15 March 2021, Volume 61 Issue 3

    

• Select all
  |

CABLE-DRIVEN ROBOTS
• Select
  Transfer characteristics of high-speed cable forces for spacecraft separation

  HOU Senhao, TANG Xiaoqiang, SUN Haining, CUI Zhiwei, WANG Dianjun

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 177-182. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.012

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Ground simulation tests of spacecraft separation require accurate force predictions as the key link in the tests. This study analyzed the cable force transfer during high-speed spacecraft separation tests. A dynamic model of the cable hoist system was developed based on Newton's laws. The second-order partial differential equation for the test function was then solved using spatial discretization. The solution predicted the force on the rope as it moved to study the factors influencing the force. The model accuracy was verified by numerical examples. The results show that the model can accurately simulate the force on the rope for high-speed spacecraft separation tests.
• Select
  Comparisons of a TBot high-speed cable-driven parallel robot with a classical Delta parallel robot

  PENG Fazhong, DUAN Jinhao, SHAO Zhufeng, ZHANG Zhaokun, WANG Daoming

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 183-192. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.027

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Parallel robots provide better stiffness, better dynamics and more compact structures than serial robots, so they are widely used in high-speed sorting. The rigid limbs in previous robot designs are replaced with cables to simplify and lighten cable-drive parallel robots (CDPR). This paper describes a high-speed CDPR tensioned by a rigid rod and a spring. The robot can execute the SCARA movement in the same way as the classical Delta parallel robot. Comparisons of these two robots show the key design issues for industrial applications of these robots. Kinematic and dynamic models are used to analyse the robot performance for typical trajectories. The results show that this cable driven robot has a simple structure, low energy consumption, large reachable workspace, and great development potential.
  
• Select
  Kinematics analysis and control of under-constrained cable-driven parallel suspension systems

  WANG Xiaoguang, WU Jun, LIN Qi

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 193-201. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.026

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Under-constrained cable-driven parallel suspension systems can be used to support wind tunnels for forced and free dynamics tests. This paper describes the motion characteristics of an under-constrained cable-driven parallel suspension system and a computed torque control method based on a nonlinear disturbance observer. An adaptive particle swarm optimization algorithm was used to solve the coupled equations for the kinematics and statics of the under-constrained system to determine the initial equilibrium state. A computed torque controller was designed based on the dynamics equations with a nonlinear disturbance observer used to compensate for the external interference. The system stability was proven by the Lyapunov function method. Then, the control system was used to control typical dynamic responses in wind tunnel tests, such as single degree-of-freedom (DOF) sinusoidal oscillations and two DOF motion. The results show that the control method effectively compensates for external disturbances with accurate tracking of the desired trajectory while always keeping the cables taut. The system can also simulate an aircraft model’s free motion in uncontrolled directions. Thus, this research describes how to use an under-constrained cable-driven parallel suspension system to improve wind tunnel tests.
• Select
  Vibration suppression in flexible parallel cable structures

  TANG Lewei, SHI Pengshuai

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 202-208. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.024

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Large trusses are widely used in the navigation and aerospace fields due to their lightweight structures, high payload-weight ratios and low costs. Since trusses with large length-diameter ratios are flexible, they have vibrations due to external excitations that impact the entire system performance. A parallel cable configuration acting as an Euler-Bernoulli beam for simplified trusses was used in this study to suppress vibrations in the truss. The control method, which is a function of the cable length, was modeled in MATLAB/Simulink. Parallel cables have active and passive dynamic modes. The results show that the vibrations far from the cable attachment points are effectively reduced by the cable configuration. The cable tension suppresses the short period deformation in the passive cable mode without tension control, while the active mode with tension control suppresses the vibrations over relatively long suppression periods.
  
• Select
  Tracking control of a two-DOF cable-suspended parallel robot using the computed torque method

  REN Ping, HE Mengyi

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 209-216. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.028

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  A computed torque controller was developed for trajectory tracking control of a two-DOF cable-suspended parallel robot. Interval arithmetic was used to analyze the cable tensions' parameter space consisting of the controller parameter and the initial states of the end-effector and to obtain the sufficient conditions of the reliable parameter space which satisfied the tension constraints. This method can track various trajectories such as periodic horizontal, vertical and circular trajectories, as well as point-to-point dynamic trajectories. When the initial state is not on the desired trajectory, the method provides a transition to the desired trajectory. Simulations show that the parameter combinations in the reliable space ensure that the end-effector converges exponentially to the desired trajectory and that the cable tensions remain positive at all times.
• Select
  Cable-driven projectile loading robot

  YANG Jifeng, YAO Rui, CHEN Jie

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 217-223. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.025

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  A three-cable parallel robot was developed for naval missile systems for automatic transfer and loading of the missile with improved loading safety and efficiency. A two-step optimization method was used to completely consider the construction difficulties, weight and cost of the missile loading system with a global search optimization and a local search optimization. The global search considers all the search parameters to determine the tower height as the key design parameter. Then, a cantilever displacement formula is used to relate the cable tower height and diameter with a dimensionless optimization function for the cable force rate of change as a function of the tower height. An optimization function is then used to minimize the maximum cable force. This design optimization method is useful for designing naval automatic projectile loading systems.
• Select
  Optimal solution method for the cable catenary between cable-driven parallel robots

  WEI Huiling, QIU Yuanying, SHENG Ying, CHEN Haichu, LU Qinghua

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 224-229. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.019

