A joint thermal perception gait cooperative control method of a robot dog in a high-temperature environment

Abstract

The application discloses a kind of high-temperature environment under robot dog joint thermal perception gait coordination control method, belong to robot control technical field, including the following steps: step 1, temperature sensing array construction, step 2, data acquisition and upload, step 3, maximum current calculation, step 4, current limiting instruction issue and execution, step 5, protection and mode switching, step 6, gait adjustment;Through joint level multi-point thermal perception and gait phase collaborative calculation, the thermal state difference of each joint of robot dog can be accurately covered, not only pay attention to motor body temperature, but also include the heat dissipation state of driver, and deeply couple gait phase torque demand, so that current limiting is more suitable for joint actual working condition, provide more accurate control basis for stable operation of robot dog under high temperature.

Description

Technical Field

[0001] This invention relates to the field of robot control technology, and more specifically, to a method for joint thermal sensing gait coordination control of a robot dog under high-temperature conditions. Background Technology

[0002] With the rapid development of automated inspection needs in high-temperature industrial scenarios such as electrolytic aluminum and steel smelting, quadruped robot dogs are gradually being introduced into high-risk areas with ambient temperatures ≥80℃ to perform tasks due to their excellent terrain adaptability. Robot dog control is essentially a motion planning and servo drive technology that enables the robot dog to maintain a stable gait in complex terrain by rapidly and collaboratively calculating the torque and position of multiple joints such as the hip, knee, and ankle. However, in high-temperature environments, the temperature rise of the joint motors and drivers can lead to problems such as encoder thermal drift and increased MOSFET on-resistance, resulting in degradation of control performance.

[0003] The prior art patent document with authorization announcement number CN116451557A discloses "a method and device for intelligent agent reinforcement learning for controlling the movement of a robot dog". During the execution process, the controlling intelligent agent receives the position and movement information of each component of the robot dog as input state, calculates and outputs the control parameters of each joint as action, and feeds back the corresponding reward value according to the movement behavior and forward distance of the robot dog. The training goal is to enable the robot dog to move forward quickly and smoothly and avoid falling.

[0004] The patent document with authorization announcement number CN120985647A discloses "a mechanical joint control method for a robot dog". This method acquires a real-time sensor data set of the whole body joints, including torque readings of multi-axis force sensors, angle data of joint encoders and body posture data of inertial measurement units. Based on a dynamic window mechanism, the sensor data is used to perform motion state compensation to generate a dynamic compensation matrix containing kinematic compensation parameters. The matrix is ​​then input into a gradient descent model under spatiotemporal constraints to complete the fusion of multimodal motion features and obtain a fused motion control dataset.

[0005] While existing technologies can improve the stability and smoothness of robot dog movement control through reinforcement learning, and achieve high-precision, adaptive joint control through multi-source sensor fusion, dynamic compensation, and virtual mapping, thus solving the problems of gait planning and joint control accuracy in the routine movement of robot dogs, they lack adaptive control strategies for the special working conditions of high-temperature industrial scenarios. They also lack precise temperature sensing mechanisms at the joint level and fail to dynamically couple joint thermal states with gait planning and joint torque control. This makes it impossible to solve problems such as sensor thermal drift and actuator performance degradation caused by high temperatures. Furthermore, they lack refined thermal management and graded protection mechanisms for the heterogeneous thermal states of multiple joints, which can easily lead to torque imbalance, gait instability, or even complete shutdown of the entire machine due to local joint overheating and derating. This makes it difficult to meet the continuous and stable operation requirements of robot dogs in high-temperature environments of 80°C and above. Summary of the Invention

