Robot control method and program product

Abstract

The application discloses a robot control method and a program product, the method comprising: receiving an operation instruction, obtaining multi-source observation information of a robot based on a multi-modal sensor deployed on the robot, inputting the multi-source observation information into an environment detection model, and obtaining an environment detection result of the robot; wherein the operation instruction comprises a target operation to be performed by the robot, and the environment detection result comprises environment information in a work scene of the robot; generating a control parameter corresponding to the robot based on the environment information and the operation instruction, generating a motion parameter sequence corresponding to the robot according to the control parameter, and controlling the robot to perform the target operation according to the motion parameter sequence; wherein the control parameter comprises a trajectory planning parameter and a force control parameter, so as to realize accurate control of the robot.

Description

Technical Field

[0001] This invention relates to the field of nuclear power plant operation and maintenance robots and embodied intelligent control technology, and in particular to a robot control method and program product. Background Technology

[0002] Robotics technology has deeply penetrated multiple industries such as manufacturing, logistics, and healthcare, playing an increasingly crucial role in scenarios such as automated production, intelligent warehousing, surgical assistance, and inspection and maintenance.

[0003] In related technologies, robots primarily rely on preset operation instructions and fixed motion libraries for control when performing tasks. Tasks are decomposed into a series of pre-programmed standardized actions stored in the motion library, and the target operation is achieved through parameterized calls. However, this control based on static motion libraries has significant limitations when facing the complexities of the real world. Preset action parameters often fail to adapt to dynamically changing environmental conditions or object characteristics, leading to execution failures or performance degradation. Furthermore, training strategies from simulation environments often experience performance degradation or even loss of control when transferred to physical robot platforms due to modeling errors, sensor noise, actuator latency, and other factors, resulting in instability in the simulation-to-real-world transfer and failing to meet the demands of complex, dynamic, or high-precision tasks. Therefore, a robot control method is urgently needed to achieve precise control of robot operations. Summary of the Invention

[0004] This invention provides a robot control method and program product to solve the problems of parameter mismatch, weak generalization ability, and unstable migration from simulation to actual machine in robot control based on fixed motion libraries in related technologies.

[0005] According to one aspect of the present invention, a robot control method is provided, the method comprising: The robot receives an operation command, acquires multi-source observation information of the robot based on multi-modal sensors deployed on the robot, inputs the multi-source observation information into an environmental detection model, and obtains environmental detection results of the robot; wherein, the operation command includes the target operation to be performed by the robot, and the environmental detection results include environmental information within the robot's operating scene; Based on the environmental information and the operation instructions, control parameters corresponding to the robot are generated. Based on the control parameters, a sequence of motion parameters corresponding to the robot is generated. Based on the sequence of motion parameters, the robot is controlled to perform the target operation. The control parameters include trajectory planning parameters and force control parameters.

[0006] According to another aspect of the present invention, a robot control device is provided, the device comprising: The acquisition module is used to receive operation instructions, acquire multi-source observation information of the robot based on multi-modal sensors deployed on the robot, input the multi-source observation information into the environmental detection model, and obtain the environmental detection results of the robot; wherein, the operation instructions include the target operation to be performed by the robot, and the environmental detection results include environmental information within the robot's operating scene; The control module is used to generate control parameters corresponding to the robot based on the environmental information and the operation instructions, generate a sequence of motion parameters corresponding to the robot based on the control parameters, and control the robot to perform a target operation based on the sequence of motion parameters; wherein, the control parameters include trajectory planning parameters and force control parameters.

[0007] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the robot control method according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the robot control method according to any embodiment of the present invention.

[0009] According to another aspect of the present invention, embodiments of this disclosure also provide a computer program product, including a computer program that, when executed by a processor, implements the robot control method as described in any of the embodiments of this disclosure.

[0010] The technical solution of this invention firstly involves receiving an operation command and acquiring multi-source observation information of the robot based on multi-modal sensors deployed on the robot. This multi-source observation information is then input into an environmental detection model to obtain environmental detection results for the robot. The operation command includes the target operation to be performed by the robot, and the environmental detection results include environmental information within the robot's work environment. Determining the target operation enables comprehensive and accurate perception of the robot's work environment, improving the integrity and reliability of robot status and environmental detection. Next, control parameters corresponding to the robot are generated based on the environmental information and the operation command. A sequence of motion parameters corresponding to the robot is then generated based on the control parameters, and the robot is controlled to perform the target operation based on the motion parameter sequence. The control parameters include trajectory planning parameters and force control parameters. Generating control parameters based on environmental information and operation commands allows the robot's control to adapt to both the task objective and the actual work environment, improving the safety and adaptability of the robot in performing the target operation, ensuring the rationality of motion planning, achieving high-precision and high-stability motion execution, and enhancing the robot's operational accuracy and robustness in complex scenarios.

[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart of a robot control method provided according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of a robot control method provided according to Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the structure of a robot control device according to Embodiment 3 of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device that implements the robot control method of this invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0017] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0018] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0019] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0020] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0021] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0022] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0023] Example 1 Figure 1 This is a flowchart of a robot control method provided in Embodiment 1 of the present invention. This embodiment is applicable to situations where robot operations need to be controlled. The method can be executed by a robot control device, which can be implemented in hardware and / or software, optionally through an electronic device, such as a mobile terminal, PC, or server. Figure 1 As shown, the method may specifically include: S110. Receive operation instructions, acquire multi-source observation information of the robot based on multi-modal sensors deployed on the robot, input the multi-source observation information into an environmental detection model, and obtain environmental detection results of the robot; wherein, the operation instructions include the target operation to be performed by the robot, and the environmental detection results include environmental information within the robot's operating scenario.

[0024] In this embodiment of the invention, the operation instruction can be understood as an instruction used to instruct the robot to perform a target operation. The target operation can be a specific task that the robot needs to complete according to the operation instruction, such as grasping or carrying.

