The objective of this thesis is to apply Lagrange equations with multipliers to study dynamics and control of Delta parallel robots. Particularly, mechanical model, mathematical model, and control algorithms for Delta parallel robots are developed as a scientific basis for the research and development of parallel Delta robots.
MINISTRY OF EDUCATION VIETNAM ACADEMY OF SCIENCE AND TRAINING AND TECHNOLOGY GRADUATE UNIVERSITY OF SCIENCE AND TECHNOLOGY …… ….***………… NGUYEN DINH DUNG INVERSE DYNAMICS AND MOTION CONTROL OF DELTA PARALLEL ROBOT Major: Engineering Mechanics Code: 9 52 01 01 SUMMARY OF THE DOCTORAL THESIS Hanoi – 2018 The thesis has been completed at Graduate University of Science and Technology, Vietnam Academy of Science and Technology Supervisor 1: Prof. Dr. Sc. Nguyen Van Khang Supervisor 2: Assoc. Prof. Dr. Nguyen Quang Hoang Reviewer 1: Prof. Dr. Dinh Van Phong Reviewer 2: : Prof. Dr. Tran Van Tuan Reviewer 3: : Assoc. Prof. Dr. Le Luong Tai The thesis is defended to the thesis committee for the Doctoral Degree, at Graduate University of Science and Technology - Vietnam Academy of Science and Technology, on Date Month Year 2018 Hardcopy of the thesis can be found at: - Library of Graduate University of Science and Technology - National Library of Vietnam 1 INTRODUCTION The rationale for the thesis Parallel robots are robots with closed kinematics structure in which the links are connected by joints. Although the parallel robot has a complex dynamic structure, and is difficult to design and control, but it has some outstanding advantages over the serial robot: high load bearing capacity, high rigidity due to configuration. They can perform complex operations and operate with high accuracy. Therefore, study on the problem of dynamics and control of the parallel robot in order to take advantage of it is a scientific and practical matter. The objective of the thesis The objective of this thesis is to apply Lagrange equations with multipliers to study dynamics and control of Delta parallel robots. Particularly, mechanical model, mathematical model, and control algorithms for Delta parallel robots are developed as a scientific basis for the research and development of parallel Delta robots. The object and the main content of the thesis Research objects: Dynamics and control of two Delta parallel robots are 3RUS robots and 3PUS robots. The main content of the thesis includes: Study of mathematical and mechanical modeling problems, study of dynamics and control algorithms for Delta parallel robot. The thesis does not study the problem of design and manufacture of Delta parallel robots. The outline of the thesis The outline of the thesis contains Introduction, Four main chapters, Conclusions and findings of the thesis. Chapter 1: Overview of the study of dynamics and control of Delta parallel robot in and outside the country is first presented. Since then, the 2 direction of the thesis has been selected to address scientific significance and practical application. Chapter 2: Presents the construction of mechanical models and application of Lagrangian equations with multipliers to formulate mathematical models for two Delta Parallel Robots. Each robot offers two mechanical models for study and comparison. Chapter 3: Presents some improvements in numerical methods to solve the inverse kinematics and inverse dynamics of parallel robots. Inverse kinematic problem is solved by applying the improved Newton-Raphson method. Inverse dynamics problem is solved by reducing Lagrange multipliers to calculate moments or driving forces in active joints. Chapter 4: Presents tracking control of parallel robot manipulators based on the mathematical model of parallel robots, which is a system of differential – algebraic equations. The trajectory of serial robots described by differential equations is often well studied. While the Delta parallel robot trajectory is based on the mathematical model, the differential – algebraic equations system is rarely studied. Control law such as PD control, PID control, sliding mode control, sliding mode control using neural network are studied in this chapter. CHAPTER OVERVIEW OF DYNAMICS AND CONTROL PARALLEL ROBOT 1.1 Parallel robot Parallel robots usually consist of a manipulator connected to a fixed frame, driven in multiple parallel branches also called legs. The number of legs is equal to the number of degrees of freedom. Each leg is controlled by the actuator on a fixed frame or on the leg. Therefore, parallel robots are sometimes referred to as platformed robots. 