Mechanical arm control method, control module and device based on force feedback and mechanical arm

Abstract

The invention provides a mechanical arm control method, control module and device based on force feedback and a mechanical arm. Spatial position information of the mechanical arm and a mechanical arm operation area is collected according to a preset sampling frequency; for any sampling moment, state information of the mechanical arm in the mechanical arm operation area at the sampling moment is determined based on the spatial position information and a Kalman filtering algorithm; based on the state information, the input target position, an A star algorithm and a reinforcement learning algorithm, a target motion path of the mechanical arm at the sampling moment is determined; generating a mechanical arm control signal based on the target motion path and sending the mechanical arm control signal to the mechanical arm; and based on the state information of the sampling moment, the target motion path of the previous sampling moment and a force feedback control algorithm, a force feedback signal of the mechanical arm at the sampling moment is determined and sent to a force feedback module, so that the force feedback module outputs corresponding feedback force. In this way, through real-time motion path planning and force feedback control, the control accuracy of the mechanical arm can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of robotic arm control, and in particular to a robotic arm control method, control module, device and robotic arm based on force feedback. Background Art

[0002] With the rapid development of robotics and remote operation technology, remote robotic arms are gradually being used in delicate operation tasks such as minimally invasive surgery and microelectronics assembly. However, existing remote robotic arm control often relies on simple PID control algorithms and preset planning paths, and the accuracy of robotic arm control is low when facing complex and changing operating environments. Summary of the invention

[0003] In view of this, the embodiments of the present application at least provide a robotic arm control method, a control module, a device and a robotic arm based on force feedback. The present application can improve the accuracy of robotic arm control through real-time motion path planning and force feedback control.

[0004] This application mainly includes the following aspects:

[0005] In a first aspect, an embodiment of the present application provides a mechanical arm control method based on force feedback, which is applied to a control module of a mechanical arm control device based on force feedback, wherein the mechanical arm control device further includes an environment perception module and a force feedback module; the control module is respectively connected to the environment perception module, the force feedback module and the mechanical arm in communication; the mechanical arm control method includes:

[0006] Collecting the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected by the environment perception module according to a preset sampling frequency;

[0007] For any sampling moment, based on the spatial position information of the robotic arm and the operating area of ​​the robotic arm and the Kalman filter algorithm, determine the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment;

[0008] For any sampling moment, based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm and the reinforcement learning algorithm, determine the target motion path of the robotic arm at the sampling moment;

[0009] For any sampling moment, a robot arm control signal is generated based on the target motion path at the sampling moment, and the robot arm control signal is sent to the robot arm so that the robot arm moves according to the target motion path;

[0010] For any sampling moment, based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment and the force feedback control algorithm, the force feedback signal of the robotic arm at the sampling moment is determined, and the force feedback signal is sent to the force feedback module so that the force feedback module outputs the feedback force corresponding to the sampling moment.

[0011] In a second aspect, an embodiment of the present application further provides a control module, which is applied to a mechanical arm control device based on force feedback, wherein the mechanical arm control device further includes an environment perception module and a force feedback module; the control module is respectively connected to the environment perception module, the force feedback module and the mechanical arm for communication; the control module includes:

[0012] A sampling unit, configured to collect the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected by the environment perception module according to a preset sampling frequency;

[0013] A first determining unit is used to determine, for any sampling moment, state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment based on spatial position information of the robotic arm and the operating area of ​​the robotic arm and a Kalman filter algorithm;

[0014] A second determination unit is used to determine, for any sampling moment, a target motion path of the robotic arm at the sampling moment based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm, and the reinforcement learning algorithm;

[0015] A first execution unit is used to generate a robot arm control signal based on a target motion path at any sampling moment, and send the robot arm control signal to the robot arm so that the robot arm moves according to the target motion path;

[0016] The second execution unit is used to determine, for any sampling moment, the force feedback signal of the robotic arm at the sampling moment based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment of the sampling moment, and the force feedback control algorithm, and send the force feedback signal to the force feedback module so that the force feedback module outputs the feedback force corresponding to the sampling moment.

[0017] In a third aspect, an embodiment of the present application further provides a mechanical arm control device based on force feedback, the mechanical arm control device comprising:

[0018] An environmental perception module, a force feedback module and the control module in the second aspect above.

[0019] In a fourth aspect, an embodiment of the present application further provides a robotic arm, comprising the robotic arm control device based on force feedback in the third aspect mentioned above.

[0020] The embodiment of the present application provides a mechanical arm control method, a control module, a device and a mechanical arm based on force feedback, wherein the mechanical arm control method is applied to the control module of a mechanical arm control device based on force feedback, and the mechanical arm control device also includes an environment perception module and a force feedback module; the control module is respectively communicated with the environment perception module, the force feedback module and the mechanical arm; the mechanical arm control method comprises: collecting spatial position information of the mechanical arm and the operating area of ​​the mechanical arm collected by the environment perception module according to a preset sampling frequency; for any sampling moment, based on the spatial position information of the mechanical arm and the operating area of ​​the mechanical arm and the Kalman filtering algorithm, determining the state information of the mechanical arm in the operating area of ​​the mechanical arm at the sampling moment; for any At a sampling moment, based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm and the reinforcement learning algorithm, the target motion path of the manipulator at the sampling moment is determined; for any sampling moment, a manipulator control signal is generated based on the target motion path at the sampling moment, and the manipulator control signal is sent to the manipulator so that the manipulator moves according to the target motion path; for any sampling moment, based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment of the sampling moment and the force feedback control algorithm, the force feedback signal of the manipulator at the sampling moment is determined, and the force feedback signal is sent to the force feedback module so that the force feedback module outputs the feedback force corresponding to the sampling moment. In this way, the accuracy of manipulator control can be improved through real-time motion path planning and force feedback control.

