain length based on environmental complexity specifically involves: when the environmental complexity index... When the value exceeds the preset high threshold, a shorter prediction window is used. When the environmental complexity index Use a longer prediction window when the value is below a preset threshold. .

[0014] Preferably, the comprehensive optimization objective function of the adaptive rolling domain planning module is: ; in, This represents the comprehensive optimization objective function value; For the sake of security; For the cost of energy consumption; This is the cost of path smoothness; Indicates the current time; The predicted time domain length is dynamically adjusted; , , These are the weighting coefficients.

[0015] Compared with existing technologies, the present invention provides an underwater snake robot path planning system based on environmental perception, which has the following advantages: 1. This invention employs an adaptive rolling time-domain planning strategy. It comprehensively and quantitatively assesses dynamic obstacle density, motion disorder, flow field drastic changes, and the proportion of risk areas using environmental complexity indicators, and dynamically adjusts the prediction time-domain length accordingly. In complex environments, short-time-domain, high-frequency replanning is used to enhance response agility, while in stable environments, long-time-domain, low-frequency replanning is employed to reduce computational overhead. This achieves an adaptive optimal match between planning accuracy and computational efficiency, effectively conserving the limited computational resources of the underwater robot. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the structure of the present invention. Detailed Implementation The technical solutions of 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 within the scope of protection of the present invention.

[0017] Please see Figure 1 An underwater snake robot path planning system based on environmental perception includes: an environmental perception module, an environmental evolution prediction module, a dynamic traversable domain generation module, a dynamic safety domain construction module, and an adaptive rolling domain planning module. The environmental perception module is used to acquire real-time information about the environment around the underwater snake robot through multi-source sensors, synchronize the data of each sensor in time and unify the coordinates, extract obstacle information in the environment and estimate the local ocean current speed, direction and intensity of environmental disturbance, and construct an environmental state model at the current moment. The environmental perception module is equipped with sensors including a forward-looking sonar, a multibeam sonar, an inertial measurement unit, a Doppler velocimeter, and a depth sensor. The forward-looking sonar is used to detect the position and contour information of obstacles in front of the robot. The multibeam sonar is used to acquire three-dimensional point cloud data of the surrounding environment. The Doppler velocimeter is used to measure the robot's motion speed relative to the seabed or water body. The inertial measurement unit is used to acquire attitude angle and angular velocity information. The depth sensor is used to acquire the water depth where the robot is located in real time.

[0018] The environmental evolution prediction module is used to predict the future location of dynamic obstacles based on the environmental state model, establish a local flow field model by combining sonar observation results and predict the future trend of ocean current direction and velocity changes, construct an environmental risk propagation model and predict the future expansion trend of risk areas, thus forming the future environmental state. In the environmental evolution prediction module, different motion models are used to predict different types of dynamic obstacles, including schools of fish, floating objects, and other autonomous underwater vehicles. The environmental evolution prediction module incorporates flow field factors as environmental constraints into the prediction model.

[0019] The dynamic traversable domain generation module is used to divide the environmental space into multiple discrete grid cells based on the future environmental state, and calculate a traversability index for each grid cell. Generate dynamic passable domains ; Accessibility indicators The expression is as follows: ; in, Represents grid cells exist The comprehensive accessibility index value at any given time; This represents the normalized obstacle distance factor; Indicates the ocean current direction fit factor; This represents the normalized risk level factor; Indicates the perceived credibility factor; , , and Indicates the weighting coefficient; This represents the row index of the discrete mesh element in the environment within a two-dimensional horizontal plane; This represents the column index of the discrete mesh element in the environment within a two-dimensional horizontal plane. Indicates the current planning time.

[0020] Dynamic passable domain The expression is as follows: ; in, Indicates time A dynamic set of passable domains; The planar spatial index representing the discrete mesh cell of the environment; Represents grid cells exist The comprehensive accessibility index value at any given time; This indicates the preset passability threshold.

[0021] The dynamic security domain construction module is used in the dynamic accessibility domain Based on this, a safety buffer zone is introduced around each passable grid. The buffer zone is dynamically adjusted according to the obstacle prediction error, the degree of ocean current fluctuation, and the robot positioning error. The safety domains at different prediction times are time-series superimposed to construct a four-dimensional dynamic safety domain. Passable domain The expression for the safety buffer radius is as follows: ; in, Indicates at time Grid cells The radius of the safety space that needs to be reserved around it; The standard deviation of the uncertainty in the predicted location of the obstacle; Indicates the degree of ocean current fluctuation; This represents the standard deviation of the robot's positioning error; , and These are the weighting coefficients; This represents the row index of the discrete mesh element in the environment within a two-dimensional horizontal plane; This represents the column index of the discrete mesh element in the environment within a two-dimensional horizontal plane. Indicates the current planning time.

[0022] The adaptive rolling domain planning module is used to calculate the environmental complexity index based on the environmental state. The prediction time domain length is dynamically adjusted according to the environmental complexity, the optimal path is solved under the constraints of the four-dimensional dynamic safety domain, and the future planned path is continuously updated through a rolling optimization mechanism.

[0023] Environmental complexity index The expression is as follows: ; in, This represents the normalized dynamic obstacle quantity factor; This represents the normalized obstacle velocity variance factor; This represents the normalized flow field rate of change factor. This represents the normalized risk area proportion factor; , , and This represents the weighting coefficient.

[0024] The prediction time domain length is dynamically adjusted based on environmental complexity, specifically: when the environmental complexity index... When the value exceeds the preset high threshold, a shorter prediction window is used. When the environmental complexity index Use a longer prediction window when the value is below a preset threshold. .

