305 results about "Underwater robotics" patented technology

The invention relates to the technical field of AUVs (Autonomous Underwater Vehicles), in particular to a real-time communication and positioning data fusion processing method for AUV (Autonomous Underwater Vehicle) collaborative operation, which comprises the following steps of: deploying a multi-modal sensor, acquiring communication data, kinematics data, acoustic positioning data and visual data required by an AUV, and preprocessing the data; the method comprises the following steps: fusing multi-modal data by using an extended Kalman filtering algorithm, calculating three-dimensional position and attitude information of the AUV in real time, carrying out dynamic path planning by using an ant colony optimization algorithm based on fused positioning data, optimizing a collaborative operation path between the AUVs, and sharing the positioning data and a path planning result in real time through an underwater communication network. And dynamically adjusting a task allocation strategy by using an improved task allocation algorithm, optimizing the cooperation efficiency, and finally, storing all data, path planning records and task allocation schemes in a database. According to the invention, the accuracy and stability of AUV collaborative operation are improved.

The invention discloses an underwater robot dynamic environment sensing and navigation device and method, and relates to the technical field of underwater robos.The device comprises an underwater robot, a sensor system, a data processing system, a control system, a navigation system and a power system, and the sensor system, the data processing system, the control system, the navigation system and the power system are arranged on the underwater robot; the surrounding environment can be comprehensively sensed in real time, and the optimal path is calculated through the path planning module according to the real-time environment data; performing obstacle detection according to the environment image to obtain an obstacle detection result; when an obstacle is detected, the control system sends out an obstacle avoidance decision instruction, after the path planning module receives the obstacle avoidance decision instruction, an adaptive algorithm is used for adjusting the optimal path according to the information of the obstacle to be detected, the adjusted path is obtained, and autonomous navigation for a complex underwater environment and autonomous navigation are achieved. And the autonomous navigation and operation capabilities of the underwater robot are obviously improved.

The invention relates to an underwater mechanical arm dragging control method and a mechanical arm system, and is suitable for the fields of ocean exploration, deep sea operation and the like. According to the method, dragging control input of an operator is received through a force feedback handle or virtual reality equipment, and input signals are converted into motion instructions of all joints of the mechanical arm through an inverse kinematics formula. The system is provided with a force sensor and an attitude sensor, feedback data of the underwater environment is collected in real time, the motion trail of the mechanical arm is dynamically optimized based on a PID control algorithm, and high precision and flexibility of the mechanical arm are kept in the complex underwater environment. The system has a multi-degree-of-freedom design and can execute diversified fine operation tasks, a force feedback mechanism ensures that an operator perceives force changes in real time, and the safety and accuracy of operation are improved. The wireless communication module ensures stable transmission of underwater signals, realizes remote operation and real-time feedback, and is suitable for complex underwater operation scenes.

The invention relates to the technical field of underwater robots, in particular to an underwater robot cluster path planning method and system based on a dung beetle algorithm, and the method comprises the steps: obtaining a coordinate division region to construct an environment model, calculating and extracting an obstacle distance to fuse ocean current interference to generate a dynamic obstacle avoidance strategy, and evaluating the path speed similarity to generate a candidate path. And acquiring an offset angle difference, normalizing environmental factors to generate a correction path, and calculating a curvature fusion collision probability to generate a global path control instruction. According to the method, a dynamic model is constructed based on Cartesian coordinate division, the spatial analysis precision is improved, a sonar distance and ocean current interference are fused to generate a dynamic obstacle avoidance strategy, the trajectory smoothness is enhanced, candidate paths are screened by using vector similarity and Monte Carlo simulation, the collaboration is optimized, disturbance is reduced through coordinate offset and ocean current correction, and the accuracy of the dynamic obstacle avoidance strategy is improved. The curvature and collision probability are combined to balance safe energy consumption, and multi-dimensional parameter fusion improves cluster adaptability, precision, efficiency and energy utilization rate.

