Dog-arm workspace expansion method based on posture linkage

Abstract

The application relates to a dog-arm workspace expansion method based on posture linkage, which is as follows: taking a robot dog body as an adjustable base in a mechanical arm motion chain, and taking body pitching, rolling, height, front-back micro-displacement and each joint variable of the mechanical arm as joint configuration variables; after receiving a target six-dimensional position and posture output by a target detection module and operation constraints, it is judged whether the target is located in a single reachable space of the mechanical arm; if the target is located in the single reachable space of the mechanical arm, a conventional mechanical arm inverse kinematics is solved; if the target is out of the single reachable space of the mechanical arm, posture linkage solving is started to compensate for insufficient mechanical arm workspace through robot dog body squatting, pitching, rolling or micro-displacement; during the solving process, stability constraints such as support polygon, body inclination, foot end contact margin and mechanical arm load limitation are introduced; during the execution process, target errors are closed-loop corrected through IMU and visual servoing. Through dog-arm joint kinematics modeling, posture linkage inverse solution, stability constraint screening and closed-loop execution correction, the application converts robot dog body posture capability into part of the mechanical arm workspace, and improves the reachability and inspection operation capability of a light mobile mechanical arm under the premise of ensuring stability and safety.

Description

Technical Field

[0001] This invention relates to the fields of robot motion control, inspection robots, mobile robotic arms, and robot motion planning technology, specifically to a method for expanding the workspace of a dog arm based on posture linkage. Background Technology

[0002] With the development of mobile robots and industrial inspection technology, robot dogs, due to their strong terrain adaptability, can perform inspection tasks in complex environments such as stairs, ramps, gravel, gratings, and narrow passages. To further improve the operational capabilities of inspection robots, existing solutions typically install a lightweight robotic arm on the back or top of the robot dog, enabling it to touch buttons, turn valves, pull switches, and approach instruments to collect images.

[0003] However, the use of robotic arms in inspection robots is limited by overall weight, center of gravity height, battery life, and dynamic stability. To ensure the robot's walking and standing stability, the robotic arm typically needs to be lightweight, have a low load capacity, and a short reach. These robotic arms are relatively inexpensive and lightweight, making them suitable for mobile platforms, but their workspace is limited. When the target is located near the top edge of a cabinet, high-level instruments, valve tops, or low-level sensors in a pit, the robotic arm often struggles to reach the target's position using only its joint movements.

[0004] In existing technologies, common methods to extend the reach of robotic arms include replacing them with robotic arms with longer arms, adding lifting mechanisms, configuring additional mobile chassis for the robot to adjust its position, or manually redeploying the robot before operation. These methods have the following drawbacks:

[0005] 1. Increased robotic arm size leads to increased overall machine load: While lengthening or adding weight to the robotic arm can expand the workspace, it will significantly increase the load on the robot dog's back and the eccentric torque, reduce its battery life, and increase the risk of imbalance during standing and walking.

[0006] 2. Adding lifting or telescopic mechanisms increases structural complexity: Adding lifting platforms, telescopic columns, and other mechanisms to the robot dog will increase equipment costs, installation space, and control complexity, and may also cause collision risks in narrow inspection environments.

[0007] 3. Low efficiency in adjusting the position of the mobile chassis: For targets with varying heights or obstructions, relying solely on the robot dog's overall movement to adjust the position is insufficient to accurately compensate for the attitude error at the end of the robotic arm, and in situations where ground space is limited, no effective adjustment margin can be obtained.

[0008] 4. Insufficient utilization of the robot dog's posture capabilities: The robot dog possesses pitch, roll, rise, squat, leg-raising support, and small-range micro-displacement capabilities. These capabilities can essentially change the position and orientation of the robotic arm base. Existing robotic arm control methods mostly treat the robot dog body as a static rigid body or a fixed base, failing to incorporate the robot dog's posture changes as additional degrees of freedom for the robotic arm in a unified solution.

[0009] 5. Lack of unified constraints on accessibility and stability: If the robotic arm is simply assisted in reaching the target by pitching or crouching the body, the center of gravity margin within the supporting polygon may be reduced, or the foot contact margin may be insufficient, resulting in instability, slippage, or overloading of the robotic arm.