1.2 Comparison between Serial and Parallel Manipulators 3 Parallel robot has high rigidity and load bearing capacity due to load sharing of each actuator operating in parallel. The accuracy of the position of the parallel robot is high because there are no cumulative joint errors as the serial robot. While kinematic chains create kinematic constraints and workspace limitations, typical designs have low inertia characteristics. The fields of parallel robot application include: CNC machine, high precision machine, automation machine in semiconductor and high speed and high acceleration electronics assembly industry. A comparison between parallel and serial robots is given in the following table: Table 1.1: Comparison between Serial and Parallel Manipulators STT Features 1 2 3 4 5 6 7 Accuracy Workspace Stiffness Payload Inertial Speed Design/control complexity Serial robot Parallel robot Lower Large Low Low Large Low Simple Higher Small High High Small High complex 8 Singularity problem Some Abundant 1.3 Research on dynamics and control of parallel robots outside of the country 1.3.1 Inverse dynamics of parallel robots On the mechanical side, parallel robots are closed-loop multibody system. Dynamic computation is essential to designing and improving the control quality of parallel robots. The literature on the theory and calculation method of robot dynamics is quite substantial [47, 73, 85-88, 96, 103]. The methods of establishing the dynamic equations of closed-loop multibody 4 system are well documented in [88, 103]. The kinematics and dynamics problems are then more specifically mentioned in the literature on parallel robots [67, 96]. In the above studies on Delta parallel robots, the methods used to establish equation of motion are Lagrange equations with multipliers, virtual work principle, Newton-Euler equation, subsystem When establishing the equation, the bar between the actuating link and the manipulator is modeled with a uniform bar or with a zero-mass bar and two masses at the ends of the bar. Up to now, there have been no comparative work on these two types of models. 1.3.2 Tracking control of parallel robots The documentation on robot control is very rich. There are various approaches to controlling robots given by Spong and Vidyasagar [90], Sciavicco and Siciliano [87]. However, these works are less focused on the specific problems of parallel robots. Recently, the works on improving the control quality of Delta robot was also published quite a lot. These works develop control law based on the equations of motion, which are obtained by simplifying the dynamics model of each parallelogram by a zero-mass bar with two mass points at both ends. Model linearization methods are used to establish simple control laws. Hemici et al. [80-82] designed PID, H controllers based on linear models to robustly control Delta robot. This model was also used by A. Mohsen [68] to establish PD and PID control laws in combination with fuzzy supervision to perform motion tracking control of manipulator. These works use different controllers for the purpose of forcing the movement of the manipulator to follow a desired trajectory. These controllers partly meet the desired requirements. However, there is a lack of 5 comparative studies of controllers and recommendations on how to choose an appropriate ones. 1.4 Studies in Vietnam The research in Vietnam mainly focuses on solving the kinematics problem, establishing the equation of motion and presenting the method of solving the equations of motion. Control problems are little researched. 