125 results about "Biped robot" patented technology

The invention belongs to the technical field of robots, and discloses a biped robot state estimation method based on multi-sensor information, and the method comprises the steps: obtaining the data of an inertial measurement unit, a joint encoder and a joint sensor on a biped robot; a self-adaptive threshold contact state detection algorithm based on current sensing information is designed, and the contact state of the sole and the environment is judged in real time; determining a supporting phase of the robot according to the contact state, and applying a dynamic reference system conversion mechanism of supporting phase perception; constructing a state estimator based on invariant extended Kalman filter IEKF, defining a state vector including the posture, position and speed of the robot, performing state prediction by using IMU data, and performing state updating in combination with contact state information, kinematics information and the like; and outputting accurately estimated robot state information. According to the method, the adaptive contact detection and the IEKF are introduced, so that the state estimation precision and robustness of the biped robot in the dynamic walking process are effectively improved.

The invention discloses a biped robot knee joint self-adaptive locking device which comprises a rotary joint module, a sliding locking module and a sensing control module, the sliding locking module comprises an electromagnetic lock, a sliding sleeve, a reset spring, a locking sliding block, a linear guide rail and a connecting block, and the sliding sleeve is connected with the electromagnetic lock; the locking sliding block is installed on the sliding sleeve through the reset spring, the electromagnetic lock and the linear guide rail are fixed to the connecting block through bolts, the sliding sleeve is installed on the linear guide rail, and the sliding sleeve and the linear guide rail form a sliding pair. By means of the ingenious mechanical structure design, in combination with a gait phase recognition control method following a natural biped walking mode, rigid locking of the knee joint in a standing phase and flexible releasing of the knee joint in a swinging phase are achieved in a proper walking stage, and the locking mechanism has the advantages of being light in weight, low in power consumption and rapid in response; compared with an existing hydraulic locking scheme, the hydraulic locking device is simpler in structure, free of complex hydraulic pipelines and driving systems, smaller in overall size and lighter in mass.

The invention discloses a biped robot gait network training method. The method comprises the following steps: constructing a dual-channel deep reinforcement learning architecture; the running states of the X biped robots in the simple terrain of the simulation environment are collected; obtaining a current reward according to the current running state, merging each piece of information into a Markov decision process, and storing the information into an experience playback area; when the number of the Markov decision-making processes in the experience playback area is greater than a preset threshold value n, randomly taking a preset number of Markov decision-making processes from the Markov decision-making processes, and updating parameters of the main network and the opponent network; disturbance is carried out on main network parameters, and a human memory curve is simulated to carry out continuous adjustment on main network hyper-parameters clip; the biped robot with the stable walking duration reaching the preset duration is moved to the terrain with the higher difficulty level, and the network parameter updating process is repeated; and course learning is continuously carried out until the accumulated reward information and the stable walking duration of all the biped robots reach preset values.

The invention relates to the technical field of biped robot reinforcement, and discloses a historical action sequence-based biped robot reinforcement learning control method and system. The biped robot reinforcement learning control method is applied to reinforcement learning control equipment and specifically comprises the following steps that S101, a floating base frame is adopted to load a biped robot URDF model in an Isaac Gym simulation platform, ten active joint freedom degrees are configured, and a control interface is constructed to output a 10-dimensional motor position command vector; the motion robustness and adaptability of the biped robot in a dynamic environment are remarkably improved by constructing a Transform time sequence modeling architecture fusing a historical state-action sequence, and the historical observation and action sequence is combined with sine-cosine position coding to be input into a multi-layer Transform encoder, so that the motion robustness and adaptability of the biped robot in the dynamic environment are improved. The self-attention mechanism is used for capturing the long time sequence dependence relation of foot-ground contact force change, topographic relief and the like, and the problem of gait instability caused by limited local observation in a traditional method is solved.

