Rare earth metal electrolytic furnace discharging robot and tail end compliance control method and system thereof

Abstract

The invention provides a rare earth metal electrolytic furnace discharging robot and a tail end compliance control method and system thereof, and the method comprises the steps: building a rare earth discharging simulation environment for simulating a discharging scene, and building a rare earth discharging control model for simulating the motion trail of a discharging tool through reinforcement learning training based on the simulation environment; contact force sense information between a discharging tool and the electrolytic furnace in the discharging process of the rare earth metal electrolytic furnace is obtained in real time and input into the rare earth discharging control model, and a real-time discharging tool movement track is obtained; and the real-time discharging tool motion trail is input into a discharging robot controller so as to achieve flexible control over the tail end of the discharging robot. According to the robot discharging operation method suitable for the rare earth metal electrolytic furnace, the robot discharging operation method suitable for the rare earth metal electrolytic furnace is achieved on the basis of reinforcement learning and force sensing, the force sensor senses the acting force of a tool and the electrolytic furnace, the working tail end of the rare earth metal electrolytic furnace is controlled through the reinforcement learning method, and then anthropomorphic discharging operation is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of special robots, and in particular to a rare earth metal electrolysis furnace discharging robot based on reinforcement learning and terminal force perception, and a terminal compliance control method and system thereof. Background Art

[0002] In rare earth metal production, traditional manual unloading methods are labor-intensive and unsafe. With rising labor costs, companies urgently need to automate production processes to reduce labor costs and improve safety. Rare earth unloading environments are characterized by high temperatures and unstructured environments. First, the electrolytic furnace temperature exceeds 1000°C, placing high demands on the electromechanical system design. Second, during the electrolysis process, the cathode and crucible at the bottom of the furnace change shape due to metal adhesion. Existing unloading robots fail to account for these changes in the internal shape of the electrolytic furnace during design, resulting in poor stability and reliability in unloading operations. The high-temperature electrolyte makes it difficult for visual sensors and other devices to perceive the contents of the electrolytic furnace. Therefore, existing unloading robots are not suitable for unloading rare earth metal electrolytic furnaces.

[0003] Currently, no description or report of similar technology to the present invention has been found, and similar information at home and abroad has not been collected. Summary of the Invention

[0004] In response to the above-mentioned deficiencies in the prior art, the present invention provides a rare earth metal electrolytic furnace discharging robot and a terminal compliance control method and system thereof. The method and system are based on reinforcement learning and terminal force perception to realize intelligent control of rare earth discharging operations, and are suitable for robot discharging operations of rare earth metal electrolytic furnaces.

[0005] According to one aspect of the present invention, a method for controlling the terminal compliance of a discharge robot of a rare earth metal electrolysis furnace is provided, comprising:

[0006] Establishing a rare earth discharging simulation environment for simulating discharging scenarios, and constructing a rare earth discharging control model for simulating the motion trajectory of the discharging tool through reinforcement learning training based on the simulation environment;

[0007] Real-time acquisition of contact force information between a discharge tool and the electrolytic furnace during the discharge process of the rare earth metal electrolytic furnace, and inputting the information into the rare earth discharge control model to obtain a real-time motion trajectory of the discharge tool;

[0008] The real-time motion trajectory of the discharging tool is input into the discharging robot controller to achieve flexible control of the end of the discharging robot.

[0009] According to another aspect of the present invention, a terminal compliance control system for a discharge robot of a rare earth metal electrolysis furnace is provided, comprising: a force sensing module, a compliance controller, and a reinforcement learning training module; wherein:

[0010] The force sensing module is provided at the end of the discharging robot and is used to obtain the contact force information between the discharging tool and the electrolytic furnace in real time during the discharging process of the rare earth metal electrolytic furnace;

[0011] The reinforcement learning training module is used to establish a rare earth discharging simulation environment for simulating a discharging scenario, and based on the simulation environment, construct a rare earth discharging control model for simulating the motion trajectory of the discharging tool through reinforcement learning training;

[0012] The compliant controller is used to carry the rare earth discharging control model, and obtain the contact force information obtained by the force sensing module in real time as the input of the rare earth discharging control model, obtain the real-time discharging tool motion trajectory and output it to the discharging robot controller, so as to realize the compliant control of the end of the discharging robot.

[0013] According to a third aspect of the present invention, a rare earth metal electrolysis furnace discharge robot is provided, comprising: a discharge robot body, wherein the discharge robot body realizes compliant control of the discharge robot end through the method or system described above.

[0014] Due to the adoption of the above technical solution, the present invention has at least one of the following beneficial effects compared with the prior art:

[0015] The rare earth metal electrolytic furnace discharging robot and its terminal compliant control method and system provided by the present invention utilize an operating mode in which the trajectory of the tool can be adjusted by the contact force between the tool and the electrolytic furnace during manual discharging. The force between the tool and the electrolytic furnace is sensed by a force sensor, and the working terminal is controlled by a reinforcement learning method to realize anthropomorphic discharging operation.

