An electric follow-up robot, a stiffness compensation control system and method

Abstract

The application relates to the technical field of electric follow-up manipulators, in particular to an electric follow-up manipulator, a stiffness compensation control system and a method. The electric follow-up manipulator comprises a mechanical arm, and the angle transmission mechanism at at least one joint of the mechanical arm is a spiral bevel gear transmission structure. According to the particularity of the electric follow-up manipulator, the angle transmission mechanism at the joint of the mechanical arm is composed of a spiral bevel gear pair with long contact line, many pairs of meshing teeth, stable transmission, low noise, strong bearing capacity, so that the time-varying meshing stiffness at the joint of the above-mentioned electric follow-up manipulator is enhanced based on the high contact coincidence degree of the spiral bevel gear, the fluctuation of the time-varying meshing stiffness is reduced, and the vibration phenomenon is reduced, so that the operation precision is ensured.

Description

Technical Field

[0001] This invention relates to the field of electric follower manipulator technology, specifically an electric follower manipulator, a stiffness compensation control system, and a method. Background Technology

[0002] Electric servo manipulators are manipulators used in the nuclear industry. They rely on gears to drive the corresponding joints of the manipulator, that is, the joints of the manipulator have angular transmission mechanisms composed of gears.

[0003] Current electrically driven manipulators mostly use spur bevel gears to form angular transmission mechanisms at the joints to handle angular transmission between intersecting axes. However, due to the short contact line and few meshing teeth of spur bevel gears, their time-varying meshing stiffness fluctuates significantly, easily causing torsional vibration and impact phenomena. If the vibration is transmitted to the end effector through the transmission chain, it will lead to uneven trajectory and increased positioning error. Usually, improving machining accuracy and configuring backlash reduction mechanisms are used to alleviate the vibration phenomenon of electrically driven manipulators, but this only provides a certain degree of relief and still cannot eliminate the aforementioned defects from the mechanism. Summary of the Invention

[0004] The technical objective of this invention is to provide an electric follower manipulator based on a spiral bevel gear pair transmission, which can enhance time-varying meshing stiffness and reduce vibration, and to provide a stiffness compensation control system and a stiffness compensation control method based on the structure of the electric follower manipulator, in view of the special characteristics of the above-mentioned electric follower manipulator and the shortcomings of the prior art.

[0005] The technical objective of this invention is achieved through the following technical solution: an electrically driven manipulator, comprising a robotic arm; The angular transmission mechanism at at least one joint of the robotic arm is a spiral bevel gear transmission structure.

[0006] Furthermore, the helix angle of the spiral bevel gear pair is 15-35°, the ratio of tooth width to module is 6-12, and the contact overlap of the tooth surface is ≥2.

[0007] Furthermore, the spiral bevel gear pair is made of low-alloy high-strength carburized steel, which is carburized, quenched, and precision ground, with a surface roughness Ra≤0.4. .

[0008] Furthermore, the angular transmission mechanism composed of the aforementioned spiral bevel gear transmission structure also includes preloaded tapered roller bearings or crossed roller bearings arranged back-to-back.

[0009] Furthermore, the angular transmission mechanism is integrated with the motor and reducer in a single structure. An adjustable clearance mechanism is provided at the input shaft end, which uses an eccentric sleeve or wedge to adjust the bidirectional micro-engagement clearance.

[0010] A stiffness compensation control system for an electrically driven manipulator, the stiffness compensation control system being based on the aforementioned electrically driven manipulator structure; The stiffness compensation control system includes a data acquisition unit, a full transmission chain stiffness modeling unit, a composite compensation control unit, and an execution drive unit. The units work together to realize the stiffness parameter identification and dynamic compensation process. The data acquisition unit is used to acquire the motor current, encoder angle, and end effector force / torque of the electric follower manipulator; The end-effector Cartesian space stiffness matrix model of the full transmission chain stiffness model is established with the following relationship: ; In the formula, Represents the end-of-course Cartesian space stiffness matrix; The Jacobian matrix representing the robotic arm; Represents the diagonal stiffness matrix of the joint space; This represents the supplementary stiffness matrix introduced by external loads and attitude. The composite compensation control unit, based on the end-effector Cartesian space stiffness matrix, superimposes the feedforward joint compensation amount and the feedback adjustment amount on the trajectory to generate the desired joint command. The feedforward joint compensation amount is obtained using the following formula: ; In the formula, This indicates the feedforward joint angle compensation value; Indicates terminal generalized force; The feedback adjustment amount is obtained based on the position error and velocity error using a control or robust control algorithm; The execution drive unit is used to receive desired joint commands, drive the motor and reducer to operate, and receive the actual joint angle from the encoder in real time.

