Abstract:
A multilayer neural nerwork model for the perception of rotational motion has been developed using Reichardt’s motion detector array of correlation type, Kohonen’s self-organized feature map and Schuster-Wagner’s oscillating neural network. It is shown that the unsupervised learning could make the neurons on the second layer of the network tend to be self-organized in a form resembling columnar organization of selective directions in area MT of the primate’s visual cortex. The output layer can interpret rotation information and give the directions and velocities of rotational motion. The computer simulation results are in agreement with some psychophysical observations of rotational perception. It is demonstrated that the temporal correlation between the oscillating neurons would be powerful for solving the “binding problem” of shear components of rotational motion.

Abstract:
This paper introduces the structure of linear motor in mines. Analyze the power relation of power-AC -linear motor – vibrant machine, based on this, count the power factor; and make mechanical analysis to the vibrancy, get the power factor, which should be: in the precondition of without collision for the top and bottom magnet, do best to decrease the δ_0 to close to ΔΧ_m (ΔΧ_m depends on the technique of the vibrant load), make Κ_δ close to 1 and λ_e close to critical maximum λ_(em) . It is significantly useful to design linear motor.

Abstract:
A hybrid finite difference--finite volume (FD-FV) approach for discretization in space is proposed to solve first-order hyperbolic conservation laws. Unlike any conventional finite difference method (FDM) or finite volume method (FVM), this approach uses both cell-averaged values and nodal values as degrees of freedom (DOF). Consequently it is inherently conservative like FVM and easy to extend to high-order accuracy in space like FDM. The proposed FD-FV approach works for arbitrary flux functions, whether convex or non-convex; and it does not require any exact or approximate Riemann solver hence it is also computationally economical. Method of lines is adopted for time integration in present work; in particular, explicit Runge-Kutta methods are employed. It is theoretically proven and numerically confirmed that in general, the proposed FD-FV methods possess superior accuracy than conventional FDM or FVM. Linear stability is studied for general FD-FV schemes -- both space-accurate and time-stable FD-FV schemes of up to fifth-order accuracy in both space and time are presented. Numerical examples show that as long as the solutions are smooth, the proposed FD-FV methods are more efficient than conventional FVM of the same order, at least when explicit time-integrators are applied.

Abstract:
Most slope limiter functions in high-resolution finite volume methods to solve hyperbolic conservation laws are designed assuming one-dimensional uniform grids, and they are also used to compute slope limiters in computations on non-uniform rectilinear grids. However, this strategy may lead to either loss of total variation diminishing (TVD) stability for 1D linear problems or the loss of formal second-order accuracy if the grid is highly non-uniform. This is especially true when the limiter function is not piecewise linear. Numerical evidences are provided to support this argument for two popular finite volume strategies: MUSCL in space and method of lines in time (MUSCL-MOL), and capacity-form differencing. In order to deal with this issue, this paper presents a general approach to study and enhance the slope limiter functions for highly non-uniform grids in the MUSCL-MOL framework. This approach extends the classical reconstruct-evolve-project procedure to general grids, and it gives sufficient conditions for a slope limiter function leading to a TVD stable, formal second-order accuracy in space, and symmetry preserving numerical scheme on arbitrary grids. Several widely used limiter functions, including the smooth ones by van Leer and van Albada, are enhanced to satisfy these conditions. These properties are confirmed by solving various one-dimensional and two-dimensional benchmark problems using the enhanced limiters on highly non-uniform rectilinear grids.

Abstract:
In this paper, a visual motinn perception neural network is presented to explore the processing of winal motion information. This model emplops Reichardt's elementary motion detectors array and Rumelhart's BP (learming by back-propagating errors) neural network. In the viewpoint of computational neuroscience, we try to elaborate the nearel mechanism of perception of two dimensional pattern motion from the detection of one dimensional motion component, and answer how the motion vector is represented in the brain. By using computer simulation through a supervised learning process, it is proved that this neural network can solve the "ambiguity" resultal from local inchon deteCtion and give the aCtual orientation, motion direction and motion sped of the pattern.

Abstract:
The perception of the visual motion information includes the processing from local motion detection to the perception of global pattern movement. Using the neural circuit - network of the figure-ground discrimination of the fly's visual system as the basic unit, and the hexagonal arrays of elementary movement detectors as the input layer, we build a simple brain model for the visual motion information perception. The motion information processing on each layer of this neural computational model is simulated by computer. This model is able to predict correctly the results of differential behavioural experiments. Finally, the neural principles underlying the spatial physiological integration are discussed.

Abstract:
A multilayer neural nerwork model for the perception of rotational motion has been developed usingReichardt's motion detector array of correlation type, Kohonen's self-organized feature map and Schuster-Wagner's oscillating neural network. It is shown that the unsupervised learning could make the neurons on the second layer of the network tend to be self-organized in a form resembling columnar organization of selective directions in area MT of the primate's visual cortex. The output layer can interpret rotation information and give the directions and velocities of rotational motion. The computer simulation results are in agreement with some psychophysical observations of rotation-al perception. It is demonstrated that the temporal correlation between the oscillating neurons would be powerful for solving the "binding problem" of shear components of rotational motion.

Abstract:
In this paper an adaptive genetic algorithms is presented. The adaptive method of probabilities of reproduction, crossover, and mutation which have selectivity about the operated solutions is adopted in the course of calculation. It makes the reproduction probability of the solution which has the similar fitness decrease, the probabilities of crossover and mutation increase, hence it maintains the diversity in the population and sustains the search capacity of the genetic algorithms. The method is tested by the genetic algorithm testing functions. The results are excellent.

Abstract:
The rule of trajectory structure for fourth-order nonlinear difference equation xn+1=(xan 2+xn 3)/(xan 2xn 3+1), n=0,1,2,…, where a∈[0,1) and the initial values x 3,x 2,x 1,x0∈[0,∞), is described clearly out in this paper. Mainly, the lengths of positive and negative semicycles of its nontrivial solutions are found to occur periodically with prime period 15. The rule is 4+,3 ,1+,2 ,2+,1 ,1+, 1 in a period. By utilizing this rule its positive equilibrium point is verified to be globally asymptotically stable.