[0010] Therefore, there is a need for a workspace expansion method that can fully utilize the robot dog's posture capabilities and solve the dog-arm cooperative motion problem under stability constraints without replacing the heavy robotic arm or adding complex external mechanisms. Summary of the Invention

[0011] This invention addresses the technical problems existing in the prior art by providing a method for expanding the workspace of a robot arm based on posture linkage. This method integrates the posture adjustment capabilities of the robot dog body, such as pitch, roll, rise, squat, and forward and backward micro-displacement, with the joint motion of the lightweight robotic arm mounted on the robot dog into a joint kinematics chain. Through target pose analysis, joint inverse kinematics solution of the robot arm, stability constraint screening, and sensor closed-loop correction during execution, the lightweight robotic arm can achieve a larger effective working range without significantly increasing arm length, weight, and power consumption.

[0012] This invention is applicable to inspection scenarios in power, petrochemical, rail transit, integrated utility tunnels, data centers and industrial plants, and is especially suitable for robot dogs carrying lightweight robotic arms to complete tasks such as reading high-level instruments, operating switches on the upper edge of cabinets, touching the top of valves, detecting low-level sensors, and pressing buttons in narrow spaces.

[0013] To achieve the above objectives, the technical solution of this invention is as follows: a method for expanding the workspace of a robot arm based on posture linkage. This method uses the robot dog itself as an adjustable base in the robot arm's kinematic chain, and combines the robot's pitch, roll, height, and forward / backward micro-displacements with the joint variables of the robot arm to form a joint configuration. After receiving the target's six-dimensional pose and operational constraints output by the target detection module, it first determines whether the target is within the robot arm's individually reachable space. If the robot arm is individually reachable, conventional inverse kinematics of the robot arm is used for solution. If the target exceeds the robot arm's individually reachable space, posture linkage solution is initiated, using the robot dog's crouching, pitching, rolling, or micro-displacement to compensate for insufficient workspace. Stability constraints such as support polygons, body tilt angles, foot contact margins, and robot arm load limits are introduced during the solution process. During execution, closed-loop correction of target errors is performed using an IMU and visual servoing.

[0014] The method of this invention includes a target pose analysis module, a dog-arm joint kinematics modeling module, an attitude linkage inverse kinematics module, a stability constraint screening module, and a closed-loop execution correction module. These modules are sequentially connected according to the task flow: the target pose analysis module provides the target pose and operational constraints; the dog-arm joint kinematics modeling module maps the target pose to a unified coordinate system; the attitude linkage inverse kinematics module generates candidate dog-arm joint configurations; the stability constraint screening module eliminates candidate solutions that do not meet safety conditions; and the closed-loop execution correction module iteratively compensates for execution errors based on sensor feedback.

[0015] Target pose analysis module:

[0016] The target pose resolution module receives the target's six-dimensional pose, including target position and orientation, from visual detection, point cloud recognition, or manual annotation modules. For different objects such as buttons, instruments, valves, and switches, the target pose resolution module further generates operational constraints, such as the end effector's approach direction, allowable touch angle, operating depth, end effector velocity limit, and obstacle avoidance boundaries around the target. This information is uniformly transformed into the world coordinate system or the robot dog's standing reference coordinate system as input for dog-arm joint kinematics modeling.

[0017] Dog-arm joint kinematic modeling module:

[0018] The dog-arm joint kinematics modeling module treats the robot dog's body posture as part of the robotic arm base transformation. The robot dog's body posture includes the body pitch angle, body roll angle, body height, and forward / backward micro-displacement, while the robotic arm's posture includes the joint angles. By sequentially multiplying the robot dog's body pose transformation, the robotic arm mounting pose transformation, and the robotic arm's forward kinematics transformation, the end effector's pose in a unified coordinate system is obtained.