The invention relates to the technical field of robot control. A humanoid robot gait generation system based on hierarchical control is characterized by comprising a multi-layer control system operating at different refresh frequencies, and the multi-layer control system comprises a high-layer gait and trajectory planning layer operating at a low frequency and used for generating a macroscopic motion instruction; the middle-layer whole-body motion optimization layer runs at medium frequency, serves as a bridge of high-layer planning and bottom-layer control and is used for converting expected plantar and trunk tracks into reference motion of all joints of the whole body; the bottom joint servo control layer runs at the highest frequency and is responsible for accurately tracking a joint-level instruction output by the whole-body motion optimization layer; the state sensing and feedback closed-loop layer is used for feeding back state information, obtained in real time, of the body state of the robot to all the layers, providing a basis for track re-planning of a high layer and providing a real-time state initial value for optimization of a middle layer; and a feedback signal is provided for servo control of a bottom layer.

The invention relates to the field of humanoid robots, and provides a control method for biped robot step-by-step gait strategy fusion in order to solve the problems that a biped robot is large in training difficulty, unnatural in gait, poor in adaptability and the like. Training various different terrain categories through an imitation learning algorithm and a reinforcement learning algorithm to obtain a gait strategy model; s2, performing pairwise fusion gait strategy switching between the gait strategy models corresponding to various different terrain categories; s3, the terrain category of the current position is recognized in real time according to an information collection unit of the biped robot, a corresponding gait strategy model is obtained through a strategy execution control unit, and the biped robot is controlled to walk in the corresponding motion state; and S4, when the terrain category of the biped robot is changed, smooth transition between the two gait strategy models is achieved through gait strategy switching, and control over gait strategy fusion of the biped robot is completed.

The invention discloses a reinforcement learning biped robot gait control method based on a multilayer adaptive exploration mechanism, and belongs to the field of biped robot reinforcement learning control. The method aims at solving the problems that in existing gait control based on a near-end strategy optimization (PPO) algorithm, strategy convergence is unstable, and exploration and utilization balance is insufficient. According to the method, a multi-layer adaptive exploration mechanism is introduced on the basis of a standard PPO framework: firstly, a state-dependent standard deviation network is designed to realize adaptive noise adjustment of a space layer for dynamically controlling action disturbance intensity; secondly, constructing an entropy regularization mechanism based on advantage weighting, and realizing exploration intensity adaptive control of a direction layer; thirdly, a dynamic entropy coefficient is designed according to the uncertainty of a value function, and exploration rhythm self-adaptive adjustment of a time layer is achieved. Through the synergistic effect of three layers of mechanisms, the strategy realizes dynamic balance of exploration and convergence in different training stages, so that the stability of the training process and the strategy convergence quality are improved.

The invention discloses a biped robot control method and device, electronic equipment and a storage medium. The method comprises the steps that actual observation data at the current moment are collected through a target sensor of a biped robot body; wherein the target sensor comprises a joint encoder, an inertia measuring instrument and a plantar force sensor; predicting a state prediction value at the current moment by combining the state prediction value at the previous moment and a motion model of the biped robot; predicting predicted observation data at the current moment by using the state predicted value at the current moment based on the relationship between the observation data and the state data; correcting the state prediction value at the current moment based on the deviation between the actual observation data at the current moment and the prediction observation data to obtain a state correction value at the current moment; selecting a control parameter in a current state for the unified reinforcement learning control strategy through the correction function based on the state correction value at the current moment, and correcting the control parameter to obtain a current control parameter; and controlling the current biped robot according to the current control parameters.

The invention relates to a method and device for a biped robot to wear common sports shoes, and belongs to the technical field of robots. The method comprises the steps that a pre-planned gait track is initialized and generated; calculating the actual centroid position of the robot at each moment; a Kalman filtering algorithm is adopted to fuse multi-source data to update the normal stiffness and the friction coefficient of the sole in real time, and the rate is updated through adaptive gain adjustment parameters; adjusting a zero moment point constraint range, and generating a centroid reference position of a preset step number; establishing a sole characteristic database, training a machine learning model, inputting shoe parameters into the machine learning model, and outputting step length, step frequency and ankle joint rigidity by the machine learning model; an ankle joint target angle and a knee joint target angle are solved through inverse kinematics, and a joint motor is driven in combination with a PD control algorithm and a self-adaptive item; when the vertical acceleration of the IMU sensor is larger than a preset acceleration threshold value, the time period before grounding is judged, and a compliant mode is triggered to control the ankle joint rigidity and the torque borne by the joint.

