摘要
重点对基于BP人工神经网络的软测量模型的建立及进行网络训练时的注意事项进行了分析和探讨。在应用时,首先建立了CO的人工神经网络软测量模型,通过对车辆排放的实际采集获取了一组实际数据,然后利用CO人工神经网络软测量模型进行模拟和验证,模拟结果非常接近实际结果,证明了软测量技术在车辆排放检测结果在线处理上的有效性和可行性。
The Paper gave a brief introduction of the concept and sort of soft-sensor and then analyzed the building of the soft-sensor model based on BP ANN and the attendant proceeding. In application, the authors built the CO ANN Soft-sensor model, got experiment data from test and received good result. It proved the validity and feasibility for on-line treatment based on ANN soft-sensor model.
出处
《中国测试技术》
2007年第3期41-43,共3页
CHINA MEASUREMENT & TESTING TECHNOLOGY
