Research on Motion and Control Embedded System of Spherical Underwater Robot
|School||Beijing University of Posts and Telecommunications|
|Course||Detection Technology and Automation|
|Keywords||Underwater spherical robot Motion Control PID neural network Embedded systems|
Underwater spherical robot is a the severe nonlinear system each degree of freedom in the direction of strong coupling, is often difficult to obtain accurate equation of motion, and motion control technology is the foundation and prerequisite for underwater spherical robot system, a stable and reliable control system underwater spherical robot to complete the premise and guarantee of the expected tasks and underwater operations. The purpose of this paper is to explore the application of neural network control technology in underwater spherical robot motion control, and as the research object to a particular underwater spherical robot design with excellent performance motion control system. The thesis is based-2 Type BYSQ underwater spherical robot hardware platform, on the basis of detailed analysis of underwater spherical robot system, PID neural network control algorithm to control the simulation, and the method analysis, design based on ARM processor AT91SAM7X256 of a set of embedded motion control system, and finally a brief introduction to the operation of the system debugging. Underwater spherical robot system is a great inertia, large delay and time-varying characteristics of complex control object, the conventional cascade PID control system is difficult to achieve a better quality of regulation, based PID neural network control strategy can be obtained good control effect. The simulation results show that, compared with the traditional PID control, PID neural network (PIDNN) control not only improves the dynamic performance of the system, and to some extent also object to overcome the time-varying control effect. Introduced in the motion control system hardware components and working principle of the control system, with a ARM7 AT91SAM7X256-as the processor, the feedback of the actual attitude and heading through accelerometers and gyroscopes, to form a neural network algorithm to adjust the closed-loop control system. In the system part of the circuit, ARM7 minimum system circuit, communication circuit, the sensor circuit part of the design, and the related modules were analyzed. Introduced in software design, the basic ARM software design techniques, given the hardware interface of the program design process, the ARM program development using C language programming, highlighted a neural network control algorithm and online learning theory, sensor data reception and Communication Program Design. Upper interface of the host computer with VC6.0 design, the use of VC MSComm control to access the serial port of the computer, using asynchronous communication between the ARM7 and the host computer.