[0006] This invention provides a method for joint thermal sensing gait coordination control of a robot dog under high temperature conditions, which can solve the problems mentioned in the background.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for joint thermal sensing gait coordination control of a robot dog under high temperature conditions, comprising: S1. Deploy miniature temperature sensors inside the motor housings and on the corresponding driver heat sinks of the hip, knee, and ankle joints of each leg of the robot dog to form a joint-level temperature sensing array. S2. Configure a local joint controller for each leg of the robot dog. Collect temperature data of each joint of the corresponding leg through the local joint controller, and upload the temperature data, joint ID and current gait phase information to the main motion controller through the bus communication module. S3. Based on the pre-calibrated joint temperature-maximum continuous torque mapping relationship and combined with the current gait phase weight function, the main motion controller calculates the maximum allowable current for each joint of the robot dog in real time. S4. The main motion controller sends the calculated maximum allowable current of each joint to the motor driver of the corresponding joint and uses it as the hard limit value of the motor driver current loop. S5. The main motion controller monitors the temperature of each joint in real time. When the temperature of any joint exceeds the first warning threshold, the main motion controller enters the thermal warning mode. When the temperature of any joint exceeds the second protection threshold, the main motion controller triggers the thermal protection mode, reduces the overall step frequency of the robot dog, and sends a thermal constraint signal to the high-level motion planning module. S6. The high-level motion planning module dynamically adjusts the robot dog's gait parameters based on the received thermal constraint signal, reducing the load on the legs where the high-temperature joints are located, thus forming a closed-loop collaborative control.

[0008] Furthermore, in S1, the miniature temperature sensor is a digital temperature sensor or a platinum resistance temperature sensor, with 1-2 miniature temperature sensors deployed inside the housing of each joint motor, and 1 miniature temperature sensor deployed on the heat sink of the driver.

[0009] Furthermore, in S2, the bus communication module uses a CANFD bus or an EtherCAT bus, and the local joint controller filters and verifies the validity of the collected temperature data before uploading it.

[0010] Furthermore, in S3, the joint temperature-maximum continuous torque mapping relationship is pre-calibrated through a high-temperature aging experiment and stored in the non-volatile memory of the main motion controller; the gait phase weight function is set with different values ​​according to whether the robot dog's joint is in the support phase or the swing phase.

[0011] Furthermore, in S4, after receiving the maximum allowable current, the motor driver strictly limits the output current to the maximum allowable current range when executing position or torque commands, and the motor driver feeds back the actual output current and voltage data to the main motion controller in real time.

[0012] Furthermore, in S5, the first warning threshold is 100℃ and the second protection threshold is 110℃; in the thermal protection mode, an emergency stop threshold is also set, which triggers the emergency braking of the whole machine when the joint temperature reaches the emergency stop threshold.

[0013] Furthermore, in S6, the gait parameters include stride length, torso height, and fuselage center of gravity position. The high-level motion planning module adjusts the gait parameters according to the position of the high-temperature joints to reduce the load proportion of the corresponding legs.

[0014] Furthermore, when switching from the thermal warning mode and thermal protection mode to the normal operation mode, temperature hysteresis conditions and duration judgment conditions are set. After the conditions are met, the main motion controller releases the thermal warning or thermal protection and restores the original control parameters of the robot dog.

[0015] The beneficial effects of this invention's joint thermal sensing and gait collaborative control method for robot dogs under high-temperature environments are as follows: By using multi-point thermal sensing at the joint level and collaborative calculation of gait phase, it can accurately cover the thermal state differences of each joint in the robot dog. It not only focuses on the temperature of the motor body but also incorporates the heat dissipation state of the driver and deeply couples the gait phase torque requirements, making the current limiting more consistent with the actual working conditions of the joint, providing a more accurate control basis for the stable operation of the robot dog under high temperatures. In addition, through hierarchical distributed acquisition and bus interaction, it realizes the effective binding of joint sensing data and gait phase information, breaking the data separation barrier, enabling each module to share thermal state and motion data in real time, improving the response speed and accuracy of high-temperature control, and enhancing the adaptability of torque limiting. At the same time, through thermal state tiered protection and dynamic gait adjustment, it provides a more refined thermal management method, which can clearly distinguish the joint thermal risk level and dynamically optimize gait parameters. Collaborative control based on real-time thermal data can quickly balance thermal safety and motion performance under high-temperature conditions, greatly improving the robot dog's operational continuity and anti-instability capability in high-temperature scenarios. Attached Figure Description

[0016] The present invention will now be described in further detail with reference to the accompanying drawings and specific implementation methods.

