722 results about "Robot control" patented technology

The invention provides a robot training data analysis method and system based on deep learning. The method comprises the steps that operation parameters reflecting the work intensity of a robot are acquired; acquiring original sensor data acquired by the robot; deducing the physical state deviation of an end effector of the robot based on the operation parameters; according to the physical state deviation, adjusting the original sensor data to obtain adjusted sensor data; and optimizing a robot control strategy based on the adjusted sensor data. The method can effectively identify and quantify the sensor data error of the robot caused by the physical state deviation, and solves the problem that the model performance is reduced due to the fact that a traditional deep learning system learns in dirty data by correcting the original sensor data, thereby improving the accuracy and efficiency of robot control strategy optimization, and improving the user experience. And invalid adjustment caused by error attribution is avoided.

The embodiment of the application provides a kind of based on multi-sensor fusion's material box sorting robot control method and device, method includes: by laser radar, vision camera and wheeled encoder in robot body deployment, real-time collection multi-source perception data;The multi-source perception data is input based on the data fusion model of extended Kalman filter, by the layered architecture of prediction layer fusion, global correction fusion and local fine-tuning fusion, determine the three-dimensional pose of robot;Real-time image data is input into preset image recognition model, and dynamic obstacle list and static environment semantics are output using cross-layer feature fusion mechanism;Robot three-dimensional pose, dynamic obstacle list, static environment semantics and preset task instruction are jointly input into dynamic time window path planning model, under the condition of meeting task instruction hard constraint, path planning is carried out using space-time search algorithm, and the optimal trajectory is determined, the application can improve material box sorting operation efficiency, realize warehouse optimization upgrade.

The device envisions a robot control unit capable of easily estimating the weight and / or center of gravity of an object. This robot control unit comprises an external force calculation unit, an estimation unit, and a determination unit. The robot control unit directs a robot to perform a predefined process on an object. The external force calculation unit calculates an external force based on the output of a torque sensor provided for the robot. The estimation unit estimates the weight and / or center of gravity of the object based on the calculated external force. The determination unit uses the calculated external force to determine whether an estimate performed by the estimation unit is feasible.If, based on the calculated external force, it is determined that an estimation performed by the estimation unit is not possible, the determination unit changes the orientation of the robot and performs calculations carried out by the unit for calculating an external force and an estimation carried out by the estimation unit.

In robots, when performing continuous movement actions, the total time required for a series of movement actions is reduced without being constrained by factors such as acceleration. [Solution] When a robot is given an action command to perform action A, action B following action A, and action C following action B, action C is started when action A enters and ends in a deceleration section, or immediately before it ends. Furthermore, action B is executed in conjunction with at least one of action A and action C, such that action B starts after the start of the deceleration section of action A and ends before the end of the acceleration section of action C.

The present application relates to a kind of four-legged robot single leg sliding mode control method based on interference observer, belong to robot control technical field.First, the trajectory of four-legged robot single leg in swing phase is established;Then, the kinematic model of four-legged robot single leg is established, to realize the conversion of foot target position and corresponding joint angle;On this basis, non-singular terminal sliding mode controller is designed to deal with the dynamics control under the condition of no interference;Finally, combined with sliding mode controller and interference estimation, the final composite control strategy is designed, the stable tracking of robot joint angle is realized, so as to improve the anti-interference performance of single leg system.The present application utilizes the combination of finite time interference observer and non-singular terminal sliding mode controller, designs a kind of composite controller, to improve the foot trajectory tracking accuracy and system robustness of four-legged robot, eliminate the uncertainty caused by external disturbance and modeling error suffered by robot.

The present application belongs to the technical field of robot control, and specifically discloses a railway inspection robot laser radar anti-collision and remote video linkage control method, aiming to solve the problems of low inspection efficiency, low sensor fusion degree and missing linkage mechanism in industrial scenarios. The method comprises: performing benchmark calibration and entering inspection line data through system initialization and parameter configuration; acquiring point cloud data by high-frequency dynamic scanning using a laser radar, and real-time matching of multi-level early warning thresholds to trigger linkage response; the power control unit executes brake or speed reduction operation according to the instruction, and resets the mode to resume work after confirming safety; the system also archives the full data of linkage events to support continuous optimization of the inspection plan. Through the deep coupling linkage of the laser radar and the video imaging module, the present application can realize all-weather, high-precision and high-safety autonomous operation of the inspection robot in complex high-dynamic environments, significantly improving the inspection efficiency and risk control level.