The invention discloses a target tracking method and system based on visual identification and an underwater robot, and relates to the technical field of underwater robots. According to the invention, automatic tracking and monitoring of an underwater target are realized by using a visual identification technology, operations such as an image identification technology of YOLOv5, a motion tracking technology of an optical flow and a Kalman filter, a target re-identification algorithm of deep learning, image enhancement and the like are specifically combined, and the intelligent level and the working efficiency of underwater operation are comprehensively improved. According to the method, the diver can be efficiently tracked in a complex underwater environment, and high recognition and tracking precision is ensured under the condition that the diver disappears or the motion trail is complex.

The invention relates to the technical field of underwater robots, discloses an underwater autonomous flight control method and system for a submersible type unmanned aerial vehicle, and aims to solve the technical defects of rigid route planning, high energy consumption and poor task adaptability of an existing submersible type unmanned aerial vehicle in complex dynamic underwater environment flight control. The method comprises the steps of executing multi-modal underwater environment perception, constructing an environment state fusion model under a unified space-time reference, generating a predictive hydrodynamic field four-dimensional model, executing multi-target route optimization based on energy and risk cost, and implementing self-adaptive flight control based on feedforward compensation. By adopting the above technical scheme, the method can effectively reduce the energy consumption of the unmanned aerial vehicle, remarkably prolong the operation endurance time, improve the intelligence and task adaptability of route planning, and improve the flight stability and trajectory tracking precision. Therefore, the autonomy, the economy and the task success rate in an unknown, unstructured and strong dynamic underwater environment are enhanced.

The invention discloses a multi-mode hybrid motion bionic underwater robot, and relates to the technical field of underwater robots, the multi-mode hybrid motion bionic underwater robot comprises a body, a suspension assembly and a driving assembly, the body comprises a bionic fish head, a bionic fish body and a bionic fish tail which are connected in sequence, and a first gap exists between the bionic fish head and the bionic fish body; a second gap is formed between the bionic fish body and the bionic fish tail; the suspension assembly comprises a flexible oil bag located in the first gap and / or the second gap, and an oil control assembly communicating with the flexible oil bag and controlling expansion and contraction of the flexible oil bag. The driving assembly comprises side fins arranged on the two sides of the bionic fish body, a first driving module connected with and driving the side fins, a fluctuation fin located at the bottom of the bionic fish body and a second driving module connected with and driving the fluctuation fin, and when the robot is in a low-electric-quantity or low-power-consumption mode, the flexible oil bag can be matched with the fluctuation fin or the side fins to achieve gliding motion. The longer endurance is ensured; and when a certain drive fails, the task can be continued to be executed, so that the interference of an underwater environment is reduced.

The invention relates to a mechanical arm end effector for underwater salvage and a using method thereof, and belongs to the technical field of underwater robots. By integrating a hydraulic-driven rotary drilling assembly and an ejection anchoring mechanism, drilling is completed on an underwater target object at a time, and an anchoring part with a lifting point is implanted; a unique pressure compensation flow channel system is arranged inside, external seawater pressure is automatically balanced through oil pressure, sealing failure is prevented, and operation reliability in a large-depth environment is ensured. The drilling and hoisting device has the functions of drilling and hoisting, a plurality of hoisting points can be continuously established by rapidly replacing the drill bit assembly, and the underwater salvage process is greatly simplified.

The invention discloses a hydraulic structure superficial defect identification method for an underwater robot, and belongs to the technical field of underwater robots, and the method comprises the steps: respectively obtaining first distance information and second distance information, fed back by a sonar sensor and a laser sensor, between the underwater robot and a hydraulic structure superficial defect in the moving process, image information, fed back by the binocular camera, of superficial defects of the hydraulic structure is obtained; inputting the obtained first distance information, the second distance information and the image information into a pre-trained deep learning network model to obtain a superficial defect identification result of the hydraulic structure; on one hand, the problems that a traditional underwater robot is low in hydraulic structure defect recognition and evaluation efficiency and low in accuracy are solved; and on the other hand, the influence of the underwater complex environment on diagnosis is reduced, and the robustness and stability of the system are improved.