[0021] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0023] Figure 1 One of the functional module diagrams of a mechanical arm control device based on force feedback provided in an embodiment of the present application is shown;

[0024] Figure 2 The second functional module diagram of a mechanical arm control device based on force feedback provided in an embodiment of the present application is shown;

[0025] Figure 3 A flow chart of a mechanical arm control method based on force feedback provided in an embodiment of the present application is shown;

[0026] Figure 4 One of the functional module diagrams of a control module provided in an embodiment of the present application is shown;

[0027] Figure 5 A second functional module diagram of a control module provided in an embodiment of the present application is shown;

[0028] Figure 6 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown.

[0029] Description of main component symbols:

[0030] In the figure: 100-robotic arm control device; 110-environmental perception module; 120-force feedback module; 130-control module; 200-robotic arm; 111-visual sensor; 112-laser scanning sensor; 113-force feedback sensor; 131-sampling unit; 132-first determination unit; 133-second determination unit; 134-first execution unit; 135-second execution unit; 136-third determination unit; 137-third execution unit; 138-fourth determination unit; 139-fourth execution unit; 600-electronic device; 610-processor; 620-memory; 630-bus. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0032] The following will be combined Figure 1 as well as Figure 2 The implementation of the embodiment of the present application is described in detail; the embodiment of the present application provides a force feedback-based robotic arm control method, which is applied to a control module of a force feedback-based robotic arm control device.

[0033] See also Figure 1 , Figure 1 This is one of the functional module diagrams of a mechanical arm control device based on force feedback provided in an embodiment of the present application. The mechanical arm control method based on force feedback provided in an embodiment of the present application can be applied to Figure 1 The control module in the mechanical arm control device based on force feedback is shown in FIG. Figure 1 As shown, the robot arm control device 100 may include: an environment perception module 110, a force feedback module 120 and a control module 130; the control module 130 is respectively communicatively connected with the environment perception module 110, the force feedback module 120 and the robot arm 200.

[0034] The environmental perception module 110 is used to collect environmental information of the robot arm 200 and the operating area of ​​the robot arm 200 in real time to provide spatial position information, ensure the control module 130's comprehensive perception of the environment, and further ensure the path planning and operation safety of the robot arm 200.

[0035] The force feedback module 120 is used to sense the interaction force between the robot arm 200 and the target object in real time, and transmit the feedback information to the operator to enhance his perception of remote operation.

[0036] The robot arm 200 is used to perform precise operation tasks and realize the predetermined workflow and motion trajectory through the instructions of the control module 130. In the embodiment of the present application, the robot arm 200 is a multi-degree-of-freedom robot arm, which adopts a modular multi-joint design and has the ability of rotation and linear motion of 6 or more degrees of freedom, ensuring that the robot arm 200 can operate freely in three-dimensional space. The joints are driven by servo motors and have high-precision angle feedback control to ensure precise control of rotation and linear motion. Each joint is connected through a standard interface for easy maintenance and replacement.

[0037] In the preferred embodiment of the present application, please refer to Figure 2 , Figure 2 This is a second functional module diagram of a mechanical arm control device based on force feedback provided in an embodiment of the present application. Figure 2 As shown, the environment perception module 110 includes a visual sensor 111, a laser scanning sensor 112 and a force feedback sensor 113, and the visual sensor 111, the laser scanning sensor 112 and the force feedback sensor 113 are respectively connected to the control module 130 for communication.

[0038] The visual sensor 111 is used to collect real-time image information.

[0039] The laser scanning sensor 112 is used to accurately measure the distance and contour of the target object.

[0040] The force feedback sensor 113 is used to monitor force changes during operation.

[0041] A mechanical arm control method based on force feedback provided in an embodiment of the present application is described in detail below. The mechanical arm control method can be applied to the control module of the mechanical arm control device based on force feedback.

[0042] See also Figure 3 , Figure 3 The following is a flow chart of a method for controlling a mechanical arm based on force feedback provided in an embodiment of the present application. Figure 3 As shown, the mechanical arm control method based on force feedback provided in the embodiment of the present application includes the following steps:

[0043] S301, collecting the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected by the environment perception module according to a preset sampling frequency.

[0044] Here, the remote manipulator first performs posture initialization before starting work, records the operating area of ​​the manipulator, sets the working boundary of the operating area, and uses the preset sampling frequency as the time node to update the motion coordinates of the manipulator in real time, and sets the overall displacement adjustment parameters to ensure the stability of the manipulator operation. In the embodiment of the present application, the preset sampling frequency is specifically set according to actual needs. The embodiment of the present application uses 0.2 seconds as the time node of the preset sampling frequency to collect the spatial position information and environmental data of the manipulator and the operating area of ​​the manipulator collected by the environmental perception module.