[0011] Furthermore, the observation data collected by the data acquisition unit is processed online to obtain time-varying parameters of meshing stiffness and bearing stiffness; The online estimation process calls either the recursive least squares algorithm or the adaptive gradient algorithm.

[0012] Furthermore, the terminal Cartesian space stiffness matrix model, in relation to the terminal Cartesian space stiffness matrix... During the update process, a forgetting factor is introduced. The parameter projection operator is used to perform optimization identification calculations to constrain the stiffness parameters within the physically feasible domain.

[0013] A stiffness compensation control method for an electrically driven manipulator, the stiffness compensation control method being based on the aforementioned stiffness compensation control system; The stiffness compensation control method specifically includes the following process: S1. After the stiffness compensation control system is started, the full transmission chain stiffness modeling unit automatically loads the kinematic parameters of the robotic arm, and combines the theoretical model of the spiral bevel gear meshing stiffness, the experimental data of bearing stiffness and the joint flexibility parameters to construct a full transmission chain stiffness model that includes time-varying meshing stiffness, bearing stiffness and joint flexibility. Initialize the end Cartesian space stiffness matrix The initial value; S2. Acquire motor current, joint angle, and end force / torque data under different postures and load conditions using either calibrated operation mode or online real-time acquisition mode, and form an identification dataset after preprocessing; S3. Invoke the recursive least squares algorithm / adaptive gradient algorithm to estimate the time-varying parameters of meshing stiffness and bearing stiffness online based on the identified dataset, substitute them into the end-effector Cartesian space stiffness matrix model, and update... ; In the During the update process, through the forgetting factor The identification results are optimized using the parameter projection operator to ensure... It can reflect stiffness fluctuations caused by changes in pose and load in real time; S4. Based on the updated With terminal generalized force Calculate end deformation ; Utilizing the mapping relationship between end-effector deformation and joint deformation Solve for the angular compensation of each joint; If the robotic arm is in a singular neighborhood, then damped least squares method is used, and a damping coefficient is introduced. Perform optimized calculations to ensure numerical stability and avoid Abnormal compensation amount; S5. The composite compensation control unit superimposes the feedforward joint angle compensation amount and the feedback adjustment amount to generate the final desired joint command, which is then sent to the execution drive unit; The drive unit drives the motor to rotate, which in turn moves the joint. At the same time, the encoder provides real-time position feedback to dynamically correct motion deviations and ensure that the end effector runs smoothly along the preset trajectory. S6. During pose or load changes, when the pose change exceeds the threshold or the load change exceeds the threshold, dynamically update the end-effector Cartesian space stiffness matrix model and controller parameters. Repeat steps S3 through S5 to update. The joint angle compensation amount helps to suppress the impact of stiffness fluctuations on the performance of the electric follower manipulator.

[0014] The beneficial technical effects of the present invention are as follows: the above-mentioned technical measures are designed for the special characteristics of the above-mentioned electric follower manipulator. The angular transmission mechanism at the joint of the manipulator is composed of a spiral bevel gear pair with a long contact line, a large number of meshing teeth, smooth transmission, low noise, and strong load-bearing capacity. Thus, based on the high contact overlap of the spiral bevel gear pair, the time-varying meshing stiffness at the joint of the above-mentioned electric follower manipulator is enhanced, the fluctuation of time-varying meshing stiffness is reduced, vibration is reduced, and the operating accuracy is ensured.