The invention discloses a paddle-leg hybrid driven underwater adhesion robot, and relates to the technical field of underwater robots, propeller thruster driving, six-foot crawling movement and hydrogel bionic suckers of foot bottom equipment respectively have the advantages of long-distance cruise driving, stable crawling in unstructured terrain and reliable adhesion detection on near wall surfaces. Therefore, the adaptability problem of a single-mode underwater robot in a complex operation scene is solved. According to the invention, the six groups of multi-degree-of-freedom mechanical feet cooperate with the control system through the modular driving units, so that the robot can realize multi-modal motion functions such as seabed crawling, obstacle crossing and climbing, posture adjustment and the like; the robot can stably crawl on a vertical surface or an inverted surface through cooperation between the hydrogel bionic suction cups arranged at the foot ends and the six-foot crawling gaits. Through vector combination of the three vertical propellers and the two horizontal propellers, the robot can meet the multi-degree-of-freedom motion control requirements of advancing, retreating, ascending, sinking, yawing, pitching, pitching and hovering. According to the paddle leg hybrid drive underwater adhesion robot, the underwater hybrid drive potential is developed to the maximum extent, the underwater maneuverability is improved in various challenging underwater environments, and the underwater robot is promoted to develop towards multiple modes and high efficiency.

The present invention relates to the technical field of underwater robots, and in particular to a modular folding antenna device for an autonomous remote-controlled underwater robot. The device comprises an integrated board, an antenna base and an integrated antenna. A folding drive mechanism is arranged in the antenna base. The folding drive mechanism comprises a power output end and a rotation drive end. The power output end drives the rotation drive end to rotate between a first position and a second position. The rotation drive end is connected to the integrated antenna. When the rotation drive end rotates to the first position, the integrated antenna is located outside the antenna base in an upright state. When the rotation drive end rotates to the second position, the integrated antenna is located inside the antenna base in a folded state. The integrated antenna of the present invention is modularly designed and is easy to repair and replace. The integrated antenna has an autonomous folding function and can effectively reduce the resistance of the autonomous remote-controlled underwater robot when it is navigating underwater.

The invention belongs to the technical field of underwater robots, and discloses an underwater robot state detection system which comprises an upper computer, a master controller, a communication module, a power distribution module, an attitude sensor module, an image acquisition module, a robot carrier and a monitoring system of the upper computer. The upper computer is arranged on the water surface and used for monitoring the underwater state of the underwater robot and interacts with the master control through the communication module. According to the underwater robot state detection system of the scheme, the lower computer is connected with the underwater power line carrier module to convert network signals into electric signals, and data interaction is carried out between the electric signals and the upper computer system through the umbilical cable and the water surface power line carrier module, so that the problems that underwater wireless signals are greatly interfered, the signal attenuation speed is high, and the reliability is high are solved. Stable and reliable signals are difficult to receive, and the stability cannot be guaranteed.

The invention provides a flexible docking device and method for an underwater unmanned vehicle, and relates to the technical field of underwater robots, the flexible docking device comprises a base, a locking piece, a traction piece and at least three groups of adjustable supporting assemblies uniformly arranged on the base; the traction piece is connected with the base and the locking piece and used for pulling the locking piece to move towards the base. The adjustable supporting assembly comprises a supporting rod piece, a telescopic element and a rope driving device, the telescopic element is connected with the supporting rod piece and the base and used for driving the supporting rod piece to be unfolded or folded relative to the base so as to adjust the area of a butt-joint enveloping area formed by the tail end of the supporting rod piece, and the rope driving device comprises a driving assembly and a flexible rope. The driving assembly is arranged on the supporting rod piece and used for driving the flexible rope to drive the locking piece to move in the butt-joint enveloping area, and the locking piece is used for being in butt-joint locking with the unmanned underwater vehicle. The butt joint device can synchronously overcome the three technical problems that the rigid butt joint precision is harsh, the compensation range of a traditional funnel is insufficient, and the buffering performance is weak.