[0017] Figure 1 This is a schematic diagram of the collaborative control operation process of a robot dog joint thermal sensing gait collaborative control method under high temperature environment according to the present invention. Figure 2 This is a schematic diagram of the system structure of a robot dog joint thermal sensing gait cooperative control method under high temperature environment according to the present invention. Figure 3 This is a schematic diagram of the control timing interaction of a robot dog joint thermal sensing gait collaborative control method under high temperature environment according to the present invention. Figure 4 This is a schematic diagram of the joint thermal state machine transition in the gait coordination control method for robot dogs under high temperature environment according to the present invention. Detailed Implementation

[0018] To make the technical solution of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Example 1

[0019] like Figures 1-4 As shown, a technical solution is provided: a method for joint thermal sensing gait collaborative control of a robot dog under high temperature conditions, comprising: Step 1: Construction of temperature sensing array Miniature temperature sensors are deployed inside the motor housings and on the corresponding driver heat sinks of the hip, knee, and ankle joints of each leg of the robot dog to form a joint-level temperature sensing array. Specifically, the miniature temperature sensor uses a digital temperature sensor or a platinum resistance temperature sensor. One or two miniature temperature sensors are deployed inside the motor housing of each joint, and one miniature temperature sensor is deployed on the heat sink of the driver.

[0020] First, based on the structural dimensions of the hip, knee, and ankle joint motors of the robot dog, mounting holes for compatible miniature temperature sensors are made in the heat-sensitive area near the windings inside the motor housing. One or two digital temperature sensors or platinum resistance temperature sensors are embedded and fixed, with the sensor probes directly attached to the metal housing on the outside of the motor windings to ensure that the sensors can accurately capture the core temperature rise data when the motor is working. At the same time, a miniature temperature sensor of the same type is attached to the center of the MOSFET heat sink of the driver corresponding to each joint using thermally conductive adhesive to ensure that the sensor is in full contact with the heat sink and to collect the temperature of the driver's heat dissipation status in real time. Then, the wiring of all deployed miniature temperature sensors is arranged, and the sensor wiring on the motor housing and the driver is unified to the local joint controller wiring terminal of the corresponding leg. Shielded wires are used to complete the connection between the sensor and the local joint controller, and the shielding layer is grounded to avoid electromagnetic interference in high-temperature industrial scenarios from affecting the temperature acquisition accuracy. At the same time, each sensor is assigned a unique hardware identifier and bound to the corresponding joint ID to ensure that the collected temperature data can be accurately matched to the specific hip, knee, and ankle joints and their respective drivers. Finally, the completed joint-level temperature sensing array was powered on, debugged, and calibrated. Under a standard environment of 25°C, the initial temperature values ​​of each sensor were collected to verify the consistency and accuracy of the sensor data. Subsequently, by simulating a high-temperature environment, the acquisition accuracy of the sensors in the range of 80°C-125°C was verified. Sensors with deviations exceeding the threshold were replaced. The sampling frequency parameters were configured through the local joint controller to ensure that the sensing array can collect the temperature data of each joint motor and driver in real time and continuously, forming a complete and accurate joint-level temperature sensing array.

[0021] Step 2: Data Collection and Upload Each leg of the robot dog is equipped with a local joint controller (LJC). The temperature data of each joint in the corresponding leg is collected through the local joint controller, and the temperature data, joint ID and current gait phase information are uploaded to the main motion controller (MMC) through the bus communication module. Specifically, the bus communication module uses CANFD or EtherCAT bus, and the local joint controller filters and verifies the validity of the collected temperature data before uploading it.

