Study Of Neural Network PID In Control System Of Folder

Synchronous speed control of the fold machine is a very important section in the whole folding process. Synchronizing speed,especially the speed matching between feeding motor and folding motor,namely the speed regulation of stepping motor mentioned in the thesis,has a determining influence to the yield of fold accuracy.Now the speed control loop generally adopts traditional PID control algorithm, PID control algorithm has some characteristics, such as simple structure,convenient realization and good adaptability, but in the speed control loop its application encounters following questions: slow regulation, bad anti-jamming capability, bad stability etc. When its set point is changed under traditional PID control algorithm, the time that the measured value levels off on the set point is very long, at the same time, the speed control loop tends to have big fluctuation because of some disturbances caused by the motor load’s change, synchronized signal’s loss and other mechanical aspect’s effects. These factors restrict the increase of speed control loop’s accuracy,even more it becomes a bottleneck that blocks speed control effect’s more improvement, if we can find an appropriate algorithm to resolve this problem,this will make a far-reaching effect in whole folding process.This thesis uses these Problems as background, through analyzing the operating principle and characteristic of stepping motor carefully, researches BP neural network PID deeply,Finally validates the feasibility and advantage of this system through simulation experiment. meanwhile introduces the design philosophy, method and content with adopting the ARM7 technology and BP neural network PID algorithm.The main job and innovations are as follow:1. Mainly introduces the folding machine’s operating principle, synchronous control strategy and software system, on the basis of predecessor’s work,analyzes the relative merits of the traditional synchronous control strategy,aiming at the instability and time lag of motor speed in traditional synchronous control strategy and combining with self project practical experience to design a kind of new synchronous control strategy.2. Makes a simulation experiment about these two control effect under the MATLAB situation, then educes simulation curves and has a analysis and sum-up about them, the outcome of the test indicates that the improved BP neural network PID control algorithm designed in this thesis can achieve a more steady control effect than traditional PID control algorithm and has a very good popularizing value.3. The thesis expatiates the application of improves BP neural network PID in ARM7 control Process carefully at HD5000 automatic folder. Overlay folder machine at different operating conditions to achieve better speed control, The width of the page overlay error can be controlled within 1mm, Meet the design requirements.