124 results about "Motion strategy" patented technology

Disclosed is technology for controlling deep reinforcement learning-based legged robot locomotion by inferring implicit terrain information. A legged robot control method may include inferring an action of a quadrupedal robot from proprioception through a deep reinforcement learning-legged robot model, and a locomotion policy that implicitly infers properties of terrains through which the quadrupedal robot moves may be learned in the legged robot model.

The invention discloses an intelligent mechanical arm control system, and relates to the technical field of intelligent mechanical arms, the system comprises a sensor data acquisition module, a multi-modal data fusion module, an environment and target prediction module, an intelligent decision and control module and a control execution and feedback module; through a cooperative working mechanism of the environment and target prediction module and the intelligent decision-making and control module, future state change of a target object can be accurately predicted, possibility is provided for pre-judgment operation of the mechanical arm, then a decision-making optimization algorithm and a multi-target optimization function are applied, an optimal motion strategy is planned for the mechanical arm, and the optimal motion strategy is provided for the mechanical arm. The system further has the real-time feedback and adjustment capacity, it is ensured that the mechanical arm can automatically adapt to changes in the complex dynamic environment, the operation precision and efficiency of the mechanical arm are improved, the capacity of the mechanical arm for adapting to the complex environment and dealing with emergencies is enhanced, and the production efficiency and safety are improved.

The invention relates to a robot perception decision execution system and method, electronic equipment, a storage medium and a computer program product, and the system comprises a perception module which is used for converting task description information and environment detection information into semantic implicit vectors through a preset visual language environment model; the decision-making module is used for fusing the semantic implicit vector, the noise action sequence and the feedback information through a diffusion model running at a time step length and then carrying out iteration denoising for multiple times to generate a target action sequence; the execution module is used for converting the target action sequence into a motor driving signal through a motion strategy model to drive a motor of the robot, perception analysis under a complex environment and a complex language environment is achieved through a perception-decision-execution framework and a visual language environment model, and the motion strategy model can be used for realizing the perception analysis under the complex environment and the complex language environment through a self-adaptive feedback mechanism. The feedback information is utilized to guide the sampling process of the diffusion model, the convergence speed of the diffusion model is accelerated, and the efficiency and the performability of action planning are improved.

The invention discloses a space floating mechanical arm motion planning method and system based on zero reaction space and reinforcement learning, and relates to the field of mechanical arm motion planning. The invention aims to solve the problem of base attitude disturbance caused by movement of an existing space floating mechanical arm. According to the technical scheme, the motion space of the space mechanical arm is designed to be the zero-reaction space, it is ensured that the motion of the mechanical arm does not generate unexpected disturbance on the base, and therefore the stability of the system and the task execution precision are improved. In addition, a reinforcement learning algorithm is combined, and the motion space is set as a zero-reaction motion space. The efficient search and optimization capability of reinforcement learning enables the mechanical arm to find an optimal motion strategy in a complex task space, and meanwhile, disturbance to the base is effectively avoided. The strategy can be continuously adjusted through multiple rounds of training and reinforcement learning so as to adapt to different task requirements, and it is guaranteed that when the mechanical arm executes space grabbing, maintenance and other tasks, high-precision operation can be kept, and it can be guaranteed that the base is stable. Experimental results show that by means of the method, deviation of the attitude of the base can be remarkably reduced, the stability and efficiency of task execution are improved, and a feasible solution is provided for on-orbit service tasks of the space mechanical arm in the future.

The invention relates to the technical field of artificial intelligence, and discloses a brain-like decision-making method, device and equipment for body intelligence and a storage medium, and is applied to a robot. The method comprises the following steps: receiving a human language instruction, and carrying out semantic understanding on the human language instruction to obtain a task description; collecting multi-modal data, and generating an environment model based on the multi-modal data; and generating a pulse event sequence by adopting a pulse neural network model based on the task description and the environment model, and generating a motion strategy of the robot based on the pulse event sequence. Through application of the bionic neuromorphic computing architecture, the decision-making speed of the robot in a complex scene is increased by 50 times, the power consumption is reduced to 1 / 10, the real-time response capability of the robot is greatly improved, the robot can quickly cope with various emergencies, for example, when encountering a sudden obstacle, the robot can quickly do an avoidance action, and the robot can be prevented from being damaged. And collision accidents are avoided.

