In cement production,the input/output model used to describe the static characteristics of the raw material mixing system is nonlinear and coupled with interactive variables.A linearizing and decoupling scheme is provided and made the system linearized and decoupled.It is very significant and convenient for analyzing,designing and implementing the control system.An adaptive decoupling control system is designed based on the improved weighted least-square method,the high accuracy and stronger robustness control is realized,the simulation results and practical operation are provided and a comparison is made with the conventional adaptive control system.

Current drive is one of the major issues in controlled nuclear fusion research field. For the steady state operation, a variety of current drive methods are under systematic investigation. Helicity injection current drive is an attractive method. The experiment on Helicity Injected Tokamak(HIT) in U.S. obtained 150 kA toroidal driven current applying the coaxial helicity injection, which is the most remarkable results. This paper reports the results about equilibrium reconstruction in the above mentioned experiment. The driven current profiles and magnetic topology are consistent with the MHD equlibrium constraint from magnetic measurement data by using EFIT equilibrium fitting code. The reconstruction shows that these plasmas have a Tokamak q profile, with a hollow toroidal current profile and up to 110 kA of closed field toroidal current in a low aspect ratio configuration. Calculation results also indicate that plasma relaxed state is a non complete relaxation state away from Taylor State.

Based on the generalized Biot's theory in anisotropic porous formation and Kuijian's theory of multipole acoustic logging, full waveform acoustic logging in transversely isotropic porous formation is discussed. The monopole and dipole acoustic logging full waveforms are numerically calculated using the extended dynamic permeability given by Johnson et al. The results show that the horizontal permeability is closely related to the acoustic logging waveforms.

The analytic solution for anisotropic steady state heat conduction is applied to analyse the anisotropic rectangular cantilever plate with one edge of nonuniformed boundary temperature and other edges of convection at the first time.The effect of piy angle on the temperature distributions is discussed.With the increasement of the anisotropic angle,the gradient of temperature in the plate is increased.Numerical results show there is the lowest gradient of temperature in the cantilever plate with 0° anisotropic angle.

For various typical loads including fan pump loads; constant torque loads and constant power loads, the optimal efficiency control equation and its constraint conditions are given using the linearizing expression of the motor. Some simulation computations are provided, and a comparison is made between the results of the new method and those of the model based method for efficiency optimization.

This paper improves the identification algorithm given in[1] for raw material blending process. The improved algorithm is very suitable for the time varying process and can track the time varying parameter with high accuracy in time. Actual computation examples are provided. Compared with the results of algorithm[1], the new algorithm has a higher identification accuracy, more rapid convergence.

A novel parameter identification algorithm is presented which is suitable for raw material blending and batching problem in the industry production. Both application examples and Aimulation results are given

Artificial-Neural Network (ANN) is a complex network system composed of many simple elements connected extensively each other and can be regarded as a simulation or abstraction of biological neural system. Multi-Layer Network (MLN) is one of the most important ANN. A essential algorithm to train MLN is BP (Back-Propagation) algorithm. Based on BP algorithm, a hgih-speed algorithm-train MLN layer-by-layer algorithm is proposed with main points:4) train MLN layer-by-layer instead of all together; b) giving instruction to hidden units; c) giving appropriate "energy function' according to specific Problems; d) preserving the merits of BP algorithm. The simulation results increase by orders of magnitude in training-speed. This algorithm applies also some research on running mechanism of MLN.