Dissertation > Excellent graduate degree dissertation topics show

Multi-Objective Performance Criteria Optimization for a Redundant Dof Robot Based on Differential Evolution

Author: AiYue
Tutor: JiangYanZuo
School: Harbin University of Science and Technology
Course: Control Theory and Control Engineering
Keywords: redundant robot gradient projection method differential evolution global performance optimization
CLC: TP242
Type: Master's thesis
Year: 2011
Downloads: 24
Quote: 0
Read: Download Dissertation

Abstract


Because of the character of the redundant degree of freedom robot which is different from the traditional non-redundant robot, the operation performance is able to be improved by using the various joint configuration produced through the self-motion, when it can make sure the kinematic law of the end effector, which is a significant implication for the redundant DOF robot. The performance optimization for a redundant robot usually adopts the approach based on gradient projection method, which is essentially a local optimization algorithm. It is easy to fall into local optimization. Consequently, it is hard to make sure that redundant robot achieves optimization globally.In view of this, this paper aiming at a planar 3 degrees of freedom redundant robot presents a single or multiple performance optimization method based on differential evolution (DE). The algorithm is improved by analyzing the basic idea and process, which is designed to realize the single objective performance optimization for the redundant degree of freedom (DOF) robot; Moreover, a method based on non-dominated sorting differential evolution (NSDE) for the multi-objective performance optimization of the redundant DOF robot is proposed. By analyzing cross operator, mutation operator, selection operator and so on, it is successful to design the step of multi-objective performance optimization for the redundant DOF robot based on NSDE.At last, this paper utilizes MATLAB as the mathematic tool to experiment. The results show that the optimization method based on DE is able to optimize the single or multiple objective performance for the redundant robot. Comparing with gradient projection method, it is so hard for DE method to fall into the local optimization that the redundant robot can achieve optimization globally, and the effect is more evident.

Related Dissertations

  1. Research on Scheduling of Whole-set Orders in JSP Based on Differential Evolution Algorithm,F273
  2. Research and Improvement on K-Means Clustering Algorithm,TP311.13
  3. Stochastic Optimal Dispatch of Power System Considering Uncertain Line Failures under Extreme Weather Disaster,TM73
  4. Research on K-means Optimization Clustering Algorithm,TP311.13
  5. Research and Implementation of Jobs Intelligent Dispatching System for a Kind of Hyrbid Assembly Lines,TH186
  6. The Modified Harmony Search Algorithm with Control Parameters Co-evolution and Its Application,TP391.3
  7. Model-based rapid test method equipment,TJ06
  8. Layout Pattern and Opposition-based Cooperative Co-evolutionary Differential Evolution Algorithm and Application,TP301.6
  9. Parallel and Dual-systems Cooperative Co-evolutionary Differential Evolution Algorithms and Their Application,TP18
  10. Improved Analysis and Applied Research of Artificial Glowworm Swarm Optimization Algorithm,TP18
  11. Nonlinear Predictive Control for Polypropylene Grade Transition Process Based on Neural Network Modeling,TQ325.14
  12. Distribution Network Planning Including Distributed Generation,TM715
  13. Solving Job Shop Scheduling Problem Based on Natural Computation,TP18
  14. Research on Improved Differential Evolutionary Algorithm for Multi-Objective Optimization,TP301.6
  15. Research on Power Control Technologies in Wireless Mesh Networks,TN929.5
  16. The Research of Scraper Selecting Based on the Differential Evolution Algorithm,TD422.4
  17. Multi-Agent Differential Evolution Algorithm and Its Applications in Optimization of Fermentation,TP18
  18. Research on Predictive Control Based on Intelligent Algorithms,TP18
  19. Study on Nonlinear System Identification of DRNN Based on Differential Evolution,N945.14
  20. Study on Fuzzy RBF Neural Network Based on RPROP-DE Algorithm,N945.14
  21. Optimization run closed-loop control system design based on the the coking production process of performance evaluation,TP273

CLC: > Industrial Technology > Automation technology,computer technology > Automation technology and equipment > Robotics > Robot
© 2012 www.DissertationTopic.Net  Mobile