[0015] For electric follower manipulators with spiral bevel gear pairs forming angular transmission mechanisms, the above-mentioned stiffness compensation control technology measures, by constructing a stiffness model of the entire transmission chain and combining online estimation and feedforward-feedback composite compensation, effectively suppress stiffness fluctuations caused by posture and load, reliably improve dynamic stiffness and trajectory tracking accuracy, and solve the technical problem of the dynamic evolution of the stiffness of the entire transmission chain with changes in posture and load in spiral bevel gear pairs, thus ensuring high-precision operation of the electric follower manipulator. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the structure of the electrically driven manipulator at the joint of the robotic arm according to the present invention.

[0017] Figure 2 This is a flowchart of the stiffness compensation control of the electrically driven manipulator according to the present invention.

[0018] The symbols in the diagram mean: 1—arm body; 2—spiral bevel gear pair; 3—bearing. Detailed Implementation

[0019] This invention relates to the field of electric follower manipulator technology, specifically an electric follower manipulator structure, and a stiffness compensation control system and stiffness compensation control method based on this electric follower manipulator structure. The following description refers to the accompanying drawings. Figure 1 and Figure 2 The technical solution of this invention will be clearly and thoroughly explained.

[0020] It should be noted that the accompanying drawings of this invention are schematic, and unnecessary details have been simplified to clarify the technical objectives of this invention, so as to avoid obscuring the technical solutions contributed by this invention to the prior art. Furthermore, the terms "approximately" or "basically" used below to refer to quantities or fit relationships mean that reasonable assembly and processing errors are allowed in the industry, and do not literally describe absolute quantities or fit relationships.

[0021] See Figure 1As shown, this invention relates to an electrically driven manipulator applied in the nuclear industry. The other structures of this electrically driven manipulator are conventional structures and are not the technical contribution of this invention. The technical contribution of this invention lies in designing the angular transmission mechanism at at least one joint of the manipulator as a helical bevel gear transmission structure. That is, the arm body 1 to be combined is connected by a helical bevel gear pair 2 and necessary bearings 3. The helical bevel gear transmission structure constitutes the angular transmission mechanism at the joint of the manipulator (the other joints of the manipulator can reuse the angular transmission mechanism of this invention as needed to ensure the consistency of the overall transmission).

[0022] Specifically, the spiral bevel gear pair is made of low-alloy high-strength carburized steel, which is carburized, quenched, and precision ground, and has the following characteristic parameters: - The surface roughness Ra of the tooth teeth ≤ 0.4 (e.g. Ra is 0.4) Or Ra is 0.22 wait); - The tooth surface is modified to obtain a contact overlap ratio ≥2 (such as a contact overlap ratio of 5, 3 or 2, etc.) to suppress the fluctuation of time-varying meshing stiffness; - The helix angle is 15 to 35°, preferably 25 to 35° (such as 15°, 25°, 30° or 35°, etc.) to balance transmission smoothness and load-bearing capacity; - The ratio of tooth width to module (b / m) is controlled to be 6 to 12 (such as 6, 8, 10 or 12, etc.). By modifying the tooth tip and tooth crown, the load is evenly distributed along the tooth length, reducing local Hertzian contact stress.

[0023] Based on the above-mentioned spiral bevel gear transmission structure, the bearings are preloaded tapered roller bearings or crossed roller bearings arranged back to back, with adjustable axial preload to suppress backlash and improve the equivalent stiffness in the torsional and bending directions, thereby reducing the elastic deformation of the transmission chain.

[0024] The angular transmission mechanism composed of the aforementioned spiral bevel gear transmission structure and bearings is integrated with the motor and reducer in a single structure. An adjustable clearance mechanism is provided at the input shaft end. This adjustable clearance mechanism uses an eccentric sleeve or wedge to achieve bidirectional micro-meshing clearance adjustment.

[0025] The angle transmission mechanism described above has a default axis intersection angle of 90° in the robotic arm of an electrically driven manipulator, which is suitable for the transmission requirements of most intersecting joints. If special joint configurations need to be adapted, the axis intersection angle can be designed within the range of 60 to 120°. By adjusting the gear meshing angle and installation position, the power transmission efficiency and transmission accuracy can be ensured to remain unaffected.