[0016] The present invention provides a rare earth metal electrolytic furnace discharge robot and its terminal compliance control method and system. A high-precision six-dimensional force-distance sensor is equipped at the robot's terminal end to sense the force information of the interaction with the electrolytic furnace in real time, including the magnitude and direction of the contact force. A reinforcement learning-based compliance controller combines this force feedback data with the robot's joint position and velocity information. After deep neural network processing, it generates the discharge tool motion trajectory in real time, thereby deriving an adjustment strategy for the end effector. This enables the robot to flexibly adjust its movements based on changes in contact force, just like a skilled worker, accurately responding to high-temperature, unstructured, and dynamically changing discharge environments. This greatly improves the intelligence level of the operation and allows the precise completion of discharge tasks. The integrated force sensor and reinforcement learning algorithm give the robot's terminal end force perception and compliance control capabilities, enabling it to adjust its movements in real time based on the contact force with the electrolytic furnace, accurately responding to the high-temperature, unstructured, and dynamically changing rare earth discharge environments. Its operational accuracy and stability far exceed those of traditional robots, improving the intelligence level of the operation.

[0017] The present invention provides a rare earth metal electrolysis furnace discharge robot and a terminal compliance control method and system thereof. The compliance controller outputs a motion control strategy in real time based on force sensor feedback data and joint status information, and plans the discharge motion trajectory according to the control strategy.

[0018] The rare earth metal electrolysis furnace discharge robot and its terminal compliance control method and system provided by the present invention can rapidly detect and respond to collision risks. By calculating the motion parameters of each joint based on the input trajectory and controlling the arm joints to complete the motion, the discharge robot effectively prevents collisions and overloads, ensuring the continuity and stability of discharge operations, avoiding production interruptions and equipment damage caused by collisions or overloads, and improving production efficiency and equipment life. The reinforcement learning-based compliance controller, relying on force sensor feedback data and robot joint status information, outputs real-time end effector adjustment strategies, effectively preventing collisions and jamming, ensuring continuous and stable discharge, significantly improving reliability, and enhancing operational stability.

[0019] The present invention provides a rare earth metal electrolysis furnace discharge robot and its terminal compliance control method and system. The automated discharge solution significantly reduces manual intervention, significantly reducing labor costs in the long term as labor costs continue to rise. This automated discharge solution reduces manual intervention, significantly reducing labor costs and production costs in the long term as labor costs continue to rise.

[0020] The present invention provides a rare earth metal electrolysis furnace discharge robot and its terminal compliance control method and system. The robot replaces manual labor in high-temperature, high-risk discharge operations, completely eliminating the safety hazards associated with manual operation. By replacing manual labor in high-temperature, high-risk discharge operations, the robot eliminates the safety hazards associated with manual operation and improves operational safety.

[0021] The present invention provides a rare earth metal electrolysis furnace discharge robot and its terminal compliance control method and system. Leveraging its high-precision force sensing and compliance control capabilities, the robot can precisely perform discharge operations, ensuring consistent rare earth metal cooling and molding quality. This reduces defective products due to human error or equipment failure, improves overall product quality and performance, and contributes to enhancing a company's market competitiveness and product added value. The improved discharge accuracy and stability ensure consistent rare earth metal cooling and molding quality, reducing defective products due to human error or equipment failure, improving overall product quality and performance, and enhancing discharge quality.

[0022] The rare earth metal electrolytic furnace discharging robot and its terminal compliant control method and system provided by the present invention are based on force feedback and reinforcement learning. The robot can autonomously adapt to the discharging requirements of different electrolytic furnace models and processes, reducing equipment replacement and debugging time. It has the advantages of wide adaptability and rapid debugging, and improves production flexibility and efficiency. At the same time, the collaborative control system can optimize the production process in real time according to the electrolytic furnace operating parameters and the robot's discharging progress, further improving production efficiency and enabling enterprises to respond to market changes and customer needs more quickly. Through force feedback and reinforcement learning, it can autonomously adapt to the discharging requirements of different electrolytic furnace models and processes, reducing equipment replacement and debugging time, and improving production efficiency. At the same time, the collaborative control system optimizes the production process and further improves efficiency.

[0023] The present invention provides a rare earth metal electrolytic furnace discharge robot and its terminal compliance control method and system, combining reinforcement learning with terminal force sensing technology to provide a new intelligent solution for rare earth metal electrolytic furnace discharge operations. This technological integration not only improves the stability and quality of the robot's discharge operations but also significantly reduces operational risks and costs. It represents a cutting-edge development direction for specialized robotics in rare earth metal production and provides valuable reference for automation upgrades in similar production scenarios.

[0024] Through the above advantages, the rare earth metal electrolytic furnace discharging robot and its terminal flexible control method and system provided by the present invention not only improve the stability and quality of the robot's rare earth discharging operation, but also significantly reduce the operation risk and cost, providing a more advanced and reliable technical solution for the rare earth electrolytic furnace discharging operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0026] Figure 1 The figure is a workflow diagram of a method for controlling the terminal compliance of a discharge robot of a rare earth metal electrolysis furnace in a preferred embodiment of the present invention.

[0027] Figure 2 This is a flowchart of rare earth discharge simulation environment training and reinforcement learning training for the terminal compliant control method of the rare earth metal electrolysis furnace discharge robot in a preferred embodiment of the present invention.