[0022] First, the local joint controller for each leg is initialized and configured to match the communication protocol with the corresponding joint temperature sensor array. At the same time, the communication parameters of the CANFD bus or EtherCAT bus are configured, including baud rate, data frame format, and communication address. A bidirectional communication link is established between the local joint controller and the main motion controller, and the joint ID encoding library of the corresponding leg is bound to the local joint controller to ensure that the collected temperature data can be accurately associated with the specific hip, knee, and ankle joints. At the same time, the broadcast receiving channel of the local joint controller to the main motion controller for gait phase information is enabled to achieve real-time synchronization of gait phase information. Then, the local joint controller reads the raw temperature data of each sensor in the joint-level temperature sensor array one by one, matches the read temperature data with the pre-bound joint IDs to form a raw data group of "joint ID-temperature value", and at the same time captures the current gait phase information (support phase / swing phase) broadcast by the main motion controller in real time, adds gait phase tags with the same timestamp to the raw data group, and constructs a data packet to be processed containing joint ID, real-time temperature data and current gait phase information; In addition, the local joint controller performs filtering and validity verification on the constructed data packets to be processed. Specifically, it uses a moving average filtering algorithm to filter the temperature values, eliminating temperature data noise caused by electromagnetic interference in high-temperature industrial scenarios and improving data smoothness. At the same time, it sets verification rules according to the sensor range (e.g., 0-150℃) and data change threshold to determine whether the temperature value is within the effective range and whether the temperature data change in adjacent sampling periods exceeds 5℃. Invalid data that exceeds the rules is removed, and reasonable supplementation is performed based on the effective sampling data before and after to ensure the accuracy and validity of the data to be uploaded. Finally, the local joint controller encapsulates the filtered and valid data packets according to the standard frame structure of the CANFD or EtherCAT bus, mapping the joint ID, temperature data, and gait phase information to the specified fields of the data frame. The encapsulated data packets are then uploaded to the main motion controller in real time via the bus communication module. At the same time, the latest few sets of valid data packets are stored in the local buffer of the local joint controller. If the main motion controller loses data reception, it can resend the data according to the retransmission command of the main controller to ensure the integrity and real-time performance of data transmission.

[0023] Step 3: Calculation of maximum current The main motion controller calculates the maximum allowable current for each joint of the robot dog in real time based on the pre-calibrated joint temperature-maximum continuous torque mapping relationship and the current gait phase weight function. Specifically, the mapping relationship between joint temperature and maximum continuous torque is pre-calibrated through high-temperature aging experiments and stored in the non-volatile memory of the main motion controller; the gait phase weight function is set with different values ​​according to whether the robot dog's joint is in the support phase or the swing phase.

[0024] First, the main motion controller parses the valid data packets uploaded by each local joint controller frame by frame, accurately extracting the unique ID of each joint, real-time temperature data, and the bound current gait phase information. At the same time, it retrieves the joint temperature-maximum continuous torque mapping table pre-calibrated by high-temperature aging test from its own non-volatile memory, accurately matches the parsed real-time joint temperature with the mapping table, obtains the maximum continuous torque value of the corresponding joint at that temperature, and simultaneously retrieves the rated current basic parameters of the joint at a reference temperature of 25℃, completing the extraction and matching of all the basic data required for calculation. The assignment rules for the gait phase weight function and the devaluation coefficient are based on the current gait phase of the joint obtained from the analysis. If the joint is in the support phase, the gait phase weight function is assigned a specific value. Assign a value of 1.0, and a reduction coefficient. Assigning a value of 0.8, if the joint is in the swing phase, that is, the gait phase weighting function... Assign a value of 0.7, and a reduction coefficient. Assign a value of 0.5 and simultaneously determine the reference temperature. 25℃, the highest withstand temperature of the joint motor. For a fixed calculation parameter of 120℃, complete the precise assignment of all calculation coefficients and fixed parameters; Next, the main motion controller calls the built-in calculation program and substitutes the preset maximum allowable current calculation formula into it. The specific formula is as follows:

[0025] In the formula, This is the maximum allowable current for the joint. The rated current of the joint. This is the gait phase derating factor. This refers to the real-time temperature of the joint. For reference temperature, For the highest tolerable temperature, The gait phase weight function is used to perform floating-point calculations by substituting the matched joint rated current, the derating coefficient assigned by orientation, the gait phase weight function, the analytically extracted real-time joint temperature, the fixed reference temperature, and the maximum tolerance temperature into the formula. This completes the real-time calculation of the maximum allowable current for each joint. During the calculation process, the calculation results are range-checked to ensure that the results are within the current controllable range of the corresponding motor driver. Finally, the main motion controller re-binds the calculated maximum allowable current value of each joint with the corresponding joint ID to form a dedicated control instruction data packet of "joint ID - maximum allowable current". The control instruction data packets of all joints are classified, organized and numbered according to the partition order of the robot dog's four legs, which prepares for the precise issuance of instructions to the motor drivers of each joint in the future. At the same time, the calculated maximum allowable current data is associated with the real-time temperature and gait phase information of the corresponding joint and stored to form traceable operating data, which provides a basis for subsequent parameter optimization and aging test data updates.

