A humanoid welding robot based on clean welding double mechanical arms

Abstract

The application provides a humanoid welding robot based on cleaning welding double mechanical arms, which comprises a robot trunk, cleaning mechanical arms and welding mechanical arms connected on both sides of the robot trunk; the cleaning mechanical arms and the welding mechanical arms each comprise two or more arm bodies connected in sequence; an executing mechanism is connected to the tail end arm body; the initial end arm body, the adjacent arm bodies and the tail end arm body and the executing mechanism are connected through joints respectively; a permanent magnet synchronous motor is arranged at each joint; target trajectories of the cleaning mechanical arms and the welding mechanical arms are planned, so that the target angles of each joint are set; the permanent magnet synchronous motors of each joint adopt a fast prediction model to set the output, so that the joints move according to the target angles. The humanoid robot has cleaning and welding functions, can clean the surface before welding and process the weld after welding, effectively improves the welding quality, and effectively improves the welding efficiency by closely linking the cleaning and welding operations.

Description

Technical Field

[0001] This invention relates to the field of humanoid welding robot technology, and more specifically, to a humanoid welding robot based on a clean welding dual-arm robotic arm. Background Technology

[0002] Welding technology, acting as the tailor of industrial manufacturing, ensures robust industrial capacity and has wide applications in sectors including aerospace, petrochemicals, and automobiles. Traditional manufacturing enterprises, lacking advanced welding equipment and technology, rely heavily on manual labor and conventional equipment, resulting in inflexible production processes, high costs, and low levels of automation. Therefore, achieving automated and intelligent welding is essential for the transformation and upgrading of the manufacturing industry.

[0003] In recent years, the construction of petrochemical plants has continued to grow, leading to a surge in demand for pipeline welding. Currently, the most common method for pipeline welding is traditional manual welding. This requires welders to hold a welding torch, and the fumes, intense arc light, and high temperatures generated during the welding process can harm their health. Manual welding suffers from low efficiency, inconsistent quality, and numerous safety hazards, making it unsuitable for the demands of modern industrial production. Therefore, combining welding technology with humanoid robots can meet the needs of operations in harsh conditions and hazardous environments.

[0004] Before welding, the welding surface needs to be cleaned, and after welding, the weld seam needs to be ground. Traditional welding robots only have welding or cleaning functions, resulting in low welding efficiency and quality. Furthermore, the joint control of traditional welding robots uses the traditional FOC drive control algorithm, which has relatively poor dynamic response performance to torque. During the welding process, due to the instability of the welding voltage, the welding torch direction is prone to pulse fluctuations, which makes the system have high requirements for the dynamic performance of the joint motors of the robotic arm. Summary of the Invention

[0005] To overcome the shortcomings and deficiencies of the existing technology, the purpose of this invention is to provide a humanoid welding robot based on a dual robotic arm for cleaning and welding. This humanoid robot has both cleaning and welding functions, and can perform surface cleaning before welding and weld treatment after welding. It can effectively improve the welding quality of the welding humanoid robot, and the cleaning and welding operations are closely integrated, which can effectively improve the welding efficiency.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solution: a humanoid welding robot based on a dual-arm cleaning welding system, comprising a robot torso, cleaning arms and welding arms connected to both sides of the robot torso; both the cleaning arms and the welding arms comprise two or more arm sections connected sequentially; an actuator is connected to the end arm section; joints connect the beginning arm section to the robot torso, adjacent arms, and the end arm section to the actuator; a permanent magnet synchronous motor is provided at each joint; the actuator of the cleaning arm is a cleaning mechanism for grinding and removing impurities; the actuator of the welding arm is a welding wire feeding mechanism for welding and wire feeding.

[0007] Based on the work area, the target trajectories of the cleaning robot arm and the welding robot arm are planned, thereby setting the target angle θ* of each joint of the cleaning robot arm and the welding robot arm at each moment; the permanent magnet synchronous motors of each joint of the cleaning robot arm and the welding robot arm use a fast prediction model to set the output, so that the joints move according to the target angle θ*.

