This dissertation focuses on the network architecture of the general regression neural network (GRNN) and the optimization of the GRNN's network parameter.
It is indicated that surface energy reduction and net structure reinforcement are the effective way to improve the properties of SiO 2 thin film electret, and there is a good prospect for the SiO 2 PbO Al 2O 3 compound film to become a new kind of inorganic electret material.
It further indicates that with the same net structure and training parameters,molecular descriptor(HOMO-LUMO)*BDE has the strongest correlation with impact sensitivity of explosives.
4. form the network configuration, the aim, the process, the methods and so on we carry through the encryption theory analysis and the implementation of the methods.
(5) Study modelling thought, network configuration, majorize GNNM(1,1) mode method and learning algorithm of GNNM(1,1) mode combined grey system theory and neural network.
In view of the requirements of monitoring and controlling system for the 21st century, the idea of design, the network configuration, the design features of hardware and software of the monitoring and controlling system of new generation STI-VC2100 in ship engine room, the design idea and the coverage of main systems of STI-VC2100 series (main diesel engine remote control, monitoring and alarm, electrical power station control) are analyzed in this paper, and it is helpful to designers and users.
2. Research the application of 100M Ethernet in substation automation system, and provide a new network configuration of substation automation system base on 100M Ethernet.
Electron probe microanalysis of the deposits showed a network structure that results from the surface defects of the cold rolled Ag substrate and the lattice mismatch between substrate and deposit.
In the presentation, the notions of synchronization of dynamical systems on networks, stability of dynamical networks, and relationships between network structure and synchronizability, will be first introduced.
The perception network is shown to distinguish types of system faults correctly, and to be an artificial neural network architecture especially well suited for fault diagnosis.
The design of an NN involves the choice of several parameters which include the network architecture, number of hidden layers, number of neurons in the hidden layers, training, learning and transfer functions.
We have studied the photoluminescence (PL) of the hole-net structure of silicon and the porous structure of SiGe where the PL intensity at 706 nm and 725 nm wavelength increases obviously.
This comparison showed that M.incognita juveniles in a narrow, fine network tended to migrate actively and, in contrast, those in a wide, coarse network were prevented from migrating by the network configuration.
If a cost/benefit analysis is coupled with the proposed model, the optimal investments and the optimal network configuration will be determined simultaneously.