[0026] Step 4: Issuance and execution of current limiting command The main motion controller sends the calculated maximum allowable current for each joint to the corresponding joint's motor driver, using it as the hard limit value for the motor driver's current loop. Specifically, after receiving the maximum allowable current, the motor driver strictly limits the output current to the maximum allowable current range when executing position or torque commands, and the motor driver feeds back the actual output current and voltage data to the main motion controller in real time.

[0027] First, the main motion controller will classify, organize, and number the dedicated control command data packets of "Joint ID - Maximum Allowable Current" and encapsulate them in a second manner according to the driver communication frame format of CANFD bus or EtherCAT bus. The joint ID will be mapped to the frame address segment and the maximum allowable current value will be mapped to the frame data segment. At the same time, the transmission priority of this type of current limiting command frame will be set higher than that of ordinary motion commands. The command will be accurately and directionally sent to the motor drivers corresponding to each joint of the robot dog through the bus communication module. A command reception confirmation mechanism will be set up. If no reception response signal is received from the driver within a preset time, the main motion controller will automatically resend the command until the driver successfully receives it. Then, after receiving the current limiting command frame, the motor driver parses and extracts the frame data to obtain the maximum allowable current value of its corresponding joint. This value is directly written into the hardware limiting register of the driver's current loop as the hard limiting value of the current loop, replacing the original fixed current limiting parameter. At the same time, the validity of the written hard limiting value is checked to determine whether it is within the current output range supported by the driver hardware. If it exceeds the range, the range threshold value is automatically taken as the actual limiting value and the abnormal information is fed back to the main motion controller. After the hard limiting value is configured, the driver normally executes the position or torque control command issued by the main motion controller. During the current loop adjustment process, the output current value is detected in real time. Once the current is detected to be about to reach the hard limiting value, the hardware current clamping mechanism is immediately triggered to forcibly limit the output current within the maximum allowable current range to avoid overcurrent output. Finally, the motor driver collects its actual output current and voltage data in real time at a sampling frequency synchronized with temperature acquisition. This data is bound to its corresponding joint ID and the currently effective maximum allowable current hard limit value to construct a feedback data packet of "joint ID-hard limit value-actual current-actual voltage". After being encapsulated according to the bus communication protocol, it is uploaded to the main motion controller in real time. After receiving the feedback data packets from all drivers, the main motion controller compares the actual output current with the issued maximum allowable current one by one to verify the actual execution effect of the current limiting command. If the actual current is found to exceed the hard limit value, a hard limit value reconfiguration command is immediately sent to the corresponding driver. At the same time, all issued and feedback data are associated and stored to form a full-process data traceability of the current limiting command from issuance to execution, realizing closed-loop control of command issuance-limit execution-data feedback.

[0028] Step 5: Protection and Mode Switching The main motion controller monitors the temperature of each joint in real time. When the temperature of any joint exceeds the first warning threshold, the main motion controller enters the thermal warning mode. When the temperature of any joint exceeds the second protection threshold, the main motion controller triggers the thermal protection mode, reduces the overall step frequency of the robot dog, and sends a thermal constraint signal to the high-level motion planning module. Specifically, the first warning threshold is 100℃, and the second protection threshold is 110℃; in the thermal protection mode, an emergency stop threshold is also set, which triggers the emergency braking of the whole machine when the joint temperature reaches the emergency stop threshold.