1.5 The research problem of the thesis From the review and evaluation of the work that scientists have been working on in Delta parallel robot, this thesis will investigate the following issues: Development of the solution for the inverse dynamics problem with the aim of improving numerical accuracy. Study and comparison of different dynamic models for a parallel robot, the complexity of the models and their effect on the computational torque moment. On that basis, it is advisable for the user to use a suitable model. Design of direct control law based on differential – algebraic equations. Research comparing the quality of the controllers using different mechanical models. Conclusions of chapter Based on the results obtained from domestic and foreign researches, the thesis has identified the need for in-depth research in order to improve the quality of control for parallel robots, mechanical and mathematical modeling and numerical algorithms for solving dynamic and control problems for two parallel robots, 3RUS and 3PUS. CHAPTER BUILDING THE MECHANICAL MODEL AND MATHEMATICAL MODEL FOR DELTA PARALLEL ROBOT In this thesis, the new matrix form of the Lagrange equations with multipliers [51] is used to establish the equation of motion of two parallel 6 robots, the 3RUS robot and the 3PUS robot. With the MAPLE or MATLAB software, we obtain the analytic form of differential – algebraic equations describing the movement of parallel robots. 2.1 Dynamic model of Delta parallel robot 2.1.1 Dynamic model of Delta parallel robot 3RUS From realistic models of robots from Figure 2.1, it can be seen that the parallelogram will make the kinematic and dynamic computation on the robot quite complex. For simplicity we build two models of robot dynamics based on real model as follows: R1 m1,L1, I1 m2, L2, I2 mp r Figure 2.1: Delta parallel robot 3RUS Model 1: The parallelogram mechanisms is modeled by a bar with a uniformly distributed mass over the length of the bar. The mass and length of the bars correspond to the mass and length of the parallelogram. Model 2: The parallelogram mechanisms is modeled by a weightless bar with a concentric mass at both ends, the mass of each bar end equals half the mass of the parallelogram. 7 2.1.2 Dynamic model of 3PUS Delta parallel robot Spatial 3PUS Delta robot is a variant of the 3RUS robot when replacing rotary actuation joints linear actuation joints as shown in Figure 2.4. The 3PUS robot is also equipped with two dynamic models similar to the 3RUS. Figure 2.4: Delta parallel robot 3PUS 2.2 Establish equations of motion of the Delta parallel robot Applying the new matrix form of the Lagrange equation with multipliers [4, 51], the equation of motion of two 3RUS and 3PUS robots is the differential - algebraic equations of the following general form: M s s C s, s s g s ΦTs s λ τ (2.20) (2.58) f s 2.3 Compare the equations of motion of robot models From the equation of motion of model 1 and model 2 of each robot we have the comparison table as follows: 8 Table 2.1 Compare the equations of motion of Models 1 and 2 Number of Degree of freedom generalized coordinates Model 1. 3 Model 2. 3 3x3 + 3 = 12 3 + 3 = 6 Constrained equations 9 3 Lagrangian multipliers 9 3 equations 21 9 M M (s ), C(s, s ) M(s) const , C(s, s) Matrices M and C From Table 2.1 we find that the equation of motion of model 2 is simpler and easier to establish than model 1, but the inertia effect is not clear. 2.3 Conclusions of chapter The establishment of analytical equation of the equation of motion of Delta parallel robot is a very complex problem. Using the symbolic programming technique, this thesis has achieved some new results as follows: Using the new matrix form of equations Lagrange with multipliers [51], the differential - algebraic equations describing the motion of the two kinds Delta parallel robot (robot 3RUS and robot 3PUS) has established analytically. In addition to establishing equations of motion in view of rigid body, the thesis also provides a simple equation for motion equation by replacing the parallelogram mechanisms by two mass points. These mechanical models are the basis for computational dynamics and control of parallel robots 3RUS and 3PUS. 9 CHAPTER NUMERICAL SIMULATION OF INVERSE KINEMATICS AND INVERSE DYNAMICS FOR DELTA PARALLEL ROBOT Based on the explicit analytical form of the differential - algebraic equations description of the motion of the Delta parallel robot set up in Chapter 2, this chapter applies and develops