[0008] Preferably, the fast prediction model includes a position loop PID controller, a speed PID controller, and a fast three-vector model; the outputs of the permanent magnet synchronous motors of each joint of the cleaning robot arm and the welding robot arm are set using the following method:

[0009] The system acquires the actual speed Speed ​​of the joint permanent magnet synchronous motor and the detection angle θ of the permanent magnet synchronous motor encoder; the position loop PID controller adjusts the speed by inputting the set target angle θ* and the detection angle θ, and outputs the target speed Speed*; the speed PID controller outputs the actual required current I* by inputting the target speed Speed* and the actual speed Speed; the fast three-vector model selects an optimal vector u from six effective voltage vectors Ui, i=[1,2,3,4,5,6] based on the actual required current I*. p_1 and a suboptimal vector u p_2 And calculate the optimal vector u p_1 and suboptimal vector u p_2 The duration of action; based on the optimal vector u p_1 and suboptimal vector u p_2 The duration of the voltage vector is applied to the inverter, causing the inverter to generate a corresponding voltage vector that acts on the permanent magnet synchronous motor.

[0010] Preferably, the speed PID controller outputs the actual required current I*, which includes the d-axis component I of the required current. d *, q-axis component I q *;

[0011] In the fast three-vector model, the method for obtaining the six effective voltage vectors Ui is as follows:

[0012] Calculate the d-axis voltage values ​​u of the six effective voltage vectors Ui in the switching state. d [i] and q-axis voltage value u q [i]:

[0013] ;

[0014] in, This refers to the bus voltage of the permanent magnet synchronous motor. For the electrical angle of a permanent magnet synchronous motor, p is the number of pole pairs of the permanent magnet synchronous motor, θ is the detection angle; S_abc[i][j] is the j-th bit of the switch state value S_abc[i].

[0015] Preferably, the step of selecting an optimal vector u from the six effective voltage vectors Ui is... p_1 and a suboptimal vector u p_2 , refers to:

[0016] First, the globally optimal voltage vector at the next time step is calculated. :

[0017] ;

[0018] ;

[0019] Among them, u p_d (k+1), u p_q (k+1) represent the globally optimal voltage components along the d-axis and q-axis at the next time step, respectively; I d (k), I q (k) represent the d-axis and q-axis components of the current detected in the permanent magnet synchronous motor at the current moment, respectively; , , They are respectively:

[0020] ;

[0021] Where R is the stator resistance; L s It is the stator inductance; ω e It is the electrical angular velocity of the permanent magnet synchronous motor; ψ f It is the magnetic flux of a permanent magnet; T s To control the cycle;

[0022] Then, the globally optimal voltage vector will be... Transformed into values ​​in a two-phase stationary coordinate system ; Judgment value Sector; based on value Select the optimal vector u for the sector. p_1 and suboptimal vector up_2 .

[0023] Preferably, the value The sector determination is achieved through the sector determination method of the SVPWM algorithm.

[0024] Preferably, the optimal vector u p_1 and suboptimal vector u p_2 The method for calculating the duration of action is:

[0025] Set the control period to T s :T s = t p1 + t p2 + t0; where t p1 For the optimal vector u p_1 Duration of action; t p2 For suboptimal vector u p_2 The duration of action; t0 is the duration of action of the zero vector u0;

[0026] Setting the optimal vector u for the permanent magnet synchronous motor p_1 Suboptimal vector u p_2 After the zero vector u0, the resulting d-axis current change rates are respectively , and The resulting q-axis current change rate are respectively , and The expression for each rate of change of current is as follows:

[0027] ;

[0028] in, and These are the optimal vectors u p_1 d-axis and q-axis voltage components; and These are the suboptimal vectors u p_2 d-axis and q-axis voltage components; I d I q These are the d-axis and q-axis components of the detected permanent magnet synchronous motor current, respectively.

[0029] Calculate the optimal vector u p_1 Suboptimal vector u p_2 The duration of action of the zero vector u0:

[0030] ;

[0031] Among them, I d *、I q* Represents the d-axis and q-axis components of the required current, respectively; K is a coefficient, expressed as follows:

[0032] .

