With a view to the data fusion of four oil analysis techniques, i.
针对4种最常用的油样分析技术(铁谱、光谱、颗粒计数及理化指标分析)的信息融合问题,依据基于规则的专家系统方法,建立了各油样分析技术的子专家诊断系统;依据专家经验建立各子专家系统的诊断结果与故障论域中各故障模式的关系,得到了用于神经网络学习的训练样本。
Taking the hydraulic system fault analysis of an 800 t/h bucket elevator in a power plant for example,this paper indicates that the equipment status,trouble causes and trouble spots of the hydraulic system can be judged effectively by oil sample analysis,pressure and flow-rate analysis as well as noise diagnosis;thereby the efficiency of repair can be improved and state maintenance realized.
通过某厂800 t/h斗式提升机液压系统故障分析的实例,说明油样分析、压力和流量测试分析、噪声诊断可以有效地判断液压系统的设备状态、故障原因和故障点,从而提高修理工作效率,达到状态检修的目的。
Finally,two time series and aero-engine spectrometric oil analysis data were used to verify this model.
针对非等间隔的受加油因素影响的光谱油样分析数据的建模预测问题,建立了BP神经网络的多变量预测模型,充分考虑了油样分析数据的非等间隔性及受加油因素影响的特点,同时,用遗传算法对网络参数进行了优化。
It s used to the identification of crude oil nature in the Tarim Basin.
对２０种油品的研究结果表明，此方法性能良好，对油样属性作出定量的描述，是油样属性分析的一种新手