The invention relates to the technical field of underwater robots, and discloses an underwater launching foundation platform which comprises a cabin body, a launching well, a launching mechanism, a cover opening air bag and an air compressor, the cabin body comprises a cabin shell and a cabin cover, an opening is formed in one end of the cabin shell, and a monitoring camera is installed on the cabin shell; the cabin cover covers the opening and is detachably connected with the cabin shell; the launching well is arranged in the cabin shell, a well opening is formed in the top of the launching well and is opposite to the opening, and a well lid is fixedly arranged at the well opening in a heat sealing mode; the launching mechanism is arranged in the launching well and is used for launching a striking instrument; the uncovering airbag is fixed in the cabin shell and is close to the opening; the air compressor is arranged in the cabin shell and communicates with the inner space of the uncovering air bag and the inner space of the launching well. The underwater launching foundation platform is higher in detection capacity, opening of the well lid and the hatch cover can be achieved through cooperation of the air compressor, the cover opening air bag and the heat sealing well lid, smoothness of a launching channel is guaranteed, energy consumption of the underwater launching foundation platform is reduced, meanwhile, fault tolerance of the hatch cover and the well lid is improved, and cost is reduced.

The invention discloses an underwater explosive ordnance disposal control method and system based on an AI large model and a robot, and relates to the technical field of underwater robotics.Underwater optical images and sonar data are collected through a sensing module and input into a pre-trained AI large model for real-time analysis, potential threats are recognized, and a decision instruction containing a path planning and operation sequence is generated; when the analysis confidence coefficient is lower than a threshold value set based on historical data, interaction information containing multi-modal data abstracts, alternative schemes and risk assessment is automatically generated and sent to an operator, instructions are executed after being received, the system comprises a sensing module, an AI processing module, a communication module, a control execution module and a state monitoring module, and a robot body is integrated with a processor to solidify the method. The autonomous decision-making precision and efficiency of the robot in a complex environment are improved through an AI large model, the man-machine cooperation response speed is optimized by using an intelligent interaction mechanism, and the reliability and environmental adaptability of the system are significantly enhanced in combination with state monitoring and edge-cloud cooperative computing.

The invention belongs to the technical field of underwater robots, and particularly relates to a joint device capable of achieving underwater autonomous butt joint and separation. Comprising an underwater robot carrier I, an underwater robot carrier II, a joint mechanism and a plug, the plug is arranged on the bow portion of the underwater robot carrier II, and the joint mechanism is arranged on the stern portion of the underwater robot carrier I. The joint mechanism comprises a driving cabin, a movable joint and a butt joint cabin which are sequentially connected. The driving cabin is connected with the stern of the underwater robot carrier I, and the butt joint cabin is used for being in butt joint with a plug of the bow of the underwater robot carrier II; the movable joint has the degree of freedom of pitching and yawing, and the driving cabin has the functions of controlling the degree of freedom of the movable joint and switching the locking or opening state of the butt joint cabin. The device is simple and reliable in structure and small in size, and the underwater robot provided with the device can be in a single-body underwater robot state, a multi-body underwater robot state and a mechanical arm state.

The invention discloses a prizing-imitating micro-plastic collecting robot and a multi-mode switching control method thereof, and relates to the technical field of underwater robotics.The prizing-imitating micro-plastic collecting robot comprises a fish head, a mouth opening and closing device, pectoral fins, a trunk, a filtering assembly, dorsal fins, a fish tail, tail fins, a fish tail driving assembly, a power source and a controller, and the fish head, the trunk and the fish tail are linearly and sequentially arranged from front to back; the mouth opening and closing device is arranged at the front end of the fish head. The two pectoral fins are symmetrically arranged on the left side and the right side of the fish head. Two water filtering mechanisms are symmetrically arranged on the left side and the right side of the fish head, a collecting bin and an equipment bin are sequentially arranged in the trunk from front to back, a rotary valve assembly is arranged on the front side of the collecting bin, a filtering assembly is located on the rear portion of the collecting bin, and the fish tail is of a three-section structure. The prizing-imitating micro-plastic collecting robot continuously collects micro-plastic in water through a set route, time and labor are saved, the robot can work in a deep water area, the collecting efficiency is high, the use cost is low, the robot is not affected by the environment and weather, safety is high, and the micro-plastic in the water area is comprehensively collected.

