针对涡街流量计高实时性、高精度的要求,提出一种基于有限域实值多重信号分类(multiple signal classification,MUSIC)法的涡街信号处理方法.该方法利用能量重心校正法对涡街信号幅度谱进行频率校正,确定频率搜索域;利用实值MUSIC法快...针对涡街流量计高实时性、高精度的要求,提出一种基于有限域实值多重信号分类(multiple signal classification,MUSIC)法的涡街信号处理方法.该方法利用能量重心校正法对涡街信号幅度谱进行频率校正,确定频率搜索域;利用实值MUSIC法快速得到谱峰频率值,通过牛顿迭代法对谱峰频率值进行计算,以达到提高频率估计精度的目的.仿真结果表明:该改进方法的频率估计精度较MUSIC法提高了0.1%,实时性高,满足涡街流量计高精度与高实时性要求.展开更多
实值处理具有降低高自由度多输入多输出(multiple-input multiple-output,MIMO)雷达角度估计大计算量的优势。但受制于阵列的共轭对称性,对于任意阵列结构的双基地MIMO雷达发射角(direction of departure,DOD)和接收角(direction of arr...实值处理具有降低高自由度多输入多输出(multiple-input multiple-output,MIMO)雷达角度估计大计算量的优势。但受制于阵列的共轭对称性,对于任意阵列结构的双基地MIMO雷达发射角(direction of departure,DOD)和接收角(direction of arrival,DOA)联合估计,若不做附加的预处理则无法实现实值操作,故将常规阵列实值处理的多重信号分类(multiple signal classification,MUSIC)超分辨算法推广至任意阵列结构的双基地MIMO雷达。首先根据MIMO雷达的导向矢量共轭与镜像的对等性,提取接收信号协方差矩阵的实部,并对其进行特征分解得到"目标加倍"的信号子空间及其应对的噪声子空间;然后利用Kronecker积的特性对其进行降维处理,得到搜索区域减半的一维半实值域MUSIC谱,取出目标DOD真值与其镜像代入降维Capon算法来剔除虚拟峰值得到目标DOD估计真值;最后利用特征矢量得到模糊DOA估计值,采用方向余弦差最小范数方法得到目标DOA无模糊估计值。本文算法估计性能与一维搜索复数域MUSIC相当,计算量约降50%,且能够实现DOD和DOA的自动配对。仿真结果证明了该算法的有效性。展开更多
Drum level sloshing is the latest discovery in the application of heat recovery steam generator (HRSG) in combined cycle, and shows certain negative influence on drum level controlling. In order to improve drum level ...Drum level sloshing is the latest discovery in the application of heat recovery steam generator (HRSG) in combined cycle, and shows certain negative influence on drum level controlling. In order to improve drum level controlling, influence factors on the drum level sloshing were investigated. Firstly, drum sub-modules were developed using the method of modularization modeling, and then the model of drum level sloshing was set up as well. Experiments were carried out on the experimental rig, and the model was validated using the obtained experimental results. Dynamic simulation was made based on the model to get a 3-D graph of drum level sloshing, which shows a vivid procedure of drum level sloshing. The effect of feed-water flow rate, main-steam flow rate and heating quantity on the drum level sloshing was analyzed. The simulation results indicate that the signals with frequency higher than 0.05 Hz are that of drum level sloshing, the signals with frequency of 0.0-0.05 Hz are that of drum level trendy and "false water level", and variation of the feed-water flow rates, main-steam flow rates and heating quantities can change the frequency of drum level sloshing, i.e., the frequency of sloshing increases with the increase of feed-water flow rate, or the decrease of the main-steam flow rate and the heating quantity. This research work is fundamental to improve signal-to-noise ratio of drum level signal and precise controlling of drum level.展开更多
文摘针对涡街流量计高实时性、高精度的要求,提出一种基于有限域实值多重信号分类(multiple signal classification,MUSIC)法的涡街信号处理方法.该方法利用能量重心校正法对涡街信号幅度谱进行频率校正,确定频率搜索域;利用实值MUSIC法快速得到谱峰频率值,通过牛顿迭代法对谱峰频率值进行计算,以达到提高频率估计精度的目的.仿真结果表明:该改进方法的频率估计精度较MUSIC法提高了0.1%,实时性高,满足涡街流量计高精度与高实时性要求.
文摘实值处理具有降低高自由度多输入多输出(multiple-input multiple-output,MIMO)雷达角度估计大计算量的优势。但受制于阵列的共轭对称性,对于任意阵列结构的双基地MIMO雷达发射角(direction of departure,DOD)和接收角(direction of arrival,DOA)联合估计,若不做附加的预处理则无法实现实值操作,故将常规阵列实值处理的多重信号分类(multiple signal classification,MUSIC)超分辨算法推广至任意阵列结构的双基地MIMO雷达。首先根据MIMO雷达的导向矢量共轭与镜像的对等性,提取接收信号协方差矩阵的实部,并对其进行特征分解得到"目标加倍"的信号子空间及其应对的噪声子空间;然后利用Kronecker积的特性对其进行降维处理,得到搜索区域减半的一维半实值域MUSIC谱,取出目标DOD真值与其镜像代入降维Capon算法来剔除虚拟峰值得到目标DOD估计真值;最后利用特征矢量得到模糊DOA估计值,采用方向余弦差最小范数方法得到目标DOA无模糊估计值。本文算法估计性能与一维搜索复数域MUSIC相当,计算量约降50%,且能够实现DOD和DOA的自动配对。仿真结果证明了该算法的有效性。
基金Project(51276023) supported by the National Natural Science Foundation of ChinaProject(09k069) supported by the Open Project Funded by Universities Innovation Platform, Hunan Province, ChinaProject(2011GK311) supported by the Office of Science and Technology of Hunan Province, China
文摘Drum level sloshing is the latest discovery in the application of heat recovery steam generator (HRSG) in combined cycle, and shows certain negative influence on drum level controlling. In order to improve drum level controlling, influence factors on the drum level sloshing were investigated. Firstly, drum sub-modules were developed using the method of modularization modeling, and then the model of drum level sloshing was set up as well. Experiments were carried out on the experimental rig, and the model was validated using the obtained experimental results. Dynamic simulation was made based on the model to get a 3-D graph of drum level sloshing, which shows a vivid procedure of drum level sloshing. The effect of feed-water flow rate, main-steam flow rate and heating quantity on the drum level sloshing was analyzed. The simulation results indicate that the signals with frequency higher than 0.05 Hz are that of drum level sloshing, the signals with frequency of 0.0-0.05 Hz are that of drum level trendy and "false water level", and variation of the feed-water flow rates, main-steam flow rates and heating quantities can change the frequency of drum level sloshing, i.e., the frequency of sloshing increases with the increase of feed-water flow rate, or the decrease of the main-steam flow rate and the heating quantity. This research work is fundamental to improve signal-to-noise ratio of drum level signal and precise controlling of drum level.