[0028] Figure 3 The figure is a schematic diagram of the component modules of the terminal compliance control system of the rare earth metal electrolysis furnace discharge robot in a preferred embodiment of the present invention.

[0029] Figure 4 The figure is a schematic diagram of the installation structure of the force sensing module of the discharging robot in a specific application example of the present invention.

[0030] Figure 5 Schematic diagram of a rare earth discharging simulation environment in a specific application example of the present invention. DETAILED DESCRIPTION

[0031] The following is a detailed description of an embodiment of the present invention. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process. It should be noted that those skilled in the art may make various modifications and improvements without departing from the scope of the present invention, and these modifications and improvements fall within the scope of protection of the present invention.

[0032] In rare earth metal production, traditional manual discharging methods have problems such as high labor intensity and poor safety. Due to the high temperature and unstructured characteristics of rare earth discharging scenes, existing discharging robots are difficult to apply to rare earth metal electrolytic furnace discharging operations, and there is no automated technology for rare earth metal electrolytic furnace discharging operations in the existing technology.

[0033] In response to the above problems, an embodiment of the present invention provides a flexible control method for the end of a rare earth metal electrolytic furnace discharging robot. The method implements a robot discharging operation method suitable for rare earth metal electrolytic furnaces based on reinforcement learning and force perception. The force between the tool and the electrolytic furnace is sensed by a force sensor and the reinforcement learning method is used to control the working end of the rare earth metal electrolytic furnace, thereby realizing an anthropomorphic discharging operation.

[0034] Specifically, if Figure 1 As shown, the embodiment provides a method for controlling the terminal compliance of a discharge robot of a rare earth metal electrolysis furnace, including:

[0035] S1, establish a rare earth discharging simulation environment for simulating the discharging scene, and based on the simulation environment, build a rare earth discharging control model for simulating the motion trajectory of the discharging tool through reinforcement learning training;

[0036] S2, real-time acquisition of contact force information between the discharge tool and the electrolytic furnace during the discharge process of the rare earth metal electrolytic furnace, and input into the rare earth discharge control model to obtain the real-time motion trajectory of the discharge tool;

[0037] S3, inputs the real-time motion trajectory of the discharge tool into the discharge robot controller to achieve smooth control of the end of the discharge robot.

[0038] In some preferred embodiments, the above S1 is to establish a rare earth discharging simulation environment for simulating the discharging scene, such as Figure 2 As shown, it can further include:

[0039] A virtual simulation environment corresponding to the actual discharge scene is established, and the following initialization is performed in the virtual simulation environment to obtain a rare earth discharge simulation environment for simulating the discharge scene:

[0040] S111, setting the position of the discharging robot arm to directly above the electrolytic furnace as the initial position;

[0041] S112, placing obstacles on the inner wall of the electrolytic furnace to simulate the change of the internal shape ratio of the electrolytic furnace during the electrolysis process, and randomly generating an initial environment;

[0042] S113, initializing various parameters of the discharging robot. In some preferred embodiments, the above S111, establishing a virtual simulation environment corresponding to the actual discharging scene, may further include:

[0043] Build a virtual simulation environment corresponding to the actual discharge scene, and construct the following key objects in the virtual simulation environment:

[0044] Build an electrolytic furnace and set parameters such as its size, shape, position, mass, and center of mass to ensure that the virtual simulation environment can truly reflect the physical relationships in the actual discharge environment;

[0045] Build a 3D model of the discharging robot and adjust and configure the 3D model according to the actual specifications of the discharging robot;

[0046] Set up the sensor and ensure that the sensor's position, posture, and parameters are consistent with the actual unloading robot so that the sensor's data output can be accurately simulated in the simulation;

[0047] Define the action space and observation space, where the action space is the joint motion space of the discharging robot, and the observation space is the position posture of the end of the discharging tool and the feedback force of the force sensor; according to the dimensions and range of the action space and observation space, set the control interface and status monitoring interface of the discharging robot to ensure that the discharging robot can receive action instructions and perform corresponding actions, and at the same time can obtain the status information in the observation space in real time, providing the necessary input and output interfaces for reinforcement learning training.

[0048] The obtained rare earth discharging simulation environment, such as Figure 5 shown.

[0049] In some preferred embodiments, the above S1, based on the simulation environment, constructs a rare earth discharge control model for simulating the tool end motion trajectory through reinforcement learning training, such as Figure 2 As shown, it can further include:

[0050] A pre-trained reinforcement learning network model (neural network model) is established and the reinforcement learning network model is trained as follows:

[0051] S114, based on the current state information of the end of the discharging robot, the next action position is randomly generated through the pre-trained reinforcement learning network model to form a planned path;

[0052] S115, using the simulator to execute the action position generated in S114, and generate reward and next state information, completing a planning path simulation process;

[0053] S116, storing the action positions, execution probabilities, and rewards during the simulation;

[0054] S117, inputting the data stored in S116 into the reinforcement learning network model, updating the reinforcement learning network model parameters, and optimizing the motion control strategy;

[0055] S118, determine whether the end of the discharging robot reaches the bottom of the electrolytic furnace or whether the maximum number of training rounds has been reached; if so, end this round of training and retain the training parameters of the current reinforcement learning network model; if not, return to restart the training process until the training is completed, and construct a rare earth discharging control model.