numerical algorithms to solve the inverse kinematic and inverse dynamic problem for parallel robots 3RUS and 3PUS. 3.1 Calculation of inverse kinematic parallel robot by improved Newton-Raphson method The constrained equations of robot are rewritten in vector form as follows: f (s) f (q , x) (3.1) r n m where: f , q , x Contents of the inverse dynamics problem: Given the motion law of the manipulator, it is necessary to find the law of motion of the driving joints. Here, we will present an improved Newton-Raphson method [4] to solve the inverse kinematic problem: Step 1: Correct the increment of the vector of generalized coordinates at 0 by drawing time t0 = 0. First, we can determine the approximate vector q method (or experiment). Then apply Newton - Raphson methods to find a better solution of q0 from nonlinear equations (3.1). Step 2: Correct the increment of of the vector of generalized coordinates at time tk+1. The approximate initial value of qk+1 is approximated by the formula: k (t ) q k 1 q k q k t q In the robot kinematics computation [87], the infinitesimals of order n≥2 are often neglected in the initial approximation of Newton-Raphson. In this thesis, we take into account the second order infinitesimals, neglecting 10 the infinitesimals of order 3 and taking the formula (3.14) as the approximation of the original Newton-Raphson loop. After each step of calculating the coordinates of the joints using the improved Newton-Raphson method, the generalized velocity and acceleration of the joints are calculated by the following formulas: q Jq1Jx x (3.4) 1 (3.6) J q J q q J x x J x q x 3.2 Numerical method for solving the inverse dynamics problem of parallel robots 3.2.1 Inverse dynamics problem The general equations of motion of the robot is as follows: + g(s) +ΦTs (s) M(s)s +C(s,s)s (3.20) f (s) (3.21) r f Let q a be the vector of coordinates of active joints, z is the vector of redundant coordinates (including passive coordinates and endeffector coordinates). Symbol: T s qTa , z T , s n , q a f , z r , n f r The inverse dynamics problem of the parallel robot is expressed as: The equation of motion of the robot is known as in (3.20), (3.21), given the m motion law of the operation x xt , x Determine τ a f the driving momen / force required to produce the desired motion. 3.2.2 Solving the inverse dynamics problem by eliminating the Lagrange multipliers [4] Through the inverse kinematic with the given trajectory of the mobile platform center, we have found the vector s t , s t , s t From this, mass matrices, centrifugal inertia and Coriolis matrices, matrices Φ s , as well as the vector g(s) have been completely determined. Thus, Equation (3.20) is a linear algebraic equation with unknown driving torque vector τa and 11 Lagrange multipliers λ with equal numbers of equations and numbers. Thus, we can directly solve this system of equations and then separate the resulting momen. In this thesis, we will not directly solve equation (3.20) but try to eliminate Lagrange multipliers λ , transforming the system of differential - algebraic equations (3.20), (3.21) into the system of equations of only unknowns of only joint moments τa as follows: We put the symbol [4, 47]: E R s R(q a , z) 1 (3.24) Φ z Φ q f f f f and where E is the unit matrix size Φ z , Φ a z T q a Left multiplying both sides of (3.20) by matrix R s and simplying it yields RT s M s s RT s C s , s s RT s g s τa (3.29) The expression in the left-hand side of equation (3.29) are known from the results of the inverse kinematics problem. Thus, active joint moments are calculated according to this equation. 3.3 Numerical simulation of inverse kinematics and inverse dynamics of Delta parllel robot 3.3.1 Numerical simulation of 3RUS inverse kinematics of robot To evaluate the correctness of algorithms and calculations of the thesis, we computed the inverse dynamics problem of 3RUS robot with the DELTA-IMECH program developed based on MATLAB software. For comparison, the robot parameter data and manipulation motion are given in [61] of Y. Li and Q. Xu. Using the DELTA-IMECH program we obtain the results of the numerical simulation of inverse kinematics and have the following comparison table: 12 The results of the thesis 100 Joint1 Joint2 The results of work [61] Joint3 [degree] 80 60 40 20 [rad/s] 0.5 t[s] 1.5 -1 -2 0.5 t[s] 1.5 [rad/s2] -2 -4 0.5 t[s] 1.5 Figure 3.11: Comparison of the results of the inverse kinematic problems against the literature [61] 13 Comment: Figure 3.11 shows that the results of the inverse kinematics of the thesis are consistent with the results of the paper [61]. 