Abstract: Based on a novel five-DOF parallel robot independently developed in this paper, the hardware of the control system, consisting of a microcomputer, a brushless DC motor, and a PCI interface control card, was designed and completed. The inverse kinematics curves of the mechanism were solved using ADAMS software, and the control software for the robot system was written in C++, realizing continuous trajectory motion of the parallel mechanism. Experimental results verify the feasibility of this parallel mechanism and its control system.
Keywords: Parallel mechanism; Brushless motor; Motion control; Position analysis
Motion control of a novel five-DOF parallel robot based on BLDC LI Shi-hua GAO Guo-qin MA Lv-zhong WANG Jin-song
Abstr act: On the foundation of novel five - DOF parallel robot developed by University, the hardware of the control system based on microcomputer, BLDC and PCI interface measure & control card was done. Inverse kinematics solution of the machine solved by ADAMS, the control software of the robotsystem was written using C++ language. Continuous track motion of the parallel robot was realized. Experiment result approved the feasibility of the parallel Mechanism and the control system.
Key words: Par allel mechanism; BLDC; Motion control; Position analysis.
1 Introduction
Parallel mechanisms are closed-loop mechanical systems composed of multiple parallel chains. Compared to serial mechanisms, because their drive units are installed in fixed locations and their positions do not change with the movement of the end effector, they can achieve high-speed and high-precision motion. Parallel mechanisms have advantages such as high stiffness, no joint error accumulation and amplification, and easy position inversion, forming a complementary relationship with serial mechanisms in applications. Currently, most research on parallel robots focuses on 6-DOF parallel robots, but in some applications, 2-5 DDOF are sufficient. These parallel robots with fewer than 6 DDOF are called low-DOF parallel robots. Low-DOF parallel robots have high practical value due to their fewer drive components, lower cost, and compact structure.
Based on the research and development of a 5-DOF parallel massage robot and its position analysis, a robot control interface was designed using VC++6.0 with microcomputer, PCI bus control card, and PCI bus data acquisition card as the hardware foundation to realize the continuous trajectory motion of the mechanism.
2. Principle of a Novel Five-DOF Parallel Robot Mechanism
The parallel mechanism studied is shown in Figure 1. A1～A4 and B1～B4 are ball joints, R1～R8 are revolute joints, and L1～L4 are electric actuators that realize telescopic motion. A1A2A3A4 form a static platform, and B1B2B3B4 form a moving platform.
In the robot system, the four electric actuators L1 to L4 and the central revolute joint (O) are the active inputs, giving the moving platform five degrees of freedom relative to the stationary platform. The corresponding control quantities are displacements l1, l2, l3, and l4, and rotation angle. During operation, the drive joints are controlled to move or rotate the workpiece in three-dimensional space, thus realizing the motion of the moving platform.
3. Hardware Components of the Control System
The control system of the parallel mechanism is shown in Figure 2. The system consists of a microcomputer, a PCI bus measurement and control card, a brushless DC motor and its driver, displacement sensors, etc.
A microcomputer serves as the processor to perform control calculations. It receives and sends control signals to the control card via time-interrupt methods, with a minimum interrupt time interval of 1ms. The control card has 5 D/A outputs, 16 A/D inputs, and 16 digital input/output channels, which can well meet the needs of practical control. The D/A output resolution is 15 bits, with an output range of DC 0–10V. The A/D sampling frequency is 120kHz, with a resolution of 12 bits and a sampling range of 0–10V. It has a built-in sample-and-hold circuit and operates in software polling mode. The digital output has a high level of +12V and a low level of 0V. The microcomputer reads the signals from the displacement sensor via the A/D converter, calculates the position of the electric actuator and the rotation angle, and then sends speed, direction, and motor operating status signals to the motor driver to control the various displacement quantities. The control system uses a permanent magnet brushless DC motor, which offers stepless speed regulation and a wide operating speed range of 0–3000 r/min. It can operate at ultra-low speeds, meeting the speed requirements of various operating modes. This motor features high low-speed torque, smooth operation, high efficiency, and low noise. The connection method between the motor, its driver, and the control board is shown in Figure 3. The driver offers three selectable speed regulation methods: internal potentiometer speed regulation, external input speed regulation, and multi-segment selection speed regulation. In practical applications, external input speed regulation is selected, meaning a D/A converted voltage (relative to COM) is input to the "AVI" terminal for speed control. The "AVI" terminal accepts DC 0V–10V, corresponding to a motor speed of 0–3000 rpm. A 200KΩ internal resistor is connected to the COM terminal; therefore, leaving it unconnected will be interpreted as a zero input.