[0019] The positive kinematic relationship of the canine arm joint is expressed as:

[0020]

[0021] in, This represents the homogeneous pose matrix of the end effector in the world coordinate system. This indicates the posture parameters of the robot dog itself. The homogeneous transformation matrix from the defined fuselage coordinate system to the world coordinate system. ,in Indicates the aircraft's pitch angle. Indicates the fuselage roll angle. Indicates fuselage height. This indicates the amount of slight forward and backward displacement of the fuselage; This represents the fixed installation transformation matrix of the robotic arm mounting base relative to the robot dog's body; Represents the variables of the robotic arm joints The defined forward kinematics transformation matrix of the robotic arm. ,in Indicates the first One robotic arm joint variable, Indicates the number of joints in the robotic arm; This represents the reference homogeneous pose in the end-effector coordinate system.

[0022] Through the above modeling, the robot dog's pose is no longer merely a state of the mobile platform, but rather an additional degree of freedom affecting the pose of the robotic arm's end effector. The target pose output by the target pose resolution module is denoted as... The attitude linkage inverse kinematics module adjusts... and ,make as close as possible .

[0023] Attitude linkage inverse kinematics module:

[0024] The attitude linkage inverse kinematics module is used to solve for the dog arm joint configuration that meets the target pose requirements. This module first attempts to optimize only the joint variables of the robotic arm. If the robotic arm can reach the target area independently and meets the constraints of joint limits, obstacle avoidance, and operation direction, then the control quantity of the robotic arm is directly output; if the robotic arm cannot reach the target area independently or the posture error exceeds the threshold, then the posture parameters of the robot dog body are output. Include optimization variables.

[0025] The goal of attitude linkage inverse kinematics is to minimize changes in the robot dog's overall posture and drastic changes in the robotic arm joints while meeting the end-effector pose error requirements. Its optimization objective is expressed as:

[0026]

[0027] in, This represents the objective function for joint inverse solution optimization. This represents the current position vector of the end effector; Represents the target position vector; This represents the current attitude vector of the end effector; Represents the target attitude vector; Indicates the current or default posture parameters of the robot dog; Indicates the current or default joint variable of the robotic arm; Indicates the position error weight. Indicates the attitude error weights. This represents the penalty weight for changes in the robot dog's posture. This indicates the penalty weight for changes in the robotic arm joints. This represents the L2 norm.

[0028] In this objective function, the position error term and attitude error term ensure that the end effector approaches the target pose, the body attitude change penalty term is used to avoid excessive pitch, roll, or height changes in the robot dog, and the joint change penalty term is used to avoid unnecessary large swings of the robotic arm. In this way, the system prioritizes utilizing the existing reachability of the robotic arm, and then gradually introduces crouching, pitch, roll, or micro-displacement compensation when the robotic arm's capabilities are insufficient.

[0029] Stability constraint filtering module:

[0030] The stability constraint screening module performs safety screening on the candidate solutions generated by the attitude linkage inverse kinematics module. While incorporating the robot dog's own attitude into the solution can expand the effective workspace of the robotic arm, it may also reduce the robot dog's standing stability. Therefore, this invention simultaneously constrains the centroid projection margin, fuselage tilt angle range, foot contact margin, and robotic arm load torque within the supporting polygon during the joint inverse kinematics process.

[0031] The candidate dog-arm joint configuration must meet the following stability constraints:

[0032]

[0033] in, This represents the minimum distance from the robot dog's center of gravity projected onto the boundary of the supporting polygon, given the current robot dog's body posture and robotic arm joint configuration. This represents the minimum allowable stability margin; Indicates the maximum allowable pitch angle of the fuselage; Indicates the maximum allowable roll angle of the fuselage; This indicates the joint configuration of the robotic arm. and end equivalent load The maximum joint torque under action; This represents the equivalent load vector formed by the reaction forces of the end-effector, the manipulated object, and the clamping load. This indicates the maximum allowable joint torque of the robotic arm.

[0034] When a candidate solution does not meet the above constraints, the stability constraint screening module discards it and requires the attitude linkage inverse solution module to reduce the amplitude of the fuselage attitude change, adjust the robot arm attitude, or change the direction of the micro-displacement to solve it again. If there are multiple candidate solutions that meet the constraints, the solution with a larger stability margin, smaller fuselage attitude change, and smoother robot arm joint change is selected first.

