Research And Application Of Micro-EDM Intelligent Control Algorithm

As an important method for manufacturing precise and micro products, micro-EDM greatly attracts academia and industry concerns. With the rapid development of micro and nano technology, micro-EDM machine is critically required to improve its working efficiency and the quality of products. Valid micro-EDM control system lays prerequisites and foundations for accurate, efficient and stable processing. However, as a non-linear system of complex object, complex environments and complex tasks, the control system is a bottleneck for the development of micro-EDM. Intelligent control method provides an efficient way to deal with such complex non-linear problems, and has become a significant research topic for micro-EDM control method. According to characteristics of micro-EDM processing, it is quite necessary to do research on intelligent control algorithms which are suitable for micro-EDM from the aspects of hardware and software.Given the features of high pulse discharge frequency, serious discharge waveform distortion and complex interface in micro-EDM, this paper proposes a two-stage fuzzy logic micro-EDM control method. In the first stage fuzzy logic, discharge states of sampling points are calculated so that the present working state of each sampling point is analyzed. By sorting and compiling all states in an analyzing cycle, the discharge state of this cycle is obtained, which is a state of a period of time. In the second stage fuzzy logic, discharge states from the previous stage are taken as the input to infer the servo feed speed. Applying the sampling data during processing as training data of BP algorithm, two-stage fuzzy logic membership functions are optimized with neural network.This micro-EDM control system adopts industrial PC (IPC) with PCI card and PMAC plugged to make detection and motion control work well together. On Windows operating system with VC++, micro-EDM control system software is designed and developed to integrate data acquisition, pulse power setting, motor motion control, machine state display, etc., making it quite convenient to operate, which guarantees the application of the two-stage fuzzy logic intelligent control method.The two-stage fuzzy logic control method is used in the micro-EDM machine which is designed and built by ourselves. The two-stage fuzzy logic membership functions are trained with a large amount of experiment data. After the first stage fuzzy logic membership functions are trained, the discharges during processing are more continuous; after the second stage fuzzy logic membership functions are trained, the valid speed of electrode is increased. The contrast experiments of process time show that the two-stage fuzzy logic control method is obviously superior to conventional control method, and applying neural network to the training of fuzzy logic membership function is an efficient way to further improve the working efficiency of micro-EDM. Thus, the two-stage fuzzy logic control method presented in this paper is quite suitable for micro-EDM, and supplies a solid way for enhancing the efficiency and stability of micro-EDM processing.