[0025] Optionally, the robot can receive input operation commands, determine the target operation to be performed by the robot based on the operation commands, and simultaneously detect the robot's own state and the working environment based on multimodal sensors deployed on the robot to acquire multi-source observation information. Multimodal sensors can be understood as a combination of sensors capable of collecting various types of information, which can be used to collect environmental information and robot's own state information from different dimensions. Multimodal sensors may include, but are not limited to, vision sensors, depth sensors, LiDAR sensors (Light Detection and Ranging Sensor), infrared radar, dose sensors, acoustic sensors, and other sensor units, avoiding the perception limitations of single sensors. Multi-source observation information can be understood as observation data collected from multiple dimensions by multiple sensors. Multi-source observation information may include, but is not limited to, image data, point cloud data, force data, position data, attitude data, velocity data, acceleration data, and other types of information.

[0026] Furthermore, after acquiring multi-source observation information, this information can be input into an environmental detection model. The model processes and analyzes this information, fusing visual and operational information to output environmental detection results. These results include State(t) of the robot's operational scenario, indicating the environmental state at multiple points in time within the scenario. The operational scenario can be understood as the spatial environment in which the robot performs its target operation. Environmental information can be related to the robot's operation, such as, but not limited to, the robot's position, obstacles within the scenario, the distance between the robot and obstacles, and the robot's pose. The environmental detection results can also include, but are not limited to, obstacle information, scene geometry, spatial constraints, and environmental stability information, providing a precise basis for subsequent robot trajectory planning.

[0027] Optionally, the environmental detection model can be a world model, which has strong environmental prediction and state maintenance capabilities. By predicting environmental information within the robot's operating scene through the environmental detection model, even under conditions where the visual sensor or LiDAR sensor is temporarily occluded, the environmental detection model can still use the State(t) predicted by internal dynamics to maintain the control loop without interruption through dead-space operations, effectively avoiding control failure due to sensor occlusion. At the same time, in occluded or low-light scenarios, the environmental detection model can maintain stable tracking of the operating scene and robot state through temporal memory and loop closure association, make short-term predictions of the current state based on historical states, and retrieve historical environmental information for loop closure matching, ultimately outputting accurate state estimates to improve the robot's robustness in complex perception degradation scenarios.

[0028] Building upon the aforementioned approach, the environmental detection model, when fusing multi-source observation information collected by multimodal sensors, can achieve dynamic separation of static and dynamic elements. For example, static elements can include fixed objects within the work environment, such as walls and stationary equipment, while dynamic elements can include movable objects within the work environment, such as mobile robots and mobile devices. It outputs a static map and a dynamic object list for trajectory prediction, respectively, further improving the accuracy of environmental detection. Furthermore, the environmental detection model can optimize the traditional occupancy grid system by not only marking each grid cell as occupied but also indicating whether the cell is passable and the passability cost, thus constructing a passable grid. The system uses a traversable grid map. For example, when the grid type is empty, the occupancy probability is 0%, and the traversability score can be 1, indicating that it is completely traversable. When the grid type is a static obstacle, the occupancy probability is 100%, and the traversability score can be 0, indicating that it is not traversable. When the grid type is a dynamic obstacle and it is predicted to leave after 500ms, the occupancy probability is 80%, and the traversability score can be 0.6, indicating that it is temporarily not traversable. This allows the robot to search for low-cost, safe, and feasible motion trajectories on the traversable grid, avoid dynamic obstacles in the spatiotemporal dimension, effectively improve the safety and efficiency of robot operations, and avoid the risk of operation interruption and collision caused by dynamic obstacles.

[0029] S120. Based on the environmental information and the operation instructions, generate control parameters corresponding to the robot, generate a sequence of motion parameters corresponding to the robot according to the control parameters, and control the robot to perform the target operation according to the sequence of motion parameters; wherein, the control parameters include trajectory planning parameters and force control parameters.

[0030] Control parameters can be understood as a set of parameters used to regulate the robot's motion and interaction behavior. These parameters can include, but are not limited to, trajectory planning parameters and force control parameters. Trajectory planning parameters are used to plan the robot's motion path and pose changes. Force control parameters are used to control the robot's interaction forces, torques, and contact states. The motion parameter sequence can be a combination of time-series parameters that transforms the control parameters into direct robot-driven actions. This sequence can be arranged in chronological order and may include, but is not limited to, continuous pose, velocity, acceleration, force, and torque data to achieve continuous and precise control of the robot's movements.

[0031] Based on the above scheme, optionally, the trajectory planning parameters include at least one of the following: initial pose, target pose, and velocity information, wherein the velocity information includes maximum velocity and / or maximum acceleration; and / or, the force control parameters include at least one of the following: contact holding duration and interaction force constraint parameters, wherein the interaction force constraint parameters include target contact pressure and / or maximum allowable torque. The initial pose can be the starting position and attitude of the trajectory planning, and the target pose can be the ending position and attitude of the trajectory planning. The velocity information can include at least the maximum velocity and / or maximum acceleration, wherein the maximum velocity can be the upper limit of the allowable robot movement speed, and the maximum acceleration can be the upper limit of the allowable robot movement acceleration. By setting the velocity parameters, instability caused by excessively fast robot movement or low work efficiency caused by excessively slow movement can be avoided. The contact holding duration can be the duration for which the robot maintains the target contact state. The interaction force constraint parameters can include at least the target contact pressure and / or maximum allowable torque, wherein the target contact pressure can be the contact pressure that the robot is expected to maintain, and the maximum allowable torque can be the upper limit of the torque allowed during robot operation. By setting interaction force parameters, the work object and robot equipment can be effectively protected, avoiding damage caused by uncontrolled force, and improving the safety and reliability of the operation.

[0032] Based on the above scheme, optionally, generating the control parameters corresponding to the robot based on the environmental information and the operation instructions includes: performing semantic parsing on the operation instructions according to a semantic analysis model to obtain multiple operation parameters; wherein, the semantic analysis model is trained on a machine learning model based on multiple sample operation instructions and their labeled operation parameters; determining multiple target operation parameters corresponding to the robot from the multiple operation parameters based on a device knowledge base; and generating the control parameters corresponding to the robot according to the environmental information and the multiple target operation parameters.