[0035] Closed-loop execution correction module:

[0036] The closed-loop execution correction module is used to compensate for modeling errors, ground contact variations, target detection errors, and robotic arm end effector deviations during actual execution. When the robot dog performs attitude adjustments, the IMU provides real-time feedback on changes in the body's pitch, roll, and altitude, while the visual servo module continuously tracks the target pose and the end effector pose. When the end effector pose error exceeds the allowable range, the closed-loop execution correction module feeds back the correction amount to the attitude linkage inverse kinematics module, recalculating small-amplitude attitude compensation and joint compensation to gradually converge the end effector to the target pose.

[0037] In terms of execution sequence, the system first controls the robot dog to enter a stable standing state, then performs low-frequency attitude adjustments such as crouching, pitching, or micro-displacement of the robot body, and then controls the robotic arm to complete end-effector approach and fine operation. In contact operations such as valves, buttons, or switches, the system reduces the speed of the robotic arm before end-effector contact and judges the contact state based on torque or current changes to avoid excessive impact caused by the superposition of changes in robot body attitude and robotic arm movements.

[0038] Compared with the prior art, the present invention has the following advantages: 1. Effectively expands the working range of the lightweight robotic arm: The present invention incorporates the robot dog's body posture into the kinematics chain of the robotic arm, so that the body pitch, roll, rise, squat and micro displacement become part of the position and posture solution of the robotic arm end effector, which can expand the effective working space without replacing the long arm or heavy arm.

[0039] 2. Reduce overall weight and energy consumption: By utilizing the robot dog's existing posture capabilities to compensate for the insufficient arm span of the robotic arm, a lightweight robotic arm can continue to be used, reducing the overall load, joint power consumption, and battery consumption, making it suitable for long-term inspection tasks.

[0040] 3. Balancing accessibility and stability: This invention introduces stability constraints such as the supporting polygon, fuselage tilt angle, and robotic arm load during the joint inverse kinematics solution process, avoiding the robot dog's instability, foot slippage, or joint torque exceeding limits due to simply pursuing end-effector accessibility.

[0041] 4. Improve adaptability to complex inspection scenarios: For targets that are difficult to reach by a robotic arm alone, such as high-level instruments, switches on the upper edge of cabinets, valve tops, and low-level sensors in pits, this invention can obtain a more suitable base position and orientation through the posture linkage of the robot dog itself, thereby improving the success rate of inspection operations.

[0042] 5. The execution process has closed-loop correction capability: By providing real-time feedback on the body posture, target posture and end effector posture through IMU and visual servo, the present invention can correct target detection error, body posture error and robotic arm end effector error, thereby improving the accuracy of actual operation.

[0043] 6. Low modification cost and easy deployment: This invention mainly expands the workspace through kinematic modeling, control algorithms and sensor closed loops, without the need to add a large lifting platform or complex external mechanism, and is suitable for deployment on existing robot dog plus lightweight robotic arm platform.

[0044] In summary, this invention transforms the robot dog's posture capabilities into part of the robotic arm's workspace through dog-arm joint kinematic modeling, posture linkage inverse kinematics, stability constraint screening, and closed-loop execution correction. This improves the accessibility and inspection capabilities of the lightweight mobile robotic arm while ensuring stability and safety. Attached Figure Description

[0045] Figure 1 This is the overall flowchart of the present invention.

[0046] Figure 2 This is a schematic diagram of the system structure. Detailed Implementation

[0047] To enhance understanding of the present invention, the embodiments will be described in detail below with reference to the accompanying drawings.

[0048] Example 1: See Figure 1 , Figure 2 The following description uses a scenario where a robotic dog carrying a six-DOF lightweight robotic arm performs operations on the upper edge switch of a cabinet to illustrate a specific embodiment of the present invention. This embodiment is only used to illustrate the method flow of the present invention and does not limit the application of the present invention to other inspection targets.

[0049] The system includes a robot dog body, a lightweight robotic arm mounted on the robot dog's back, an end effector, a vision detection unit, an IMU, a foot contact detection unit, and a control and computing unit. The robot dog body can perform movements such as pitch, roll, rise, squat, and micro-displacement forward and backward; the lightweight robotic arm is used to perform operations such as approaching, pressing, flicking, or twisting; the vision detection unit is used to identify the target's position and posture; the IMU is used to measure the robot's posture; the foot contact detection unit is used to determine the support state of each foot; the control and computing unit performs target pose analysis, dog-arm joint kinematic modeling, posture linkage inverse kinematics solution, stability constraint screening, and closed-loop execution correction.