[0033] Preferably, the optimal vector u is calculated. p_1 Suboptimal vector u p_2 After the interaction time of the zero vector u0, it is also determined whether... If so, then control period T will be... s The value is equal to the inverter's switching cycle, and the operating time is reconstructed; the reconstruction formula is:

[0034] .

[0035] Preferably, the cleaning robotic arm and the welding robotic arm are respectively connected to the robot body via a base; the base is provided with a first rotary joint; the starting arm is connected to the first rotary joint, and adjacent arms are connected via hinge joints; the end arm is connected to the actuator via a second rotary joint.

[0036] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0037] 1. The humanoid welding robot of this invention adopts dual robotic arms, which have both cleaning and welding functions. Compared with the traditional welding method of a single robotic arm, the dual robotic arms can perform surface cleaning before welding and weld treatment after welding, which can effectively improve the welding quality of the humanoid welding robot. The close integration of cleaning and welding operations can also effectively improve welding efficiency.

[0038] 2. This invention employs a predictive current control method based on a fast predictive model to control the current (torque) of each joint of the dual robotic arm. Compared with the current PID closed-loop in the traditional vector control algorithm FOC, the method of this invention has better dynamic response performance, can effectively cope with torque changes, can better adapt to the pulse fluctuations generated during the welding process, ensure the stability of the welding process, and does not require repeated adjustment of parameters in the controller, thus simplifying the controller and facilitating control. Compared with the traditional model predictive control algorithm, the method of this invention does not need to calculate the value function to predict the voltage vector to be applied at the next moment. Instead, it generates the voltage vector required at the next moment through deadbeat control of the dq axis current. This reduces the number of predictions, thereby reducing calculation time and alleviating the computational burden. Attached Figure Description

[0039] Figure 1 This is a structural schematic diagram of the humanoid welding robot of the present invention;

[0040] Figure 2This is a schematic diagram of the welding robotic arm of the humanoid welding robot of the present invention;

[0041] Figure 3 This is a schematic diagram of the cleaning robotic arm of the humanoid welding robot of the present invention;

[0042] Figure 4 This is a control block diagram of each joint of the present invention based on a fast prediction model. Detailed Implementation

[0043] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0044] Example

[0045] This embodiment describes a humanoid welding robot based on a clean welding dual-arm robotic arm, such as... Figures 1 to 3 As shown, the robot includes a robot chassis 1 for movement, a robot torso 2, a robot head 3, and cleaning robotic arms 5 and welding robotic arms 4 connected to both sides of the robot torso 2. Both the cleaning robotic arm 5 and the welding robotic arm 4 include two or more arm sections 7 connected sequentially; an actuator 8 is connected to the end arm section 7; the beginning arm section 7 is connected to the robot torso 2, adjacent arm sections 7 are connected to each other, and the end arm section 7 is connected to the actuator 8 via joints; each joint is equipped with a permanent magnet synchronous motor.

[0046] Specifically, the cleaning robotic arm 5 and the welding robotic arm 4 are connected to the robot body 2 via a base 6; a rotary joint 9 is provided on the base 6; the starting arm 7 is connected to the rotary joint 9, and adjacent arms 7 are connected via hinge joints 10; the ending arm 7 is connected to the actuator 8 via a rotary joint 11. The actuator 8 of the cleaning robotic arm 5 is a cleaning mechanism for grinding and removing impurities; the actuator 8 of the welding robotic arm 4 is a welding wire feeding mechanism for welding and wire feeding.

[0047] Based on the work area, the target trajectories of the cleaning and welding robotic arms are planned, and then the target angles θ* of each joint of the cleaning and welding robotic arms at each moment are set. Specifically, the motion trajectories of the end effectors of the cleaning robotic arm 5 and the welding robotic arm 4 are interpolated linearly or circularly to obtain several discrete positions and orientations of the actuators. Inverse kinematics is then performed on these positions and orientations to calculate the angles rotated by the joint axes corresponding to these orientations. The inverse kinematics solution of the robotic arms can be divided into two parts: the first part calculates the rotation angles of the first few joint axes based on the coordinates of the wrist point of the robotic arm; the second part calculates the rotation angles of the subsequent joint axes using the calculated rotation angles of the first few joint axes and the given end effector orientation matrix of the robotic arm. The first part uses a planar geometry solution method, and the second part uses an Euler angle transformation method.