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Accurate dynamic models are difficult to develop for long-span cable-driven parallel robots. This paper presents a reduced dimension optimization method for designing nonlinear cable catenaries. This paper gives the differential equation for the catenary and its boundary conditions. The dimension of the transcendental equation of the catenary model was then reduced according to the boundary conditions with the analytical solution obtained using the substitution method and the Taylor expansion method. Then, the Newton method was used to numerically solve the catenary model to study the range and characteristics of the numerical solution. Finally, the effectiveness of the reduced dimension optimization method are verified by examples. This research provides a theoretical basis for accurate dynamic modeling and real-time motion stability control strategies for cable-driven parallel mechanisms.
AEROSPACE AND ENGINEERING MECHANICS
• Select
  Onboard guidance algorithm for the powered landing phase of a reusable rocket

  SONG Yu, ZHANG Wei, MIAO Xinyuan, ZHANG Zhiguo, GONG Shengping

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 230-239. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.015

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Precise soft rocket recovery landings require precise guidance, navigation, and control. A convex optimization guidance algorithm was developed for soft vertical rocket recovery landings and tested in a closed-loop simulation system. A lossless convex model was combined with successive convex iterations to transform the rocket recovery stage guidance problem into a convex optimization problem using the interior point method. The algorithm robustness was evaluated for various factors including random atmospheric disturbances, engine throttling characteristics, random navigation system deviations, and system delays. The simulations show that the onboard guidance algorithm has millisecond convergence and is very robust. Various simulations show that the closed-loop simulation results can provide precise soft landings for rocket recovery even with the combined effects of various disturbances including dynamic environments, control and navigation system deviations, and system delays.
• Select
  Observation mission planning for maneuverable satellite constellations towards multiple targets

  GAN Lan, GONG Shengping

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 240-247. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.013

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Satellite constellations require rapid maneuvering and mission planning for Earth observation missions. This paper presents a mission planning algorithm for a constellation of maneuverable satellites in circular Sun-synchronous orbits. A method is developed to determine the visibility of a single target from a single satellite. Analytical and numerical methods are combined to predict the target visibility and determine the maneuvering and return strategies. The observation mission planning towards multiple targets generates a set of maneuvering strategies with a genetic algorithm used to improve the mission planning efficiency. The numerical results show that the algorithm is accurate and improves the satellite constellation observation ability as well as the average number of observed targets.
ELECTRONIC ENGINEERING
• Select
  EEG classification based on a match filter and metric learning

  LIU Hongma, WANG Shengjin

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 248-253. https://doi.org/10.16511/j.cnki.qhdxxb.2020.22.024

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Brain signal analyses and brain-computer interfaces are key topics in human-computer interaction research. Current electroencephalography (EEG) signal classification methods are complicated and difficult to apply in practice. This paper presents a match filter based classification framework using a hypothesis testing model and a match filter which is a linear function of the signal due to the Gaussian noise assumption. A metric learning based method is then used to estimate the principle component and the optimal covariance matrix to further enhance the model discrimination. The results show that this method provides better recognition accuracy with less computational complexity than other algorithms which makes it more practical.
• Select
  Q-learning environment recognition method based on odor-reward shaping

  RUAN Xiaogang, LIU Pengfei, ZHU Xiaoqing

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 254-260. https://doi.org/10.16511/j.cnki.qhdxxb.2020.25.036

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Q-learning is a model-free iterative reinforcement learning algorithm that is widely used for navigating mobile robots in unstructured environments. However, the exploration and utilization of the environmental data limits the Q-learning convergence speed for mobile robot navigation. This study used the Q-learning algorithm and the fact that rodents use olfactory cues for spatial orientation and navigation to develop a Q-learning environmental cognitive strategy based on odor-reward shaping. This algorithm reduces useless exploration of the environment by improving the Q-learning action selection strategy. Environmental odor information is integrated into the algorithm with the olfactory factor used to weight the Q-learning and the odor-reward shaping in the action selection strategy. The algorithm effectiveness is evaluated in a simulation of movement in the labyrinth environment used in the Tolman mouse experiment. The results show that the Q-learning algorithm with odor-reward shaping reduces useless exploration of the environment, enhances cognitive learning of the environment, and improves the algorithm convergence speed.
• Select
  Stepped sound diffuser arrangement and diffusion analysis

  WANG Jundong, ZHAO Yuezhe

  Journal of Tsinghua University(Science and Technology). 2021, 61(3): 261-268. https://doi.org/10.16511/j.cnki.qhdxxb.2020.26.020

  Abstract ( ) Download PDF ( ) HTML ( )   Knowledge map   Save

  Diffuser sound scattering was predicted using the finite difference time domain (FDTD) method. There were no obvious differences in the diffusion coefficient when changing the distance between the diffuser and the receivers or source in the approximate far field. Then, FDTD was combined with the immune genetic algorithm to optimize the shape of a 6th order stepped sound diffuser which had better diffusion than the original diffuser. The sound scattering from the two diffusers became less diffuse when repeated over many periods. An aperiodic modulation method was used to combine a diffuser and its inverse diffuser using a binary code. Tests show that aperiodic modulated diffusers have better diffusion than periodic diffusers; however, this modulation cannot produce a flat diffusion coefficient curve, so the best diffusion is obtained by arranging the optimized stepped diffuser with the original diffuser together using a binary code.