3.3.1 Numerical simulation of inverse dynamics of robot 3PUS The robot parameter data and movement of the manipulator as follows: L 0.242, R 0.16, r 0.029(m), m1 0.12, m2 0.15, mP 0.2(kg) 2 2 xP 0.05cos t ; yP 0.05sin t ; zP 0.5 (m) T T Numerical simulation results were computed based on models 1 and 2 of the 3PUS robot using the DELTA-IMECH program. Model 1 Model 1 T = 1 (s), (Fast motion manipulator) -4 -4.4 [N] -4.5 -4.6 [N] -5 -5.5 -6 -4.8 -5 0.2 0.4 0.6 0.8 -5.2 0.2 0.4 t[s] 0.6 0.8 t[s] T = 10 (s), (Slow motion manipulator) -3 -3 -4 [N] [N] -4 -5 -6 -7 -5 -6 -7 10 t[s] 10 t[s] Figure 3.22: Results of numerical simulation inverse dynamics robot 3PUS 14 Comment: When motion of the manipulator is fast, the results of the two models are different. When motion of the manipulator is slow, the results of the two models are the same Conclusions of chapter The contribution of the thesis in this chapter is: 1. Develop a program, called DELTA-IMECH program, to calculate numerically inverse kinematics and inverse dynamics problems of 3RUS and 3PUS. The results computed by this program are consistent with the literature [61], [92]. This proves that the equations of motion of robots that have been established and the algorithms and programs in DELTA-IMECH are correct. 2. The numerical simulation results show that when the movement of the manipulator is not fast, a simple robot model can be used to compute the dynamics of two types of research robots. However, when using simple models the inertial effects of spatial rigid bodies are not reflected in the equation. That is the limitation that should be considered. CHAPTER TRAJECTORY TRACKING CONTROL OF THE DELTA PARALLEL ROBOT BASED ON MECHANICAL MODELS The use of inverse dynamic methods to control position of serial robot has been discussed extensively in engineering [1, 87]. In this chapter, based on the differential - algebraic equations written explicitly in Chapter 2 and the numerical method for solving the inverse dynamics problem in Chapter 3, the PD, PID, Sliding mode control, Sliding mode control using neural network controller is built for 3RUS and 3PUS Delta parallel robots. 4.1 Overview of the tracking control of the manipulator The task of the trajectory tracking control problem of the manipulator: To guarantee that the end-effector moves along the desired trajectory in the 15 d work space. Given the desired trajectory x (t) , it is required to control the actual trajectory x to satisfy the following condition: || x xd || (4.1) 4.2 Trajectory tracking control of the parallel robots in joint space based on Lagrange equations with multipliers 4.2.1 Background of dynamics of closed-loop multibody systems Using the Lagrange equations with multipliers, equation of motion of parallel robots in the form of (3.20) and (3.21) and equation (3.20) is transformed into (3.29), we proceed to modify this equation in the following form: a C s , s q a g s d(s, s) τa M s q (4.12) M s : RT (s)M s R (s) T (s, s) C s , s R (s) Where : C s , s : R (s) M s R T g s : R (s)g s d(s, s ) : RT (s)d(s, s) Equation (4.12) is the basis for establishing the control law for parallel robots. 4.2.2 Development of control algorithms In this thesis the trajectory tracking control algorithms are built based on the differential - algebraic equations describing the motion of parallel robots. The stability and tracking properties of the control algorithms are well proven. 