160 results about "Gait planning" patented technology

The invention discloses a gait planning method for walking of a biped robot along a slope, and belongs to the technical field of robots. The method contains two core theories, namely, non-orthogonal decomposition and synthesis of a three-dimensional linear inverted pendulum and gait planning based on a biped long linear inverted pendulum. The method comprises the steps of gait parameter configuration, non-orthogonal decomposition of a three-dimensional linear inverted pendulum, foot trajectory planning based on three times of spline interpolation, centroid trajectory planning of a sagittal plane and a coronal plane based on a biped long linear inverted pendulum, non-orthogonal synthesis of the centroid trajectories of the sagittal plane and the coronal plane, and solving of joint trajectories of a robot based on inverse kinematics. By adopting the method of the invention, stable walking of a biped robot along all directions of a slope can be effectively achieved. The method is universal, simple in algorithm and high in practicality.

The invention provides a hexapod-robot real-time gait planning method based on deep reinforcement learning. The method comprises the following steps of using a hexapod robot to acquire environment road condition information and making an integral movement track; through a camera, acquiring an environment photograph, according to the photograph, using a binocular range finding method to calculate road condition information of a target track, and using the calculated road condition information of the track in robot center of mass movement track navigation; in a foot end swinging space range of robot legs, taking photographs of a road condition environment, and through a trained deep reinforcement learning network based on a depth determinacy policy gradient (DDPG), carrying out data dimension reduction and feature extraction on the photograph; and according to a feature extraction result, acquiring a control policy of a hexapod robot, wherein the hexapod robot controls foot laying of a robot according to the control policy so that real-time walking of the hexapod robot is realized. By using the gait planning method, a complex and non-structural environment of a road condition can be planned in real time. The method has an important meaning for increasing an environmental adaptive capacity of the hexapod robot.

The invention discloses a previewing control humanoid robot gait planning method based on deep reinforcement learning. The method comprises steps of 1) acquiring state information through a sensor assembled on a humanoid robot; 2) improving a current deep reinforcement learning network, and defining totally new states, action vectors and awarding functions; 3) using the defined action vectors to correct output of a previewing controller, calculating the angle of each steering engine of the feet of the humanoid robot and guiding the humanoid robot to walk; and 4) during walking processes of thehumanoid robot, using values of the states, the action vectors and the awarding functions to update the improved deep reinforcement learning network. According to the invention, the walking problem of the humanoid robot under a complex environment can be effectively solved; and a test is carried out on a simulation platform and a real robot to verify the validity of the method.

The invention discloses a gait planning method for a biped robot based on deep reinforcement learning, and the method utilizes the stability and flexibility of human gait and combines deep reinforcement learning to effectively control the gait of the biped robot. The method includes the following steps: 1), establishing a passive biped robot model; 2), acquiring and processing human gait data andtarget gait data; 3), extracting the implicit features of the gait data of the biped robot and human gait data through using a noise reduction auto-encoder; 4), learning the gait characteristics of ahuman body through using deep reinforcement learning, and then planning the gait of the biped robot. The method combines deep reinforcement learning with human gait data to control the biped robot towalk as stable and compliant as a human.

The invention provides a hexapod robot gait planning method based on depth reinforcement learning. The method comprises the steps of environment information acquiring, environment modeling, anchor point screening, upper-level exercise policy formulating, lower-level policy executing and robot movement drive. Through the hexapod robot gait planning method, a hexapod robot uses a depth learning andreinforcement learning algorithm to acquire the optimal path by solving a quincuncial pile maze problem abstracted from an environment, and can select the suitable anchor point according to the optimal path to realize efficient walking in a non-structural environment.