In various examples, policy prediction-based motion planner systems and methods for autonomous and semi-autonomous systems and applications are provided. A scenario tree structure may be generated that represents potential behaviors of one or more peripheral agents based on perception data of a scene within which an ego vehicle operates. A joint MPC algorithm may optimize the motion of an ego vehicle within the context of the scenario tree structure to produce a policy tree structure. An MPC policy prediction model may be trained to predict the policy tree structures that a joint MPC algorithm would produce, given a set of environmental perception data. An ego vehicle may comprise a trained MPC policy prediction model that receives perception data, and based on that input predicts a policy tree structure that may be used to define a motion policy for navigating the ego vehicle through the scene.

The present application relates to the technical field of robot control. Provided are a robot obstacle avoidance control method and a related device. In the present application, an RMP mapping tree is constructed on the basis of individual actual spatial positions of all obstacles currently present in an operating environment where a target robot is located, such that a root node task of the RMP mapping tree corresponds to a robot joint space, and corresponding leaf node tasks comprise a position motion task and a pose motion task of a robot tail end executing an expected operation, and comprise obstacle avoidance motion tasks of a plurality of robot key parts of the robot tail end respectively performing obstacle avoidance on the obstacles; and then, geometric dynamical systems involving speed information are respectively constructed for various motion tasks, and an expected joint acceleration is solved on the basis of an RMP push-forward operation and an RMP pull-back operation, so as to control the target robot to move, thereby enabling the robot to achieve expected operation execution effects while avoiding dynamic obstacles with high agility and real-time performance.

The invention discloses a quadruped robot motion control method based on topographic map reinforcement learning. The quadruped robot motion control method comprises the following steps: S1, building a quadruped robot model; s2, constructing a topographic map confidence estimation model; s3, constructing a state estimation model; s4, constructing an actor network and a commentator network to train the quadruped robot model; and S5, acquiring data of the laser radar and the inertial measurement unit by using a sensor fusion technology, generating a height map around the quadruped robot in real time, and updating environmental terrain information. By introducing the topographic map confidence estimation model and combining topographic map generation and topographic uncertainty estimation, the motion strategy of the quadruped robot in a complex terrain can be adjusted in real time, the adaptability of the robot to irregular terrains, obstacles and topographic changes is remarkably improved, and the reliability of the quadruped robot is improved. And control errors and falling risks caused by topographic changes or sensor noise are reduced.

The invention provides a quadruped robot fault-tolerant control method based on explicit predictive feedback, and aims to solve the problem that a transferable robot fault-tolerant control model cannot be obtained due to the fact that an explainable and transferable representation containing robot motion model information cannot be obtained in an existing method. The state of the next moment is explicitly predicted by using historical observation information and the current action, and a prediction error is obtained by using an observation true value at the next moment. Prediction errors containing robot motion model information are introduced into input of a motion strategy, a feedback structure is formed, the perceptual ability for the motion model and the environment is provided for a motion network, and a brand-new high-performance fault-tolerant control model which does not need explicit damage judgment and has mobility and interpretability is obtained.

The invention discloses a football motion recognition method and system based on deep learning. The method comprises the steps of data measurement and acquisition, optimized data processing, sensing motion recognition, motion strategy optimization and adaptive feedback adjustment. The invention relates to the technical field of football motion analysis, in particular to a football motion recognition method and system based on deep learning, and the scheme adopts a method of combining basic preprocessing with dynamic time window feature extraction and fusion to carry out data optimization processing; constructing a recognition model by adopting an output activation removal bidirectional long-short-term memory neural network combined with a self-attention mechanism; an integrated differential evolution algorithm combined with reinforcement learning is adopted, an initial prediction model is optimized, and an action strategy optimization model is further obtained through training.