[0033] The semantic analysis model can be understood as a model used to perform natural language understanding and parameter parsing of operation instructions. It is trained on a machine learning model based on multiple sample operation instructions and their labeled operation parameters. Operation parameters can be action elements parsed from the operation instructions, and can include, but are not limited to, parameters of various types such as target object, action type, target position, force, and speed. The device knowledge base is a database containing operation parameters, operation constraints, and executable actions corresponding to various devices. Through the device knowledge base, target operation parameters suitable for the current robot can be selected from general operation parameters. Target operation parameters can be operation parameters that match the robot, determined by combining the device knowledge base.

[0034] Operation instructions can take the form of Standard Operating Procedures (SOPs) or operation tickets in natural language. To achieve accurate understanding of operation instructions, a semantic analysis model can be used to segment, identify intent, and semantically parse the instructions, breaking them down into multiple operation parameters and filling them into corresponding parameter templates. The output parameter templates should include at least the target pose, pose tolerance, maximum torque / pressure, hold time, upper bound of velocity / acceleration, dose threshold and budget, reset strategy, and abnormal handling method. This provides a standardized basis for the generation of subsequent control parameters, ensuring a consistent parameter format after parsing different operation instructions and improving the standardization of robot control.

[0035] Considering the differences in equipment parameters and operational capabilities among different robots, to avoid issues such as parameter mismatch and operational anomalies caused by mismatched operating parameters and robot specifications, multiple target operating parameters corresponding to the robot can be determined from a range of operating parameters using the equipment knowledge base. This allows for the adaptation and verification of units, tolerances, and tools, avoiding control deviations caused by geometric differences between units with the same name. Simultaneously, the target operating parameters can be adaptively corrected based on the equipment knowledge base to avoid instability during the transition from simulation to the actual machine and large end-to-end performance fluctuations. Key parameters such as force limits, torque limits, upper limits for speed and acceleration, instantaneous dose and dose budget, minimum gap, and minimum inter-robot distance are all issued within parameter templates according to task type to meet different operational requirements.

[0036] After determining the target operation parameters, trajectory planning and force control can be performed based on environmental information and multiple target operation parameters to generate the corresponding control parameters for the robot. Then, a sequence of motion parameters ordered by time is generated based on the control parameters, and the trajectory planning parameters and force control parameters corresponding to the robot at multiple moments are determined based on the sequence of motion parameters, thereby controlling the robot to execute the target operation step by step.

[0037] To further enhance the flexibility and adaptability of robot control and meet the needs of different operational scenarios, multiple control strategies and models can be flexibly switched. Robot control can be flexibly switched between Model Predictive Control (MPC), Iterative Linear Quadratic Regulator (iLQR), and Model Predictive Path Integral Control (MPPI), allowing the selection of the optimal control strategy based on operational accuracy and real-time requirements. The perception backbone can be replaced between Vision Transformer (ViT), Swing Transformer, ConvNeXt, and sparse voxel encoding to adapt to different perception data types and computing power requirements. The map service supports dense / sparse representations, which can be flexibly selected based on scene complexity and computing resources, optimizing computing power consumption while ensuring positioning accuracy.

[0038] To ensure the real-time performance and consistency of robot control, point-to-point (PTP) or hardware triggering methods can be used to guarantee the time synchronization of actions in each stage. Using a priority message queue and bidirectional asynchronous buffer design for control and perception effectively meets the end-to-end P99 latency target, ensuring rapid response to operational commands. Based on actual operational needs, the outer loop control cycle can be set to 10-20ms, and the inner loop control cycle to 1-5ms. The end-to-end P99 latency is controlled within the range of 50-80ms, and the perception fusion P99 latency is controlled within the range of 60-100ms, achieving an optimal balance between real-time performance and control accuracy.

[0039] Based on the above scheme, optionally, after controlling the robot to perform the target operation according to the motion parameter sequence, the method includes: obtaining the actual operating parameters of the robot when performing the target operation, and generating an audit log based on the motion parameter sequence and the actual operating parameters. The audit log can be a robot operation record generated based on the motion parameter sequence and the actual operating parameters. The audit log may include parameters such as pose, force, velocity, and acceleration of the robot at multiple moments to ensure traceability of the entire robot operation process.

[0040] To ensure the robustness and adaptability of robot control under complex perception degradation conditions, optional multi-dimensional combined test scenarios can be constructed, such as single or combined working conditions including but not limited to low illumination, strong reflection, visual occlusion and high radiation dose, to simulate common sensor performance degradation in real working environments, so as to ensure the robot's stable operation under non-ideal perception conditions. Meanwhile, for each type of test scenario, multiple sets of baseline control strategies can be set for comparative evaluation. The baselines may include rule-based controllers built based on expert experience, traditional proportional-integral-derivative (PID) controllers, classical model predictive controllers (MPC), and simplified impedance closed-loop controllers without mass terms. Furthermore, the test evaluation can adopt a multi-dimensional quantitative index system, such as task success rate, positioning / alignment error, torque / pressure overshoot, holding time deviation, dose / speed / torque over-limit rate, energy consumption and P95 / P99 delay, number of realignments and total task duration, etc., to ensure the fairness and comparability of the evaluation results. If a certain index exceeds the corresponding threshold, realignment, path fine-tuning or process resampling / retightening will be automatically triggered.

[0041] Optionally, it can also support canary release and fast rollback mechanisms to ensure that the updated control strategy is deployed safely and controllably. During the robot's control operation, the distribution drift of the state estimation residual and the abnormal score is continuously monitored. Once a significant shift is detected, a small-sample online retraining process is triggered to maintain the model's effectiveness. At the same time, all operation logs and control decisions are written to read-only audit storage to ensure traceability throughout the entire lifecycle. Meanwhile, the hardware deployment must meet the real-time requirements of an inner-loop control frequency of not less than 1kHz and an end-to-end latency of not more than 80ms. Through communication link redundancy and uninterruptible power supply (UPS) configuration, it is ensured that the robot can safely dock in the event of a sudden failure.

