Font Size: a A A

Control System Research And Software Designfor The Underwater Self-Reconfigurable Robot

Posted on:2008-08-14Degree:MasterType:Thesis
Country:ChinaCandidate:R Y GaoFull Text:PDF
GTID:2132360212476673Subject:Ships and marine structures, design of manufacturing
Abstract/Summary:PDF Full Text Request
The underwater self-reconfigurable robot is a novel concept in the field of underwater robot design. It consists of a quantity, but only a few types of intelligent modules linked with each other. Although each module has simple functions such as motion and environment perceiving, they can achieve complicated global functions by modules'collaboration. The robot can change its configuration and take different gait to adapt to the environment and missions.On the base of the predecessors'researches, we investigated the problem of module design, control system design, communication realization, configuration changing, then, the top layer control software was completed, the software realize the function of locomotion control.The intelligent modules are basic units of the self-reconfigurable robot, the structure and modularized method is employed in the modules design, which enhances the system's transplant ability and compatibility. On analysis of the reconfigurable method, the mechanical structure of helm mechanism, buoyancy adjusting mechanism, connect mechanism, photoelectric sensor lead system, intelligent unit, and sensor system design have been discussed in detail. Motion control of the module is based on the PID technique. Considering the requirement of coordinated movement, the USR motion pattern has been investigated.Control strategy algorithm, which can be classified into two categories: distributed algorithm and centralized algorithm, is an import part of self-reconfiguration research. Furthermore, it that Agent has special automation, collaboration and so on, gives a new idea for the control strategy of the self-reconfigurable robot. Therefore, in combination with multi-Agent system's principle, this paper constructs a motion planning model based on multi-Agent. This model can divide complicated task into small scale, coefficient and distributed sub-task in order to decrease calculation and complexity.The modules control network is built on the CAN bus technique, and the advanced communication protocol was designed to meet USR'S communication needs, which provide both the point to point transmission mode and the broadcasting one by message filtering. The modules can not only feedback message passively, but also feedback some given message actively according to the motion status of itself.The top layer software is designed to harmonize modules'movement, here, after analysis the requirement of system function, the software data structure and program structure is detailed discussed, then, the program run process is present.Then, the paper presents the principle of configuration switching, demonstrates theory of some typical locomotion, such as eel-like motion, worm motion and octopus motion. And, the thesis provides mathematic methods to solute the joint angle while system configuration is changing from a chain to a circle or micro-adjusting the mechanism. At last, USR's capability of locomotion and self-reconfiguration is examined through experiments.
Keywords/Search Tags:self-reconfiguration, Agent, CAN bus, communication protocol, configuration switching, gait control
PDF Full Text Request
Related items