[0029] First, the main motion controller synchronizes with the frequency of joint temperature data acquisition and upload, and receives valid temperature data packets uploaded by each local joint controller in real time. After parsing the data, it constructs a real-time thermal status monitoring list containing all joint IDs, real-time temperature values, and gait phases. At the same time, it presets three-level temperature thresholds and completes parameter configuration in the main controller. For example, the first warning threshold is set to 100℃, the second protection threshold is set to 110℃, and the emergency stop threshold is set to 125℃. The main controller compares the real-time temperature of each joint in the monitoring list with the three-level thresholds one by one according to the preset frequency, accurately marks the joint ID, corresponding temperature value, and over-threshold type that exceed the threshold, and forms a thermal status abnormality record. Then, the main controller executes the corresponding mode switching and control actions according to the thermal state abnormality record. If the temperature of any joint is detected to be ≥100℃ and <110℃, it immediately enters the thermal warning mode, focuses on monitoring the high-temperature joint and marks the warning level in the thermal state record. If the temperature of any joint is detected to be ≥110℃ and <125℃, it immediately triggers the thermal protection mode. The built-in step frequency adjustment program reduces the overall step frequency of the robot dog by a preset ratio. At the same time, the high-temperature joint ID, temperature value and thermal protection level are encapsulated into a thermal constraint signal and sent to the high-level motion planning module in real time through the bus. If the temperature of any joint is detected to be ≥125℃, it immediately triggers the emergency stop mechanism, sends a stop command to all joint motor drivers, cuts off the power supply to the motors and triggers the joint locking protection to realize the emergency braking of the whole machine. Finally, in thermal warning and thermal protection modes, the main motion controller continuously monitors the temperature changes of the corresponding high-temperature joints and executes mode recovery operations according to preset temperature hysteresis and duration judgment conditions. For example, in thermal warning mode, the temperature of the corresponding joint must be monitored to be <95℃ and this state must last for 5 seconds; in thermal protection mode, the temperature of the corresponding joint must be monitored to be <90℃ and this state must last for 10 seconds. After the corresponding conditions are met, the main controller automatically releases the corresponding thermal warning or thermal protection mode, restores the overall gait frequency of the robot dog to the original set value, and stops sending thermal constraint signals to the high-level motion planning module. It also restores the original maximum allowable current calculation and amplitude limit rules of each joint. In addition, the main controller stores the complete data of this mode switching process, including information on joints exceeding the threshold, mode duration, temperature change curve, gait frequency adjustment parameters, etc., to form a thermal protection operation record, providing data support for subsequent threshold optimization and gait adjustment strategy improvement.

[0030] Step 6: Gait Adjustment The high-level motion planning module dynamically adjusts the robot dog's gait parameters based on the received thermal constraint signal, reducing the load on the legs where the high-temperature joints are located, thus forming a closed-loop collaborative control. Specifically, gait parameters include stride length, torso height, and fuselage center of gravity position. The high-level motion planning module adjusts gait parameters according to the location of high-temperature joints to reduce the load on the corresponding legs.

[0031] First, the high-level motion planning module receives the thermal constraint signal sent by the main motion controller in real time, parses and extracts the signal frame, and accurately obtains the unique ID of the high-temperature joint, the position of the leg to which it belongs, the thermal protection level and the current real-time temperature value of the joint. At the same time, it retrieves the gait parameter benchmark values ​​currently executed by the robot dog from the parameter library built into the module, including the benchmark stride, benchmark torso height, the original three-dimensional coordinates of the body center of gravity, and the calculation model of the load ratio of each leg matched with the current gait, and completes all the basic data preparation before gait adjustment. Then, the high-level motion planning module, based on the analyzed high-temperature joints belonging to the legs (left front leg / right front leg / left hind leg / right hind leg), combined with the load ratio calculation model, performs directional gait parameter adjustment operations. First, it reduces the robot dog's overall stride by a preset ratio to reduce the range of motion and torque output requirements of the leg joints. Then, it appropriately increases the torso height to reduce the bending angle of the high-temperature leg joints to reduce the load. Finally, based on the position of the high-temperature leg offsetting the robot's center of gravity, it moves the center of gravity horizontally towards the non-high-temperature leg direction to achieve a precise reduction in the load ratio of the high-temperature leg. All adjustment actions are performed based on the robot dog's motion stability model to ensure that the robot body remains balanced after parameter adjustment. In addition, the high-level motion planning module integrates the adjusted stride length, torso height, and new coordinates of the fuselage center of gravity into a new gait parameter set, encapsulates it according to the bus communication protocol, and sends it to the main motion controller in real time. After receiving it, the main motion controller updates the calculation benchmark of the gait phase synchronously, so that the new gait parameters are coordinated and matched with the maximum allowable current limiting strategy of each joint. At the same time, the high-level motion planning module monitors the actual load change of the high-temperature leg in real time through the torque data of each leg fed back by the main motion controller, and verifies the actual effect of the gait parameter adjustment. If the load reduction does not reach the preset target, a second fine adjustment is performed. Finally, the high-level motion planning module maintains continuous two-way data interaction with the main motion controller, receiving real-time data on high-temperature joint temperature changes and the load percentage of each leg uploaded by the main controller. When the high-temperature joint temperature drops to the corresponding hysteresis threshold and meets the duration judgment condition, the module automatically restores all gait parameters to the original baseline values ​​before adjustment when the main controller deactivates the thermal protection / thermal warning mode. At the same time, it associates and stores the entire process data of this gait adjustment, including parameter values ​​before and after adjustment, adjustment duration, high-temperature joint temperature change curve, and load change data of each leg, with the thermal protection operation record of the main motion controller, forming a complete closed-loop collaborative control of "thermal constraint signal transmission - gait parameter orientation adjustment - load monitoring - parameter recovery", realizing the adaptation of thermal state and gait movement.

