CAJ | 학술논문

采用双DSP设计完成了交流位置伺服系统的神经网络实时控制系统，其中一片DSP作为系统的小波神经网络控制器( NNC )，另一片 DSP 作为系统的 BP 神经网络在线辨识器( NNI)。由NNI对被控对象进行在线辨识的基础上，通过对NNC的权系进行实时调整，使系统具有自适应性。通过对交流位置伺服系统的控制实验，验证了本系统的实用性。
채용쌍DSP설계완성료교류위치사복계통적신경망락실시공제계통，기중일편DSP작위계통적소파신경망락공제기( NNC )，령일편 DSP 작위계통적 BP 신경망락재선변식기( NNI)。유NNI대피공대상진행재선변식적기출상，통과대NNC적권계진행실시조정，사계통구유자괄응성。통과대교류위치사복계통적공제실험，험증료본계통적실용성。
This paper uses the dual DSP, one as wavelet neural network control er( NNC), the other as BP neural network identifier ( NNI) to design the AC position servo neural network realtime control system.Due to the real time adjustment of NNC weight and the NNI on-line identification of the control ed object, the system is more available. Its practical practicability is verified by the control sim-ulation of AC position servo system.