The invention belongs to the related technical field of biped robot gait planning, and discloses a method, application and equipment for generating a biped robot anthropomorphic foot end trajectory based on a full-connection network, and the method comprises the steps: (1) grouping the collected foot end trajectory data according to a starting step, an intermediate circulation step and a termination step; (2) mapping each group of foot end track data into foot end track data which can be used by a robot based on a kinematics mapping relation between a human body and the robot, and carrying out conformal interpolation; (3) training a full-connection neural network model by adopting the interpolated foot end trajectory data, further generating a foot end trajectory of each intermediate step by adopting the full-connection neural network model, and mapping the directly generated foot end trajectories of the starting step and the ending step to jointly form an anthropomorphic foot end trajectory; and the main loss of back propagation of the full-connection neural network model is punished by introducing the trajectory smoothness and the prediction precision of the start end and the tail end. According to the method, the anthropomorphism of the foot end track is improved.

The invention discloses a hip structure, a leg structure and a biped robot, and relates to the technical field of robotics.The hip structure comprises a hip lateral extension assembly fixing base, a hip lateral extension assembly, a hip pitching assembly and a hip turnover assembly, the hip lateral extension assembly comprises a hip lateral extension connecting piece and a hip lateral extension joint, and the hip pitching assembly comprises a hip pitching connecting piece and a hip pitching joint; the hip turnover assembly comprises a hip turnover connecting piece and a hip turnover joint; the hip lateral extension connecting piece is fixedly installed on the hip lateral extension assembly fixing seat, the output end of the hip lateral extension joint is connected with the output end of the hip pitching joint, the hip pitching connecting piece is connected with the output end of the hip turnover joint, and the hip turnover connecting piece is used for being connected with a thigh skeleton. According to the hip structure, the tail end mass of the leg structure can be reduced, the rotational inertia is reduced, and therefore the flexibility and cruising ability of the biped robot are improved.

The invention discloses a biped robot high-dynamic action control method based on reinforcement learning, and the method comprises the steps: setting reference trajectories of a leg joint pitch angle and a shoulder joint pitch angle of a simulation robot for two processes of running and jumping, and setting a pose reference trajectory of the simulation robot and a reference sequential logic of foot contact; according to the biped robot high-dynamic action control method based on reinforcement learning, simple actions such as running and jumping do not need to depend on complex demonstration data or a fine action template, and by designing a reference track of a pitch angle of each joint and setting a pose reference track of a simulation robot and a reference sequential logic of foot contact, the high-dynamic action control of the biped robot can be realized. Through reinforcement learning, expected values of all degrees of freedom of each joint are learned, control over all degrees of freedom of each joint is achieved, then motion control over the robot is achieved, the learning process can be guided only through simple and visual motion priori, the exploration space of reinforcement learning is reduced, and the learning complexity is reduced.

The invention relates to a biped robot gait planning method. The invention relates to the technical field of robot gait planning, and the method comprises the steps: taking a kinetic model of a robot as input, so as to achieve the neural network reasoning of simulation; describing the robot by using a URDF file, and enabling the robot to be equivalent to a kinematic chain formed by connecting rigid bodies in series; through accurate description, a robot consistent with that in reality is established in a simulation environment, the robot is enabled to carry out gait learning in a simulation physical environment, and supervised random divergence trial and error are carried out in a mode of setting initial conditions, termination conditions and reward and punishment conditions so as to realize gait learning. According to the invention, through repeated modification of parameters and structural adjustment of adaptability, a stable, low-energy-consumption and high-robustness control strategy is finally obtained through training.

1. Name of the product in this design: Bipedal Robot. 2. Application of this design: Robots used for inspection, security, exploration, handling, industrial operations, surveying, scientific research, service, education, companionship, entertainment, and performance. 3. The key design feature of this product is its shape. 4. The image or photograph that best illustrates the design's key points: 3D view 1.