The invention discloses a humanoid robot gait planning method based on human body teaching. Human body movement information is used for gait planning of a humanoid robot. The joint driving angle of the robot is obtained through inverse operation of skeleton information during human body movement. The mass center track information of the robot is estimated through human body teaching data in orderto meet the stability during movement of the humanoid robot. Compensation control is conducted on the driving angle of the robot by combining a ZMP criterion and a robot mass center-angle jacobi matrix. In selection of supporting feet, switching of the supporting feet during human body teaching is judged through a hysteretic curve, so that selection of the supporting feet of the robot is determined. The humanoid robot gait planning method based on human body teaching greatly simplifies the complexity of humanoid robot gait planning, and the movement process of the robot is more personified through the gait planning based on human body teaching.

The invention discloses a self-adaptive control method for diagonal gait of a four-footed robot, which is used for self-adaptively controlling the walking of the four-footed robot in the nonstructural terrain with the diagonal gait. The self-adaptive control method bases on a fuzzy reasoning learning method and a foot trajectory real-time correction method and takes the diagonal gait as the motion mode so as to comprehensively control a robot body to adapt for the nonstructural environment. The gait planning is carried out on the environmental information collected by the sensor through the fuzzy reasoning learning method; the stability judgment is carried out on the gait planning information; if the planned four-footed robot has stable diagonal gait, the joint corner information is sent to a control system; if not, the supporting time of the diagonal feet is corrected by the foot trajectory real-time correction method so that the four-footed robot has better stability when walking under the changed gait in the nonstructural terrain. The self-adaptive control method is used for correcting foot trajectory planning and walking stability under changed gait when the four-footed robot walks diagonally in a dynamic way, thus realizing the self-perception, self-correction and self-adjustment of the four-footed robot.

The invention provides a lower limb rehabilitation robot gait planning method based on lower limb walking characteristics. Firstly, parameterized modeling is performed based on the lower limb walking characteristics, wherein parameterized mathematic models of the lower limb hips, knees and ankle turn angles of the human body are established, and the forms of function curves of the models and required parameter number are determined; secondly, a mathematic inverse solution equation set of curve parameters is established; thirdly, the linear and nonlinear mapping relation of human body structure parameters and characteristic key point positions is found; finally, human body structure parameters are measured, the characteristic key points of a patient are found according to the mapping relation between the human body structure parameters and the characteristic key points, the curve parameters are inversely solved to obtain a specific curve expression, kinematic positions and corresponding lower angles of a robot are solved according to the specific structure and configuration of the lower limb rehabilitation robot to generate a joint movement instruction corresponding to the robot. The lower limb rehabilitation robot gait planning method conforms to a human body lower limb movement rule and can also reveal the mechanism of human body lower limb movements.

The invention discloses a biped robot climbing stair gait planning method and device and a robot. The biped robot climbing stair gait planning method comprises the steps that in the stair climbing process, a variable-length first-order inverted pendulum model is adopted to carry out trajectory planning on a forward gait, the motion trajectory of an ankle joint is constrained according to preset constraint conditions, and a first-order linear inverted pendulum model is adopted to carry out trajectory planning on a lateral gait to realize stair climbing of a biped robot; the process is that whenfacing the first stair, the center of gravity of a body is controlled to move to the left; a right foot of a swing leg is lifted to drop the right foot onto the stair; the center of gravity of the body is moved to the right foot, and the left foot of the swing leg is lifted so that the left foot falls onto the stair; the steps are repeated until the first stair is the last stair; whether the center of gravity falls within a stable region or not is judged; if yes, the walking is finished; and if not, the two legs are controlled to close and return to the starting state. Based on the biped robot climbing stair gait planning method and device and a robot, an effective application of a robot walking mode can be realized, so that the robot can walk on the stairs as freely as a human.