[0025] The comprehensive optimization objective function of the adaptive rolling domain planning module is: ; in, This represents the comprehensive optimization objective function value; For the sake of security; For the cost of energy consumption; This is the cost of path smoothness; Indicates the current time; The predicted time domain length is dynamically adjusted; , , These are the weighting coefficients.

[0026] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A path planning system for an underwater snake-like robot based on environmental perception, characterized in that: It includes an environmental perception module, an environmental evolution prediction module, a dynamic traversable domain generation module, a dynamic safety domain construction module, and an adaptive rolling domain planning module; The environmental perception module is used to acquire real-time information about the environment around the underwater snake robot through multi-source sensors, synchronize the data of each sensor in time and unify the coordinates, extract obstacle information in the environment and estimate the local ocean current speed, direction and intensity of environmental disturbance, and construct an environmental state model at the current moment. The environmental evolution prediction module is used to predict the future location of dynamic obstacles based on the environmental state model, establish a local flow field model by combining sonar observation results and predict the future trend of ocean current direction and velocity changes, construct an environmental risk propagation model and predict the future expansion trend of risk areas, thus forming the future environmental state. The dynamic traversable domain generation module is used to divide the environmental space into multiple discrete grid cells based on the future environmental state, and calculate a traversability index for each grid cell. Generate dynamic passable domains ; The dynamic security domain construction module is used in the dynamic accessibility domain Based on this, a safety buffer zone is introduced around each passable grid. The buffer range is dynamically adjusted according to the obstacle prediction error, the degree of ocean current fluctuation and the robot positioning error. The safety domains at different prediction times are superimposed in time sequence to construct a four-dimensional dynamic safety domain. The adaptive rolling domain planning module is used to calculate the environmental complexity index based on the environmental state. The prediction time domain length is dynamically adjusted according to the environmental complexity, the optimal path is solved under the constraints of the four-dimensional dynamic safety domain, and the future planned path is continuously updated through a rolling optimization mechanism.

2. The underwater snake robot path planning system based on environmental perception according to claim 1, characterized in that: The environmental perception module is equipped with sensors including a forward-looking sonar, a multibeam sonar, an inertial measurement unit, a Doppler velocimeter, and a depth sensor. The forward-looking sonar is used to detect the position and contour information of obstacles in front of the robot. The multibeam sonar is used to acquire three-dimensional point cloud data of the surrounding environment. The Doppler velocimeter is used to measure the robot's motion speed relative to the seabed or water body. The inertial measurement unit is used to acquire attitude angle and angular velocity information. The depth sensor is used to acquire the water depth where the robot is located in real time.

3. The underwater snake robot path planning system based on environmental perception according to claim 1, characterized in that: In the environmental evolution prediction module, different motion models are used to predict different types of dynamic obstacles, including schools of fish, floating objects, and other autonomous underwater vehicles.

4. The underwater snake robot path planning system based on environmental perception according to claim 1, characterized in that: The environmental evolution prediction module incorporates flow field factors as environmental constraints into the prediction model.

5. The underwater snake robot path planning system based on environmental perception according to claim 1, characterized in that: The accessibility index The expression is as follows: ； in, Represents grid cells exist The comprehensive accessibility index value at any given time; This represents the normalized obstacle distance factor; Indicates the ocean current direction fit factor; This represents the normalized risk level factor; This represents the perceived credibility factor; , , and Indicates the weighting coefficient; This represents the row index of the discrete mesh element in the environment within a two-dimensional horizontal plane; This represents the column index of the discrete mesh element in the environment within a two-dimensional horizontal plane. Indicates the current planning time.

6. The underwater snake robot path planning system based on environmental perception according to claim 1, characterized in that: The dynamic passable domain The expression is as follows: ； in, Indicates time A dynamic set of passable domains; The planar spatial index representing the discrete mesh cell of the environment; Represents grid cells exist The comprehensive accessibility index value at any given time; This indicates the preset passability threshold.

7. The underwater snake robot path planning system based on environmental perception according to claim 1, characterized in that: The passable area The expression for the safety buffer radius is as follows: ； in, Indicates at time Grid cells The radius of the safety space that needs to be reserved around it; The standard deviation of the uncertainty in the predicted location of the obstacle; Indicates the degree of ocean current fluctuation; This represents the standard deviation of the robot's positioning error; , and These are the weighting coefficients; This represents the row index of the discrete mesh element in the environment within a two-dimensional horizontal plane; This represents the column index of the discrete mesh element in the environment within a two-dimensional horizontal plane. Indicates the current planning time.

8. The underwater snake robot path planning system based on environmental perception according to claim 1, characterized in that: The environmental complexity index The expression is as follows: ； in, This represents the normalized dynamic obstacle quantity factor; This represents the normalized obstacle velocity variance factor; This represents the normalized flow field rate of change factor; This represents the normalized risk area proportion factor; , , and This represents the weighting coefficient.

9. The underwater snake robot path planning system based on environmental perception according to claim 1, characterized in that: The dynamic adjustment of the prediction time domain length based on environmental complexity specifically refers to: when the environmental complexity index... When the value exceeds the preset high threshold, a shorter prediction window is used. When the environmental complexity index Use a longer prediction window when the value is below a preset threshold. .

10. The underwater snake robot path planning system based on environmental perception according to claim 1, characterized in that: The comprehensive optimization objective function of the adaptive rolling domain planning module is: ； in, This represents the comprehensive optimization objective function value; For the sake of security; For the cost of energy consumption; This is the cost of path smoothness; Indicates the current time; The predicted time domain length is dynamically adjusted; , , These are the weighting coefficients.

Images