[0056] In some preferred embodiments, the above S2, which acquires contact force information between the discharge tool and the electrolytic furnace in real time during the discharge process of the rare earth metal electrolytic furnace and inputs it into the rare earth discharge control model to obtain the real-time motion trajectory of the discharge tool, may further include:

[0057] The force sensor feedback data between the discharge tool and the electrolytic furnace during the actual discharge process of the rare earth metal electrolytic furnace is obtained in real time, and combined with the joint position and speed information of the discharge robot, it is used as the input of the rare earth discharge control model. The corresponding motion control strategy (i.e., the direction and speed of tool movement) is output through the rare earth discharge control model, and the joint control parameters are solved according to the motion control strategy to obtain the real-time discharge tool motion trajectory.

[0058] In some preferred embodiments, the above S3, inputting the real-time motion trajectory of the discharge tool to the discharge robot controller to achieve compliant control of the end of the discharge robot, may further include:

[0059] The real-time motion trajectory of the discharge tool is input into the discharge robot controller. The discharge robot calculates the motion parameters of each joint according to the input trajectory and controls the joints of the robotic arm to complete the movement. It simulates manual discharge and corrects the action in real time based on the contact force to achieve smooth control of the end of the discharge robot.

[0060] The above-mentioned embodiment of the present invention provides a method for flexible control of the terminal of a rare earth metal electrolytic furnace discharging robot. Aiming at the electrolytic furnace discharging operation scene and operation process, a force sensing system adapted to high-temperature scenes is designed. The discharging scene is simulated to construct a rare earth discharging simulation environment for reinforcement learning training of the rare earth discharging control model. During the operation, the trained rare earth discharging control model is installed in the flexible controller. The flexible controller obtains the force information detected by the force sensing system in real time, inputs it into the rare earth discharging control model, calculates the trajectory of the tool end movement through the model, and outputs it to the robot controller. Flexible control of the terminal is achieved. This method aims to address the problems existing in the discharge of rare earth metal electrolytic furnaces, and with the help of reinforcement learning and terminal force sensing technology, realizes the intelligence, stabilization and reliability of the robot discharging operation, significantly improves the safety and efficiency of the operation, and reduces labor costs.

[0061] Based on the same inventive concept, an embodiment of the present invention also provides a terminal compliance control system for a rare earth metal electrolysis furnace discharge robot.

[0062] Specifically, if Figure 3 As shown, the terminal compliance control system of the rare earth metal electrolysis furnace discharge robot provided in this embodiment may include: a force sensing module, a compliance controller, and a reinforcement learning training module; wherein:

[0063] The force sensing module is installed at the end of the discharging robot and is used to obtain the contact force information between the discharging tool and the electrolytic furnace in real time during the discharging process of the rare earth metal electrolytic furnace;

[0064] A reinforcement learning training module is used to establish a rare earth discharging simulation environment for simulating discharging scenarios, and based on the simulation environment, a rare earth discharging control model for simulating the motion trajectory of the discharging tool is constructed through reinforcement learning training;

[0065] The compliant controller is used to carry the rare earth discharging control model and obtain the contact force information obtained by the force sensing module in real time as the input of the rare earth discharging control model, obtain the real-time motion trajectory of the discharging tool and output it to the discharging robot controller to achieve compliant control of the end of the discharging robot.

[0066] The specific implementation of each functional module constituting the system provided by the above embodiment of the present invention is further described in detail below in conjunction with the preferred implementation manner.

[0067] In some preferred embodiments, the force sensing module may further include any one or more of the following:

[0068] High-precision six-dimensional force-distance sensors are used and distributed on the end effector of the discharging robot;

[0069] The EtherCAT bus is used for data acquisition, and the collected analog signals are converted into digital signals through transmitters; the data synchronization period is <2ms, preferably 1ms;

[0070] Use heat-insulating material to install between the discharging tool.

[0071] In some preferred embodiments, the reinforcement learning training module may further include: a simulation environment construction unit, an environment training unit, and a reinforcement learning training unit.

[0072] A simulation environment construction unit is used to establish a virtual simulation environment corresponding to the actual discharge scene.

[0073] The environment training unit is used to perform the following initialization in the virtual simulation environment:

[0074] Set the discharging robot arm position directly above the electrolytic furnace as the initial position;

[0075] By placing obstacles on the inner wall of the electrolytic furnace to simulate the change of the internal shape ratio of the electrolytic furnace during the electrolysis process, the initial environment is randomly generated;

[0076] Initialize various parameters of the discharge robot.