The invention discloses a robot motion strategy model optimization method and a related device, and relates to the field of artificial intelligence, and the method comprises the steps: carrying out the preliminary training of a robot motion strategy model based on an autonomous interaction playback data memory library, and obtaining a preliminarily trained robot motion strategy model; according to the situation when the second robot executes the task, whether a scene which does not meet the requirement exists or not is judged; if yes, retraining the preliminarily trained robot motion strategy model on the basis of a hybrid interaction playback data memory bank for a scene which does not meet the requirement to obtain a trained robot motion strategy model; according to the method and the device, the demonstration data of the person is collected for a scene in which task execution of the preliminarily trained robot motion strategy model does not meet requirements, so that post-mixed training is performed in a targeted manner, the robot motion strategy model can quickly meet training requirements, the training efficiency of the robot motion strategy model is improved, the training period is shortened, and the training efficiency of the robot motion strategy model is improved. The training cost is reduced.

The invention discloses a quadruped robot anti-disturbance motion control method based on reinforcement learning and application, and belongs to the field of robot motion control. Firstly, a self-adaptive propulsion control frame is provided, and a propeller and a leg movement system of the quadruped robot are deeply fused. And secondly, through cooperative training of a double-encoder system (a teacher encoder and a student encoder) and a predictor, hidden state extraction and environment disturbance reconstruction are realized, so that the robot adjusts a motion strategy in real time under complex scenes such as actuator faults and water flow interference. And finally, in combination with multi-stage reinforcement learning training and privilege information utilization, strategy network output is optimized, and a speed and trajectory tracking task is completed. The motion robustness of the robot under unknown interference and damage conditions is remarkably improved through cooperative control of the propellers and the legs, anti-disturbance decision of hidden state driving and dynamic reward function design. The method is suitable for amphibious environments, can adapt to diversified disturbance scenes without additional fine adjustment, and has high practicability and reliability.

The invention belongs to the technical field of three-dimensional scanning path planning and industrial automatic detection, and particularly relates to a full-coverage scanning path planning method for a large complex structural part. The revolutionary optimization of the three-dimensional scanning path of the large structural member is realized by constructing a comprehensive technical system of curvature self-adaptive sampling, pose collaborative optimization, a three-stage Weibull motion strategy, real-time collision detection and whole-process efficiency improvement. According to the scheme, geometric feature perception, motion continuity guarantee, algorithm adaptability enhancement and operation safety control are creatively integrated, a technical breakthrough is formed in three dimensions of non-blind area coverage, high motion efficiency and zero collision risk, and meanwhile, through parameterization design and a cross-platform compatible framework, the operation safety is improved. Deployment flexibility and system robustness in an industrial scene are remarkably improved, and a complete solution with high precision, high efficiency and high reliability is provided for automatic detection of a complex curved surface.

The invention discloses a robot motion control optimization method, device, equipment and medium, and the method comprises the steps: inputting task description in a natural language form, robot state information, a historical execution track and environment feedback information into a preset large language model; through the large language model, selecting a target motion strategy of the robot according to the task description; when it is detected that the task of the robot fails or the action of the robot is abnormal, a reward function used for controlling a deep reinforcement learning controller of the robot is reconstructed; and when the state of the robot is abnormal, parameters of the deep reinforcement learning controller are adjusted and optimized, the motion stability, the task completion efficiency and the strategy adaptability of the robot in a high-risk, high-temperature and high-complexity scene are remarkably improved, and the self-adaptive adjusting and optimizing capability and the abnormity recovery capability of the robot in a dynamic environment are enhanced.

The invention discloses a multi-unmanned aerial vehicle cooperative efficient deployment method in a dynamic environment, and the method comprises the steps: S1, constructing a multi-target optimization model based on a genetic algorithm, and solving the multi-target optimization model, so as to plan a feasible path with the least number of relay unmanned aerial vehicle I nodes between a base station and a nest; and S2, constructing an intelligent agent control model based on a deep reinforcement learning algorithm, so that each relay unmanned aerial vehicle II can adjust a motion strategy according to the real-time position of the task unmanned aerial vehicle, the communication requirement and the environment change in a dynamic task scene to guarantee the stability of a communication link between the aircraft nest and the task unmanned aerial vehicle. According to the two-stage unmanned aerial vehicle cooperative efficient optimization deployment scheme provided by the invention, the efficiency and the stability of an air communication link are remarkably improved in a complex three-dimensional space environment.