[0042] Taking a precise valve operation task in an industrial scenario as an example, the system receives the natural language operation command "turn clockwise to 90 degrees, maximum 40 N•m, hold for 3 seconds," which is parsed by a language analysis model to generate a structured parameter template and skill map. Next, an environmental detection model estimates the valve stem and handwheel pose, reachable operating range, and other environmental information based on multimodal sensor data, and outputs an environmental representation including the environmental detection results and corresponding uncertainty scores. The environmental information and operation command are input into the outer loop controller to generate a trajectory parameter sequence and control parameter sequence that satisfy kinematic and force control constraints, and allocates a 3-second hold time and torque-time dose budget. The inner loop controller uses an impedance control model to track the trajectory parameter sequence. Simultaneously, a torque closed-loop mechanism ensures that the output torque does not exceed the preset maximum limit of 40 N·m. When the valve stem engages and jams, a micro-oscillation action is actively triggered to assist in decoupling. If an instantaneous zero angular velocity, torque exceeding the limit without rotation, or approaching the speed / torque safety boundary is detected, a retreat and realignment strategy is triggered. If the control barrier function determines that the operation is infeasible, an emergency stop and safe docking are executed. After completing the target operation, key indicators are evaluated to ensure that parameters such as angular displacement error, attitude deviation, torque overshoot, hold time deviation, and dose curve all meet the preset thresholds in the parameter template, and that the end-to-end control delay P99 does not exceed the corresponding threshold. Finally, all process data is written to a read-only audit log to complete the closed-loop quality verification. This robot control effectively solves problems such as low torque control accuracy, weak jamming handling capability, and lack of traceability in valve operation, improving the accuracy and safety of valve operation.

[0043] Taking the precise sample collection and sealing task as an example, key process parameters are first extracted from the operating procedures, which may include the target sampling volume, expected contact pressure, contact duration, sealing steps, and strategies for handling non-conforming samples. Next, based on the output of the environmental detection model, the outer loop planner generates a safe motion trajectory X(t) and end-effector contact posture to avoid high-dose areas and narrow channels. The inner loop controller employs an impedance control strategy to adjust the end-effector force in real time, ensuring that the actual contact force stably tracks the target value, with the steady-state error not exceeding the preset tolerance (Tolerance Differential Pressure, Delta). The robot minimizes shear force through an adaptive normal alignment mechanism to ensure sample integrity is not disturbed during sampling. After sampling, weighing or volume estimation is performed, and process verification is conducted using cleanliness sensors. If any indicator fails to meet the standard, a resampling process is triggered. The operation automatically terminates and an anomaly is reported after the number of resampling attempts exceeds a preset limit. After confirmation of compliance, automated sealing and barcode affixing are performed, embedding timestamps and geographic / pose markers. All process data is written to read-only audit storage. The task is considered successful if the closed-loop control error of contact pressure and duration is within the threshold range, the sample volume and cleanliness compliance rate meet process requirements, the cumulative radiation dose throughout the process does not exceed the budget, and the recovery time after an anomaly is within an acceptable window. The above robot control effectively solves the problems of sample susceptibility to disturbance, low sampling accuracy, and difficulty in quality control during sample collection, achieving full automation of sample collection, verification, sealing, and traceability, thus improving operational efficiency and sample quality.

[0044] The technical solution of this invention firstly involves receiving an operation command and acquiring multi-source observation information of the robot based on multi-modal sensors deployed on the robot. This multi-source observation information is then input into an environmental detection model to obtain environmental detection results for the robot. The operation command includes the target operation to be performed by the robot, and the environmental detection results include environmental information within the robot's work environment. Determining the target operation enables comprehensive and accurate perception of the robot's work environment, improving the integrity and reliability of robot status and environmental detection. Next, control parameters corresponding to the robot are generated based on the environmental information and the operation command. A sequence of motion parameters corresponding to the robot is then generated based on the control parameters, and the robot is controlled to perform the target operation based on the motion parameter sequence. The control parameters include trajectory planning parameters and force control parameters. Generating control parameters based on environmental information and operation commands allows the robot's control to adapt to both the task objective and the actual work environment, improving the safety and adaptability of the robot in performing the target operation, ensuring the rationality of motion planning, achieving high-precision and high-stability motion execution, and enhancing the robot's operational accuracy and robustness in complex scenarios.

[0045] Example 2 Figure 2 This is a flowchart illustrating a robot control method according to Embodiment 2 of the present invention, further describing the specific implementation of constructing a sequence of motion parameters. Specific implementation details can be found in the description of this embodiment. Technical features that are the same as or similar to those in the foregoing embodiments will not be repeated here. Figure 2 As shown, the method may specifically include: S210. Receive operation instructions, acquire multi-source observation information of the robot based on multi-modal sensors deployed on the robot, input the multi-source observation information into an environmental detection model, and obtain environmental detection results of the robot; wherein, the operation instructions include the target operation to be performed by the robot, and the environmental detection results include environmental information within the robot's operating scenario.

[0046] S220. Based on the environmental information and the operation instructions, generate control parameters corresponding to the robot; wherein, the control parameters include trajectory planning parameters and force control parameters.

[0047] S230. Generate a trajectory parameter sequence based on the trajectory planning parameters; wherein the trajectory parameter sequence includes at least a first pose, a first velocity, and a first acceleration.

[0048] In this embodiment of the invention, the trajectory parameter sequence can be a set of motion parameters arranged in chronological order, generated from trajectory planning parameters. The trajectory parameter sequence includes at least parameters such as a first pose, a first velocity, and a first acceleration. The first pose can be understood as the desired pose output by trajectory planning. The first velocity can be understood as the desired velocity output by trajectory planning. The first acceleration can be understood as the desired acceleration output by trajectory planning.

[0049] Optionally, the outer loop controller can generate a trajectory parameter sequence X(t) based on the trajectory planning parameters and allocate hold time and dose budget to achieve coordinated adaptation between trajectory planning and task resource constraints. Specifically, based on the initial pose, target pose, and maximum velocity / maximum acceleration in the trajectory planning parameters, a continuous and smooth trajectory parameter sequence X(t) with continuous position, velocity, and acceleration can be generated using fifth-order polynomial interpolation or an S-curve velocity profile planning algorithm (S-curve). This suppresses mechanical vibrations caused by abrupt acceleration and deceleration, protects the robot joints and end effector, and improves the stability of the operation process.