[0077] A reinforcement learning training unit is used to perform the following reinforcement learning training process in a virtual simulation environment to build a rare earth discharge control model for simulating the tool end motion trajectory:

[0078] Build a pre-trained reinforcement learning neural network model;

[0079] Based on the current state information of the discharge robot end, the pre-trained neural network model randomly generates the next action position to form a planned path;

[0080] The simulator executes the action position generated, generates rewards and next state information, and completes a planning path simulation process;

[0081] Store action positions, execution probabilities, and rewards during simulation;

[0082] Input the stored data into the neural network model, update the neural network model parameters, and optimize the control strategy;

[0083] When the end of the discharging robot reaches the bottom of the electrolytic furnace or the maximum number of training rounds has passed, the current round of training ends, the training parameters of the current neural network model are retained, and a rare earth discharging control model is constructed.

[0084] In some preferred embodiments, the simulation environment construction unit, which establishes a virtual simulation environment corresponding to the actual discharge scene, may further include:

[0085] Build a virtual simulation environment corresponding to the actual discharge scene, and construct the following key objects in the virtual simulation environment:

[0086] Build an electrolytic furnace and set parameters such as its size, shape, position, mass, and center of mass to ensure that the virtual simulation environment can truly reflect the physical relationships in the actual discharge environment;

[0087] Build a 3D model of the discharging robot and adjust and configure the 3D model according to the actual specifications of the discharging robot;

[0088] Set up the sensor and ensure that the sensor's position, posture, and parameters are consistent with the actual unloading robot so that the sensor's data output can be accurately simulated in the simulation;

[0089] Define the action space and observation space, where the action space is the joint motion space of the discharging robot, and the observation space is the position posture of the end of the discharging tool and the feedback force of the force sensor; according to the dimensions and range of the action space and observation space, set the control interface and status monitoring interface of the discharging robot to ensure that the discharging robot can receive action instructions and perform corresponding actions, and at the same time can obtain the status information in the observation space in real time, providing the necessary input and output interfaces for reinforcement learning training.

[0090] In some preferred embodiments, the compliance controller may further include:

[0091] The control unit is used to carry the rare earth discharging control model and obtain the contact force information obtained by the force sensing module in real time as the input of the rare earth discharging control model, obtain the real-time motion trajectory of the discharging tool and output it to the discharging robot controller to achieve flexible control of the end of the discharging robot.

[0092] It should be noted that the steps in the method provided by the present invention can be implemented by using the corresponding components in the system. Those skilled in the art can refer to the technical solution of the system to implement the step flow of the method, and can also refer to the technical solution of the method to implement the composition of the system. That is, the embodiments in the system and the embodiments in the method can be understood as preferred examples of each other, and will not be elaborated here.

[0093] Based on the same inventive concept, an embodiment of the present invention further provides a rare earth metal electrolysis furnace discharge robot.

[0094] Specifically, the rare earth metal electrolysis furnace discharging robot provided in this embodiment includes: a discharging robot body, and the discharging robot body realizes flexible control of the end of the discharging robot through the method or system provided in the above embodiment of the present invention.

[0095] In this example:

[0096] Force sensing module: A high-precision six-dimensional force-distance sensor is distributed and embedded in the robot's end effector to obtain real-time force information on the interaction between the robot and the electrolytic furnace, covering the magnitude and direction of the contact force, providing data support for subsequent compliant control.

[0097] Compliant controller: Built based on a reinforcement learning algorithm, it takes force sensor feedback data and robot joint position and velocity information as input, processes them through a rare earth discharge control model based on a deep neural network, outputs the motion trajectory of the discharge tool, and outputs the value to the robot control module;

[0098] Discharging robot controller: This controller adjusts the control strategy based on the motion trajectory of the discharging tool, including motion direction correction and force fine-tuning. This simulates the flexible operation of manual discharging, which allows for real-time correction of movements based on contact force, enhancing the robot's adaptability to complex environments.

[0099] Reinforcement Learning Training Module: Build a virtual simulation platform that simulates the discharge scenario of a rare earth metal electrolytic furnace, incorporating key elements such as high temperature, unstructured environment, and dynamic changes in the internal shape of the electrolytic furnace. The robot is allowed to learn through trial and error in this environment, and the control strategy is optimized based on reward function feedback until high-precision and high-stability discharge effects are achieved in the simulated scenario. The mature strategy is then migrated to actual production to ensure the intelligence level of the robot's discharge operation.

[0100] Discharging Operation Module: This module activates the compliant control function after the discharging tool enters the electrolytic furnace space. The robot moves along the discharging tool's trajectory, gradually approaching the bottom of the electrolytic furnace. During the approach process, the end force sensing module continuously monitors the contact force between the discharging tool and the electrolytic furnace. The rare earth discharging control model corrects the tool trajectory based on the end contact force to ensure that the end of the tool maintains no contact or low contact force with the inner wall of the electrolytic furnace. When the end of the tool reaches the lowest position set by the trajectory or the bottom contact force is too great to reach downward, the robot stops for a period of time and then moves upward. During movement, the system detects the contact force and adjusts the trajectory according to the rare earth discharging control model.

[0101] Force Sensing and Feedback Adjustment Module: Throughout the discharge process, force sensors at the robot's end-effector monitor the interaction forces between the discharge tool, the electrolytic furnace, and the crucible in real time. This force information is transmitted at high frequency to the compliance controller, which rapidly adjusts the robot's end-effector based on a pre-set force feedback control algorithm.