The motor's forward/reverse, direction, and run/stop control terminals are pulled up to 12V by internal resistors, and are all high-level when there is no input. The motor's operation and stop can be controlled by switching the "R/S" terminal relative to "COM". When "R/S" is disconnected from terminal "COM", the motor stops; conversely, the motor runs. When using the run/stop terminal to stop the motor, it stops naturally, and its motion pattern is related to the load inertia. The motor's direction of rotation can be controlled by switching the "DIR" terminal relative to terminal "COM". When "DIR" is not connected to terminal "COM", the motor runs clockwise (facing the motor shaft), which is considered forward rotation; conversely, it runs counterclockwise, which is considered reverse rotation. To avoid damaging the driver, direction control should be avoided while the motor is running. The driver can control the rapid stop of the brushless motor via terminals BRK~COM. Braking uses controlled energy consumption braking, which is much faster than the free stop via R/S, but the specific time is affected by the user's system (especially system inertia).
4. Planning the motion trajectory of the moving platform
The trajectory parameters used in the actual control of this parallel mechanism were obtained through simulation in the ADAMS environment. Some simulation data results are shown in Figure 4. The ADAMS software uses an interactive graphical environment and libraries of parts, constraints, and forces to create a fully parameterized geometric model of the mechanical system. Its solver employs the Lagrange equation method from the dynamics theory of multi-rigid-body systems to establish the system's dynamic equations and perform static, kinematic, and dynamic analyses on the virtual mechanical system. It can output displacement, velocity, acceleration, and reaction force curves.
ADAMS is a virtual prototype analysis application software that allows users to easily perform static, kinematic, and dynamic analyses on virtual mechanical systems. This software can obtain inverse solution curves of branch variables; by discretizing these curves, the desired control variables can be obtained, with the position given as the desired target position after discretization.
5. Software Design of the Control System
In the Windows environment, the control program was designed using Visual C++. The closed-loop control of the displacement l1 is shown in Figure 5. The control time interval T = 10 ms, the position is the discretized desired trajectory, and the position feedback reads the signal from the displacement sensor through A/D conversion, and calculates the current value of the controlled variable after digital filtering.
The program uses `SetTimer(nIDEvent, time, NULL)` to set up interrupts, where `nIDEvent` is the interrupt number and `time` is the interrupt time interval. The flowchart of the interrupt handling function is shown in Figure 6. Because the parallel mechanism has a certain degree of coupling between its branches during operation, large-scale independent operation of each branch should be avoided. At program startup, each control port must be initialized, all analog outputs cleared, and the digital outputs set to enable the motor's stop and rapid braking terminals, ensuring the safety of the entire system at program startup. To ensure the synchronous operation of the five branches of the parallel mechanism, five corresponding interrupt handling functions are set up in the program.
In addition, a timer was set up to periodically change the desired position at intervals of t. The movement speed of the moving platform was adjusted by changing the value of t. The given position r(kt) was obtained from ADAMS simulation, and the discretized time interval was 0.05s.
The current position c(KT) of the controlled variable is obtained by A/D sampling, and the average value filtering is used. The number of sampling times is 20.
Displacement deviation: e(KT) = r(kt) - c(KT)
(1) Experimental verification shows that when e(KT)>0, the F/R terminal of the motor driver is at a high level, that is, the motor rotates forward; when e(KT)<0, the F/R terminal of the motor driver is at a low level, that is, the motor rotates in reverse. This condition is a necessary condition for the stability of the closed-loop system. The control quantity u(KT) is calculated by the controller, and its value is output to the speed terminal "AVI" of the motor driver by D/A conversion to adjust the speed of the brushless DC motor.
6. Experimental Verification of the Control System
During the experiment, the branches of the parallel mechanism were first adjusted to position the moving platform at the zero coordinate. Then, the moving platform was made to perform the following composite motion: a 100mm reciprocating translation along the Y-axis and a ±15° rotation along the X-axis. (The composite method diagram is omitted.) The corresponding inverse kinematics solution was obtained using ADAMS software. The discretized result was used in the control program to make the moving platform repeat the reciprocating motion. During this process, the parallel mechanism operated smoothly, and the motion trajectory of the moving platform showed good repeatability.
7. Conclusion
The hardware of the system was constructed using a brushless DC motor as the drive component, a microcomputer as the processor, and a PCI bus control card as the data interface. The control software was written in C++. The inverse kinematics curves of the parallel mechanism were solved using ADAMS software and applied to the experiment.
Experimental results show that:
(1) Due to the strong driving capability of the brushless DC motor, the system response and operating performance are improved.
(2) The control accuracy of each branch of the parallel mechanism can meet the motion requirements of the platform.
(3) The position curve obtained by ADAMS software has been well applied in the verification of physical evidence.
References
1 Kim HS, Tsai L W.Kinematic Synthesis of a Spatial 3-RPS Parallel Manipulator.Journal of Mechanical Design, Transactions of the ASME, 2003,125(1):92～97
2 Cai Zixing. Robotics [M]. Beijing: Tsinghua University Press, 2000.