[0045] S302, for any sampling moment, based on the spatial position information of the robotic arm and the operating area of ​​the robotic arm and a Kalman filter algorithm, determine the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment.

[0046] Here, for any sampling moment, the Kalman filter algorithm is used to fuse the data of the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected above to obtain the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment, thereby forming an accurate geometric and spatial description of the operating environment of the robotic arm.

[0047] S303, for any sampling moment, based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm and the reinforcement learning algorithm, determine the target motion path of the robot arm at the sampling moment.

[0048] Here, for any sampling moment, based on the precise geometric and spatial description of the robot arm operating area environment formed above, that is, the state information corresponding to the sampling moment, and the target position of controlling the movement of the robot arm in response to user input, the adaptive path planning algorithm that combines the A* algorithm and the reinforcement learning (Q-Learning) algorithm is used to automatically adjust the motion path of the robot arm to determine the target motion path of the robot arm at the sampling moment, so as to cope with the path deviation caused by changes in the environment of the robot arm operating area.

[0049] S304: For any sampling moment, generate a robot arm control signal based on the target motion path at the sampling moment, and send the robot arm control signal to the robot arm, so that the robot arm moves according to the target motion path.

[0050] Here, for any sampling moment, the control module generates a robot arm control signal according to the target motion path at the sampling moment, including a position signal, a velocity signal, and a force signal, and sends the robot arm control signal to each joint and end effector of the robot arm to achieve precise motion control and drive the robot arm to move according to the target motion path.

[0051] S305, for any sampling moment, based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment of the sampling moment, and the force feedback control algorithm, determine the force feedback signal of the robotic arm at the sampling moment, and send the force feedback signal to the force feedback module, so that the force feedback module outputs the feedback force corresponding to the sampling moment.

[0052] Here, force feedback control is to enable the operator to feel the interactive force of the robotic arm when remotely operating the robotic arm. For any sampling moment, the force feedback data of the robotic arm at the sampling moment, i.e., the force feedback signal, is determined by the force feedback control algorithm according to the state information corresponding to the sampling moment and the target motion path determined at the previous sampling moment of the sampling moment, and the force feedback signal is sent to the force feedback module, so that the force feedback module converts the interactive force between the instrument installed on the robotic arm and the environment of the robotic arm operation area into feedback force through the nonlinear force feedback function, so that the operator can feel the real sense of contact with the instrument on the robotic arm in the environment of remotely operating the robotic arm, thereby improving the accuracy and safety of the operation.

[0053] In a preferred solution of the embodiment of the present application, collecting the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected by the environment perception module according to a preset sampling frequency includes:

[0054] Step a1: collecting image information of the robotic arm and the operating area of ​​the robotic arm collected by the visual sensor according to a preset sampling frequency.

[0055] Here, the visual sensor is a camera installed near the operating area of ​​the robotic arm, which is used to collect image information of the robotic arm and the operating area of ​​the robotic arm according to a preset sampling frequency.

[0056] Step a2: collecting the contour of the target object in the operating area of ​​the robotic arm and the distance from the robotic arm to the target object collected by the laser scanning sensor according to a preset sampling frequency.

[0057] Here, the laser scanning sensor is a laser scanner mounted on the robot arm, which is used to scan the contour of the target object in the operation area of ​​the robot arm and accurately measure the distance from the robot arm to the target object.

[0058] Step a3: collecting the force signal of the movement of the robotic arm collected by the force feedback sensor according to a preset sampling frequency.

[0059] Here, force feedback sensors are installed on each joint and end effector of the robotic arm. The contact force between the robotic arm and the outside world is collected through strain gauges or piezoelectric sensors, the force changes of the robotic arm during operation are monitored, and the force signal of the robotic arm movement is obtained.

[0060] Step a4, determining the image information of the robotic arm and the operating area of ​​the robotic arm, the outline of the target object in the operating area of ​​the robotic arm, the distance from the robotic arm to the target object, and the force signal of the movement of the robotic arm as the spatial position information of the robotic arm and the operating area of ​​the robotic arm.

[0061] Here, the environmental perception module integrates visual sensors, laser scanning sensors and force feedback sensors. By collecting image information of the robotic arm and the operating area of ​​the robotic arm, the outline of the target object in the operating area of ​​the robotic arm, the distance from the robotic arm to the target object and the force signal of the robotic arm's movement, the data of each sensor can be integrated together through data fusion technology to obtain the spatial position information of the robotic arm and the operating area of ​​the robotic arm, so as to subsequently build a three-dimensional model of the robotic arm's operating area or working environment in real time, thereby realizing accurate perception of the operating environment.

[0062] Further, for any sampling moment, the determining of the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment based on the spatial position information of the robotic arm and the operating area of ​​the robotic arm and the Kalman filter algorithm includes:

[0063] Step b1, determining a predicted value of the state information at the sampling moment according to the state information of the robot arm in the operating area of ​​the robot arm determined at the previous sampling moment of the sampling moment.

[0064] Here, according to the Kalman filter algorithm, the state information of the robot arm in the operating area determined at the previous sampling moment is used to predict the state and prediction uncertainty at the current sampling moment using the state transition model. The specific formula is as follows:

[0065] State prediction formula:

[0066] Covariance prediction formula: P k|k-1 =AP k-1|k-1 A T +Q;

[0067] In the formula, is the state prediction at this sampling moment relative to the previous sampling moment; P k|k-1 is the predicted covariance; A is the state transfer matrix; B is the control input matrix; u k is the control input at time step k; Q is the process noise covariance matrix.