[0050] Furthermore, in contact operations or restricted scenarios, the outer loop controller can also simultaneously complete task resource allocation. According to the process requirements defined in the operation instructions, it can embed holding time in specific stages of the robot trajectory, such as the contact segment, to ensure the integrity of the operation. It can also dynamically allocate the dose budget based on the preset dose model. If the preliminary planning results have risks of exceeding the limits in dose, speed, etc., it will automatically adjust the trajectory shape, such as extending the transition segment, reducing the approach speed, or reducing the contact depth, to ensure that the final output trajectory parameter sequence simultaneously meets the requirements of motion performance, safety compliance, and resource constraints.

[0051] S240. Construct a piecewise function based on the force control parameters to generate a sequence of control parameters; wherein the sequence of control parameters includes at least a first contact force and a first torque.

[0052] The control parameter sequence can be a set of force and torque parameters arranged in time sequence based on force control parameters. The control parameter sequence includes at least a first contact force and a first torque. The first contact force can be understood as the desired contact force between the robot and the environment / workpiece. The first torque can be understood as the desired output torque of the robot joint or end effector.

[0053] Optionally, a control parameter sequence U(t) can be generated by the outer loop controller based on the force control parameters. Specifically, based on constraints such as the target contact pressure, maximum allowable torque, and contact holding time in the force control parameters, a piecewise force / torque function is constructed to generate a smooth and process-compliant control parameter sequence U(t). This allows the robot's target operation process to be divided into multiple stages, with different mathematical expressions used in each stage to achieve smooth transition and precise control of force / torque. For example, in the approach stage, the contact force and output torque can be set to zero to avoid premature contact and impact; in the contact stage, a linear ramp function can be used to gradually increase the contact force / torque from zero to the target value to suppress transient overshoot; in the holding stage, a constant target contact force and torque can be maintained for a preset holding time to meet process requirements; and in the withdrawal stage, a reverse ramp function can be used to linearly unload the force / torque to zero to ensure smooth disengagement. This approach can adapt to the force control requirements of different operation stages, improving operation quality and safety.

[0054] S250. Construct the motion parameter sequence based on the trajectory parameter sequence and the control parameter sequence, and control the robot to perform the target operation according to the motion parameter sequence.

[0055] Based on the obtained trajectory parameter sequence and control parameter sequence, a motion parameter sequence can be constructed. This involves aligning the generated trajectory parameter sequence and control parameter sequence on the time axis to form a multidimensional motion parameter sequence. The multidimensional motion parameter sequence includes at least multiple parameters such as the robot's position, velocity, acceleration segment, contact force, and torque at multiple moments, enabling comprehensive control of the robot's motion and interaction states.

[0056] Optionally, the following function can be used as an evaluation criterion to ensure that the generated robot trajectory balances motion smoothness and mechanical compliance, i.e. ,in, Indicates the first The actual state estimate of the robot's end effector within a control cycle is usually output by the environmental detection model; For the first in the trajectory parameter sequence The element is used to represent the first element. The desired motion state output by the inner and outer loop controllers in the first control cycle, i.e., the first... The trajectory parameter sequence corresponding to each control cycle; Indicates the first The actual force / torque output applied by the robot within each control cycle can be calculated by the inner loop controller. The first of the control parameter sequence The element, i.e., the th element The sequence of control parameters corresponding to each control cycle is used to represent the desired mechanical command output by the outer loop controller. The state weight matrix is ​​a symmetric positive definite matrix used to adjust the relative importance of errors in various dimensions such as position, velocity, and acceleration. This represents the control weight matrix, which is a symmetric positive definite matrix used to balance the changes in control input, suppress sudden changes in force / torque, and improve operational smoothness. The number of control cycles planned for the outer loop controller; The time step index indicates discrete time points in the sequence of action parameters; the total cost function is formed by summing the deviation costs across all time steps. It is used to evaluate the quality of the entire motion parameter sequence, ensuring that the motion parameter sequence meets both motion smoothness requirements and mechanical compliance constraints, and provides high-quality parameter support for subsequent motion execution.

[0057] Based on the above scheme, optionally, after constructing the motion parameter sequence based on the trajectory parameter sequence and the control parameter sequence, the method includes: updating the motion parameter sequence based on the robot's actual operating parameters and impedance control model. Through real-time feedback and dynamic correction, the problem of motion deviation caused by environmental changes, mechanical errors, etc., during robot execution is solved, thereby improving the robustness and accuracy of robot control.

[0058] Among them, the actual operating parameters can be understood as the real-time state parameters collected by the robot during the execution of actions, which can truly reflect the robot's actual motion and interaction state. The impedance control model can be understood as a dynamic model based on mass, damping, and stiffness characteristics to achieve compliant interactive control of the robot, which can be used to adjust the output force and pose according to the error between the desired state and the actual state.

[0059] Furthermore, after constructing the initial motion parameter sequence based on the trajectory parameter sequence and the control parameter sequence, the inner loop controller can continuously monitor the deviation between the actual operating parameters and the motion parameter sequence output by the outer loop controller, and update the motion parameter sequence according to the impedance control model, dynamically adjusting the motion parameter sequence of subsequent time steps to ensure that the robot's movements always conform to the desired trajectory and force control requirements.

[0060] Based on the above scheme, optionally, the actual operating parameters of the robot include at least the second pose, second velocity, second acceleration, and second contact force, and the impedance control model is represented by the following formula, including: ; in, This indicates the mass parameters of the robot; This represents the damping parameters of the robot; This represents the rigidity parameter of the robot; This represents the error between the robot's second pose and the first pose in the sequence of motion parameters; This represents the error between the robot's second velocity and the first velocity in the sequence of motion parameters; This represents the error between the robot's second acceleration and the first acceleration in the sequence of motion parameters; This represents the first contact force in the sequence of action parameters; This represents the second contact force of the robot.

