摘要
讨论了近年来神经网络软测量技术的一些研究新进展 ,并介绍了其与控制技术、计算机通信、虚拟仪器及www结合的应用实例。目前 ,在构造软仪表方面 ,成功地应用于实际生产过程中的神经网络主要有前向BP网络和RBF网络 ,此外还有融合模糊技术的神经网络算法。数据处理仍是一个十分重要的问题 ,此外 ,现有的在线校正方法十分有限 ,应发展更多更新的方法以适应复杂工业系统控制的要求。同时 ,将软测量技术与系统调优结合起来 ,变开环指导为闭环控制 ,使其有更广泛的应用推广。
This paper discusses the development of soft measurement based on neural networks of recent years. And also introduces its implementation which is integrated with control technology, computer communication system, virtual instruments, and www system together. At present, Back-Propagation Algorithm, RBF Algorithm and Neural Networks Algorithm with fuzzy technology have gained widely success in the application of soft measurement. And the date processing is still the big problem. Besides, more new compensation methods should be developed in order to adapt for the demands of complex process control system. At the same time, soft measurement will be more widely used if it can be combined with optimal control system in order to turn open loop control into closed loop control.
出处
《控制工程》
CSCD
2003年第1期15-17,61,共4页
Control Engineering of China