4.2.2.1 PD control u t M s ν C s, s q a g s d s, s da KDe a KPea with: ν q (4.18) (4.19) 16 K P diag kP1 , kP2 ,, kPna , K D diag kD1 , kD ,, kDna , kPi 0 , kDi 4.2.2.2 PID control u t M s ν C s, s q g s d s, s t a da K De a K Pea K I ea ( )d With: ν q K diag k P1 , k P ,, k Pna , 0K D diag k D1 , k D ,, k Dna , P K I diag k I , k I ,, k Ina where: kDi 0, kPi 0, kIi 0, kDi kPi kIi 0; i 1,2, , na (4.26) (4.27) 4.2.2.3 Sliding Mode control a C(s, s)q a g d M(s)Λe a u t M(s)q C(s, s ) Λe a K PD ν K S sign ν T where: sign ν sign v1 , sign v2 ,, sign vna K diag kPD1 , kPD ,, kPDna , kPDi PD K diag kS1 , kS ,, kSna , kSi S d d (4.45) (4.46) (4.47) 4.2.2.4 Sliding mode control using neural network da C(s, s )q da g d M (s ) Λe a C(s, s ) Λe a Kν u M (s )q ν 1 Wσ ν i i ν w (4.62) (4.63) in which K is positive definite symmetric matrix of size na , and , 4.3 The numerical simulation of the control law of the Delta parallel robot based on the mechanical models 4.3 The numerical simulation of the control laws of the robots 3RUS and 3PUS Delta parallel 17 Table 4.1: Comparison of robot trajectory errors Using model in controller Using model in controller PD control, the robot errors and disturbance -3 0.015 0.01 x 10 0.01 0.7 [m] 0.005 -2 0.6 0.8 -0.005 -0.01 ex x 10 -2 -4 0.6 0.015 0.005 [m] -3 0.2 ey 0.4 0.6 -0.005 ez 0.8 -0.01 0.7 0.8 ey ez ex 0.2 0.4 t[s] 0.6 0.8 t[s] PID control, the robot errors and disturbance 0.015 0.01 -2 0.6 0.005 [m] x 10 0.7 0.005 0.8 [m] 0.01 -0.005 -0.01 e e x -3 0.015 -3 0.2 0.4 0.6 z 0.8 -0.01 0.7 ex ey 0.8 -0.005 e y x 10 -2 -4 0.6 0.2 0.4 0.6 t[s] ez 0.8 t[s] Sliding mode control, the robot errors and disturbance -4 0.01 0.005 -5 x 10 0.005 0.7 0.8 -0.01 ex ey -0.005 ez 0.2 0.4 0.6 0.8 0.7 ex 0.8 0.2 ey 0.4 0.6 t[s] t[s] 0.6 -0.01 x 10 -5 0.01 0.6 -0.005 -4 0.015 [m] [m] 0.015 ez 0.8 18 Sliding mode control using neural network, the robot errors and disturbance 0.015 -4 0.01 0.015 x 10 0.01 -2 -4 0.6 0.7 0.8 -0.005 -0.01 e e x 0.2 e y 0.4 0.6 0.8 -0.01 0.7 0.8 ex -0.005 z x 10 -2 0.6 0.005 [m] [m] 0.005 -4 0.2 t[s] ey 0.4 0.6 ez 0.8 t[s] 4.3.3 Comments on numerical simulation results It is easier to design the control law based on the model 1 than on the model 2 because the model 1 has larger number of equations and is more complicated than model 2 (see Table 2.1) When using the PD, PID to control the real robot with exact knowledge of the dynamic parameters and no disturbance, the use of model to design the control law produces less accurate results compared to when using model 1. When using PD and PID control laws to control real robots without knowing exactly the dynamic and disturbance parameters, both models give inaccurate results (~10-3 m see Figures 4.11, 4.12). , 4.19 and 4.20). When using the sliding mode controller, controller based on the principle of sliding using neural network for the robot, with exact knowledge of the dynamic parameter and without disturbances, and without exact knowledge of the dynamic parameter and with disturbances, same good results are obtained (accuracy of ~10-4 m is shown in Figure 4.27 to Figure 4.36). So when using the sliding mode control law and sliding mode control using neural network, we only need to use simple model to design the 19 control rules. The control law will be very simple but still give good results like when we use complex models. The comments for control law design for the 3PUS robot are similar to those of the 3RUS robot when simulating the tracking trajectory. Conclusions of chapter The contribution of the thesis in this chapter is: 1. To prove theoretically the stability of PD, PID, sliding mode control and sliding mode control using neural networks laws of the parallel robot based on the differential - algebraic equations describing the motion of the robot. 2. When the mechanical model of a robot is correctly constructed and there is no force disturbance during operation, the PD and PID control laws can be used but must be set up from a complex mechanical model (model 1), so that it still ensures the desired trajectory of operation. 3. When the mechanical model of the robot is not properly constructed and in the presence of disturbance during the working process, modern control rules such as sliding mode control sliding mode control using neural network are used. It is required to design the controller only from a simple mechanical model (model 2) but the controller still ensures the desired traction of the operation. CONCLUSIONS AND FINDINGS OF THE THESIS The findings of the thesis 1) Applying the new matrix form of the Lagrangian Equation with multipliers establishes differential - algebraic equations describing the movement of 3RUS and 3PUS Delta parallel robots. Equations are explicit analytical equations but are quite complex. 