This application belongs to the field of collaborative robot control technology, specifically disclosing a human-machine compliant interactive stability control method and system for achieving stiffness jumps. The method includes: real-time acquisition of the collaborative robot's end-effector interaction force and end-effector Cartesian velocity input to a fuzzy inference function to obtain human intention coefficients; calculation of the desired total stiffness and desired total damping at the current moment based on the human intention coefficients, and decomposition them into time-varying components and time-invariant basis components, respectively; employing an energy tank method to ensure system stability, calculating the filling rate of the energy tank based on the power dissipated by the basis damping when the energy tank is not full; if the energy tank is not depleted, using an energy exchange control method to replenish the collaborative robot with the energy required for time-varying stiffness and time-varying damping, and obtaining the desired velocity, desired position, and joint angle of the collaborative robot at the next moment. This method can dynamically and rapidly adjust admittance parameters according to the operator's intentions while ensuring the collaborative robot system remains strictly passive.

The application provides a robot dance automatic generation method and system based on music feature analysis, and relates to the technical field of robot control. First, a music feature spectrum covering rhythm, energy and other dimensions is constructed to convert abstract music into structured information understandable by machines, and then a dance semantic label sequence bound to a time axis is generated to provide accurate basis for action selection. By combining with a forward-looking window to predict future music features, the current and future candidate actions are filtered through a multi-objective model to realize the dual matching of action and music micro-rhythm and macro-structure. The finally generated control instruction sequence directly drives the physical robot, which breaks the dependence on video materials, realizes the automatic generation of dance with creativity and artistic expression, and improves the generation efficiency and adaptive flexibility.

An embodiment of the present disclosure provides a method of controlling a robot based on a simplified map and a natural-language command. The method may include acquiring an image of a space, acquiring a simplified map of the space, the simplified map including a current position of a robot and a destination position of the robot and not including at least some elements describing the space, acquiring a natural-language command specifying compliance requirements that the robot is to comply with during an operation of the robot, generating input tokens based on at least one of the image, the simplified map, and the natural-language command, and generating waypoint tokens by processing the input tokens with a trained vision-language-action model.

The application relates to the technical field of robot control, and particularly discloses an embodied intelligent robot autonomous decision-making method and system based on environment interaction perception, which comprises the following steps: determining a precision map according to the cargo information, determining the reference working parameters of the robot based on the precision map, sending the reference working parameters to the robot, receiving the collection information fed back by the robot, creating and updating a real scene model according to the collection information and a preset BIM model, cutting out a sub-area in the real scene model, updating the precision map according to the information state of the sub-area in the model sequence, and determining the floating working parameters of the robot based on the updated precision map. Through random inspection of the whole scene, compared with global inspection, the method of randomly inspecting only part of important areas greatly reduces the workload while reducing the operation effect by only a little, and prolongs the endurance time.

This application belongs to the field of household appliance technology, specifically relating to a floor cleaning device, a cleaning robot, a control method, and a cleaning system. The floor cleaning device is installed on the main unit of the cleaning robot and includes a housing and a cleaning component. The housing reciprocates between a first position and a second position. The first position protrudes from the bottom surface of the main unit, and the second position is located inside the main unit. The housing has a communicating cleaning port and a water inlet, with the cleaning port located at the bottom of the housing. The cleaning component is rotatably disposed within the housing, with at least a portion of the cleaning component protruding from the cleaning port. This application effectively cleans stubborn stains on the floor and improves the cleaning ability of the cleaning robot.

The present application relates to the field of device for supporting, operating or replacing casting nozzle of a container for casting a melt, in particular to a submerged nozzle quick-change and crystallizer slagging robot system and a use method thereof. The submerged nozzle quick-change and crystallizer slagging robot system comprises an execution module (1), a visual positioning module (2) and a centralized control system (3), characterized in that: the execution module (1) comprises an inverted track robot (11) and an end effector (12); the visual positioning module (2) comprises at least one binocular camera (21); and the centralized control system (3) comprises a programmable logic controller (31), an industrial computer (32) and a robot control cabinet (33). The present application has high automation degree, reduces labor intensity and improves production efficiency.

A robot control device includes a robot language program execution section capable of executing a robot language program described by a dedicated robot language, an application program execution section capable of executing an application program described by a language different from the robot language, and an I / O input / output section capable of exchanging data using I / O data between the robot language program execution section and the application program execution section, the robot language program execution section and the application program execution section each converting data to be transferred to the other side into I / O data.

A robot system includes: a robot; a measurement apparatus; and a control apparatus. The robot includes: a robotic arm; an end effector; a movement apparatus to which the end effector is attached and which is configured to move the end effector along a translation axis; and a base to which the movement apparatus is attached and which is located at a position that is fixed relative to the robotic arm. The base includes a reflective member. The measurement apparatus measures a position of the base. The control apparatus controls at least one of the robotic arm and the movement apparatus based on a measured result by the measurement apparatus, the base is detachable from the robotic arm.