[0068] Step b2, based on the predicted value of the state information at the sampling moment and the spatial position information of the robotic arm and the operating area of ​​the robotic arm at the sampling moment, update the predicted value to obtain the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment; the state information includes the position information and speed information of the robotic arm in the operating area of ​​the robotic arm.

[0069] Here, in the update phase of the Kalman filter, the predicted state at the current sampling moment and the new sensor sampling data are combined to update the state estimate, taking the actual sensor data into reference to optimize the state prediction and adjust the uncertainty, which specifically includes the following steps:

[0070] Calculate the Kalman gain: K k =P k|k-1 H T (HP k|k-1 H T +R) -1 ; Among them, K k is the Kalman gain, which determines the weight of the measurement data in the state update; P k|k-1 is the predicted covariance matrix; H is the measurement matrix, which maps the state space to the measurement space; R is the measurement noise covariance matrix.

[0071] Update the state estimate: in, is the updated state estimate at the current sampling time, z k is the measured value at the current sampling moment.

[0072] Update the covariance matrix: P k|k =(IK k H)Pk|k-1 ; Among them, P k|k is the updated covariance matrix, reflecting the uncertainty of the updated state estimate; I is the identity matrix.

[0073] As new sensor data is continuously acquired, prediction and update steps are continuously applied to maintain an accurate estimate of the state information. Through these steps, the Kalman filter is able to provide precise position and velocity information of the robot arm in the operating area, which is essential for subsequent path planning and force feedback control.

[0074] Further, for any sampling moment, determining the target motion path of the robot arm at the sampling moment based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm and the reinforcement learning algorithm includes:

[0075] Step c1, determining the current position of the robot arm according to the state information corresponding to the sampling moment.

[0076] Here, the current position of the robot arm at each sampling moment is determined through the state information obtained at each sampling moment.

[0077] Step c2, determining the shortest path from the current position to the target position according to the current position, the input target position and the A-star algorithm.

[0078] Here, the input of the A* algorithm includes a starting point, a target point and map / environmental information. In the embodiment of the present application, for any sampling moment, the current position of the robot arm at the sampling moment is taken as the starting point, the target position pre-input by the user is taken as the target point, and the layout and obstacle information of the robot arm operating area are taken as map / environmental information, which are jointly input into the A* algorithm model to output the shortest path from the starting point to the target point and the shortest path from the current position to the target position.

[0079] Specifically, the steps of the A* algorithm include:

[0080] First, initialize the open list and the closed list. The open list is used to store the nodes to be checked, and the closed list is used to store the checked nodes.

[0081] Next, add the starting point to the open list.

[0082] Then, loop through the following steps until the target point is added to the open list:

[0083] In the first step, the node with the lowest f value (f=g+h) is selected from the open list as the current node; where g is the cost from the starting point to the current node, and h is the heuristic estimated cost from the current node to the target point.

[0084] The second step is to move the current node to the closed list.

[0085] The third step is to check the neighbor nodes of the current node and update their g and h values.

[0086] Step 4: If the neighbor node is not in the open list, add it to the open list.

[0087] Finally, once the target point is added to the open list, the algorithm is complete and traces back from the target point to the starting point to form a complete path.

[0088] Step c3, optimizing the shortest path according to the reinforcement learning algorithm and the spatial position information of the operating area of ​​the robotic arm to obtain the target motion path of the robotic arm at the sampling moment.

[0089] Here, for any sampling moment, after the shortest path is calculated by the A* algorithm, the shortest path is optimized by the reinforcement learning (Q-Learning) algorithm to obtain the target motion path of the robot at the sampling moment.

[0090] Specifically, the input of the reinforcement learning algorithm includes the current state, action space and reward function; the current state is the current position of the robot and the state of the environment; the action space is all the actions that the robot can perform; the reward function is the reward and penalty given according to the actions and results of the robot. The output of the reinforcement learning algorithm is the optimal path strategy, Q is based on the learned value, selects the optimal action sequence, and forms the target motion path of the robot from the starting point to the target point at the sampling time.

[0091] Specifically, the steps of the reinforcement learning algorithm include:

[0092] First, the Q-table is initialized by setting the Q-values ​​of all state-action pairs to zero.

[0093] Next, at each state, an action is chosen according to an epsilon-greedy strategy that balances exploration (trying new actions) and exploitation (selecting the best known action).

[0094] Then, perform the chosen action and observe the outcome and reward received.

[0095] Finally, the Q value is updated using the following formula:

[0096] Continue to repeat the above process until the Q value converges, that is, the strategy is stable.

[0097] Q(s,a)=Q(s,a)+α[r+γmax a′ Q(s′, a′)-Q(s, a)]; where α is the learning rate, γ is the discount factor, r is the reward, s′ is the new state, and a′ is the new action.

[0098] Further, for any sampling moment, the force feedback signal of the manipulator at the sampling moment is determined based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment, and the force feedback control algorithm, including:

[0099] Step d1, determining the actual position of the robot arm at the sampling moment according to the state information corresponding to the sampling moment.