The invention relates to the technical field of underwater robots, in particular to an ROV autonomous recovery and energy charging system based on an underwater base station, which comprises an ROV body and an underwater base station used for ROV recovery and energy charging. A recovery sensing module, a PID (Proportion Integration Differentiation) controller and a receiving coil are arranged in the ROV body; the recycling sensing module comprises a downward-looking camera and an ultrasonic transducer; a transmitting coil and a lithium battery cell voltage-sharing module are arranged in the underwater base station; a positioning two-dimensional code is arranged on the upper surface of the underwater base station; the system operation logic comprises the following steps: after the ROV enters a recovery guide stage, an ROV body recognizes a positioning two-dimensional code on the surface of an underwater base station through a downward-looking camera; then, the ROV body obtains coordinate and distance data of the underwater base station through a downward-looking camera and an ultrasonic transducer, horizontal deviation correction is achieved through a PID controller till the ROV body is attached to the surface of the underwater base station, and the underwater autonomous recovery process is completed; and the unattended cruise recovery capability and the operation cruising ability of the ROV are obviously improved.

The invention discloses a suction catching type underwater sea cucumber catching robot, which belongs to the technical field of underwater robots and comprises a frame, and a plurality of vertical propellers and a plurality of horizontal propellers are arranged on the frame; floating bodies are arranged on the top of the frame; a ginseng storage mechanism is in sliding fit in the frame, a suction-catching pipe is arranged at the inlet end of the ginseng storage mechanism, a suction-catching opening mechanism is arranged at the end, away from the ginseng storage mechanism, of the suction-catching pipe, and the suction-catching opening mechanism comprises a suction-catching horn mouth and a rotating mechanism controlling the suction-catching horn mouth to rotate; the frame is provided with a notch for the suction catching opening mechanism to stretch out or retract back. And a suction power mechanism is arranged on the ginseng storage mechanism. The rotating mechanism can adjust the angle of the sucking and catching horn mouth, the rotating mechanism controls the sucking and catching horn mouth to rotate between 0 degree and 90 degrees, the sucking and catching horn mouth is in a horizontal state relative to the water bottom when the rotating mechanism is 0 degree, the sucking and catching horn mouth is perpendicular to the water bottom when the rotating mechanism is 90 degrees, and therefore conversion from horizontal sucking and catching to vertical sucking and catching is completed.

The invention discloses an underwater robot wall-attached walking attitude auxiliary adjustment method, and relates to the technical field of underwater robots, and the method comprises the steps: obtaining a plurality of groups of altimeter data and inertial navigation data, and processing the altimeter data and inertial navigation data to obtain optimized altimeter data; comparing differences among the plurality of groups of optimized altimeter data, and when the differences are smaller than a threshold value, executing wall attachment preparation operation; calculating a roll angle and a pitch angle of the robot by using data differences of the altimeters and geometric parameters of the robot; the attitude of the underwater robot is accurately controlled through fusion processing of the altimeter and inertial navigation data, and stable wall-attached walking and operation execution are achieved.

The utility model discloses a physical simulation platform for underwater robot function test, which belongs to the technical field of underwater detection equipment and is characterized by comprising a substrate, a master control module, a voltage-stabilized source, a communication module, a posture adjusting unit, a buoyancy unit, an oil tank, an underwater detection unit, a power unit and a display module. The base plate forms a platform, and the main control module, the voltage-stabilized source, the communication module, the posture adjusting unit, the buoyancy unit, the oil tank, the underwater detection unit, the power unit and the display module are fixedly installed on the base plate in a detachable mode. The physical simulation platform for the function test of the underwater robot realizes flexible and convenient operation through tiled layout, supports batch test, greatly improves the test efficiency, has the advantages of real-time data acquisition, modular design, reduction of labor cost, enhancement of safety, convenience in fault diagnosis and the like, and is suitable for popularization and application. And research and development and application of an underwater robot technology are greatly promoted.

