The embedded machine arm controller design of type MRC in neural network application

At present, industry uses robots for better control. The program database needed is extremely huge, so it is not easy to learn and operate by the personnel. The use of robots turns the humanization operation simple, and price rationalization can be applied more extensively. This thesis draws up to put forward a type of the Embedded MRC network application of nerve theories. It goes into the type Atmega16 of IC microprocessor control of machine arm design, by the mode reference control (MRC) the structure of a nerve network. Through training we can drive the machine arm to make the accurate position control. This paper describes, the design of a PC appearance to make the sketch editor that the diagram control, to revise each axial position control in machine arm through the diagram easily. Design principles that control ascends make valid Embedded go into type design. In actual situation the design ability promote and simplify the control of the robot order through the RS232.