[0100] Here, for any sampling moment, the actual position of the robot arm at the sampling moment is determined according to the state information corresponding to the sampling moment.

[0101] Step d2, determining the predicted position of the robot arm at the sampling moment according to the target motion path determined at the previous sampling moment of the sampling moment.

[0102] Here, based on the target motion path determined at the previous sampling moment of the sampling moment, the predicted position of the robot arm at the sampling moment is calculated when the robot arm moves along the motion path to the sampling moment.

[0103] Step d3, determining the force feedback signal of the robotic arm at the sampling moment according to the deviation between the actual position and the predicted position of the robotic arm at the sampling moment and the force feedback control algorithm.

[0104] Here, the force feedback signal of the manipulator at the sampling moment is calculated by a proportional-differential controller. Specifically, the input of the proportional-differential controller is the force signal collected by the force feedback sensor, and the output is the force feedback signal of the manipulator at the sampling moment. The specific steps are:

[0105] The first step is to obtain the error between the current position of the robot and the expected position;

[0106] The second step is to calculate the rate of change of the error;

[0107] In the third step, the output force is calculated through the PD controller, and a force feedback signal is generated, which is applied to the operator through the force feedback module.

[0108] The specific formula is:

[0109] Among them, F feedback is the feedback force, e(t) is the position error (i.e., the difference between the desired position and the actual position), Kp is the proportional gain, K d is the differential gain.

[0110] This force feedback control algorithm adjusts the feedback force through two parts: proportional and differential. The proportional part adjusts the force according to the size of the position error, while the differential part predicts the error trend based on the error change rate and makes corresponding adjustments to improve the response speed and stability of the system. In this way, the operator can feel the force interaction between the robot arm and the object in real time, thereby controlling the movement of the robot arm more accurately.

[0111] Further, for any sampling moment, after generating a robot arm control signal based on the target motion path at the sampling moment, the robot arm control method further includes:

[0112] Step e1, inputting the robot arm control signal into a trained support vector machine model or a neural network model, and outputting an optimized robot arm control signal.

[0113] Here, in order to dynamically adjust the control parameters of the robotic arm during the robotic arm control process, a machine learning model can be used to optimize the control parameters of the robotic arm. Specifically, the machine learning model includes a support vector machine (SVM) model and a neural network model.

[0114] In this step, the robot arm control signal is input into the trained support vector machine model or neural network model, and the optimized robot arm control signal is output.

[0115] Specifically, the support vector machine model is trained according to the following steps:

[0116] The first step is to collect data during the robot's operation and mark it as success or failure. This data may include the robot's position, speed, acceleration, force feedback signal, and other key parameters.

[0117] In the second step, the data is input into the SVM model for training so that the model can identify the key factors that affect the success of the robot arm's work.

[0118] The third step is to use the trained model to make real-time predictions during the operation of the robotic arm and dynamically adjust the control parameters of the robotic arm.

[0119] Specifically, the neural network model is trained according to the following steps:

[0120] The first step is to build a multi-layer neural network that can process multi-dimensional input data and output suggestions for parameter adjustment. The network inputs various data during the working process of the robotic arm, including but not limited to the robotic arm's operation data, environmental perception data, force feedback data, etc.

[0121] In the second step, the multi-layer neural network is trained using historical robotic arm work data so that it can learn how to predict the optimal parameter combination based on the input feature data.

[0122] The third step is to input the multi-dimensional feature data collected in real time into the trained neural network model during the operation of the robotic arm, and dynamically adjust the robotic arm control parameters according to the output of the neural network to achieve the optimal control effect.

[0123] Through these two dynamic parameter optimization algorithms, this scheme can realize real-time optimization of control parameters in the robot arm control process, improve the automation level and adaptability of the robot arm control, and thus achieve higher robot arm working efficiency and safety in a complex and dynamic robot arm working environment.

[0124] Step e2, sending the optimized robotic arm control signal to the robotic arm, so that the robotic arm moves according to the target motion path.

[0125] Here, the optimized robot arm control signal is sent to the robot arm, so that the robot arm can move better according to the target motion path, thereby improving the accuracy of the robot arm control.

[0126] Further, after collecting the force signal of the movement of the mechanical arm collected by the force feedback sensor according to a preset sampling frequency, the mechanical arm control method further includes:

[0127] Step f1: if the value of the feedback force corresponding to the force signal is greater than a preset feedback force threshold, it is determined that an abnormal collision occurs to the robotic arm.

[0128] Here, the robotic arm control method of the embodiment of the present application also includes an emergency collision detection and stop mechanism. When it is detected that the robotic arm has an abnormal collision with an external object or exerts excessive pressure, the control module will automatically trigger an emergency stop to protect the equipment and the operating object. Specifically, the force feedback sensor monitors the force condition of the robotic arm in real time. If the value of the feedback force corresponding to the force signal is greater than the preset feedback force threshold, it is determined that the robotic arm has an abnormal collision. The preset feedback force threshold is set according to the actual situation and is not limited here.

[0129] Step f2, generating a robot arm locking signal, and sending the robot arm locking signal to the robot arm, so that the robot arm is controlled to stop moving and lock the position according to the robot arm locking signal.