The invention discloses an underwater robotic fish full-coverage path planning method and system based on multi-strategy collaborative optimization, and belongs to the technical field of underwater robot path planning. Aiming at the problems of low path planning efficiency, poor dynamic environment adaptability and difficulty in algorithm parameter adjustment in the prior art, a dynamic grid coding and multi-resolution collaborative modeling method is provided, an obstacle probability region is divided through sonar point cloud data, the grid refining depth is dynamically adjusted in combination with a greedy optimization algorithm, and a multi-layer environment representation map is generated; an improved genetic algorithm framework is adopted, path fragments are recombined through self-adaptive variation rate adjustment and a three-point crossover operator, a conflict resolution mechanism is introduced to correct a countercurrent path direction, and a motion trail is optimized in combination with path smoothness constraint. According to the invention, high-efficiency and low-energy-consumption full-coverage path planning can be realized in a complex underwater environment, and the method is suitable for scenes of ocean exploration, pipeline inspection, ecological monitoring and the like.

The invention discloses a marine organism sweeping device of an underwater robot, and relates to the technical field of underwater robots, the marine organism sweeping device comprises an underwater robot main body, and further comprises a sweeping mechanism and a winding mechanism, the sweeping mechanism comprises a controller and a first motor, the controller is installed on one side of the top of the underwater robot main body, and the first motor is installed on the bottom of the underwater robot main body; the first motor is installed on the outer wall of one side of the underwater robot body, a waterproof cable is arranged between the controller and the first motor, the outer side of an output shaft of the first motor is sleeved with a shifting fork, an eccentric bearing is arranged on the inner side of the shifting fork, a fixing shaft is arranged on the inner side, away from the first motor, of the shifting fork, and a scraper knife is installed at the bottom of the fixing shaft. The first motor works, under the action of the eccentric bearing, the shifting fork can drive the fixing shaft to rotate by a small angle, through rapid rotation, the scraper knife can achieve high-frequency vibration, and then the scraper knife can conveniently clean firm marine organisms on the surface of a ship.

The invention relates to the technical field of underwater robots, and discloses an underwater propeller intelligent obstacle avoidance system and a control method thereof, and the system comprises an environment collection module, an obstacle fitting module, a path generation module, a response duration calculation module and a path optimization selection module. The environment acquisition module acquires environment images, ultrasonic waves and water flow information; the obstacle fitting module establishes an obstacle distribution model based on the image and ultrasonic information; the path generation module generates a feasible path based on a model and a planning algorithm; the response duration calculation module calculates the time required for adjusting to the path starting point based on the dynamic model; and the path optimization selection module selects an optimal path based on a comprehensive evaluation strategy of the response duration, the obstacle avoidance capability and the energy consumption and outputs the optimal path to the driving module. According to the invention, multi-source information can be fused, water flow and dynamic constraints are considered, and safe, efficient and low-consumption intelligent obstacle avoidance is realized.

The invention discloses a submarine topography surveying and mapping method and system based on an underwater robot, relates to the technical field of underwater robots, and provides reliable basic data for measurement based on an original measurement point sound wave data set S, and extracts a multipath disturbance feature set F of each sonar scanning point; the multi-path interference in the underwater environment is effectively identified and represented, data are clustered through density clustering analysis, dense clusters and discrete points are successfully identified, then a disturbance residual Cc is obtained and compared with a preset interference judgment threshold Omap, each sonar scanning point is endowed with a multi-path label, and the multi-path interference in the underwater environment is effectively identified and represented. According to the method, the time sequence and the amplitude sequence of sonar scanning points are combined, the main path echo time of each measuring point is accurately obtained, the continuity of landforms is effectively judged, and high precision and stability of underwater topographic survey are ensured, especially in the aspects of multi-path interference elimination, main path reconstruction judgment and final accurate topographic reflection.