This application provides a robot control method and a robot that can introduce an operation stage prediction mechanism. The action generation model can predict the operation stage based on observation data containing the current operation stage identifier, current state information, and current scene image, combined with operation task information. This allows for accurate judgment of the transition timing of each operation stage in the operation task. When the predicted operation stage identifier differs from the current operation stage identifier, an operation stage jump is automatically triggered, and a predicted action that conforms to the behavioral logic of the operation stage is generated. This achieves a smooth transition in action generation, avoiding problems such as action switching stuttering and action jitter during robot control, and significantly improving the smoothness and stability of robot control.

The application discloses a multi-modal perception robot elevator control method and related equipment, the method comprises the following steps: obtaining an elevator task and current radar data, performing path planning according to the current radar data and preset floor map data to obtain an elevator path; controlling the robot to go to a floor elevator position according to the elevator path, performing multi-modal scanning on the floor elevator position, obtaining multi-modal environment data and performing fusion analysis to determine elevator state information; performing decision analysis according to the elevator state information and the obtained elevator task to generate robot control instructions, and controlling the robot to execute the elevator task according to the generated robot control instructions; obtaining multi-modal data through multi-modal scanning to perform fusion analysis, determine the elevator state information, and generate control instructions based on the elevator state information, and control the robot to execute the elevator task. The application embodiment can improve control reliability and reduce implementation cost. The application can be widely applied in the field of robot control technology.

The robot control device comprises: a first wireless device that performs wireless communication with a second wireless device enclosed in an external unit; a receive level monitoring unit that monitors the receive level between the first wireless device and the second wireless device; a receive level notification unit that notifies the receive level when it is determined that the receive level is lower than a predetermined value; and a change notification unit that notifies a request to change the position of at least one of the robot, the first wireless device, and the second wireless device.

The application provides a kind of foot type robot control method, device, electronic equipment and storage medium, it is related to robot control technical field, the foot type robot control method includes: obtaining the state vector of foot type robot, form parameter coding and control condition, input training good generative motion prior model, generate multiple candidate action sequences;Again candidate action sequence is substituted into the dynamics model that is constructed in advance, corresponding prediction state is obtained;Then risk assessment is carried out in combination with candidate action sequence and prediction state, obtains each sequence evaluation value, selects the minimum as optimal action sequence;Then the action residual correction quantity of trained residual reinforcement learning model is output, and the optimal action sequence is corrected to obtain the modified action sequence;Finally, the robot executable action sequence is obtained by the preset projection constraint module projection processing.The application can effectively improve the efficiency of foot type robot control.

The application relates to the technical field of robot control, and discloses a robot multi-mechanical arm cooperative anti-collision control method and system, which comprises the following steps: a three-dimensional global reference coordinate system with a kinematic pivot point as the origin is established, an inside and outside work area is divided, and an outside virtual node is defined; a virtual safety monitoring section is constructed in the outside driving area, a dynamic repulsion spot is generated, collision early warning is generated through position and trend compound judgment; the virtual repulsion force is mapped into equivalent joint damping torque, and is superimposed into a servo driver current control loop; the application realizes mathematical reconstruction of the outside non-perception blind area state without an external sensor through kinematic reverse deduction and virtual force field construction, and effectively prevents implicit physical collision of external connecting rods caused by the 'cohesion and external dispersion' motion coupling of the multi-mechanical arms through force feedback of the current control loop level.

One embodiment of a method for training a machine learning model to control a robot includes causing a model of the robot to move within a simulation based on one or more outputs of the machine learning model, computing an error within the simulation, computing at least one of a reward or an observation based on the error, and updating one or more parameters of the machine learning model based on the at least one of a reward or an observation.

The embodiment of the application provides a mobile fire duty robot self-navigation target control method and system, and relates to the technical field of fire duty robot control technology.The method comprises the following steps: obtaining a target operation instruction, and controlling a fire duty robot to move to a target operation area; obtaining image information of the target operation area, and determining a target two-dimensional coordinate based on the image information; when a visual confidence of a target operation component in a depth direction is lower than a preset confidence threshold, controlling an end positioning touch rod of the fire duty robot to perform physical detection in the depth direction, so as to obtain physical contact information of the target operation component; determining a target three-dimensional coordinate based on the target two-dimensional coordinate and the physical contact information; and controlling an end effector of the fire duty robot to move to the target three-dimensional coordinate to perform a preset operation.Through the application, the problem of low robot control precision is solved, and the effect of improving robot control precision is achieved.