[0130] Here, when an abnormal collision of the robotic arm is detected, the control module generates a robotic arm locking signal and sends the robotic arm locking signal to the robotic arm, so that the robotic arm controls the robotic arm to stop moving and lock the position according to the robotic arm locking signal. Specifically, the electromagnetic brake of the robotic arm is controlled to immediately lock the position of the robotic arm to ensure the safety of the equipment and the operation object.

[0131] The embodiment of the present application provides a method for controlling a manipulator based on force feedback, which is applied to a control module of a manipulator control device based on force feedback, wherein the manipulator control device further comprises an environment sensing module and a force feedback module; the control module is respectively connected to the environment sensing module, the force feedback module and the manipulator in communication; the method for controlling the manipulator comprises: collecting spatial position information of the manipulator and the operating area of ​​the manipulator collected by the environment sensing module at a preset sampling frequency; at any sampling moment, determining the state information of the manipulator in the operating area of ​​the manipulator at the sampling moment based on the spatial position information of the manipulator and the operating area of ​​the manipulator and a Kalman filtering algorithm; at any sampling moment, determining the state information of the manipulator in the operating area of ​​the manipulator based on the sampling frequency; The state information corresponding to the sampling moment, the input target position, the A-star algorithm and the reinforcement learning algorithm are used to determine the target motion path of the manipulator at the sampling moment; for any sampling moment, a manipulator control signal is generated based on the target motion path at the sampling moment, and the manipulator control signal is sent to the manipulator to make the manipulator move according to the target motion path; for any sampling moment, the force feedback signal of the manipulator at the sampling moment is determined based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment of the sampling moment and the force feedback control algorithm, and the force feedback signal is sent to the force feedback module so that the force feedback module outputs the feedback force corresponding to the sampling moment. In this way, the accuracy of manipulator control can be improved through real-time motion path planning and force feedback control.

[0132] Based on the same application concept, the embodiment of the present application also provides a control module corresponding to the robot arm control method based on force feedback provided in the above embodiment. Since the principle of solving the problem by the control module in the embodiment of the present application is similar to the robot arm control method in the above embodiment of the present application, the implementation of the control module can refer to the implementation of the method, and the repeated parts will not be repeated.

[0133] See also Figure 4 , Figure 4 This is one of the functional module diagrams of a control module provided in an embodiment of the present application. Figure 4 As shown, the control module 130 includes:

[0134] A sampling unit 131 is used to collect the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected by the environment perception module according to a preset sampling frequency;

[0135] A first determining unit 132 is used to determine, for any sampling moment, state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment based on the spatial position information of the robotic arm and the operating area of ​​the robotic arm and a Kalman filter algorithm;

[0136] A second determination unit 133 is used to determine, for any sampling moment, a target motion path of the robotic arm at the sampling moment based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm, and the reinforcement learning algorithm;

[0137] The first execution unit 134 is used to generate a manipulator control signal based on the target motion path at any sampling moment, and send the manipulator control signal to the manipulator so that the manipulator moves according to the target motion path;

[0138] The second execution unit 135 is used to determine, for any sampling moment, the force feedback signal of the robotic arm at the sampling moment based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment of the sampling moment, and the force feedback control algorithm, and send the force feedback signal to the force feedback module so that the force feedback module outputs the feedback force corresponding to the sampling moment.

[0139] In a preferred implementation of the embodiment of the present application, when the sampling unit 131 is used to collect the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected by the environment perception module according to a preset sampling frequency, the sampling unit 131 is specifically used to:

[0140] Collecting image information of the robotic arm and the operating area of ​​the robotic arm collected by the visual sensor according to a preset sampling frequency;

[0141] According to a preset sampling frequency, the contour of the target object in the operating area of ​​the robotic arm and the distance from the robotic arm to the target object collected by the laser scanning sensor are collected;

[0142] Collecting the force signal of the movement of the robotic arm collected by the force feedback sensor according to a preset sampling frequency;

[0143] The image information of the robotic arm and the operating area of ​​the robotic arm, the outline of the target object in the operating area of ​​the robotic arm, the distance from the robotic arm to the target object, and the force signal of the movement of the robotic arm are determined as the spatial position information of the robotic arm and the operating area of ​​the robotic arm.

[0144] Further, when the first determining unit 132 is used to determine, for any sampling moment, based on the spatial position information of the robotic arm and the operating area of ​​the robotic arm and the Kalman filter algorithm, the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment, the first determining unit 132 is specifically used to:

[0145] Determine a predicted value of the state information at the sampling moment according to the state information of the robot arm in the operation area of ​​the robot arm determined at the previous sampling moment of the sampling moment;

[0146] According to the predicted value of the state information at the sampling moment and the spatial position information of the robotic arm and the operating area of ​​the robotic arm at the sampling moment, the predicted value is updated to obtain the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment; the state information includes the position information and speed information of the robotic arm in the operating area of ​​the robotic arm.

[0147] Further, when the second determination unit 133 is used to determine the target motion path of the manipulator at any sampling moment based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm, and the reinforcement learning algorithm, the second determination unit 133 is specifically used to:

[0148] Determine the current position of the robotic arm according to the state information corresponding to the sampling moment;

[0149] Determine the shortest path from the current position to the target position according to the current position, the input target position and the A-star algorithm;

[0150] The shortest path is optimized according to the reinforcement learning algorithm and the spatial position information of the operating area of ​​the robotic arm to obtain the target motion path of the robotic arm at the sampling moment.