[0061] Wherein, the second pose can be the actual position and orientation measured during the actual operation of the robot; the second velocity can be the actual motion velocity measured during the actual operation of the robot; the second acceleration can be the actual motion acceleration measured during the actual operation of the robot; the second contact force can be the actual contact force measured during the actual operation of the robot; the mass parameter can be the parameter characterizing the robot's inertial characteristics in the impedance control model; the damping parameter can be the parameter characterizing the robot's damping and oscillation suppression in the impedance control model; the stiffness parameter can be the parameter characterizing the robot's stiffness and determining its compliance in the impedance control model; the pose error can be the deviation between the robot's second pose and the first pose; the velocity error can be the deviation between the robot's second velocity and the first velocity; and the acceleration error can be the deviation between the robot's second acceleration and the first acceleration.

[0062] Optionally, real-time force-position hybrid control can be achieved through an inner-loop controller to realize stable and compliant contact between the robot and the environment. Specifically, based on the sequence of motion parameters output by the outer-loop controller from the inner-loop controller, combined with the actual operating parameters provided by the world model, the joint torque commands can be dynamically adjusted through an impedance control model to ensure that the robot can accurately track the desired trajectory while maintaining stable contact force during movement. Furthermore, various fine interactive micro-movements can be realized based on the inner-loop controller, including but not limited to micro-swinging for surface exploration or friction detection, micro-insertion for alignment and pressing in precision assembly, and micro-scanning for cleaning or tactile scanning. This is achieved by superimposing high-frequency, small-amplitude perturbation signals to the sequence of motion parameters output by the outer-loop controller in a local coordinate system, combined with adaptive stiffness adjustment. The amplitude and frequency of these perturbations must be constrained by the current contact state and uncertainty score to avoid instability or exceeding limits.

[0063] To ensure operational safety, the inner-loop controller requires dual closed-loop limiting: force limiting ensures that the six-dimensional contact force or normal component at the end effector does not exceed a preset threshold, preventing workpiece damage or sensor overload; torque limiting ensures that the output torque of the joint or end effector does not exceed mechanical and electrical safety boundaries, avoiding motor saturation or excessive structural stress. By embedding a torque generation loop with hard limiting or a control barrier function (CBF), it is ensured that even in the event of sensing anomalies or planning deviations, the physical output remains within a safe envelope, reducing safety risks.

[0064] Based on the above scheme, optionally, the environmental detection result also includes an uncertainty score corresponding to the environmental information in the robot's operating scenario; the method further includes: adjusting the impedance control model based on the uncertainty score, and updating the motion parameter sequence according to the environmental information in the robot's operating scenario and the adjusted impedance control model; wherein, the uncertainty score has a negative correlation function relationship with the rigidity parameter of the robot.

[0065] Among them, the uncertainty score can be understood as a score used to quantify the reliability, stability and unknown degree of environmental information. Its value can be calculated based on the consistency measure of multimodal sensors, feature confidence, occlusion ratio, lighting conditions or the internal state estimation covariance of the environmental detection model, and accurately reflects the credibility of the current environmental information.

[0066] Optionally, the impedance control model of the inner loop can be dynamically adjusted based on the uncertainty score. For example, the stiffness parameter, damping parameter, or mass parameter of the impedance control model can be adjusted to achieve adaptive optimization of the control strategy. The uncertainty score and the stiffness parameter have a negative correlation function. That is, when the uncertainty score increases, the stiffness parameter can be automatically reduced to make the robot exhibit more compliant mechanical characteristics; conversely, in high-confidence environments, the stiffness parameter is increased to enhance trajectory tracking accuracy, ensuring the robot's safety in perception-degraded scenarios such as smoke obstruction, strong glare, or low illumination, and avoiding hard collisions or overloads caused by incorrect state estimation.

[0067] Furthermore, the constructed sequence of motion parameters can be corrected online based on the updated impedance control model and environmental information within the current work scenario. For example, if a geometrical abrupt change in the contact area is detected and the uncertainty score increases, the speed can be reduced, the holding time shortened, and the target contact force reduced in subsequent time steps, thereby generating a more conservative robot operation. The above process can be completed collaboratively through inner-loop feedback and outer-loop replanning to maintain a dynamic balance between operational intent and execution safety.

[0068] In addition, dynamic weight adjustment can be performed, and dynamic weight adjustment coefficients can be defined. When the uncertainty score is high, the control input penalty weight can be increased in the outer-loop controller cost function. This makes the optimization results tend to generate trajectories with gentle changes and low speeds, reducing the risk of actions. At the same time, the stiffness parameter can be reduced in the impedance control model of the inner loop controller, making the robot arm respond more smoothly. The two work together to achieve a dual safety strategy of conservative upper-level planning and compliant lower-level execution, ensuring that the robot can complete tasks or exit in an orderly manner with safety as the top priority even under extreme perception conditions, further improving the robustness and safety of robot control.

[0069] Optionally, based on the above scheme, the method further includes: in response to the uncertainty score being greater than a safety threshold, determining response information corresponding to the target operation based on safety constraint information in a control barrier function pre-set for the robot; the response information includes whether to execute or not execute the target operation.

[0070] The safety threshold refers to a pre-set critical value used to determine whether environmental uncertainty exceeds a safe range. The control barrier function can be understood as a mathematical function used to ensure that the robot control process meets safety constraints; it contains built-in safety constraints to limit dangerous robot actions.

[0071] Optionally, the output uncertainty score can be continuously monitored and compared with a pre-set safety threshold. When the uncertainty score exceeds the safety threshold, it indicates that the current environmental state has significant unknowns or unreliability, which may endanger the robot's execution safety. In response to the uncertainty score being greater than the safety threshold, a safety check is performed based on a pre-configured control barrier function (CBF). The control barrier function encodes multi-dimensional safety constraint information related to the target operation, including but not limited to dose constraints, mechanical constraints such as contact force / torque, kinematic constraints such as velocity, acceleration, and joint limits, and reachability constraints. These constraints are uniformly represented as... Non-negativity constraints, To comprehensively consider safety constraint information, if the safety constraint information is not met at any time, the current operation is determined to be infeasible.