A mechanical arm (5) motion generation method and apparatus, a computer-readable storage medium, and a mechanical arm (5). The mechanical arm (5) motion generation method comprises: decomposing a mechanical arm (5) grabbing task oriented to a goods shelf environment into a plurality of sub-tasks; on the basis of a geometric dynamical system, separately determining a Riemannian motion policy of each sub-task; and, on the basis of graph calculation processes of the Riemannian motion policies, integrating the Riemannian motion policy of each sub-task to obtain a global motion policy of a mechanical arm (5).

The invention provides a motion control method and device of a foot type robot, equipment and a medium. The method comprises the steps that state information, coding information and historical actions of the foot type robot at the previous moment and first sensing information at the current moment are obtained; inputting the state information of the previous moment, the coding information and the historical actions into a world model to obtain the state information of the current moment output by the world model; inputting the state information at the current moment and the first sensing information into a motion strategy model to obtain a target action at the current moment output by the motion strategy model; and controlling the movement of the foot type robot according to the target action at the current moment. Any terrain environment can be accurately sensed, so that the stability of the foot type robot in the movement process is ensured, the problems of falling or movement faults and the like are avoided, and the movement control effect of the foot type robot is improved.

The invention relates to a quadruped robot active compliance control method based on hierarchical reinforcement learning. The method comprises a bottom-layer motion strategy network and an upper-layer compliance control strategy network, the bottom layer motion strategy network comprises a basic encoder, a force estimation head, a speed estimation head and a basic strategy network module; the basic encoder is used for outputting a low-dimensional feature vector; the force estimation head is used for estimating the current external force borne by the quadruped robot; the speed estimation head is used for estimating the current trunk speed of the quadruped robot; the basic strategy network module is used for generating a current expected joint position of the quadruped robot; and the upper-layer compliance control strategy network is used for generating a current residual speed instruction and adding the current residual speed instruction and the current trunk speed instruction to obtain a corrected current trunk speed instruction. By adopting the method, high coupling of disturbance estimation and a compliance control module can be realized, and active compliance adaptive to continuous interference can be achieved.

The invention provides a quadruped robot wet-slippery rugged terrain motion control method based on explicit and implicit estimation, and aims to solve the problem that the motion ability of a quadruped robot under the wet-slippery rugged terrain is limited due to the fact that the existing method cannot fully extract environment and robot motion state information. Historical observation information and a current action are utilized to explicitly estimate an environment friction factor, a four-foot contact state and a robot centroid velocity, and current motion model information of the robot is implicitly estimated through a method of predicting a motion state at a next moment. Environmental information, robot motion information and implicit vectors containing robot motion model information are introduced into input of a motion strategy, higher sensing capacity for the motion state and the environment is provided for a motion network, and a novel control method suitable for the quadruped robot to move in the slippery terrain is obtained.

The invention relates to the technical field of intelligent robot control and artificial intelligence, and discloses a robot dynamic trajectory control method based on reinforcement learning and multi-sensor fusion, and the method comprises the steps: firstly fusing multi-modal sensor data through extended Kalman filtering, and predicting the time sequence distribution of dynamic obstacles through a long-short-term memory network; meanwhile, the dynamic operability and singularity risk factors of the robot are calculated, on this basis, a reinforcement learning network based on an attention mechanism is constructed to generate a high-level motion strategy, the high-level motion strategy is converted into a bottom-layer execution instruction through a variable impedance controller and a convex optimization torque distribution module, and singular points are avoided by using null-space characteristics. In addition, according to the method, model parameters are finely adjusted in real time by monitoring and predicting errors and the safety intervention degree on line. According to the method, organic fusion of logic decision and physical execution is realized through a layered architecture, and the operation safety, robustness and control precision of the robot in a dynamic environment are remarkably improved.