[0050] First, the vision detection unit identifies the switches along the top edge of the cabinet in the inspection scenario, outputting the switch's center point position, normal direction, and permissible approach direction. The target pose parsing module then converts this information into the target pose. It also generates operational constraints, including that the end effector should approach along the normal direction of the switch, the approach speed should not exceed the preset speed, and the end attitude deviation should not exceed the preset angle.

[0051] Subsequently, the dog-arm joint kinematic modeling module reads the current body posture of the robot dog and the current joint angles of the robotic arm. The robot dog's current posture is then formed. The current joint angle of the robotic arm is formed The control calculation unit determines the position of the robotic arm based on its mounting posture. Establish a joint kinematics chain with the structural parameters of the robotic arm, and apply the formula...

[0052]

[0053] Calculate the current pose of the end effector in the world coordinate system.

[0054] In the initial solution phase, the system first fixes the robot dog's posture and solves only the robot arm joint variables. If the end effector cannot reach the switch position at this time, or the end effector posture cannot meet the approach direction requirements, the posture linkage inverse kinematics module initiates a joint solution for the robot dog and arm. For the switch on the upper edge of the cabinet, the system first attempts to moderately raise the robot's height or adjust the robot's pitch angle to move the robot arm mounting base towards the target direction; if the target is located far in front of the robot, further micro-displacements are introduced. These adjustments are not independent actions, but are solved together with the robot arm joint variables in the same optimization objective function.

[0055] The attitude linkage inverse kinematics module is based on the formula

[0056]

[0057] Candidate configurations are generated. During implementation, the position error weight and attitude error weight are set according to the task type. For switch-pressing tasks, the position error weight is higher to ensure the end effector accurately contacts the switch center; for instrument-approaching-shooting tasks, the attitude error weight is higher to ensure the camera or end effector's observation direction meets the acquisition requirements. In this embodiment, the robot dog's body attitude change penalty weight is used to limit excessive body pitch, and the robotic arm joint change penalty weight is used to reduce joint abrupt changes, making the operation process smoother.

[0058] After candidate configurations are generated, the stability constraint screening module filters them. The control calculation unit forms a support polygon based on the robot dog's foot contact points, estimates the projected position of the robot's center of gravity by combining the robot dog's body posture, robotic arm posture, and end effector load, and calculates the minimum distance from the center of gravity projection to the boundary of the support polygon. Simultaneously, the pitch and roll angles measured by the IMU must not exceed preset upper limits, and the joint torques of the robotic arm must not exceed permissible values. The screening process is based on the formula...

[0059]

[0060] The process continues. If a candidate configuration, while enabling the robotic arm's end effector to reach the target, results in an excessively large pitch angle or insufficient stability margin, that candidate configuration is eliminated. The system then either reduces the amplitude of the robotic arm's attitude change or slightly increases the extension of the robotic arm's joints until a combined configuration that meets the target's pose and stability requirements is obtained.

[0061] After obtaining the final joint configuration, the closed-loop execution correction module executes in the following order: first the robot body, then the robotic arm, and finally the end effector fine-tuning. The robot dog first enters a stable standing state and performs crouching, rising, falling, pitching, or forward and backward micro-displacements according to the calculation results. When the IMU feedback indicates that the robot body posture has reached the target range, the robotic arm moves to the vicinity of the switch according to the planned trajectory. The visual servo module detects the deviation between the switch center point and the end effector in real time and feeds the deviation back to the posture linkage inverse kinematics module for minor correction. If the deviation mainly comes from target recognition error, the end effector posture of the robotic arm is adjusted first; if the deviation comes from the robot body posture not meeting expectations, minor compensation is simultaneously performed on the robot dog's body posture and the robotic arm joints.

[0062] Once the end effector enters the switch contact range, the system reduces the end effector speed and determines whether it has touched the target based on the end effector torque, motor current, or contact sensor signal. After completing the switch operation, the robotic arm first retracts to a safe posture, and then the robot dog body returns to its default standing posture. Through this execution process, the robot dog body posture compensation and the robotic arm end effector operation form a continuous closed loop, avoiding the sudden changes and instability risks caused by the robotic arm operation first and the body compensation later.

