Dissertation > Excellent graduate degree dissertation topics show

Research on Parallel Route Planning Algorithm for Air Vehicle

Author: ChenQianYang
Tutor: ZhouChengPing
School: Huazhong University of Science and Technology
Course: Pattern Recognition and Intelligent Systems
Keywords: Route planning Parallel arithmetic Sparse A* Search OPEN and CLOSED lists Neural network Motion equation
CLC: V249.1
Type: Master's thesis
Year: 2005
Downloads: 566
Quote: 11
Read: Download Dissertation

Abstract


Route planning is a novelty of information times, which is on the upgrade with the development of information technology, computer, automation control and artificial intelligence technology. Because the flying environment aircraft confronted with is very large and complex, and many kinds of constraints should be considered in the process of route planning, route search algorithm becomes the most challenging problem in route planning. Now , the route planning arithmetics include optimization planning algorithm, deterministic planning algorithm and random planning algorithm. There are always such-and-such problems when using these arithmetics. Many route constraints are over simplified in optimization, particularly the aircraft constraints. And also the search time of this kind of algorithms increases rapidly with the extension of planning environment. The deterministic algorithms may encounter combination-exploding problem. The random arithmetics cannot repeat the planning result and its convergence time is not determinate. For large planning space, huge data and complex constraints, most planning methods have two shortcomings. One is planning time is too long, another is their expansibility isn’t sufficient. Therefore, This paper will research parallel route planning to decrease the planning time and improve the applicability. The time complexity of 3D Sparse A* Search (SAS) arithmetic show its practicality is not enough good, Though this approach efficiently prunes the search space and shorten the search time by incorporating route constraints into search algorithm. After analyzing parallelism of 3-D SAS completely, we find how to deal with the OPEN list and CLOSED list is crucial. If the OPEN and CLOSED lists are global, there is a bottleneck of shared resource. This paper established an applied formula to reduce the influence of this bottleneck. When parallel arithmetic uses distributed OPEN and CLOSED lists, it should prune duplicate nodes in different processors and balance processors’load. This paper also presented solution, but the experiment result is not satisfacfory because the SAS arithmetic is a local search process. The experiment results also demonstrated parallel planning using global OPEN and CLOSED lists shorten the search time obviously, but the expansibility needs to be improved. Further more, this paper presented an approach of route planning based on artificial neural network. This approach has great potential for parallelism and built a neural network for every route constraint. The neural networks are used for energy punishment of constraits. If a route satisfies route constraints, its energy is low, otherwise the energy is high. A motion equation was difined for route nodes and decreased the energy of the planning route. Finally, the result route satisfies all the constraints. A series of nodes were studied, the motion process of these nodes only need local information. Thus, these nodes can be divided into several segments in parallel arithmetic. The experiment results indicated the neural network algorithm can avoid terrain and threat. Its parallel arithmetic is with strong expansibility and satisfies the requirement in application.

Related Dissertations

  1. Development of the Platform for Compressor Optimization Design and Aerodynamic Optimization Design in the Transonic Compressor,TH45
  2. Research on Feature Extraction and Classification of Tongue Shape and Tooth-Marked Tongue in TCM Tongue Diagnosis,TP391.41
  3. Research on Visual Servo System of Mechanical ARM,TP242.6
  4. Municipal tourism land use planning environmental impact assessment,X820.3
  5. Research on the Inteligent System of High Performance Concrete Mix Design in Zhujiang Triangle Area,TU528
  6. Study on Taste Characteristic of Taste Peptide Enzymatic Production from Oyster Base on A Neural Network Method,TS254.4
  7. BP network optimization based on genetic algorithm optimization of the biodiesel process,TE667
  8. The Research on Evaluation of Living Status Systems of Expressway Relocated People,D523
  9. The Identification of Fault Type in Transmission Lines Based on Neural Network,TP183
  10. Enterprise Security Benefit Evaluation and Development Strategies,F272;F224.5
  11. Mine Risk Information Integration and Intelligent Early Warning,X936
  12. Ketones 3D-QSPR study,O641
  13. Study on a New Method for Digital Transformer Protection Based on Wavelet Neural Network and FPGA,TM407
  14. The Optimization Research of Feedforward Neural Network Based on Genetic Algorithms,TP183
  15. Research of Orange Quality Classification Technology Based on Computer Vision,TP391.41
  16. Study on Risk Identification and Evaluation of Manufacturing Green Products R & D,F205;F224
  17. Optimization Study on Gating System and Molding Process Parameters of Injection Mold Based on Simulation,TQ320.662
  18. Research on Automatic Control System of Hump Push Peak Speed Based on Fuzzy Neural Network,U284.6
  19. Prediction of Binding Affinity of Human Transporter Associated with Antigen Processing,R392.1
  20. Software Project Risk Assessment Based on BP Neural Networks,TP183
  21. Microstructural Features and Their Influence on the Properties of Ti-10V-2Fe-3Al Titanium Alloy,TG146.23

CLC: > Aviation, aerospace > Aviation > Aircraft instrumentation,avionics, flight control and navigation > Flight control system and navigation > Flight control
© 2012 www.DissertationTopic.Net  Mobile