[0102] High-temperature protection design: Insulation and heat dissipation measures are taken for the electronic components and connecting lines of the force sensor and compliance controller. High-temperature resistant materials are used to create the protective shell. The circuit layout is optimized to ensure that key components can operate stably under the high-temperature radiation of the electrolytic furnace, avoid performance degradation or failure caused by overheating, and maintain the continuity of the discharge operation.

[0103] The technical solution provided by the above embodiment of the present invention is further described in detail below with reference to a specific application example.

[0104] like Figure 4 The figure shows the installation structure diagram of the force sensing module of the discharging robot. Figure 4 Among them, 101-102 are force sensor modules, 103-104 are mechanical grippers designed with heat-insulating materials, and 105 is a discharging tool designed with high-temperature-resistant and corrosion-resistant metal materials.

[0105] Based on the method and system provided by the above embodiment of the present invention, in this specific application example, the following is adopted: Figure 3 In the architecture shown, a high-real-time compliance controller receives force sensor data via the EtherCAT bus and processes it to generate motion data. The force sensor analog signal is collected by a transmitter and converted into a digital signal, which is then sent to the compliance controller via the EtherCAT bus. The three-dimensional force sensor measurement data is directly collected by the transmitter.

[0106] Reinforcement learning training: Initialize the training scenario, with the robot arm's position initialized to its initial position, directly above the electrolytic furnace. A random initial environment is generated, simulating changes in the furnace's internal shape during electrolysis by placing obstacles on the furnace's inner wall. Initialize the agent, i.e., the initial robot parameters. Training ends when all simulation scenarios are complete. The next action position is randomly generated based on the robot's current position, speed, and acceleration limits. The simulator executes the generated action position and generates a reward and the next observed state. The action positions, execution probabilities, and observed rewards from the simulation are stored. The observed data is fed into the neural network to update the execution strategy. Training ends when the robot reaches the bottom of the electrolytic furnace or after the maximum number of rounds. A training round consists of the tool descending from the top of the electrolytic furnace to the bottom.

[0107] Further:

[0108] Force sensing system design: To address the high temperatures of the electrolytic furnace, the torque sensor was installed away from the tool end (i.e., the end entering the electrolytic furnace). To address the issue of excessive torque measurement, a distributed measurement method was adopted. This involves two 3D force sensors placed in a dispersed manner to enable high torque measurement. To improve real-time measurement, EtherCAT was used for data acquisition, with a data synchronization cycle of 1ms. The force sensor was selected as a heat-resistant model, with a tolerance of 80°C, to minimize the impact of high temperatures on sensor measurement accuracy. Insulating material was used between the sensor and the discharge tool.

[0109] Simulation model construction: Use simulation software (such as a combination of Gymnasium and Mujoco) to build a virtual simulation platform that resembles the actual production environment, including key objects such as the electrolytic furnace, robots, and tools. Accurately set parameters such as the size, shape, position, mass, and center of mass of the electrolytic furnace to ensure that the simulation environment accurately reflects the physical relationships in the real scene.

[0110] Configure the robot model: Import the 3D model of the robot into the simulation software and adjust and configure it according to the actual robot's specifications. Install and configure the sensors defined above, ensuring their positions, postures, and parameters are consistent with the actual robot so that the sensor data output can be accurately simulated in the simulation.

[0111] Define the action space and observation space: Define the action space as the robot joint motion space in the simulation software, and define the observation space as the tool end position and posture, and the force sensor feedback force. Based on the dimensions and range of the action and observation spaces defined above, set up the robot's control interface and status monitoring interface. Ensure that the robot can receive action commands and execute the corresponding actions, while also obtaining real-time status information in the observation space, providing the necessary input and output interfaces for the reinforcement learning algorithm.

[0112] Defining reinforcement learning rewards: Reinforcement learning rewards include contact force rewards, approach rewards, and control rewards. The smaller the contact force detected by the sensor between the tool and the external environment, the greater the contact force reward; the closer the tool tip is to the bottom of the electrolytic furnace, the greater the approach reward; and the smaller the change in joint movement, the greater the control reward.

[0113] Reinforcement learning training: The DQN (Deep Q-Network) model is used for training. At each step, the robot selects an action (such as the linear velocity and angular velocity of the end effector) through the DQN network based on the current state (such as joint position, end force information, etc.). After executing the action in the simulation environment, the new state and reward signal are collected (calculated based on the task completion and action rationality). The state transition information is stored in the experience replay buffer, and then small batches of data are randomly sampled from it to train the DQN network, and the network parameters are updated to optimize the strategy. The training ends after the preset number of training steps or conditions are reached. The training process is as follows Figure 2 shown.

[0114] Algorithm Deployment: In the robot control system, write code to load the saved model parameters and integrate the model into the robot's control program. Ensure that the model can receive robot status information and output corresponding control actions. Test the deployed system in a simulated or real environment to verify that the model's decisions meet expectations and that the robot can stably and accurately complete the discharge task. Also, check performance indicators such as system response speed, stability, and reliability.