[0072] Furthermore, a safety assessment is performed based on the control barrier function, generating corresponding response information to indicate whether to continue executing the target operation. If all safety constraints can be satisfied, the response information is to execute the target operation and allow the outer loop to continue outputting the action parameter sequence. If any constraint is not satisfied, the response information is not to execute the target operation and triggers a safe stop, allowing the robot to smoothly decelerate to the preset safe posture.

[0073] In addition, infeasibility evidence can be retained simultaneously, and key information that leads to the violation of safety constraints can be automatically recorded, such as the specific constraint item violated, the time of occurrence, the corresponding robot state, uncertainty score and environmental information. This information is written to the audit log in read-only form and can be used for subsequent accident attribution, control strategy improvement or small sample retraining, so that the robot control has continuous learning and compliance traceability capabilities and continuously optimizes control performance.

[0074] The technical solution of this invention firstly generates a trajectory parameter sequence based on the trajectory planning parameters; wherein the trajectory parameter sequence includes at least a first pose, a first velocity, and a first acceleration; the trajectory planning parameters are refined into a time-sequential trajectory parameter sequence to ensure the continuity and executability of motion commands in the time domain, achieving smooth trajectory tracking and improving motion accuracy and mechanical safety; nextly, a control parameter sequence is generated by constructing a piecewise function based on the force control parameters; wherein the control parameter sequence includes at least a first contact force and a first torque; modeling the force control parameters using a piecewise function can flexibly adapt to the differentiated requirements of interactive force / torque at different operation stages, achieving precise physical interaction; finally, by constructing the action parameter sequence based on the trajectory parameter sequence and the control parameter sequence, motion control and interactive control are unified and integrated into the action parameter sequence, realizing a structured expression of position-force hybrid control, simultaneously satisfying geometric path constraints and physical contact constraints, significantly enhancing the robot's operational adaptability and task success rate.

[0075] Example 3 Figure 3 This is a schematic diagram of a robot control device according to Embodiment 3 of the present invention. This device is used to execute the robot control method provided in any of the above embodiments. This device and the robot control methods of the above embodiments belong to the same inventive concept. Details not described in detail in the embodiments of the robot control device can be found in the embodiments of the above robot control methods. Figure 3 As shown, the device includes: a data acquisition module 310 and a control module 320.

[0076] The acquisition module 310 is used to receive operation instructions, acquire multi-source observation information of the robot based on multi-modal sensors deployed on the robot, input the multi-source observation information into an environmental detection model, and obtain environmental detection results of the robot. The operation instructions include the target operation to be performed by the robot, and the environmental detection results include environmental information within the robot's operating scenario. The control module 320 is used to generate control parameters corresponding to the robot based on the environmental information and the operation instructions, generate a sequence of motion parameters corresponding to the robot based on the control parameters, and control the robot to perform the target operation based on the sequence of motion parameters. The control parameters include trajectory planning parameters and force control parameters.

[0077] The technical solution of this invention firstly involves receiving an operation command through a data acquisition module 310, acquiring multi-source observation information of the robot based on multi-modal sensors deployed on the robot, and inputting the multi-source observation information into an environmental detection model to obtain the robot's environmental detection results. The operation command includes the target operation to be performed by the robot, and the environmental detection results include environmental information within the robot's work environment. Determining the target operation enables comprehensive and accurate perception of the robot's work environment, improving the integrity and reliability of robot status and environmental detection. Next, a control module 320 generates control parameters corresponding to the robot based on the environmental information and the operation command, generates a sequence of motion parameters corresponding to the robot based on the control parameters, and controls the robot to perform the target operation based on the sequence of motion parameters. The control parameters include trajectory planning parameters and force control parameters. Generating control parameters based on environmental information and operation commands allows the robot's control to adapt to both the task objective and the actual work environment, improving the safety and adaptability of the robot in performing the target operation, ensuring the rationality of motion planning, achieving high-precision and high-stability motion execution, and improving the robot's operational accuracy and robustness in complex scenarios.

[0078] Based on the above scheme, optionally, the control module 320 includes a trajectory parameter sequence generation submodule, a control quantity parameter sequence generation submodule, and an action parameter sequence construction submodule. The trajectory parameter sequence generation submodule is used to generate a trajectory parameter sequence based on the trajectory planning parameters; wherein the trajectory parameter sequence includes at least a first pose, a first velocity, and a first acceleration. The control quantity parameter sequence generation submodule is used to construct a piecewise function based on the force control parameters to generate a control quantity parameter sequence; wherein the control quantity parameter sequence includes at least a first contact force and a first torque. The action parameter sequence construction submodule is used to construct the action parameter sequence based on the trajectory parameter sequence and the control quantity parameter sequence.

[0079] Optionally, based on the above scheme, the control module 320 further includes an action parameter sequence update submodule. The action parameter sequence update submodule is used to update the action parameter sequence based on the robot's actual operating parameters and impedance control model after the action parameter sequence has been constructed based on the trajectory parameter sequence and the control quantity parameter sequence.

[0080] Based on the above scheme, optionally, the actual operating parameters of the robot include at least the second pose, second velocity, second acceleration, and second contact force, and the impedance control model is represented by the following formula, including: ; in, This indicates the mass parameters of the robot; This represents the damping parameters of the robot; This represents the rigidity parameter of the robot; This represents the error between the robot's second pose and the first pose in the sequence of motion parameters; This represents the error between the robot's second velocity and the first velocity in the sequence of motion parameters; This represents the error between the robot's second acceleration and the first acceleration in the sequence of motion parameters; This represents the first contact force in the sequence of action parameters; This represents the second contact force of the robot.

[0081] Optionally, based on the above scheme, the environmental detection result may also include an uncertainty score corresponding to the environmental information within the robot's operating scenario; the device may also include an impedance control model adjustment module. The impedance control model adjustment module is used to adjust the impedance control model based on the uncertainty score, and to update the motion parameter sequence according to the environmental information within the robot's operating scenario and the adjusted impedance control model; wherein the uncertainty score has a negative correlation with the robot's rigidity parameters.