[0048] The permanent magnet synchronous motors of each joint of the cleaning and welding robotic arms use a fast predictive model to set the output, so that the joints move according to the target angle θ*.

[0049] Specifically, the rapid prediction model includes a position loop PID controller, a speed PID controller, and a rapid three-vector model; the outputs of the permanent magnet synchronous motors at each joint of the cleaning and welding robotic arms are set using the following methods: Figure 4 As shown:

[0050] The system acquires the actual speed Speed ​​of the joint permanent magnet synchronous motor and the detection angle θ of the permanent magnet synchronous motor encoder; the position loop PID controller adjusts the speed by inputting the set target angle θ* and the detection angle θ, and outputs the target speed Speed*.

[0051] The speed PID controller takes the target speed Speed* and the actual speed Speed ​​as inputs and outputs the actual required current I*; the output actual required current I* includes the d-axis component of the required current I. d *, q-axis component I q *

[0052] The position loop PID controller and speed PID controller use existing PID algorithms; this embodiment can use PI control.

[0053] The fast three-vector model selects an optimal vector and a suboptimal vector from six effective voltage vectors Ui, i=[1,2,3,4,5,6] based on the actual required current I*, and calculates the action time of the optimal and suboptimal vectors.

[0054] In the fast three-vector model, the six effective voltage vectors Ui are obtained as follows:

[0055] Calculate the d-axis voltage values ​​u of the six effective voltage vectors Ui in the switching state. d [i] and q-axis voltage value u q [i]:

[0056]

[0057] in, This refers to the bus voltage of the permanent magnet synchronous motor. For the electrical angle of a permanent magnet synchronous motor, p is the number of pole pairs of the permanent magnet synchronous motor, θ is the detection angle; S_abc[i][j] is the j-th bit of the switch state value S_abc[i]. The switch state value S_abc[i] is shown in Table 1:

[0058] Table 1. Values ​​of switch state S_abc[i]

[0059]

[0060] Selecting one optimal vector and one suboptimal vector from the six effective voltage vectors Ui means:

[0061] First, based on the deadbeat control principle of dq-axis current, the globally optimal voltage vector for the next moment is calculated from the full voltage vector range. :

[0062] ;

[0063] ;

[0064] Among them, u p_d (k+1), u p_q (k+1) represent the globally optimal voltage components along the d-axis and q-axis at the next time step, respectively; I d (k), I q (k) represent the d-axis and q-axis components of the current detected in the permanent magnet synchronous motor at the current moment, respectively; , , They are respectively:

[0065] ;

[0066] Where R is the stator resistance; L s It is the stator inductance; ω e It is the electrical angular velocity of the permanent magnet synchronous motor; ψ f It is the magnetic flux of a permanent magnet; T s To control the cycle;

[0067] Then, the globally optimal voltage vector will be... Transformed into values ​​in a two-phase stationary coordinate system ; Judgment value Sector; Value The sector determination is achieved through the sector determination method of the SVPWM algorithm;

[0068] According to the value Select the optimal vector for the sector. and suboptimal vector The optimal and suboptimal vectors can be selected by looking up the correspondence table between sectors and optimal and suboptimal vectors, as shown in Table 2.

[0069] Table 2. Correspondence between sectors and optimal and suboptimal vectors

[0070]

[0071] Optimal vector u p_1 and suboptimal vector u p_2 The method for calculating the duration of action is:

[0072] Based on dq-axis current deadbeat control, the output voltage is determined by the optimal vector u. p_1 Suboptimal vector u p_2 and zero vector Composition. Therefore, the control period is set to T. s :T s = t p1 + t p2 + t0; where t p1 For the optimal vector Duration of action; t p2 Suboptimal vector The duration of action; t0 is the zero vector. Duration of action;

[0073] Setting the optimal vector u for the permanent magnet synchronous motor p_1 Suboptimal vector u p_2 and zero vector Afterwards, the resulting d-axis current change rates are respectively , and The resulting q-axis current change rate are respectively , and The expression for each rate of change of current is as follows:

[0074] ;

[0075] in, and These are the optimal vectors u p_1 d-axis and q-axis voltage components; and These are the suboptimal vectors u p_2 d-axis and q-axis voltage components; I d I q These are the d-axis and q-axis components of the detected permanent magnet synchronous motor current, respectively.