The invention belongs to the technical field of underwater robots, and particularly relates to a micro optical fiber active compensation device for an autonomous remote control underwater robot. Comprising a box body and a control box, a fiber winding assembly, a damping assembly and a tension detection assembly which are arranged in the box body, the fiber winding assembly is used for winding a micro optical fiber, the micro optical fiber penetrates out of the box body after sequentially passing through the damping assembly and the tension detection assembly, and the damping assembly is used for providing damping for the micro optical fiber. The tension detection assembly is used for detecting the tension of the micro optical fiber and outputting tension information, the control box receives the tension information output by the tension detection assembly and outputs a control instruction to the fiber winding assembly and the damping assembly according to the tension information, and the fiber winding assembly performs actions of winding and unwinding the micro optical fiber according to the control instruction. The damping assembly changes damping according to the control instruction for tension compensation. According to the invention, real-time active compensation and tension control of the optical fiber in water are realized, optical fiber consumption can be saved, and intelligence of an optical fiber management system is enhanced.

An underwater robotic device includes a housing unit, a control unit and a propelling unit. The housing unit includes a base seat and an upper cover in liquid-tight engagement with the base seat. The control unit is disposed within the housing unit and includes a circuit module and a center-of-gravity transferring module which is electronically connected with the circuit module. The center-of-gravity transferring module has a movable weight member and a transfer driving mechanism which drives movement of the weight member so as to vary a position of a center of gravity of the underwater robotic device and to control downward and upward moving directions of the underwater robotic device in the water. The propelling unit is connected with the housing unit and is electronically connected with the control unit to produce a propelling force to move the underwater robotic device forward in the water.

The invention relates to an underwater robot technology, and aims to provide an underwater robot fault diagnosis method based on an online fine-tuning hybrid neural network. Comprising the following steps: performing normalization processing on collected state data in a navigation process of an underwater robot, and inputting the state data into a deep convolutional neural network to perform spatio-temporal feature extraction; the output data combination is converted into a pulse sequence in a Poisson coding mode, and the pulse sequence serves as the input of a pulse neural network; counting the excitation number of neurons in the spiking neural network, and determining and outputting a fault diagnosis result according to the fault mark type matched with the excitation number; parameters of the deep convolutional neural network are kept unchanged in the whole diagnosis process, and online fine tuning of the whole hybrid network model is realized based on unsupervised adaptive updating of the pulse neural network model. The method is suitable for low-frequency sampling of underwater robot data, and the accuracy of fault diagnosis is improved; the model can be dynamically adjusted according to actual working conditions, and the environmental adaptability of diagnosis is improved.

The invention belongs to the technical field of underwater robots, and particularly relates to an underwater deformable recycling robot based on an extensible mechanism. The first propelling device is mounted at one end of the robot body; the folding bracket is mounted at the other end, far away from the first propelling device, of the robot body; the second propelling device is mounted at the end part of the outer side of the folding bracket; the folding and unfolding intelligent structure is assembled on the folding bracket; the folding and unfolding intelligent structure drives the folding support to be opened or closed. According to the invention, autonomous recovery of the underwater AUV is realized, and the operation difficulty and cost are reduced. By reducing the dependence on manual operation, the requirements for professional operators and high-tech experience are reduced, meanwhile, the labor cost and time are reduced, and the overall operation efficiency is improved.

The invention provides a bionic jellyfish robot, a posture control method and a driving leg assembly manufacturing method, relates to the technical field of underwater robots, and aims to flexibly adjust the posture of the robot and improve the adaptability to the underwater environment. The bionic jellyfish robot comprises a robot body, a driving leg assembly, a power module, a control module and the like. According to the bionic jellyfish robot and the posture control method, the robot can be selectively in the machine body linear advancing posture, the machine body steering posture or the clamping posture so as to flexibly adjust the posture of the robot, and the driving leg assembly is simple in structure, is of a quasi-soft structure, is good in overall flexibility and can bear water impact; and the adaptability to the underwater environment is good. According to the manufacturing method of the driving leg assembly of the bionic jellyfish robot, the manufacturing process is simple, the manufacturing cost is low, and the production input cost of products is reduced.