[0151] Further, when the second execution unit 135 is used to determine the force feedback signal of the manipulator at any sampling moment based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment of the sampling moment, and the force feedback control algorithm, the second execution unit 135 is specifically used to:

[0152] Determine the actual position of the robot arm at the sampling moment according to the state information corresponding to the sampling moment;

[0153] Determine the predicted position of the robot arm at the sampling moment according to the target motion path determined at the previous sampling moment of the sampling moment;

[0154] The force feedback signal of the robotic arm at the sampling moment is determined according to the deviation between the actual position and the predicted position of the robotic arm at the sampling moment and the force feedback control algorithm.

[0155] For further information, see Figure 5 , Figure 5 This is a second functional module diagram of a control module provided in an embodiment of the present application. Figure 5 As shown, the control module 130 also includes:

[0156] The third determination unit 136 is used to input the robot arm control signal into a trained support vector machine model or a neural network model, and output an optimized robot arm control signal.

[0157] The third execution unit 137 is used to send the optimized robot arm control signal to the robot arm, so that the robot arm moves according to the target motion path.

[0158] Furthermore, if Figure 5 As shown, the control module 130 also includes:

[0159] The fourth determining unit 138 is configured to determine that an abnormal collision occurs to the robotic arm if the value of the feedback force corresponding to the force signal is greater than a preset feedback force threshold.

[0160] The fourth execution unit 139 is used to generate a robot arm locking signal, and send the robot arm locking signal to the robot arm, so that the robot arm is controlled to stop moving and lock the position according to the robot arm locking signal.

[0161] The control module provided in the embodiment of the present application is applied to a mechanical arm control device based on force feedback, and the mechanical arm control device also includes an environment perception module and a force feedback module; the control module is respectively connected to the environment perception module, the force feedback module and the mechanical arm in communication; the control module includes: a sampling unit for collecting spatial position information of the mechanical arm and the operating area of ​​the mechanical arm collected by the environment perception module according to a preset sampling frequency; a first determination unit for determining, at any sampling moment, the state information of the mechanical arm in the operating area of ​​the mechanical arm based on the spatial position information of the mechanical arm and the operating area of ​​the mechanical arm and the Kalman filter algorithm; a second determination unit for determining, at any sampling moment, the state information of the mechanical arm in the operating area of ​​the mechanical arm based on the spatial position information of the mechanical arm and the operating area of ​​the mechanical arm and the Kalman filter algorithm; The state information, the input target position, the A-star algorithm and the reinforcement learning algorithm are used to determine the target motion path of the manipulator at the sampling moment; the first execution unit is used to generate a manipulator control signal based on the target motion path at the sampling moment at any sampling moment, and send the manipulator control signal to the manipulator to make the manipulator move according to the target motion path; the second execution unit is used to determine the force feedback signal of the manipulator at the sampling moment based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment of the sampling moment and the force feedback control algorithm at any sampling moment, and send the force feedback signal to the force feedback module so that the force feedback module outputs the feedback force corresponding to the sampling moment. In this way, the accuracy of manipulator control can be improved through real-time motion path planning and force feedback control.

[0162] Based on the same application concept, an embodiment of the present application also provides a robotic arm, which includes a robotic arm control device based on force feedback as introduced above, and uses a robotic arm control method based on force feedback as introduced above to control the robotic arm, and the repeated parts will not be repeated.

[0163] Based on the same application idea, please refer to Figure 6 , Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 6 As shown, the electronic device 600 includes a processor 610 , a memory 620 and a bus 630 .

[0164] The memory 620 stores machine-readable instructions executable by the processor 610. When the electronic device 600 is running, the processor 610 communicates with the memory 620 through the bus 630. When the processor 610 is running, the machine-readable instructions are executed by the steps of the force feedback-based robotic arm control method provided in the above embodiment. The specific implementation method can be found in the method embodiment, which will not be repeated here.

[0165] An embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the force feedback-based robotic arm control method provided in the above embodiment are executed. The specific implementation method can be found in the method embodiment, which will not be repeated here.

[0166] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0167] In the embodiments provided in the present application, it should be understood that the disclosed methods, modules and devices can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0168] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0169] In addition, each functional unit in the embodiments provided in the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0170] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0171] It should be noted that similar numbers and letters represent similar items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are only used to distinguish the description and are not to be understood as indicating or implying relative importance.

[0172] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application. They should all be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A mechanical arm control method based on force feedback, characterized in that: A control module applied to a mechanical arm control device based on force feedback, wherein the mechanical arm control device further comprises an environment perception module and a force feedback module; The control module is communicatively connected with the environment perception module, the force feedback module and the robotic arm respectively; The robotic arm control method comprises: Collecting the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected by the environment perception module according to a preset sampling frequency; For any sampling moment, based on the spatial position information of the robotic arm and the operating area of ​​the robotic arm and the Kalman filter algorithm, determine the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment; For any sampling moment, based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm and the reinforcement learning algorithm, determine the target motion path of the robotic arm at the sampling moment; For any sampling moment, a robot arm control signal is generated based on the target motion path at the sampling moment, and the robot arm control signal is sent to the robot arm so that the robot arm moves according to the target motion path; For any sampling moment, based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment and the force feedback control algorithm, the force feedback signal of the robotic arm at the sampling moment is determined, and the force feedback signal is sent to the force feedback module so that the force feedback module outputs the feedback force corresponding to the sampling moment.