[0063] For high-level instrument reading scenarios, the implementation method is similar to the aforementioned switch operation, the difference being that the target pose constraint focuses on the observation direction of the end-effector camera or sensor. The system can improve the observation angle of the end-effector camera by raising the body height and adjusting the tilt. For low-level sensor detection scenarios in pits, the system can lower the height of the robotic arm base by crouching and tilting the body, allowing the lightweight robotic arm to reach lower positions. All of the above scenarios employ the same target pose analysis, dog-arm joint kinematic modeling, attitude linkage inverse kinematics, stability constraint screening, and closed-loop execution correction process, without introducing additional independent control logic.

[0064] It should be noted that the above embodiments are not intended to limit the scope of protection of the present invention. Equivalent transformations or substitutions made based on the above technical solutions all fall within the scope of protection of the claims of the present invention.

Claims

1. A method for expanding the workspace of a dog arm based on posture linkage, characterized in that, The method is as follows: The robot dog body is used as an adjustable base in the kinematic chain of the robotic arm. The pitch, roll, height, and forward / backward micro-displacement of the robot body are combined with the joint variables of the robotic arm to form a joint configuration quantity. After receiving the target's six-dimensional pose and operation constraints output by the target detection module, it is first determined whether the target is located in the space that the robotic arm can reach alone. If the robotic arm can reach it alone, the conventional inverse kinematics of the robotic arm is used to solve the problem. If the target exceeds the space that the robotic arm can reach alone, the attitude linkage solution is initiated, and the robot dog body squats, pitches, rolls, or micro-displaces to compensate for the insufficient working space of the robotic arm. Stability constraints such as supporting polygons, fuselage tilt angle, foot contact margin, and robotic arm load limit are introduced during the solution process; during execution, closed-loop correction of target error is performed through IMU and visual servo.

2. The method for expanding the workspace of a dog arm based on posture linkage according to claim 1, characterized in that, This can be achieved through the following modules. The system consists of a target pose analysis module, a dog-arm joint kinematics modeling module, an attitude linkage inverse kinematics module, a stability constraint screening module, and a closed-loop execution correction module. These modules are connected in the order of the task flow: the target pose analysis module provides the target pose and operation constraints; the dog-arm joint kinematics modeling module maps the target pose to a unified coordinate system; the attitude linkage inverse kinematics module generates candidate dog-arm joint configurations; the stability constraint screening module eliminates candidate solutions that do not meet safety conditions; and the closed-loop execution correction module iteratively compensates for execution errors based on sensor feedback.

3. The method for expanding the workspace of a dog arm based on posture linkage according to claim 2, characterized in that, The target pose resolution module receives the target's six-dimensional pose, including the target position and target posture, from the visual detection, point cloud recognition, or manual annotation modules. For different work objects such as buttons, instruments, valves, and switches, the target pose resolution module further generates operational constraints, such as the end effector approach direction, allowable touch angle, operation depth, end effector speed limit, and obstacle avoidance boundaries around the target. The above information is uniformly converted into the world coordinate system or the robot dog's standing reference coordinate system as input for dog-arm joint kinematics modeling.

4. The method for expanding the workspace of a dog arm based on posture linkage according to claim 3, characterized in that, The dog-arm joint kinematics modeling module treats the robot dog's body posture as part of the robotic arm base transformation. The robot dog's body posture includes the body pitch angle, body roll angle, body height, and forward / backward micro-displacement. The robotic arm part includes the joint angles. By sequentially multiplying the robot dog's body pose transformation, the robotic arm mounting pose transformation, and the robotic arm's forward kinematics transformation, the pose of the end effector in a unified coordinate system is obtained. The positive kinematic relationship of the canine arm joint is expressed as: in, This represents the homogeneous pose matrix of the end effector in the world coordinate system. This indicates the posture parameters of the robot dog itself. The homogeneous transformation matrix from the defined fuselage coordinate system to the world coordinate system. ,in Indicates the aircraft's pitch angle. Indicates the fuselage roll angle. Indicates fuselage height. This indicates the amount of slight forward and backward displacement of the fuselage; This represents the fixed installation transformation matrix of the robotic arm mounting base relative to the robot dog's body; Represents the variables of the robotic arm joints The defined forward kinematics transformation matrix of the robotic arm. ,in Indicates the first One robotic arm joint variable, Indicates the number of joints in the robotic arm; This represents the baseline homogeneous pose in the end-tool coordinate system. Through the above modeling, the robot dog's pose is no longer merely a state of the mobile platform, but rather an additional degree of freedom affecting the pose of the robotic arm's end effector. The target pose output by the target pose resolution module is denoted as... The attitude linkage inverse kinematics module adjusts... and ,make as close as possible .