The invention discloses a robot whole body control method and system based on real-time inertia motion capture and reinforcement learning, and belongs to the field of robot control. The system comprises a motion capture module, a data processing and forward kinematics module, a motion redirection and optimization module, a general motion strategy control module and a real-time control and driving module which are connected in sequence, and an end-to-end closed-loop control assembly line is formed through a shared memory and a high-speed communication protocol. According to the method, human body joint data are collected through inertia motion capture equipment, high-quality mapping from a human body posture to a robot model is achieved through posture calculation, non-uniform scaling and two-stage differential inverse kinematics optimization, and a stable joint instruction is generated in combination with a reinforcement learning strategy trained by a PPO algorithm. The method solves the problems of lack of feedback closed loop, joint deadlock, contradiction between fidelity and stability and the like in the prior art, has the characteristics of high robustness and high expansibility, and can be widely applied to scenes of remote operation, action reproduction and the like.

The invention relates to the technical field of track obstacle avoidance, and discloses a track robot path obstacle avoidance data updating method and system, and the method comprises the steps: carrying out the vector assembly of multi-dimensional features in a real-time environment data flow of a track robot, and obtaining a primary feature vector, carrying out abnormal mode identification on the primary feature vector and a historical normal mode library to obtain a dynamic abnormal feature; according to the dynamic abnormal features, reconstructing a local topology of a pre-stored path knowledge graph to generate an obstacle influence domain; analyzing the field intensity distribution change trend of the obstacle influence domain to obtain a gradient field, and mapping the dominant direction of the gradient field into a conflict resolution strategy; correcting a predetermined path of the orbital robot to obtain a conflict-free path trajectory fragment; performing efficiency judgment on the conflict-free path trajectory fragments to obtain a multi-dimensional evaluation result; fusing multi-dimensional evaluation results to obtain an emergence type motion strategy; the efficiency of updating the path obstacle avoidance data of the rail robot can be improved.

The invention discloses a humanoid robot control method and system, the system comprises a sensor module, a sensing module, a state representation module, a decision module, a motion control module, an actuator module and a feedback module, and the method comprises the following steps: step 1, data acquisition; step 2, environment perception; step 3, state representation; step 4, action selection; step 5, motion control; step 6, performing feedback and adjustment; according to the method, the hybrid model combining the ViT and the CNN is adopted for environment perception, the environment perception precision is improved, the error recognition rate is reduced, motion planning is carried out through reinforcement learning, the robot can autonomously learn the optimal motion strategy, decision making and adjustment can be autonomously carried out according to the environment change and the task target, and the robot experience is improved. The problem that a traditional motion planning algorithm is high in calculation complexity is solved, the robot motion planning efficiency, the autonomous decision-making ability and the environment adaptability are improved, and safe, efficient and autonomous motion of the humanoid robot in a complex dynamic environment is achieved.

The invention relates to the field of servo control systems, in particular to an azimuth-pitch axis system cooperative control method of an aviation platform swing sweeping device. According to the method, a swing sweeping device is included, and the swing sweeping device comprises an azimuth axis system, a pitch axis system, an azimuth angle acquisition module, a pitch angle acquisition module, an azimuth controller and a pitch controller. The azimuth angle acquisition module and the pitching angle acquisition module are designed to acquire angle information respectively, information interaction is performed between the azimuth driver and the pitching driver, and cooperative information is transmitted between the azimuth driver and the pitching driver. The two controllers drive the azimuth axis system and the pitch axis system to complete corresponding motion according to a designed and planned motion strategy through information transmitted to each other, the limitation that the two controllers independently control axis system motion is broken through, the accuracy of a platform scanning function and the integrity of a scanning area are improved, and complete and accurate scanning of a two-dimensional target area is completed.