[0026] The electrically driven manipulator based on the above structure achieves advantages such as smooth transmission, low noise, and high load-bearing capacity due to the helical curve of the spiral bevel gear pair, the gradual engagement-disengagement process, and the simultaneous engagement of more teeth and a longer contact line. However, the stiffness of the entire transmission chain formed by the coupling of the spiral bevel gear pair with bearings and joint flexibility dynamically evolves with changes in posture and load. Without online estimation and compensation, residual errors will still limit the performance of the electrically driven manipulator. Therefore, this invention designs the following stiffness compensation control system for the electrically driven manipulator based on the above structure.

[0027] The stiffness compensation control system of the present invention includes a data acquisition unit, a full transmission chain stiffness modeling unit, a composite compensation control unit, and an execution drive unit. The units work together to realize the stiffness parameter identification and dynamic compensation process.

[0028] The data acquisition unit is used to collect data such as motor current, encoder angle, and end effector force / torque (achieved through sensors) of the electric servo manipulator. The data collected by the data acquisition unit serves as observations for online identification. After online estimation processing, time-varying parameters of meshing stiffness and bearing stiffness are obtained. This online estimation processing calls either a recursive least squares algorithm or an adaptive gradient algorithm.

[0029] The end-effector Cartesian space stiffness matrix model of the entire transmission chain stiffness model is established with the following relationship: ; In the formula, Represents the end-of-course Cartesian space stiffness matrix; The Jacobian matrix representing the robotic arm (calculated through kinematic modeling); express The inverse of a matrix and its transpose; This represents the diagonal stiffness matrix of the joint space (obtained from the stiffness parameters of the joint components). express The inverse of a matrix; This represents the supplementary stiffness matrix introduced by external loads and attitude; supplementary stiffness matrix The calculation is performed by differentiating the attitude variables through the generalized force F at the end point. And in areas of deteriorated motion performance (such as joint limit positions and heavy load conditions) Weighted processing is applied to enhance the compensation effect in this area and avoid performance degradation.

[0030] To suppress noise interference and parameter drift, the aforementioned end-Cartesian space stiffness matrix model, after adjusting the end-Cartesian space stiffness matrix... During the update process, a forgetting factor is introduced. The optimization identification calculation is performed with the parameter projection operator to constrain the stiffness parameters within the physically feasible domain (e.g., the meshing stiffness is not lower than the lower limit of the material's allowable stiffness), ensuring the validity of the identification results. The aforementioned forgetting factor... The value is usually between 0.95 and 0.99, such as 0.95, 0.97 or 0.99.

[0031] The composite compensation control unit includes a feedforward compensation channel based on the aforementioned model (i.e., the end-effector Cartesian space stiffness matrix model) and an error-based feedback channel. The composite compensation control unit, based on the end-effector Cartesian space stiffness matrix, superimposes the feedforward joint compensation amount and the feedback adjustment amount on the trajectory to generate the desired joint command. The feedforward joint compensation amount is obtained using the following formula: ; In the formula, This indicates the feedforward joint angle compensation value; Indicates terminal generalized force; The feedback adjustment amount is based on the position error. With speed error ,use The output adjustment amount of the control or robust control algorithm.

[0032] The actuator is used to receive desired joint commands, drive the motor and reducer to operate, and drive the spiral bevel gear pair to realize joint movement; at the same time, it receives the actual joint angle feedback from the encoder in real time to form closed-loop control, ensuring joint movement accuracy and trajectory tracking performance.

