针对转弯机动目标,提出了一种两层模型嵌套的跟踪算法。算法的内层模型由基于函数的"当前"统计(Function based current statistic,FB-CS)模型构成,该模型在"当前"统计(Current statistic,CS)模型的基础上,通过加...针对转弯机动目标,提出了一种两层模型嵌套的跟踪算法。算法的内层模型由基于函数的"当前"统计(Function based current statistic,FB-CS)模型构成,该模型在"当前"统计(Current statistic,CS)模型的基础上,通过加权一个以新息方差之迹为参数的活化函数,对加速度方差和机动频率进行自适应处理,再针对目标的速度方向角进行滤波,将获得的角速度估计值作为外层的输入;外层模型由曲线模型构成,其方向角度、角速度和角加速度由内层模型提供。由于既准确估计目标的角速度,又设计了合理的运动模型,算法显著提高了转弯机动目标的跟踪精度.仿真实验验证了算法的有效性。展开更多
The neural network partial least square (NNPLS) method was used to establish a robust reaction model for a multi-component catalyst of methane oxidative coupling. The details, including the learning algorithm, the num...The neural network partial least square (NNPLS) method was used to establish a robust reaction model for a multi-component catalyst of methane oxidative coupling. The details, including the learning algorithm, the number of hidden units of the inner network, activation function, initialization of the network weights and the principal components, are discussed. The results show that the structural organizations of inner neural network are 1-10-5-1, 1-8-4-1, 1-8-5-1, 1-7-4-1, 1-8-4-1, 1-8-6-1, respectively. The Levenberg-Marquardt method was used in the learning algorithm, and the central sigmoidal function is the activation function. Calculation results show that four principal components are convenient in the use of the multi-component catalyst modeling of methane oxidative coupling. Therefore a robust reaction model expressed by NNPLS succeeds in correlating the relations between elements in catalyst and catalytic reaction results. Compared with the direct network modeling, NNPLS model can be adjusted by experimental data and the calculation of the model is simpler and faster than that of the direct network model.展开更多
Differential scanning calorimetric （DSC） study was carried out at different heating rates to examine the solid state reactions in a 7150 A1-Zn-Mg alloy in water-quenched （WQ） state, naturally and artificially aged...Differential scanning calorimetric （DSC） study was carried out at different heating rates to examine the solid state reactions in a 7150 A1-Zn-Mg alloy in water-quenched （WQ） state, naturally and artificially aged tempers. The exothermic and endothermic peaks of the thermograms indicating the solid state reaction sequence were identified. The shift of peak temperatures to higher temperatures with increasing heating rates suggests that the solid state reactions are thermally activated and kinetically controlled. The artificial aging behaviour of the alloy was assessed by measuring the variations of hardness with aging time. The fraction of transformation （Y）, the rate of transformation （dY/dt）, the transformation functionflY）, and the kinetic parameters such as activation energy （Q） and frequency factor （k0） of all the solid state reactions in the alloy were determined by analyzing the DSC data, i.e. heat flow involved with the corresponding DSC peaks. It was found that the kinetic parameters of the solid state reactions are in good agreement with the published data.展开更多
Thermodynamic analysis of the possibility of silver nanocrystalline preparation by high energy milling silver oxide was investigated. The molar Gibbs free energy function of mechanically activated samples was calculat...Thermodynamic analysis of the possibility of silver nanocrystalline preparation by high energy milling silver oxide was investigated. The molar Gibbs free energy function of mechanically activated samples was calculated from the structural defects such as amorphization, dislocation and surface energy. According to the molar Gibbs free energy function, the equilibrium temperature of mechanical reduction of silver oxide milled for 21 h was estimated at about 304 K. Consequently, at this temperature silver oxide cannot be stable and will transform to silver during the milling.展开更多
A discussion is given on the convergence of the on-line gradient methods for two-layer feedforward neural networks in general cases. The theories are applied to some usual activation functions and energy functions.
文摘针对转弯机动目标,提出了一种两层模型嵌套的跟踪算法。算法的内层模型由基于函数的"当前"统计(Function based current statistic,FB-CS)模型构成,该模型在"当前"统计(Current statistic,CS)模型的基础上,通过加权一个以新息方差之迹为参数的活化函数,对加速度方差和机动频率进行自适应处理,再针对目标的速度方向角进行滤波,将获得的角速度估计值作为外层的输入;外层模型由曲线模型构成,其方向角度、角速度和角加速度由内层模型提供。由于既准确估计目标的角速度,又设计了合理的运动模型,算法显著提高了转弯机动目标的跟踪精度.仿真实验验证了算法的有效性。
文摘The neural network partial least square (NNPLS) method was used to establish a robust reaction model for a multi-component catalyst of methane oxidative coupling. The details, including the learning algorithm, the number of hidden units of the inner network, activation function, initialization of the network weights and the principal components, are discussed. The results show that the structural organizations of inner neural network are 1-10-5-1, 1-8-4-1, 1-8-5-1, 1-7-4-1, 1-8-4-1, 1-8-6-1, respectively. The Levenberg-Marquardt method was used in the learning algorithm, and the central sigmoidal function is the activation function. Calculation results show that four principal components are convenient in the use of the multi-component catalyst modeling of methane oxidative coupling. Therefore a robust reaction model expressed by NNPLS succeeds in correlating the relations between elements in catalyst and catalytic reaction results. Compared with the direct network modeling, NNPLS model can be adjusted by experimental data and the calculation of the model is simpler and faster than that of the direct network model.
文摘Differential scanning calorimetric （DSC） study was carried out at different heating rates to examine the solid state reactions in a 7150 A1-Zn-Mg alloy in water-quenched （WQ） state, naturally and artificially aged tempers. The exothermic and endothermic peaks of the thermograms indicating the solid state reaction sequence were identified. The shift of peak temperatures to higher temperatures with increasing heating rates suggests that the solid state reactions are thermally activated and kinetically controlled. The artificial aging behaviour of the alloy was assessed by measuring the variations of hardness with aging time. The fraction of transformation （Y）, the rate of transformation （dY/dt）, the transformation functionflY）, and the kinetic parameters such as activation energy （Q） and frequency factor （k0） of all the solid state reactions in the alloy were determined by analyzing the DSC data, i.e. heat flow involved with the corresponding DSC peaks. It was found that the kinetic parameters of the solid state reactions are in good agreement with the published data.
文摘Thermodynamic analysis of the possibility of silver nanocrystalline preparation by high energy milling silver oxide was investigated. The molar Gibbs free energy function of mechanically activated samples was calculated from the structural defects such as amorphization, dislocation and surface energy. According to the molar Gibbs free energy function, the equilibrium temperature of mechanical reduction of silver oxide milled for 21 h was estimated at about 304 K. Consequently, at this temperature silver oxide cannot be stable and will transform to silver during the milling.
基金Supported by the Natural Science Foundation of China
文摘A discussion is given on the convergence of the on-line gradient methods for two-layer feedforward neural networks in general cases. The theories are applied to some usual activation functions and energy functions.