2. The robot arm control method according to claim 1, characterized in that: The environment perception module includes a visual sensor, a laser scanning sensor and a force feedback sensor; the visual sensor, the laser scanning sensor and the force feedback sensor are respectively connected to the control module for communication; the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected by the environment perception module according to a preset sampling frequency includes: Collecting image information of the robotic arm and the operating area of ​​the robotic arm collected by the visual sensor according to a preset sampling frequency; According to a preset sampling frequency, the contour of the target object in the operating area of ​​the robotic arm and the distance from the robotic arm to the target object collected by the laser scanning sensor are collected; Collecting the force signal of the movement of the robotic arm collected by the force feedback sensor according to a preset sampling frequency; The image information of the robotic arm and the operating area of ​​the robotic arm, the outline of the target object in the operating area of ​​the robotic arm, the distance from the robotic arm to the target object, and the force signal of the movement of the robotic arm are determined as the spatial position information of the robotic arm and the operating area of ​​the robotic arm.

3. The robot arm control method according to claim 2, characterized in that: For any sampling moment, the determining of the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment based on the spatial position information of the robotic arm and the operating area of ​​the robotic arm and the Kalman filter algorithm includes: Determine a predicted value of the state information at the sampling moment according to the state information of the robot arm in the operation area of ​​the robot arm determined at the previous sampling moment of the sampling moment; According to the predicted value of the state information at the sampling moment and the spatial position information of the robotic arm and the operating area of ​​the robotic arm at the sampling moment, the predicted value is updated to obtain the state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment; the state information includes the position information and speed information of the robotic arm in the operating area of ​​the robotic arm.

4. The robot arm control method according to claim 1, characterized in that: For any sampling moment, the target motion path of the robot arm at the sampling moment is determined based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm and the reinforcement learning algorithm, including: Determine the current position of the robotic arm according to the state information corresponding to the sampling moment; Determine the shortest path from the current position to the target position according to the current position, the input target position and the A-star algorithm; The shortest path is optimized according to the reinforcement learning algorithm and the spatial position information of the operating area of ​​the robotic arm to obtain the target motion path of the robotic arm at the sampling moment.

5. The robot arm control method according to claim 2, characterized in that: For any sampling moment, determining the force feedback signal of the manipulator at the sampling moment based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment, and the force feedback control algorithm includes: Determine the actual position of the robot arm at the sampling moment according to the state information corresponding to the sampling moment; Determine the predicted position of the robot arm at the sampling moment according to the target motion path determined at the previous sampling moment of the sampling moment; The force feedback signal of the robotic arm at the sampling moment is determined according to the deviation between the actual position and the predicted position of the robotic arm at the sampling moment and the force feedback control algorithm.

6. The robot arm control method according to claim 1, characterized in that: For any sampling moment, after generating a robot arm control signal based on the target motion path at the sampling moment, the robot arm control method further includes: Inputting the robot arm control signal into a trained support vector machine model or a neural network model, and outputting an optimized robot arm control signal; The optimized robot arm control signal is sent to the robot arm, so that the robot arm moves according to the target motion path.

7. The robot arm control method according to claim 2, characterized in that: After collecting the force signal of the movement of the mechanical arm collected by the force feedback sensor according to the preset sampling frequency, the mechanical arm control method further includes: If the value of the feedback force corresponding to the force signal is greater than a preset feedback force threshold, it is determined that an abnormal collision occurs to the robotic arm; A robot arm locking signal is generated and sent to the robot arm, so that the robot arm is controlled to stop moving and lock the position according to the robot arm locking signal.

8. A control module, characterized in that: Applied to a mechanical arm control device based on force feedback, the mechanical arm control device further includes an environment perception module and a force feedback module; the control module is respectively connected to the environment perception module, the force feedback module and the mechanical arm in communication; the control module includes: A sampling unit, configured to collect the spatial position information of the robotic arm and the operating area of ​​the robotic arm collected by the environment perception module according to a preset sampling frequency; A first determining unit is used to determine, for any sampling moment, state information of the robotic arm in the operating area of ​​the robotic arm at the sampling moment based on spatial position information of the robotic arm and the operating area of ​​the robotic arm and a Kalman filter algorithm; A second determination unit is used to determine, for any sampling moment, a target motion path of the robotic arm at the sampling moment based on the state information corresponding to the sampling moment, the input target position, the A-star algorithm, and the reinforcement learning algorithm; A first execution unit is used to generate a robot arm control signal based on a target motion path at any sampling moment, and send the robot arm control signal to the robot arm so that the robot arm moves according to the target motion path; The second execution unit is used to determine, for any sampling moment, the force feedback signal of the robotic arm at the sampling moment based on the state information corresponding to the sampling moment, the target motion path determined at the previous sampling moment of the sampling moment, and the force feedback control algorithm, and send the force feedback signal to the force feedback module so that the force feedback module outputs the feedback force corresponding to the sampling moment.

9. A mechanical arm control device based on force feedback, characterized in that: The mechanical arm control device comprises: An environmental perception module, a force feedback module and a control module as claimed in claim 8.

10. A robotic arm, characterized in that: It comprises the mechanical arm control device based on force feedback as claimed in claim 9.

Images