The application relates to the technical field of robots, and discloses a kind of two-degree-of-freedom parallel ankle joint mechanism of humanoid robot and kinematics and mechanics solving method, including seven-link transmission structure, which is composed of two groups of closed-chain transmission of parallel four-bar linkages.Motor I and II are installed on the upper end of the lower leg, and the rotation drives link IV and III to move through link I, II, V and VI, and then coupled control ankle joint pitch and lateral freedom degree.The mechanism realizes load balancing of double motors through parallel transmission, improves equivalent torque, and can combine kinematics and mechanics solving method, so that precise control can be realized only by motor feedback, sensor cost is reduced, and is suitable for high stability motion demand of biped robot.

The variable-stiffness self-adaptive toe joint comprises a sole plate and toes located on the front side of the sole plate, the bottom of the sole plate and the bottoms of the toes are located on the same horizontal plane, one side of each toe is installed on the sole plate through a titanium steel plate spring, a pressing wheel is arranged above the titanium steel plate spring, and the other side of each toe is installed on the sole plate. The variable stiffness of the titanium steel plate spring can be adjusted by adjusting the position of the pressing wheel above the titanium steel plate spring. One side of the toe is connected with the sole plate through the titanium steel plate spring, the pressing wheel is located above the titanium steel plate spring and can press the titanium steel plate spring, and the driving mechanism can drive the pressing wheel to move along the titanium steel plate spring; the bending radius when the toes are stressed is changed through different contact positions of the pressing wheels and the titanium steel plate springs on the toes, variable-rigidity self-adaptive movement of the toes in the movement process is achieved, and the walking speed, the stride height and stride of the robot in the walking process and the walking stability of the robot are improved.

A robot step control method, a robot control apparatus, and a storage medium are provided. The method includes: determining an expected support force of two legs of a biped robot according to zero-moment point planning data and actual position data of the two legs at a current moment, and determining a current desired joint posture angle of ankle joints of the two legs and a desired joint position matching an actual leg support state using a compliance control algorithm based on an expected support force of the two legs, and centroid movement planning data, centroid actual movement data, step planning data and actual force data of the two legs at the current moment. In such manner, all-direction compliant controls can be performed on a desired leg pose condition according to the actual motion status of the biped robot, thereby improving the walking stability and terrain adaptability of the biped robot.

The invention relates to a biped robot approximate period control method and system based on a self-adaptive period control obstacle function. The method comprises the steps that a state space model of approximate periodic walking of the biped robot is established, the measurement instantaneous frequency of a system state is obtained in real time through short-time Fourier transform, and approximate periodic characteristics of an accumulated phase time representation system are introduced; acquiring an estimated instantaneous frequency of a system state in real time by using a self-adaptive extended Kalman filter, adjusting the frequency of a control barrier function, and constructing a self-adaptive periodic discrete time control barrier function to obtain a function-based security set forward invariance guarantee; the accumulated phase time error is analyzed, three reset strategies are used for relieving the accumulated phase time error, and a proper reset strategy is selected according to the actual situation; under a model prediction control framework of single-step prediction, a control obstacle function is used as a constraint, and a walking task of the biped robot is completed by solving an optimization problem. The method adapts to the approximate periodic characteristics of the system by estimating the system period in real time.

The utility model relates to the technical field of robot equipment, and provides a humanoid foot plate assembly of a biped robot, which comprises a main body assembly, and the main body assembly comprises a foot plate, a heel shell, a front foot sole shell and a gravity sensor; the heel shell is mounted at the rear end of the top of the sole plate, the front sole shell is mounted at the front end of the top of the sole plate, and the gravity sensor is mounted in the heel shell; the maintenance assembly comprises a placement groove, an insertion frame, a rotating gear and a threaded rod; the placing groove is formed in the foot bottom plate, the inserting frame is fixedly connected to the inner bottom wall of the foot bottom plate, and the rotating gear is rotationally connected to the inner side wall of the foot bottom plate. According to the utility model, through the auxiliary cooperation between the tooth block and the insertion frame, the heel shell can be pulled out upwards, and the two groups of screw blocks release the clamping of one end of the gravity sensor at the same time, so that a worker can be assisted to quickly disassemble, assemble and take the gravity sensor, and a more abundant maintenance disassembly and assembly space is provided for the worker.