[0076] The application time is allocated based on the deadbeat principle of dq-axis current, as expressed below:

[0077]

[0078]

[0079] By transforming the above formula, the optimal vector u can be obtained. p_1 Suboptimal vector u p_2The formula for calculating the duration of action of the zero vector u0 is as follows:

[0080] ;

[0081] Among them, I d *、I q * Represents the d-axis and q-axis components of the required current, respectively; K is a coefficient, expressed as follows:

[0082] .

[0083] Calculate the optimal vector u p_1 Suboptimal vector u p_2 After the interaction time of the zero vector u0, it is also determined whether... If so, then control period T will be... s The value is equal to the inverter's switching cycle, and the operating time is reconstructed; the reconstruction formula is:

[0084] .

[0085] Compared to the current PID closed-loop control in the traditional voltage vector control algorithm (FOC), this predictive current control method has the advantage of not requiring repeated adjustment of the controller parameters and being able to more easily constrain the state variables in the system. This makes the controller simpler and easier to control.

[0086] According to the optimal vector u p_1 Suboptimal vector u p_2 The voltage vector, acting on the inverter for a specified duration, is applied to the permanent magnet synchronous motor. The inverter consists of a driver chip DRV8301 and an inverter circuit topology connected in parallel with it.

[0087] To achieve control of the permanent magnet synchronous motor, the following modules are also used:

[0088] The sampling module, including the sampling resistor of the lower bridge arm of the inverter circuit and the differential operational amplifier in the driver chip DRV8301, is used to collect inputs for processing by the MCU.

[0089] The feedback module includes a magneto-electric relative encoder installed on the permanent magnet synchronous motor. The pulse signal generated by the rotation of the permanent magnet synchronous motor is input to the MCU for processing, which can obtain the mechanical angle of the current motor rotor.

[0090] The coordinate transformation module consists of Clarke and Park in the software and their inverse transformation. It takes the three-phase current of the permanent magnet synchronous motor collected by the sampling module as input and performs coordinate transformation to obtain the current in two-phase stationary coordinates and the current in two-phase rotating coordinates, namely the αβ axis current and the dq axis current.

[0091] By inputting the outputs (current, speed, angle) of the sampling module and the feedback module into the fast prediction model, online control of the permanent magnet synchronous motor can be achieved.

[0092] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A humanoid welding robot based on clean welding dual robotic arms, characterized in that: The robot includes a torso, cleaning robotic arms and welding robotic arms connected to both sides of the torso; both the cleaning robotic arms and the welding robotic arms consist of two or more arm sections connected in sequence; the end arm section is connected to an actuator; the beginning arm section is connected to the robot torso, adjacent arms are connected to each other, and the end arm section is connected to the actuator via joints; each joint is equipped with a permanent magnet synchronous motor; the actuator of the cleaning robotic arm is a cleaning mechanism for grinding and removing impurities; the actuator of the welding robotic arm is a welding wire feeding mechanism for welding and wire feeding. Based on the work area, the target trajectories of the cleaning robot arm and the welding robot arm are planned, thereby setting the target angle θ* of each joint of the cleaning robot arm and the welding robot arm at each moment; the permanent magnet synchronous motors of each joint of the cleaning robot arm and the welding robot arm use a fast prediction model to set the output, so that the joints move according to the target angle θ*. The rapid prediction model includes a position loop PID controller, a speed PID controller, and a rapid three-vector model; the outputs of the permanent magnet synchronous motors at each joint of the cleaning and welding robotic arms are set using the following method: The system acquires the actual speed (Speed) of the joint permanent magnet synchronous motor and the detection angle (θ) of the permanent magnet synchronous motor encoder. The position loop PID controller adjusts the speed based on the input target angle (θ*) and detection angle (θ), outputting the target speed (Speed*). The speed PID controller, based on the input target speed (Speed*) and actual speed (Speed), outputs the actual required current (I*). The output actual required current (I*) includes the d-axis component of the required current (Id). d *, q-axis component I q *; The fast three-vector model is based on the required current d-axis component I. d *, q-axis component I q *, The global optimal voltage vector at the next time step is calculated. : ； ； Among them, u p_d (k+1), u p_q (k+1) represent the globally optimal voltage components along the d-axis and q-axis at the next time step, respectively; I d (k), I q (k) represent the d-axis and q-axis components of the current detected in the permanent magnet synchronous motor at the current moment, respectively; , , They are respectively: ； Where R is the stator resistance; L s It is the stator inductance; ω e It is the electrical angular velocity of the permanent magnet synchronous motor; ψ f It is the magnetic flux of a permanent magnet; T s To control the cycle; Then, the globally optimal voltage vector will be... Transformed into values ​​in a two-phase stationary coordinate system ; Judgment value Sector; based on value Select the optimal vector u for the sector. p_1 and suboptimal vector u p_2 ; Calculate the optimal vector u p_1 and suboptimal vector u p_2 The duration of action; based on the optimal vector u p_1 and suboptimal vector u p_2 The duration of the voltage vector is applied to the inverter, causing the inverter to generate a corresponding voltage vector that acts on the permanent magnet synchronous motor.