The invention relates to the technical field of electric power inspection and maintenance robot control, and discloses a power transmission line crawling robot lifting self-adaptive shake suppression control method and system. The method comprises the following steps: acquiring three-axis acceleration and three-axis angular velocity data of a robot platform, calculating a yaw angle and an angular velocity of the robot platform relative to a vertical balance position, and constructing an evaluation index representing the shaking strength of the platform; according to the shake strength evaluation index, driving parameters of a lifting motor are dynamically adjusted through a nonlinear mapping relation; and the swing period and phase of the robot platform are monitored, and when the platform swings to a specific phase point, an instantaneous driving pulse is applied to a lifting motor to consume swing energy. By means of the method, the maximum swing angle of the robot platform can be reduced by 60% or above, the static time after the robot platform reaches the target position is shortened by 70% or above, and the working efficiency is remarkably improved.

The application discloses a kind of based on active learning and man-machine collaborative optimization's massaging robot control method, it is related to artificial intelligence, medical rehabilitation robot and complex contact type operation control cross technical field, this method relies on multimodal perception system, core computing hub and remote man-machine collaborative interface control platform, in turn complete the initial massaging strategy model construction of multi-source heterogeneous expert data, the uncertainty measurement based on bayesian neural network, man-machine collaborative key frame active learning trigger, reinforcement learning reward function design of fusion patient biological feedback and incremental learning of impedance control parameter and force-position hybrid output;The application greatly improves the clinical safety of massaging robot, exponentially reduces data acquisition cost, realizes individual physiological closed-loop rigid-flexible massaging, also overcome the catastrophic forgetting problem of neural network, endow system lifelong learning ability.

This invention relates to the field of master-slave robot control, specifically providing a motion mapping method, apparatus, electronic device, and storage medium for master-slave robots. The method includes: acquiring the position data and movement speed of the master hand, as well as the shortest movement distance from the hand to the microscope boundary; determining a preliminary mapping ratio based on the movement speed; the preliminary mapping ratio being the ratio of the master hand displacement to the slave hand displacement; determining a safety damping coefficient based on the shortest movement distance and a preset safety distance; multiplying the preliminary mapping ratio by the safety damping coefficient to obtain a target mapping ratio; determining a cutoff frequency based on the movement speed, and filtering the position data based on the cutoff frequency to determine the target filtered displacement increment of the master hand; determining the target displacement of the slave hand based on the target filtered displacement increment, and mapping the target displacement to the movement of the slave hand. The technical solution provided by this invention can improve the accuracy of master-slave motion control.

This invention provides an autonomous management and scheduling system for high-throughput chemical experimental robots based on digital twins, relating to the field of robot control technology. It includes a digital twin unit and a production execution unit. The production execution unit inputs production instructions and outputs a first production instruction signal. Based on the first and second production instruction signals, the production execution unit controls the robot to perform chemical experimental operations. The production execution unit feeds back the production result signal to the digital twin unit. This invention continuously collects overall operational and simulation data of the high-throughput chemical experiment through the digital twin unit, continuously improves the equipment twin model related to the entire experiment, and then interacts with the production execution unit. This provides data support for improving the efficiency of high-throughput chemical experiments and the output of experimental samples, and also provides a foundation for decision optimization of the entire system.

A robot control device configured to control operation of a robot configured to transfer a substrate while holding the substrate is provided. The substrate becomes in a first state where an end effector holds the substrate and the substrate is not placed at an installation position, when the end effector positions at a first teaching point above the installation position. The substrate becomes in a second state where the end effector does not hold the substrate and the substrate is placed at the installation position, when the end effector positions at a second teaching point below the installation position. The first and second states can be switched by causing the robot to perform a first operation to move the end effector from either one of the first teaching point and the second teaching point to the other one of the first teaching point and the second teaching point.

The application relates to the technical field of robot control, and discloses a terminal position dynamic compensation method based on fusion of visual servoing and force feedback, which comprises the following steps: collecting a top view image, segmenting an irregular residual hole contour, and calculating a dynamic alignment center coordinate; then, a reverse conical spiral descending trajectory is generated in combination with a lateral force vector to guide a fiber tail fiber connector into a guide port; in the insertion stage, the curvature radius characteristic of a force-displacement curve resistance peak point is used to determine the authenticity of bottom touching; for the case that the bottom touching is determined to be false, the side projection chord length is used to deduce an under-insertion depth, and a high-frequency micro-oscillation compensation insertion is driven until a real buckle locking feature is captured; the application realizes accurate alignment and reliable positioning of fiber plug-in and plug-out actions under complex working conditions such as shielding and mechanical wear.