The application discloses a biped robot with a quick-change hand and a motion analysis method under special working conditions, and belongs to the field of robot technology.The biped robot comprises a trunk unit, an arm unit, a hip joint unit and a leg unit; the arm unit is provided with a mechanical hand through a quick-change structure, and the mechanical hand is divided into a clamping mechanical hand, a scissors mechanical hand and a lock twisting mechanical hand.The robot has the advantages that each movable joint of the robot is designed in a module mode, the integration and the modularization of the robot are greatly improved, a connecting rod mechanism is arranged at the rear side of a knee joint, a design of moving outward of a hinge point is adopted, the motion dead point of the knee joint is offset, the joint module can more smoothly exert force, and the complexity of the joint connecting part is reduced.

This invention relates to the field of machine control, and in particular to a method and apparatus for controlling the jumping and somersaulting motion of a biwheeled legged robot. The method includes: simplifying the biwheeled legged robot into a structure comprising a base, legs, and wheels, as the target structure; constructing dynamic models of the target structure during the standing phase and the flight phase; determining the states of the target structure at the instant of takeoff and landing based on the target jump height and target jump distance; determining a reference trajectory during the jumping and somersaulting process based on dynamic constraints, torque constraints, joint limitations, and the dynamic model; designing corresponding controllers during the takeoff preparation phase and the landing recovery phase to track the motion trajectory and maintain balance; and designing an attitude adjustment controller based on wheel speed adjustment during the flight phase to ensure the smooth completion of the jumping and somersaulting motion. This invention ensures that the robot can complete the somersaulting motion.

The application discloses a biped robot form optimization method based on metric index identification variable additive grouping, which comprises an additive structure grouping stage and a Bayesian optimization stage based on additive decomposition; in the additive structure grouping stage, first, the interaction strength between variables is calculated through the designed variable interaction metric index, and the interaction strength is converted into a distance matrix; subsequently, using the distance matrix, a K-Means clustering algorithm with capacity limitation is used to cluster high-dimensional variables to obtain grouping, so that an additive grouping structure is obtained; in the Bayesian optimization stage based on additive decomposition, an additive kernel GP proxy model is constructed based on the obtained additive grouping structure, the next evaluation point is obtained by maximizing the obtained function, and the form parameters of the evaluation point are used as the initial form parameters of the biped robot for final application. Compared with the existing method, the application is more simple and robust, and by limiting the maximum dimension of each sub-group, the optimization efficiency and grouping identification accuracy are balanced.

The application relates to the field of robot walking control, and discloses a biped robot, a walking control method and device thereof and a readable storage medium. The method comprises the following steps: acquiring the center-of-mass state of the robot in each control period, wherein the center-of-mass state comprises a planned center-of-mass state and a measured center-of-mass state; calculating the capture point error of the robot in each control period based on the planned center-of-mass state and the measured center-of-mass state; accumulating the capture point error of each control period to obtain the total capture point cumulative error; when the capture point error or the total capture point cumulative error exceeds a preset threshold range, adjusting the landing point position of the swing leg of the robot based on the total capture point cumulative error; and controlling the swing leg to move so that the robot lands at the adjusted landing point position. The method ensures that the biped robot can still walk stably without falling down when subjected to thrust disturbance.

The application discloses a truss environment-oriented two-legged robot crawling landing point sequence planning method and belongs to the technical field of robots; the method first constructs a truss undirected graph topology model based on the spatial connection relationship of truss joints and members, takes nodes to represent the truss joints, takes edges to represent the members and takes the member length as the edge weight; the A star algorithm is adopted on the topology model to search for the node sequence and the member sequence of the robot from the starting member to the target member; subsequently, the axial position of the member is described through a normalized member parameter model, and the non-connected feasible landing area set of the robot foot end is represented in combination with One-Hot coding; on this basis, the least number of steps required for the robot to reach the target position is taken as an optimization target, a mixed integer quadratic constraint programming model is established, the landing point sequence is solved, and the optimal landing point sequence satisfying the truss topology constraint and the robot motion constraint is obtained. The application can realize the landing point sequence planning of a two-legged truss crawling robot on a complex truss structure.