[0033] Based on the aforementioned stiffness compensation control system, the control method for stiffness compensation of the electrically driven manipulator specifically includes the following technological processes: Figure 2 As shown: S1. After the stiffness compensation control system is started, the full transmission chain stiffness modeling unit automatically loads the kinematic parameters of the robotic arm (such as link length and joint parameters), and combines the theoretical model of spiral bevel gear meshing stiffness, experimental data of bearing stiffness and joint flexibility parameters to construct a full transmission chain stiffness model that includes time-varying meshing stiffness, bearing stiffness and joint flexibility. Initialize the end Cartesian space stiffness matrix The initial value; S2. Acquire motor current, joint angle, and end effector force / torque data under different postures and load conditions using either a calibrated operation mode (e.g., the robotic arm moves along a preset trajectory, covering typical poses and load conditions) or an online real-time acquisition mode. After preprocessing (denoising and data alignment), an identification dataset is formed for subsequent online estimation of stiffness parameters. S3. Invoke the recursive least squares algorithm / adaptive gradient algorithm to estimate the time-varying parameters of meshing stiffness and bearing stiffness online based on the identified dataset, substitute them into the end-effector Cartesian space stiffness matrix model, and update... ; In the During the update process, through the forgetting factor The identification results are optimized using the parameter projection operator to ensure... It can reflect stiffness fluctuations caused by changes in pose and load in real time; S4. Based on the updated With terminal generalized force Calculate end deformation ; Utilizing the mapping relationship between end-effector deformation and joint deformation Solve for the angular compensation of each joint; If the robotic arm is in a singular neighborhood (the Jacobian matrix is ​​close to singular), then damped least squares method is used, and a damping coefficient is introduced. Perform optimized calculations to ensure numerical stability and avoid Abnormal compensation amount; S5. The composite compensation control unit superimposes the feedforward joint angle compensation amount and the feedback adjustment amount to generate the final desired joint command, which is then sent to the execution drive unit; The drive unit drives the motor to rotate, which in turn moves the joint. At the same time, the encoder provides real-time position feedback to dynamically correct motion deviations and ensure that the end effector runs smoothly along the preset trajectory. S6. Continuously monitor the pose and load changes of the robotic arm (determined by end-effector force / torque sensors and encoders). During pose or load changes, when the pose change exceeds the threshold (e.g., joint angle change ≥ 5°) or the load change exceeds the threshold (e.g., end-effector load change ≥ 10% of rated load), automatically trigger the stiffness model and controller parameter update - i.e. dynamically update the end-effector Cartesian space stiffness matrix model and controller parameters. Repeat steps S3 through S5 to update. The joint angle compensation amount helps to suppress the impact of stiffness fluctuations on the performance of the electric follower manipulator.

[0034] Finite element and simulation analysis shows that, under the same conditions of module, number of teeth, and pressure angle, the maximum contact stress and comprehensive elastic deformation of the tooth surface of the spiral bevel gear pair can be reduced by about 1 / 3 compared with the straight bevel gear, the end repeatability error is reduced by about 30%, and the fluctuation of time-varying meshing stiffness is significantly reduced. This is beneficial for reducing vibration and ensuring the operating accuracy of the electric follower manipulator.

[0035] The above specific technical solutions are only used to illustrate the present invention, and are not intended to limit it.

[0036] Although the present invention has been described in detail with reference to the specific technical solutions described above, those skilled in the art should understand that modifications can still be made to the specific technical solutions described above, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the present invention.

Claims

1. An electrically servo-driven robotic arm, comprising a robotic arm; Its features are: The angular transmission mechanism at at least one joint of the robotic arm is a spiral bevel gear transmission structure.

2. The electro-servo manipulator according to claim 1, characterized in that: The spiral bevel gear pair has a helix angle of 15-35°, a tooth width to module ratio of 6-12, and a tooth surface contact overlap of ≥2.

3. The electro-servo manipulator according to claim 1 or 2, characterized in that: The spiral bevel gear pair is made of low-alloy high-strength carburized steel, which is carburized, quenched, and precision ground. The surface roughness Ra is ≤ 0.

4. .

4. The electro-servo manipulator according to claim 1, characterized in that: The angular transmission mechanism composed of the aforementioned spiral bevel gear transmission structure also includes preloaded tapered roller bearings or crossed roller bearings arranged back-to-back.

5. The electro-servo manipulator according to claim 1 or 4, characterized in that: The angle transmission mechanism is integrated with the motor and reducer in a single structure. An adjustable clearance mechanism is provided at the input shaft end, which uses an eccentric sleeve or wedge to adjust the bidirectional micro-engagement clearance.