[0082] Optionally, based on the above scheme, the device further includes a safety constraint module. The safety constraint module is used to determine response information corresponding to the target operation based on safety constraint information in a pre-set control barrier function for the robot in response to the uncertainty score exceeding a safety threshold; the response information includes whether to execute or not execute the target operation.

[0083] Based on the above scheme, optionally, the control module 320 includes an operation instruction parsing submodule and a control parameter generation submodule. The operation instruction parsing submodule is used to perform semantic parsing on the operation instructions according to a semantic analysis model to obtain multiple operation parameters; wherein the semantic analysis model is trained on a machine learning model based on multiple sample operation instructions and their labeled operation parameters. The control parameter generation submodule is used to determine multiple target operation parameters corresponding to the robot from the multiple operation parameters based on a device knowledge base, and generate control parameters corresponding to the robot based on the environmental information and the multiple target operation parameters.

[0084] Optionally, based on the above solution, the device further includes an audit module. The audit module is used to obtain the actual operating parameters of the robot when performing the target operation, after the robot is controlled to perform the target operation according to the action parameter sequence, and to generate an audit log based on the action parameter sequence and the actual operating parameters.

[0085] Based on the above scheme, optionally, the trajectory planning parameters include at least one of the initial pose, target pose, and velocity information, wherein the velocity information includes maximum velocity and / or maximum acceleration; and / or, the force control parameters include at least one of the contact holding time and interactive force constraint parameters, wherein the interactive force constraint parameters include target contact pressure and / or maximum allowable torque.

[0086] The robot control device provided in the embodiments of the present invention can execute the robot control method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0087] Example 4 Figure 4 A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0088] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0089] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0090] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as robot control methods.

[0091] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication unit 19, or installed from storage unit 18, or installed from ROM 12. When the computer program is executed by processor 11, it performs the functions defined in the methods of the embodiments of the present invention.

[0092] In some embodiments, the robot control method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the robot control method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to execute the robot control method by any other suitable means (e.g., by means of firmware).

[0093] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0094] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0095] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0096] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0097] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0098] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0099] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0100] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A robot control method, characterized in that, include: The robot receives an operation command, acquires multi-source observation information of the robot based on multi-modal sensors deployed on the robot, inputs the multi-source observation information into an environmental detection model, and obtains environmental detection results of the robot; wherein, the operation command includes the target operation to be performed by the robot, and the environmental detection results include environmental information within the robot's operating scene; Based on the environmental information and the operation instructions, control parameters corresponding to the robot are generated. Based on the control parameters, a sequence of motion parameters corresponding to the robot is generated. Based on the sequence of motion parameters, the robot is controlled to perform the target operation. The control parameters include trajectory planning parameters and force control parameters.

2. The robot control method according to claim 1, characterized in that, The step of generating the sequence of motion parameters corresponding to the robot based on the control parameters includes: A trajectory parameter sequence is generated based on the trajectory planning parameters; wherein the trajectory parameter sequence includes at least a first pose, a first velocity, and a first acceleration; A piecewise function is constructed based on the force control parameters to generate a sequence of control parameters; wherein the sequence of control parameters includes at least a first contact force and a first torque; The motion parameter sequence is constructed based on the trajectory parameter sequence and the control quantity parameter sequence.

3. The robot control method according to claim 2, characterized in that, After constructing the motion parameter sequence based on the trajectory parameter sequence and the control parameter sequence, the following is included: The sequence of motion parameters is updated based on the robot's actual operating parameters and impedance control model.

4. The robot control method according to claim 3, characterized in that, The actual operating parameters of the robot include at least the second pose, second velocity, second acceleration, and second contact force, and the impedance control model is represented by the following formula: ； in, This indicates the mass parameters of the robot; This represents the damping parameters of the robot; This represents the rigidity parameter of the robot; This represents the error between the robot's second pose and the first pose in the sequence of motion parameters; This represents the error between the robot's second velocity and the first velocity in the sequence of motion parameters; This represents the error between the robot's second acceleration and the first acceleration in the sequence of motion parameters; This represents the first contact force in the sequence of action parameters; This represents the second contact force of the robot.

5. The robot control method according to claim 3, characterized in that, The environmental detection results also include uncertainty scores corresponding to environmental information within the robot's operating scenario; The robot control method further includes: The impedance control model is adjusted based on the uncertainty score, and the action parameter sequence is updated according to the environmental information in the robot's operating scenario and the adjusted impedance control model; wherein, the uncertainty score has a negative correlation with the robot's rigidity parameter.

6. The robot control method according to claim 5, characterized in that, The robot control method further includes: In response to the uncertainty score being greater than a safety threshold, response information corresponding to the target operation is determined based on the safety constraint information in the control barrier function pre-set for the robot; the response information includes whether to execute or not execute the target operation.

7. The robot control method according to claim 1, characterized in that, The process of generating control parameters for the robot based on the environmental information and the operation instructions includes: The operation instructions are semantically parsed using a semantic analysis model to obtain multiple operation parameters; wherein, the semantic analysis model is trained on a machine learning model based on multiple sample operation instructions and their labeled operation parameters. Based on the device knowledge base, multiple target operation parameters corresponding to the robot are determined from multiple operation parameters, and control parameters corresponding to the robot are generated according to the environmental information and the multiple target operation parameters.

8. The robot control method according to claim 1, characterized in that, After controlling the robot to perform the target operation according to the sequence of motion parameters, the following steps are included: Obtain the actual operating parameters of the robot when it performs the target operation, and generate an audit log based on the action parameter sequence and the actual operating parameters.

9. The robot control method according to claim 1, characterized in that, The trajectory planning parameters include at least one of the following: initial pose, target pose, and velocity information; the velocity information includes maximum velocity and / or maximum acceleration; and / or... The force control parameters include at least one of the contact holding time and the interactive force constraint parameters, wherein the interactive force constraint parameters include the target contact pressure and / or the maximum permissible torque.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the robot control method as described in any one of claims 1-9.

Images