Development Of Control System Of MPF And The Modeling And Control Algorithms Of Plate Forming

The technique of multi-point forming(MPF) capitalizes on the envelope surface of a series of regular array,tunable height base bodies forming "flexibility multi-point dies"replacing of the traditional dies. How to accurately control the thousands of base bodies's height form the except enveloping surface is an important problem for MPF technique; moreover, during the shape forming, because of effects that come under all kinds of forming complication, actually , it demands forming many times to reach the appeal, but which control algorithm to choose can let the process automatically achieve the optimum efficency is another important problem.The paper is based on the national 211 project fund and the sixth framework program in European engineering project "the digitalization adjusting device of aircraft skin fabrication based on MPF". Using RS-485 communication bus and distributed computer control system to design the MPF control system that with the compound manners; using neural network self-adapt control for the closed-loop control system, and has obtained a favourable impression in the experimentation.1. in the old MPF control system, it does't use the serial way, but the parallel way; The serial form adjustment manners costs less but it takes too much time, and adjustment efficiency is too low; in the parallel form adjustment manners, every base body uses one electromotor to set position, it has cut down the time dumplicately that it takes to set position, but the cost is increasing along with the electromotors. Further more, due to the number of base bodies increasing, there is no room left for the electromotors. If there are too much base bodies and their volume are small,the electromotor will not be installed because of no room left, so the parallel form adjustment manners can not adapt when the base bodies are small and the equipment asks for high accuracy. In my paper, the compound manners is used, it does not only gives consideration to the time of form adjustment, but also resolves the problem of accuracy, so it is especially acceptable for the micromodule equipment with small base bodies.2. Research on closed loop control of the MPF equipment based on distributed control system, and use RS-485 communication bus to realize the communication between the upper subsystem and lower subsystem, also adjust the position of base bodies in the way of using the direct current electromotor as executive device. Design the electric circuit of control unit, which is used to adjust the position of every group of base bodies, and look the control unit as basic unit of the control system。The control unit is regarded microcontroller (the AT89C51 microcontroller) as the core plan, consisting communication interface circuit,clock circuit,watch dog circuit,drive circuit,select circuit,testing circuit and reactive circuit etc. It must be provided with data transmission,data acquisition,data handing and controlling the actuator etc function, and adhibit several anti interference measure in allusion to the environment of the bad practical problem. Moreover , import the Three-Dimensional Laser measurement to measure the shape of the product ,and transfer the measuring result to the upper subsystem as feedback quantity ,so now it can achieve the closed loop control.3. Model the forming process of MPF based on the characteristics of MPF ;describe the discrete fourier transform of the 3-DM surface shape; the MPF sampling theorem is proposed as the Shannon sampling theorem, and research the relationship between the center distance, which is the distance of two adjacent base bodies, and the maximum frequency component based on the theorem.The feasibility and efficiency of MPF is proved by the power spectrum density theoretic, and the robustness of the random measurement error, which is coming after discrete fourier transforming, is proved also by the power spectrum density theoretic。Apply the window function method to design FIR Filter, which is used to reduce the random measurement error that comes forth in the process of the measurement, and make sure the equipment form the product rapidly. Also do some Z-transform with the forming procedure function, then the function will be used in the difference equation.4. the MPF is a nonlinear complex process, is a discrete process caused by another event, which is caused by the interaction according to some rules. because of the characteristic of the MPF, nonlinear and time-varying, design the control system with the neural network adaptive control algorithm, and adopt serial-parallel approaches to be the neural network identification unit, using one stair time-delay link z^-1 and RBF network N_1,5,1; using RBF and time delay achieve neural network controller。Last, simulate the control system in Matlab.5. Check up the experimental result of the control unit, what is the basic composition of the MPF control system, to identify whether the position error of the base body meets demand, and analyses the reason why it comes out. According to experiments and research, the accuracy of examination can satisfy the engineering requirements.With the rapid de velopment of modern industry, classes of product are more and more, and the improvement of the product is faster and faster, so small-batch product is more popular, the improvement of the technology of MPF is needed, because of the development of automobile industry and aerospace industry, too. The control system of MPF is very important for the equipment of MPF, and its research degree directly restrict the development of MPF, but it is a complicated technology, involving many aspects such as computer,electron and machinery etc, so a lot of problems needed to research more.