[0115] Discharging process: After the tool enters the electrolytic furnace, the compliant control function is activated. The robot moves along the planned path, gradually approaching the bottom of the electrolytic furnace. During this approach, the end-point force sensing system continuously monitors the contact force between the robot and the electrolytic furnace. The algorithm model corrects the robot's trajectory based on the end-point contact force, ensuring that the end of the tool maintains no contact or low contact force with the inner wall of the electrolytic furnace. If the end of the tool reaches the lowest position set in the trajectory or the contact force at the bottom is too great to reach downward, the robot stops for a while and then moves upward. During this movement, the system detects the contact force and adjusts the trajectory according to the algorithm model.

[0116] Force Sensing and Feedback Adjustment: Throughout the discharge process, force sensors at the robot's end-effector monitor the interaction forces between the robot and the electrolytic furnace and crucible in real time. This force information is transmitted at high frequency to the compliance controller, which rapidly calculates an adjustment strategy for the robot's end-effector based on a pre-set force feedback control algorithm.

[0117] The rare earth metal electrolysis furnace discharge robot and its terminal compliance control method and system provided by the above embodiments of the present invention have the following significant technical effects:

[0118] In terms of intelligent operation, the robot's end-of-line is equipped with a high-precision six-dimensional force-distance sensor, which provides real-time information on the force interaction with the electrolytic furnace, including the magnitude and direction of the contact force. A reinforcement learning-based compliant controller combines this force feedback data with information on the robot's joint position and velocity. After deep neural network processing, it generates real-time adjustment strategies for the end-effector. This enables the robot to flexibly adjust its movements based on changes in contact force, just like a skilled worker, precisely adapting to the high-temperature, unstructured, and dynamically changing discharge environment. This significantly enhances the intelligent operation and enables precise completion of discharge tasks.

[0119] In terms of operation stability: the flexible controller outputs the end effector adjustment strategy in real time based on the force sensor feedback data and joint status information, can quickly detect and respond to collision risks, and the discharge robot calculates the motion parameters of each joint according to the input trajectory and controls the robotic arm joints to complete the movement, effectively preventing the occurrence of collisions and overloads, ensuring the continuity and stability of the discharge operation, avoiding production interruptions and equipment damage caused by collisions or overloads, and improving production efficiency and equipment service life.

[0120] Reduced labor costs: The automated discharging solution significantly reduces manual intervention. As labor costs continue to rise, in the long run, it can significantly reduce the company's labor cost expenditure and bring considerable economic benefits to the company.

[0121] In terms of safe operation: robots replace manual labor to complete high-temperature and high-risk discharging operations, completely eliminating the safety hazards of manual operation.

[0122] Regarding discharge quality: The robot, with its high-precision force sensing and compliant control capabilities, accurately discharges materials, ensuring consistent rare earth metal cooling and forming quality. This reduces defective products caused by human error or equipment failure, improves overall product quality and performance, and contributes to enhancing the company's market competitiveness and product added value.

[0123] Adaptability and flexibility: Based on force feedback and reinforcement learning, the robot can autonomously adapt to the discharge requirements of different electrolytic furnace models and processes, reducing equipment replacement and commissioning time and improving production flexibility and efficiency. Furthermore, the collaborative control system optimizes the production process in real time based on the electrolytic furnace operating parameters and the robot's discharge progress, further improving production efficiency and enabling the company to more quickly respond to market changes and customer needs.

[0124] Technological advancement: The combination of reinforcement learning and end-point force sensing technology provides a new intelligent solution for the discharge of rare earth metal electrolytic furnaces. This technological integration not only improves the stability and quality of robotic discharge operations, but also significantly reduces operational risks and costs. It represents the forefront of specialized robotics technology in rare earth metal production and provides valuable reference for automation upgrades in similar production scenarios.

[0125] Matters not mentioned in the above embodiments of the present invention are well known in the art.

[0126] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art may make various variations or modifications within the scope of the claims, which do not affect the essence of the present invention.

Claims

1. A method for controlling the terminal compliance of a discharge robot of a rare earth metal electrolysis furnace, characterized in that: include: Establishing a rare earth discharging simulation environment for simulating discharging scenarios, and constructing a rare earth discharging control model for simulating the motion trajectory of the discharging tool through reinforcement learning training based on the simulation environment; Real-time acquisition of contact force information between a discharge tool and the electrolytic furnace during the discharge process of the rare earth metal electrolytic furnace, and inputting the information into the rare earth discharge control model to obtain a real-time motion trajectory of the discharge tool; The real-time motion trajectory of the discharging tool is input into the discharging robot controller to achieve flexible control of the end of the discharging robot.

2. The method for controlling the terminal compliance of a discharge robot of a rare earth metal electrolysis furnace according to claim 1, characterized in that: The establishment of a rare earth discharging simulation environment for simulating a discharging scenario includes: A virtual simulation environment corresponding to the actual discharge scene is established, and the following initialization is performed in the virtual simulation environment to obtain a rare earth discharge simulation environment for simulating the discharge scene: Set the discharging robot arm position directly above the electrolytic furnace as the initial position; By placing obstacles on the inner wall of the electrolytic furnace to simulate the change of the internal shape ratio of the electrolytic furnace during the electrolysis process, the initial environment is randomly generated; Initialize various parameters of the discharge robot.