The invention provides an action mask assisted obstacle avoidance target tracking method based on deep reinforcement learning, and the method comprises the steps: constructing simple to complex obstacle-target combinations based on an obstacle distribution type and a target motion type, and each combination corresponds to a training stage; in each training stage, according to an obstacle distribution type and a target motion type corresponding to the obstacle-target combination, determining an environment mode of the current stage, and further constructing a task execution map; the input of the main strategy network and the target network is robot observation information introduced with the action masks, and the optimal action output is generated after the action probability distribution output by the main strategy network and the target network is corrected by utilizing the action masks. By using the method, a reasonable motion strategy can be provided for the robot in an environment with a dynamic and static mixed obstacle and a variable-motion-mode target, the risk that the robot collides with the obstacle is reduced, and effective tracking of the target is realized and maintained.

The invention discloses a lithium battery recycling and disassembling method based on AI + 3D vision, and relates to the technical field of battery recycling, and the method comprises the steps of S1, multispectral polarization image acquisition, S2, polarization parameter dynamic calibration, S3, image fusion, S4, deep learning denoising optimization, S5, 3D point cloud reconstruction and feature extraction, and S6, robot motion planning and execution. According to the method, a dynamic situation field model is combined with an impedance control principle, so that the robot can sense obstacle distribution in real time in a dynamic environment and autonomously adjust a motion strategy, and meanwhile, the contact force between an end effector and a target object is accurately controlled; a rejection and attraction field is modeled in real time, so that the sudden mechanical interference risk in the disassembling process is effectively avoided; the impedance coupling mechanism ensures that the robot maintains stable interaction force in contact operation through force feedback closed-loop adjustment, and battery damage or tool failure caused by sudden change of external force is avoided.

The invention provides an anti-falling control method and device for a quadruped robot in a tunnel scene. According to the method, based on point cloud depth data acquired by a laser radar, an advancing strategy of the quadruped robot is obtained by inputting a first motion strategy model trained by adopting a DQN algorithm and a second motion strategy model comprising an SLAM positioning and navigation module. The problem that in the prior art, in a power cable tunnel scene, due to a tunnel and a high platform, a quadruped robot cannot work after falling off is solved.

The invention discloses a salamander-imitating robot omni-directional motion control method and system based on reinforcement learning, and belongs to the field of electromechanical system control. The method comprises the following steps: acquiring robot state information and a motion instruction; a periodic phase variable is generated through a phase integration module, and target actions of all joints are output through a track generator; designing a reward function based on behavior constraint, and constraining motion directivity and attitude stability; the morphological symmetry of the body structure of the salamander-imitating robot is utilized to symmetrically enhance the state-action sample so as to expand training data; and a reinforcement learning algorithm is adopted, and a stable and efficient omnidirectional motion control model is obtained according to a reward feedback and symmetric enhancement data iteration updating control strategy. The method can realize the generation of the omni-directional motion strategy driven by reinforcement learning without presetting a gait model, has the advantages of natural motion mode, high control precision and strong adaptability, and can be widely applied to the fields of complex terrain detection, intelligent inspection and the like.

In various examples, policy prediction-based motion planner systems and methods for autonomous and semi-autonomous systems and applications are provided. A scenario tree structure may be generated that represents potential behaviors of one or more peripheral agents based on perception data of a scene within which an ego vehicle operates. A joint MPC algorithm may optimize the motion of an ego vehicle within the context of the scenario tree structure to produce a policy tree structure. An MPC policy prediction model may be trained to predict the policy tree structures that a joint MPC algorithm would produce, given a set of environmental perception data. An ego vehicle may comprise a trained MPC policy prediction model that receives perception data, and based on that input predicts a policy tree structure that may be used to define a motion policy for navigating the ego vehicle through the scene.

The invention belongs to the technical field of multi-robot path planning, and particularly relates to a distributed multi-robot path planning method based on safety reinforcement learning, which provides a distributed motion strategy for each robot, effectively reduces the risk of conflict and collision among the robots and improves the operation efficiency. According to the method, a state space is set as a small-scale three-channel local observation map which comprises a local distance cost map, a neighbor relative position map and a neighbor target relative position map, information of a robot and the environment is represented in a local map form, the scale and scene generalization is effectively improved, and the real-time performance of the robot is improved based on a breadth-first search algorithm. And a distance cost map is constructed, so that global information is introduced into local observation, and an effective solution is provided for the problem which is easy to fall into local optimum.