[0032] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for joint thermal sensing gait collaborative control of a robot dog under high-temperature conditions, characterized in that: S1. Deploy miniature temperature sensors inside the motor housings and on the corresponding driver heat sinks of the hip, knee, and ankle joints of each leg of the robot dog to form a joint-level temperature sensing array. S2. Configure a local joint controller for each leg of the robot dog. Collect temperature data of each joint of the corresponding leg through the local joint controller, and upload the temperature data, joint ID and current gait phase information to the main motion controller through the bus communication module. S3. Based on the pre-calibrated joint temperature-maximum continuous torque mapping relationship and combined with the current gait phase weight function, the main motion controller calculates the maximum allowable current for each joint of the robot dog in real time. S4. The main motion controller sends the calculated maximum allowable current of each joint to the motor driver of the corresponding joint and uses it as the hard limit value of the motor driver current loop. S5. The main motion controller monitors the temperature of each joint in real time. When the temperature of any joint exceeds the first warning threshold, the main motion controller enters the thermal warning mode. When the temperature of any joint exceeds the second protection threshold, the main motion controller triggers the thermal protection mode, reduces the overall step frequency of the robot dog, and sends a thermal constraint signal to the high-level motion planning module. S6. The high-level motion planning module dynamically adjusts the robot dog's gait parameters based on the received thermal constraint signal, reducing the load on the legs where the high-temperature joints are located, thus forming a closed-loop collaborative control.

2. The method for joint thermal sensing gait coordination control of a robot dog under high temperature environment according to claim 1, characterized in that: In S1, the miniature temperature sensor is a digital temperature sensor or a platinum resistance temperature sensor. One or two miniature temperature sensors are deployed inside the housing of each joint motor, and one miniature temperature sensor is deployed on the heat sink of the driver.

3. The method for joint thermal sensing gait coordination control of a robot dog under high temperature conditions according to claim 1, characterized in that: In S2, the bus communication module uses CANFD bus or EtherCAT bus, and the local joint controller filters and verifies the validity of the collected temperature data before uploading it.

4. The method for joint thermal sensing gait coordination control of a robot dog under high temperature environment according to claim 1, characterized in that: In S3, the joint temperature-maximum continuous torque mapping relationship is pre-calibrated through a high-temperature aging experiment and stored in the non-volatile memory of the main motion controller; The gait phase weighting function is set with different values ​​depending on whether the robot dog's joint is in the support phase or the swing phase.

5. The method for joint thermal sensing gait coordination control of a robot dog under high temperature environment according to claim 1, characterized in that: In S4, after receiving the maximum allowable current, the motor driver strictly limits the output current within the maximum allowable current range when executing position or torque commands, and the motor driver feeds back the actual output current and voltage data to the main motion controller in real time.

6. The method for joint thermal sensing gait coordination control of a robot dog under high temperature environment according to claim 1, characterized in that: In S5, the first warning threshold is 100℃ and the second protection threshold is 110℃; in the thermal protection mode, an emergency stop threshold is also set, which triggers the emergency braking of the whole machine when the joint temperature reaches the emergency stop threshold.

7. The method for joint thermal sensing gait coordination control of a robot dog under high temperature environment according to claim 1, characterized in that: In S6, gait parameters include stride length, torso height, and fuselage center of gravity position. The high-level motion planning module adjusts the gait parameters according to the position of the high-temperature joints to reduce the load on the corresponding legs.

8. The method for joint thermal sensing gait coordination control of a robot dog under high temperature environment according to claim 1, characterized in that: When switching from the thermal warning mode and thermal protection mode to the normal operation mode, temperature hysteresis conditions and duration judgment conditions are set. After the conditions are met, the main motion controller releases the thermal warning or thermal protection and restores the original control parameters of the robot dog.

Images