3. The method for controlling the terminal compliance of a discharge robot of a rare earth metal electrolysis furnace according to claim 2, characterized in that: The step of establishing a virtual simulation environment corresponding to the actual discharge scene includes: Build a virtual simulation environment corresponding to the actual discharge scene, and construct the following key objects in the virtual simulation environment: Constructing an electrolytic furnace and setting the parameters of the electrolytic furnace to ensure that the virtual simulation environment can truly reflect the physical relationship in the real discharge environment; Constructing a three-dimensional model of the discharging robot and adjusting and configuring the three-dimensional model according to the actual specifications of the discharging robot; Set up sensors and ensure that the position, posture, and parameters of the sensors are consistent with those of the actual unloading robot so that the data output of the sensors can be accurately simulated in the simulation; Define the action space and observation space, where the action space is the joint motion space of the discharging robot, and the observation space is the position posture of the end of the discharging tool and the feedback force of the force sensor; according to the dimensions and range of the action space and observation space, set the control interface and status monitoring interface of the discharging robot, so that the discharging robot receives action instructions and performs corresponding actions, and at the same time obtains the status information in the observation space, providing the necessary input and output interfaces for reinforcement learning training.

4. The method for controlling the terminal compliance of a discharge robot of a rare earth metal electrolysis furnace according to claim 1, characterized in that: The method of constructing a rare earth discharge control model for simulating the tool end motion trajectory based on the simulation environment through reinforcement learning training includes: A pre-trained reinforcement learning network model is established, and the reinforcement learning network model is trained as follows: Using the simulation environment, real-time status information of the end of the discharging robot is obtained; Based on the current state information of the end of the discharging robot, the next action position is randomly generated through the reinforcement learning network model to form a planned path; The simulator generates action positions, rewards, and next state information to complete a planning path simulation process. Storing action positions, execution probabilities, and rewards during the simulation; Inputting the stored data into the reinforcement learning network model, updating the reinforcement learning network model parameters, and optimizing the motion control strategy; Determine whether the end of the discharging robot reaches the bottom of the electrolytic furnace or whether the maximum number of training rounds has been reached; if so, end this round of training and retain the training parameters of the current reinforcement learning network model; if not, return to restart the training process until the training is completed, and construct a rare earth discharging control model.

5. The method for controlling the terminal compliance of a discharge robot of a rare earth metal electrolysis furnace according to claim 1, characterized in that: The real-time acquisition of contact force information between the discharge tool and the electrolytic furnace during the discharge process of the rare earth metal electrolytic furnace and inputting it into the rare earth discharge control model to obtain the real-time motion trajectory of the discharge tool includes: The force sensor feedback data between the discharge tool and the electrolytic furnace during the actual discharge process of the rare earth metal electrolytic furnace is obtained in real time, and combined with the joint position and speed information of the discharge robot, it is used as the input of the rare earth discharge control model. The corresponding motion control strategy is output through the rare earth discharge control model, and the real-time discharge tool motion trajectory is obtained according to the motion control strategy.

6. The method for controlling the terminal compliance of a discharge robot of a rare earth metal electrolysis furnace according to claim 1, characterized in that: The step of inputting the real-time motion trajectory of the discharge tool into the discharge robot controller to achieve compliant control of the end of the discharge robot includes: The real-time motion trajectory of the discharging tool is input into the discharging robot controller. The discharging robot controller calculates the motion parameters of each joint according to the input trajectory and controls the joints of the robotic arm to complete the movement. It simulates manual discharging and corrects the action in real time according to the contact force to achieve smooth control of the end of the discharging robot.

7. A terminal compliance control system for a discharge robot of a rare earth metal electrolytic furnace, characterized in that: include: Force sensing module, compliant controller, reinforcement learning training module; among them: The force sensing module is provided at the end of the discharging robot and is used to obtain the contact force information between the discharging tool and the electrolytic furnace in real time during the discharging process of the rare earth metal electrolytic furnace; The reinforcement learning training module is used to establish a rare earth discharging simulation environment for simulating a discharging scenario, and based on the simulation environment, construct a rare earth discharging control model for simulating the motion trajectory of the discharging tool through reinforcement learning training; The compliant controller is used to carry the rare earth discharging control model, and obtain the contact force information obtained by the force sensing module in real time as the input of the rare earth discharging control model, obtain the real-time discharging tool motion trajectory and output it to the discharging robot controller, so as to realize the compliant control of the end of the discharging robot.

8. The terminal compliance control system of the rare earth metal electrolysis furnace discharging robot according to claim 7, characterized in that: The force sensing module may further include any one or more of the following: Use six-dimensional force distance sensors, which are distributed on the end effector of the discharging robot; The EtherCAT bus is used for data acquisition, and the collected analog signals are converted into digital signals through transmitters; the data synchronization cycle is <2ms; Use heat-insulating material to install between the discharging tool.

9. A rare earth metal electrolysis furnace discharge robot, comprising: The discharging robot body is characterized in that the discharging robot body realizes flexible control of the end of the discharging robot through the method described in any one of claims 1-6 or the system described in any one of claims 7-8.

Images