5. The method for expanding the workspace of a dog arm based on posture linkage according to claim 4, characterized in that, The attitude linkage inverse kinematics module is used to solve for the dog arm joint configuration that can meet the target pose requirements. This module first attempts to optimize only the joint variables of the robotic arm. If the robotic arm can reach the target area independently and meets the constraints of joint limits, obstacle avoidance, and operation direction, then the control quantity of the robotic arm is directly output; if the robotic arm cannot reach the target area independently or the posture error exceeds the threshold, then the posture parameters of the robot dog body are output. Incorporate optimization variables, The goal of attitude linkage inverse kinematics is to minimize changes in the robot dog's posture and drastic changes in the robotic arm joints while meeting the end-effector pose error requirements. Its optimization objective is expressed as: in, This represents the objective function for joint inverse solution optimization. This represents the current position vector of the end effector; Represents the target position vector; This represents the current attitude vector of the end effector; Represents the target attitude vector; Indicates the current or default posture parameters of the robot dog; Indicates the current or default joint variable of the robotic arm; Indicates the position error weight. Indicates the attitude error weights. This represents the penalty weight for changes in the robot dog's posture. This indicates the penalty weight for changes in the robotic arm joints. This represents the L2 norm.

6. The method for expanding the workspace of a dog arm based on posture linkage according to claim 5, characterized in that, The stability constraint screening module performs a safety screening of the candidate solutions generated by the attitude linkage inverse solution module. The candidate dog-arm joint configuration must meet the following stability constraints: in, This represents the minimum distance from the robot dog's center of gravity projected onto the boundary of the supporting polygon, given the current robot dog's body posture and robotic arm joint configuration. This represents the minimum allowable stability margin; Indicates the maximum allowable pitch angle of the fuselage; Indicates the maximum allowable roll angle of the fuselage; This indicates the joint configuration of the robotic arm. and end equivalent load Maximum joint torque under action; This represents the equivalent load vector formed by the reaction forces of the end-effector, the manipulated object, and the clamping load. This indicates the maximum allowable joint torque of the robotic arm.

7. The method for expanding the workspace of a dog arm based on posture linkage according to claim 6, characterized in that, The closed-loop execution correction module is used to compensate for modeling errors, ground contact changes, target detection errors, and robotic arm end effector deviations during actual execution. When the robot dog performs attitude adjustments, the IMU provides real-time feedback on changes in the body's pitch, roll, and altitude angles. The visual servo module continuously tracks the target pose and the end effector pose. When the end effector pose error exceeds the allowable range, the closed-loop execution correction module feeds back the correction amount to the attitude linkage inverse kinematics module, recalculating small-amplitude attitude compensation and joint compensation, so that the end effector gradually converges to the target pose. In terms of execution sequence, the system first controls the robot dog to enter a stable standing state, then performs low-frequency posture adjustments such as crouching, pitching, or micro-displacement of the robot body, and then controls the robotic arm to complete end-effector approach and fine operation. In contact operations such as valves, buttons, or switches, the system reduces the speed of the robotic arm before end-effector contact and judges the contact state based on torque or current changes to avoid excessive impact caused by the superposition of robot body posture changes and robotic arm movements.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the dog arm workspace expansion method based on posture linkage as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing computer instructions thereon, characterized in that, When executed by the processor, the computer instructions implement the dog arm workspace expansion method based on posture linkage as described in any one of claims 1-7.

Images