The application discloses a kind of biped robot control method, electronic equipment and storage medium.Therein, the method includes: in response to receiving the stride instruction of biped robot, the stride trajectory of biped robot is acquired, wherein, stride trajectory is used to indicate the movement trajectory of swing foot of biped robot from current posture to desired posture, biped robot also includes support foot and body, support foot is used to support body to be in stable state;Based on the current center of mass position of body and the current position of support foot, the body is controlled to move along the first direction, and the first moving position is obtained, wherein, the first direction is used to indicate the direction of support foot relative to swing foot;Stride trajectory and first moving position are used to control the body to move along the second direction, and the second moving position is obtained, wherein, the second direction is used to indicate the moving direction of swing foot in stride trajectory.The application solves the technical problem that the control effect of related biped robot is poor.

The application relates to a control method for dynamic walking of a biped robot and the biped robot, and belongs to the technical field of robot control, which comprises the following steps: 1, a self-recurrent cerebellar model neural network is used to establish a dynamic model of the biped robot with a disturbance term, and dynamic robust walking of the biped robot is converted into a problem of realizing stability of a multi-input multi-output nonlinear system with a bounded uncertain term; 2, an adaptive self-recurrent cerebellar model neural network error observer is designed to estimate an error upper limit; 3, an adaptive law of network weight is designed to realize real-time updating of the network weight space and to adjust parameters of each learning factor; and 4, a boundary value estimation algorithm is used to compensate for an estimation error and feedback to a robot walking system, so that the biped robot can realize asymptotic stable walking. The application enables the control system to adapt to time-varying characteristics of the biped walking system on line, and has continuous learning and adaptive capacity for unknown dynamics.

This application provides a cooperative control method and system for two humanoid robots, relating to the field of robot control technology. The method includes acquiring interaction data from two demonstrators; performing interactive perception motion redirection on the two humanoid robots based on the interaction data to generate a reference trajectory and topological prior that conform to robot kinematic constraints. Specifically, kinematic conflicts arising from heterogeneous redirection are resolved through dual-scale manifold optimization, accurately preserving the interaction intent; using the reference trajectory as the tracking target and guided by the topological prior, reinforcement learning is performed to obtain a cooperative control strategy for the two humanoid robots; the trained cooperative control strategy is deployed to the two humanoid robots, which exchange state information through real-time communication to perform physical cooperative tasks; this significantly improves the robustness of control and training efficiency; and ultimately achieves zero-sample transfer of the cooperative strategy from simulation to real-world operation.

The invention discloses a flexible control method and system for actions of a biped robot, and relates to the technical field of robot control, and the method comprises the following steps: setting the movement speed of each movement unit as a preset maximum speed value; acquiring the current position of each motion unit as an initial position; receiving a new target position instruction for at least one target motion unit; for the target motion unit, calculating a target distance between the initial position and the new target position; determining a microstep interval K and a microstep number N; calculating delay time L after execution of each microstep; and circularly executing the control operation and the delay operation for N times. The method has the advantages that on the basis of the motor subdivision control theory and the interpolation algorithm, the joint movement of the robot is no longer the previous mechanical impression, the impression is flexible and rapid, the robot is separated from the previous action image of the robot, and the posture is close to the posture of a real person.

The application relates to a complex terrain adaptive gait planning method for a biped robot and the biped robot, and belongs to the technical field of robots, and comprises the following steps: a fractional order PID controller is constructed, the fractional order PID controller comprises a multi-sensor network module, an adaptive module and a fractional order calculus control module; a reinforcement learning model for gait planning of the biped robot is constructed, a state, an action, a decision and a reinforcement learning reward function of the reinforcement learning are set, a policy gradient method is used to continuously iterate and optimally adjust the reinforcement learning model, policy parameters are updated, and control parameters of the biped robot are output; and the control parameters output by the fractional order PID controller and the reinforcement learning model are decisionally fused through a comprehensive loss function, so that globally optimal control parameters of the biped robot are obtained. The application can provide more accurate gait planning for the biped robot in complex terrain, and better walking flexibility and adaptability are achieved.

1. The name of the design product: biped robot. 2. The use of the design product: used in the fields of display, service, interactive entertainment, education, industrial manufacturing and logistics storage, etc. 3. The design points of the design product: in shape. 4. The picture or photo that best indicates the design points: perspective view.