20 2) Transformation the differential - algebraic equations is transformed into a system of ordinary differential equations based on the idea of W. Schiehlen and colleagues [28]. Then the driving moment/ force is calculated. This method is used to solve the inverse dynamic problem of parallel robot 3RUS and 3PUS. The simulation results obtained according to the proposed algorithm are consistent with known results. 3) Transformation the differential- algebraic equations of motion of parallel robots into the ordinary differential equation with redundant coordinates the new mass matrices M(s) , the inertial force matrix and the new coriolis and centrifugal matrix C(s, s ) are obtained. The PD, PID, sliding mode control, and sliding mode control using neural network algorithms developed for serial robots then can be applied to control parallel robots. Stability of PD, PID, sliding mode control, sliding mode control using neural network is proven based on mathematical model of robot which is the system of differential- algebraic equations 4) Writing a program named DELTA-IMECH for calculating inverse kinematics, inverse dynamics, and motion control of two parallel robots 3RUS and 3PUS. Some examples calculated by the DELTA-IMECH program are consistent with known results. The algorithm and program of DELTA-IMECH are correct and reliable. 5) In engineering, the rigid body moving in space is sometimes replaced by a model of mass points. In this thesis, the dynamic and control calculations with this model are presented for the Delta parallel robot 3RUS and 3PUS. Caution should be taken, because in the simplest model the inertial effect of the system is not reflected accurately. Suggestions Passive-based control of parallel robots. Dynamics and control of parallel robots taking into account the elasticity of the links. LIST OF PUBLISHED WORKS Nguyen Van Khang, Nguyen Quang Hoang, Nguyen Duc Sang, Nguyen Dinh Dung (2015), “A comparison study of some control methods for Delta spatial parallel robot”. Journal of computer science and cybernetics, VAST, Vol. 31, pp 71-81. Nguyen Van Khang, Nguyen Quang Hoang, Nguyen Dinh Dung, Nguyen Van Quyen (2016), “Model-based Control of a 3-PRS Spatial Parallel Robot in The Space of Redundant Coordinates”. Journal of Science and Technology Technical Universities, Vol. 112, pp. 49-53. Nguyen Quang Hoang, Nguyen Van Khang, Nguyen Dinh Dung (2015), Influence of models on computed torque of delta spatial parallel robot. Proceedings of the 16th Asia Pacific Vibration Conference, Hanoi, pp. 791-798 Nguyen Dinh Dung, Nguyen Van Khang, Nguyen Quang Hoang (2016), Modelling and sliding mode control based models of a 3RUS spatial parallel. Proceedings of International Conference on Engineering Mechanics and Automation (ICEMA4), Ha Noi, pp. 198-205. Nguyen Van Khang and Nguyen Dinh Dung (2013), Về một dạng thức mới phương trình chuyển động của robot song song, The 2nd Vietnam Conference on Control and Automation VCCA-2013, Da Nang, pp. 457466. Nguyen Van Khang, Nguyen Quang Hoang, Nguyen Dinh Dung, Mai Trong Dung (2015), Xây dựng mơ hình cơ học cho robot song song Delta không gian 3PUS. National Conference on Engineering Mechanics, Da Nang, pp. 398-406. Nguyen Van Khang, Nguyen Dinh Dung, Nguyen Van Quyen (2016), Điều khiển bám quỹ đạo robot song song Delta khơng gian 3-PRS dựa trên mơ hình hệ các phương trình vi phân-đại số. The Vietnam Conference on Mechatronics 2016 (VCM 2016), Can Tho, pp. 830 – 840 ... study of dynamics and control algorithms for Delta parallel robot. The thesis does not study the problem of design and manufacture of Delta parallel robots. The outline of the thesis The ... of the thesis Research objects: Dynamics and control of two Delta parallel robots are 3RUS robots and 3PUS robots. The main content of the thesis includes: Study of mathematical and ... The inverse dynamics problem of the parallel robot is expressed as: The equation of motion of the robot is known as in (3.20), (3.21), given the m motion law of the