2. The humanoid welding robot based on clean welding dual robotic arms according to claim 1, characterized in that: The speed PID controller outputs the actual required current I*, including the d-axis component I of the required current. d *, q-axis component I q *; In the fast three-vector model, the method for obtaining the six effective voltage vectors Ui is as follows: Calculate the d-axis voltage values ​​u of the six effective voltage vectors Ui in the switching state. d [i] and q-axis voltage value u q [i]: ； in, This refers to the bus voltage of the permanent magnet synchronous motor. For the electrical angle of a permanent magnet synchronous motor, p is the number of pole pairs of the permanent magnet synchronous motor, θ is the detection angle; S_abc[i][j] is the j-th bit of the switch state value S_abc[i].

3. The humanoid welding robot based on clean welding dual robotic arms according to claim 1, characterized in that: The value The sector determination is achieved through the sector determination method of the SVPWM algorithm.

4. The humanoid welding robot based on clean welding dual robotic arms according to claim 1, characterized in that: The optimal vector u p_1 and suboptimal vector u p_2 The method for calculating the duration of action is as follows: Set the control period to T s :T s = t p1 + t p2 + t0; where t p1 For the optimal vector u p_1 Duration of action; t p2 For suboptimal vector u p_2 The duration of action; t0 is the duration of action of the zero vector u0; Setting the optimal vector u for the permanent magnet synchronous motor p_1 Suboptimal vector u p_2 After the zero vector u0, the resulting d-axis current change rates are respectively , and The resulting q-axis current change rate are respectively , and The expression for each rate of change of current is as follows: ； in, and These are the optimal vectors u p_1 d-axis and q-axis voltage components; and These are the suboptimal vectors u p_2 d-axis and q-axis voltage components; I d I q These are the d-axis and q-axis components of the detected permanent magnet synchronous motor current, respectively. Calculate the optimal vector u p_1 Suboptimal vector u p_2 The duration of action of the zero vector u0: ； Among them, I d *、I q * Represents the d-axis and q-axis components of the required current, respectively; K is a coefficient, expressed as follows: 。 5. The humanoid welding robot based on clean welding dual robotic arms according to claim 4, characterized in that: Calculate the optimal vector u p_1 Suboptimal vector u p_2 After the interaction time of the zero vector u0, it is also determined whether... If so, then control period T will be... s The value is equal to the inverter's switching cycle, and the operating time is reconstructed; the reconstruction formula is: 。 6. The humanoid welding robot based on clean welding dual robotic arms according to claim 1, characterized in that: The cleaning robotic arm and the welding robotic arm are respectively connected to the robot body via a base; the base is provided with a first rotary joint; the starting arm is connected to the first rotary joint, and adjacent arms are connected via hinge joints; the end arm is connected to the actuator via a second rotary joint.

Images