The invention discloses a double leg strength test experimental platform of a foot type walking robot. The experimental platform is composed of a robot double leg experimental platform frame, robot single leg assemblies, a robot double leg connector, a robot double leg experimental mechanism and a robot double leg controller. The two robot single leg assemblies are connected through the robot double leg connector, and the robot double leg experiment mechanism is connected with the robot double leg connector through a sliding block on a vertical guide rail. The robot double leg control system is used for collecting and processing data of the robot double leg experiment platform by setting communication of an upper computer and a lower computer and performing planning control over movement strategies. The double leg strength test experimental platform is suitable for controlling single leg strength and double leg strength of the foot type walking robot, walking tests on different terrain, load distribution and gait planning and suitable for early-stage study and control strategy verification of researchers of the foot type walking robot and has the advantages of being low in cost, easy to manufacture and accurate and reliable in experimental result.

The invention discloses a linear inverted pendulum model-based robot gait planning method. The linear inverted pendulum model-based robot gait planning method comprises the following steps of building a linear inverted pendulum model, in which a proportion factor Gamma is introduced on the basis of a dual-connection rod inverted pendulum model (DLIPM) to build a new linear inverted pendulum model; and performing robot gait planning, in which zero-torque point track, trunk mass center motion track and free foot planning track are generated on the basis of the linear inverted pendulum model, and each joint gait motion curve is solved by inverse kinematics. The linear inverted pendulum model built by the method more conforms to actual mass distribution condition of a humanoid robot, the zero-torque point track, the trunk mass center motion track and the feet foot planning track which are generated on the basis of the model and the each solved joint gait motion curve linear are more reasonable and more conform to the gait process of a natural person dual-foot robot, and the gait planning accuracy is high.

The invention discloses a rugged-terrain-oriented four-foot robot double-layer structure gait planning method. During upper layer movement planning, landing points of swing legs under each swing phaseare planned according to a gait sequence, and centroid track planning of a robot trunk is carried out on each supporting phase to form the optimal target movement state of each step till the finishing point of a given planned path is reached, wherein the landing points are planned in the mode that a sector search area of the current swing legs in the four-foot robot is built in a grid map, and coordinates of the optimal target landing points are found in a passable area of the sector search area. By using method, the calculation efficiency and the terrain adaptability of the robot are improved under the condition that the movement speed and stability of the robot are not reduced.

The invention discloses a humanoid gait planning method for a biped robot, and the method enables the movement characteristics of human walking to be integrated into the gait planning of the biped robot, so as to improve the walking energy efficiency of the robot and improve the endurance of the robot. The humanoid gait planning method comprises double-foot support phase planning and single-foot support phase planning, in the biped support phase, the robot enables the mass center of the robot to move forwards by means of the thrust of the ankle, meanwhile the mass center of the robot gets close to the supporting legs in a forward plane, so that the landing time is provided to the single-foot supporting phase; and in the single-foot supporting phase, the degree of freedom of the single-footsupporting phase and the degree of freedom of the lateral plane of the robot supporting ankle joint are in a passive state, the mass center of the robot moves forwards naturally under the action of gravity, and the swinging foot moves to the foot falling point quickly within the estimated landing time.

The invention relates to a human body motion capture data-based humanoid robot gait planning method. The method includes the following steps that: human body gait modes are extracted, angle curves of the six degrees of freedom of a single leg are obtained, wherein the human body gait modes include hip joint roll, hip joint pitch, hip joint yaw, knee joint pitch, ankle joint pitch and ankle joint roll; format conversion and mathematical optimization are performed on the human body gait mode data, so that the data can be applied to a human simulating robot; and a ZMP control algorithm is introduced to finely adjust the joint angles of a humanoid robot to improve the robustness of the humanoid robot. With the human body motion capture data-based humanoid robot gait planning method of the invention adopted, the gait planning difficulty of the humanoid robot can be simplified, the humanoid property and stability of the gait of the humanoid robot can be improved, and a foundation can be provided for the complex action planning of the humanoid robot. Compared with a traditional humanoid robot gait planning method, the human body motion capture data-based humanoid robot gait planning method has outstanding advantages and has a bright prospect.