6. A stiffness compensation control system for an electrically driven manipulator, characterized in that: The stiffness compensation control system is based on the electric follow-up manipulator structure according to any one of claims 1 to 5; The stiffness compensation control system includes a data acquisition unit, a full transmission chain stiffness modeling unit, a composite compensation control unit, and an execution drive unit. The units work together to realize the stiffness parameter identification and dynamic compensation process. The data acquisition unit is used to acquire the motor current, encoder angle, and end effector force / torque of the electric follower manipulator; The end-effector Cartesian space stiffness matrix model of the full transmission chain stiffness model is established with the following relationship: ； In the formula, Represents the end-of-course Cartesian space stiffness matrix; The Jacobian matrix representing the robotic arm; Represents the diagonal stiffness matrix of the joint space; This represents the supplementary stiffness matrix introduced by external loads and attitude. The composite compensation control unit, based on the end-effector Cartesian space stiffness matrix, superimposes the feedforward joint compensation amount and the feedback adjustment amount on the trajectory to generate the desired joint command. The feedforward joint compensation amount is obtained using the following formula: ； In the formula, This indicates the feedforward joint angle compensation value; Indicates terminal generalized force; The feedback adjustment amount is obtained based on the position error and velocity error using a control or robust control algorithm; The execution drive unit is used to receive desired joint commands, drive the motor and reducer to operate, and receive the actual joint angle from the encoder in real time.

7. The stiffness compensation control system for the electrically driven manipulator according to claim 6, characterized in that: The observation data collected by the data acquisition unit is processed online to obtain time-varying parameters of meshing stiffness and bearing stiffness. The online estimation process calls either the recursive least squares algorithm or the adaptive gradient algorithm.

8. The stiffness compensation control system for the electrically driven manipulator according to claim 6, characterized in that: The terminal Cartesian space stiffness matrix model, in relation to the terminal Cartesian space stiffness matrix During the update process, a forgetting factor is introduced. The parameter projection operator is used to perform optimization identification calculations to constrain the stiffness parameters within the physically feasible domain.

9. A stiffness compensation control method for an electrically driven manipulator, characterized in that: The stiffness compensation control method is based on the stiffness compensation control system according to any one of claims 6 to 8; The stiffness compensation control method specifically includes the following process: S1. After the stiffness compensation control system is started, the full transmission chain stiffness modeling unit automatically loads the kinematic parameters of the robotic arm, and combines the theoretical model of the spiral bevel gear meshing stiffness, the experimental data of bearing stiffness and the joint flexibility parameters to construct a full transmission chain stiffness model that includes time-varying meshing stiffness, bearing stiffness and joint flexibility. Initialize the end Cartesian space stiffness matrix The initial value; S2. Acquire motor current, joint angle, and end force / torque data under different postures and load conditions using either calibrated operation mode or online real-time acquisition mode, and form an identification dataset after preprocessing; S3. Invoke the recursive least squares algorithm / adaptive gradient algorithm to estimate the time-varying parameters of meshing stiffness and bearing stiffness online based on the identified dataset, substitute them into the end-effector Cartesian space stiffness matrix model, and update... ; In the During the update process, through the forgetting factor The parameter projection operator optimizes the identification results to ensure... It can reflect stiffness fluctuations caused by changes in pose and load in real time; S4. Based on the updated With terminal generalized force Calculate end deformation ; Utilizing the mapping relationship between end-effector deformation and joint deformation Solve for the angular compensation of each joint; If the robotic arm is in a singular neighborhood, then damped least squares method is used, and a damping coefficient is introduced. Perform optimized calculations to ensure numerical stability and avoid Abnormal compensation amount; S5. The composite compensation control unit superimposes the feedforward joint angle compensation amount and the feedback adjustment amount to generate the final desired joint command, which is then sent to the execution drive unit; The drive unit drives the motor to rotate, which in turn moves the joint. At the same time, the encoder provides real-time position feedback to dynamically correct motion deviations and ensure that the end effector runs smoothly along the preset trajectory. S6. During pose or load changes, when the pose change exceeds the threshold or the load change exceeds the threshold, dynamically update the end-effector Cartesian space stiffness matrix model and controller parameters. Repeat steps S3 through S5 to update. The joint angle compensation amount helps to suppress the impact of stiffness fluctuations on the performance of the electric follower manipulator.

Images