摘要
根据CMAC神经网络的工作原理和内部结构特点,提出了基于FPGA软件硬化技术设计CMAC硬件控制芯片的技术路线,详细讨论了基于FPGA芯片将CMAC神经网络软件硬化的方法,给出了仿真运行波形。构建出用于数控机床直线电动机驱动伺服系统的CMAC复合控制策略。以美国Kollmorgen公司的IC22-050A2P1型直线电动机作为试验对象,对硬化实现的CMAC控制器进行了实际应用,以验证方案的正确性。
The thought of designing CMAC （Cerebellar Model Articulation Controller） hardware control chip which bases on software hardening technology is put forward according to the working theories and inner structure characteristics of CMAC neural network. The method of software hardening based on FPGA （Field Programmable Gate Array） is discussed in detail. The corresponding simulation waveforms are given. The CMAC compound control strategy used in PMSLM （Permanent Magnet Synchronous Linear Motor） is designed. As the test object, IC22-050A2P1 PMLSM made by Kollmorgen company is used in the actual application of hardening CMAC control chip to verify the correctness of the program.
出处
《科技通报》
北大核心
2011年第5期661-665,670,共6页
Bulletin of Science and Technology
基金
江苏省自然科学基金(BK2008367)
江苏省"六大人才高峰"高层次人才资助项目(2008163)