The invention provides a biped walking simulation evaluation system and method of a robot based on six-degree-of-freedom legs. The system comprises a biped robot model and a gait planning algorithm, the biped robot model-1 and the gait planning algorithm-2 are input of the simulation evaluation system, the biped robot model contains the shape and dynamics parameters of the robot, and the gait planning algorithm can be used for planning motion trajectories of more than 20 joints of the whole body according to expected staying footprints; the simulation evaluation system outputs a simulation evaluation result used for evaluating the robot structure and the performance of the gait planning algorithm. Through a computer simulation mode, the purpose of omnidirectional walking motion optimization with more complex parameters and more involved factors is achieved.

The invention discloses a shape memory alloy-driving bionic wall climbing robot leg unit. The shape memory alloy-driving bionic wall climbing robot leg unit comprises bionic legs, a foot module, a shock absorption spring, a bi-directional shape memory alloy driving device and a single-directional shape memory alloy driving device, wherein the front-back and left-right movement of the bionic legs are driven by virtue of the bi-directional shape memory alloy driving device and the single-directional shape memory alloy driving device. The invention also discloses a shape memory alloy-driving bionic wall climbing robot driven by shape memory alloy. The bionic wall climbing robot comprises the bionic wall climbing robot leg unit. The shape memory alloy driving device adopts a bionic structure,so that the wall climbing movement of the robot on a medium surface is realized; the driving device is simple in structure, light in weight and flexible in movement; and the foot module adopts an ironmagnetic ball body, no surplus leg raising and re-absorption action is needed in the movement process of the bionic leg, the leg can provide a stable supporting force and absorption adsorption force,the supporting point is stable, the center of gravity of the robot is slight in change, and the gait planning is simple.

The invention discloses a quadruped robot and a gait planning method thereof. The quadruped robot comprises a robotic trunk, wherein the upper part of the front end of the robotic trunk is connected with a robotic head through a robotic neck, and the lower part of the robotic trunk is connected with four legs of a first leg, a second leg, a third leg and a fourth leg; the first leg and the second leg are mounted at the front end of the robotic trunk, the third leg and the fourth leg are mounted at the rear end of the robotic trunk, and each leg comprises four parts of a shank part, a thigh part, a vertical rod part and a horizontal rod part which are sequentially connected from bottom to top; a control device is arranged on the trunk and used for controlling the robot to walk. The quadruped robot disclosed by the invention has the functions of walking smoothly and automatically avoiding obstacles.

The invention discloses a humanoid robot walking control method, belongs to the field of humanoid robots, and provides the following technical scheme in order to solve the problem of improving the stability and balance ability of a humanoid robot in the walking process: the humanoid robot walking control method performs offline gait planning on the humanoid robot, so a joint movement track generated by the humanoid robot in an off-line tracking mode has the walking capacity; state information in the walking process of the humanoid robot is collected in real time, the walking posture of the humanoid robot is adjusted on line in real time in response to a stability controller of the state information, the humanoid robot can walk on an uneven road surface, and the effect is that walking control over the humanoid robot is achieved in the mode of combining offline gait planning and line stability control, and the humanoid robot has the effect that the humanoid robot has the capability of adapting to complex environments.

The invention discloses a pipeline detection robot. The pipeline detection robot comprises a robot head, a robot middle portion connected with the robot head through a first elastic connecting body and a robot tail connected with the robot middle portion through a second elastic connecting body. A bionic leg for traveling is arranged on each of the left sides and the right sides of the robot head, the robot middle portion and the robot tail. A camera for camera-shooting and multiple ultrasonic sensors for detecting obstacles are arranged on the robot head. A circuit control board for controlling the traveling gait of the robot and receiving detection signals is arranged on the robot middle portion. A winder for take-up and pay-off of cables is arranged at the robot tail. According to the pipeline detection robot, a body of the pipeline detection robot has certain scalability and flexibility, meanwhile has flexible gait planning capability and can adapt to the application needs of cornering, crawling, obstacle avoidance and detection in different pipeline environments; and due to the fact that detection signal transmission is stable and reliable, the pipeline failure detection efficiency of the robot is higher.