A three-dimensional space interpolation method of grey-depth image sequence is presented. The way breaks away from the limit of original practical photographing route. The pictures can cruise at will in space. By using space sparse sampling, the memorial capacity can be decreased greatly and the reproduced scenes can be controlled. To solve complex computations in three-dimensional interpolation algorithm, a fast and practical algorithm of scattered space lattice and the Warp algorithm with proper depth are studied. By several simple aspects of the three dimensional space interpolation, some simple and practical algorithms are developed. Some results of simulated experiments with computers show that the new method is absolutely feasible.

This paper discussed the possibility and key problem of constructing the neural network time series model, and three time series neural network forecasting methods has been proposed. That is, the neural network nonlinear time series model, the neural network multi-dimension time model and the neural network com bining predictive model. These three methods are applied 10 real prcblems. The results show that these methods are better than the traditional ones. Furthermore, the comparison of the neural network with the traditional methods and the constructed model of intellectual information forecasting system are given.

This paper proves that a good linear block error-correcting code implements a good linear local-random sequence generator. Therefore the close relation between error-correcting coding theory and modern cryptography is discovered once again.

A statistical noise model and a mathematical model for real speckle pattern are presented, and then, in view of the models, a new adaptive suboptimal image filtering approach is proposed. The proposed approach, with the local direction features of speckle pattern, combines the characteristcs of optimal linear filter with non-linear filter and is an adaptive approximation to linear minimum mean square error filter. Experimental results show that the proposed approach has fairly good edge-preserved performance, compared with other present image filters, as well as much better filtering performance and robustness for speckle pattern.

A new approach with high resolution 2-D spectrum estimation is presented. Based on the array signal processing and the eigenstructure decomposition method, the signal subspace and the noise subspace are achieved by reconstructing and decomposing the data matrix collected by the array sensors. By using the orthogonal properties of the two subspaces, the information about the velocities and bearings of spatial moving targets can be exactly obtained. This method is applied to phased array radars. The computer simulation results demonstrate its availability.

The principle and performance cf Synthetic Impulse and Antenna Ra-der (SIAR) are analyzed with the concept of time-space 3-D matched filtering. The discussion here centres on the characteristics of synthetic impulse and antenna in the case of three dimension. The results obtained are helpful for designing this new style radar.

A new approach is proposed which can satisfy the range ambiguity requirement while keep the antenna width not to be enlarged and get the high quality image also. Here the chirp slope of the transmitted linear frequency modulated pulses is reversed alternately. Therefore, during the range compression significant portion of the ambiguous return is suppressed due to the mismatching with the signal reference function. The suggested method provides significant improvement in RAR and can be utilized in attaining wider swath.

An integral equation is derived from the electric field integral equation (EFIE) for plane wave illumination onto the metallic circular disc-ring structures by using Fourier transformation and separating the induced charge s contribution from the EFIE. For normal incidence, the equation is extremely simple and easy to be solved numerically regardless of the electric dimension of the structures. Comparison between RCS by this method and that by analytical solution for a disc shows the effectiveness of this approach. Numerical results are also given for the current distribution and the far radiation patterns of a disc-ring structure.

The building blocks of current-mode CC1I are proposed. The leapfrog-type CCII filter structures and their design based on the building blocks are discussed. The filters yield low sensitivities, and they are suitable for fully integrated continuous-time filters.

On the basis of psychological acoustic theories and experiments, this paper proposes an acoustic model which is based on acoustic perceptual feature. Compared with the physiological acoustics based acoustic model, this model car. more true represent human s perceptual features of continuous speech, so it is more suitable for recognition of continuous speech.

The picosecond photoconductive switches are developed and used to detect the pulse laser waveform. By using the photoconductive switches, a novel lab model of an ultrawide band (UWB) radar is also developed. The experimental results are given to show the performances of the swithches and the UWB radar.

On the basis of the experimental element pattern and GTD, an efficient numerical method for analyzing 3-D radiation field of antennas on the truncated cone-cylinder is presented. These formulas are also applicable to the truncated cylinder and cone. Examples are given on the calculation of 3-D radiation fields of microstrip antennas on the truncated cone. The results of calculations agree well with the experimental ones.

One dimensional Fresnel zone phase corrected plate (1-DFP) can be made by a 1/P wavelength phase correcting along one dimension upon the plate. When it is used as a reflector, it gives an alternative to parabolic cylinder reflector in application on fan-beam antenna. In this paper, Physical Optics (PO) is used in analyzing the caustic characteristics of 1-DFP reflector with plane wave illumination. Numerical results of a series of fields in caustic are also presented.

The development of a compact high repetition rate electron beam accelerator is reported. The specifications of the device are output electron energy 120--150keV, beam current 5--50A, pulse duration 1 μs, beam cross-section 5×50cm² and repetition frequence 100Hz. The key part of the accelerator is a wire-ion-pla-sma (WIP) electron gun of particular advantage. The principle of operation of the WIP electron gun and its physical properties are described. The caculation formula of output parameter is given and preliminary experimental results are presented.

The migration of fluorine atoms in poly-silicon and SiO₂ of BF⁺₂ implanted Si-gate before and after annealing has been analyzed using SIMS. The results show that a part of fluorine atoms of BF⁺₂ implanted Si-gate with an energy of 80keV and doses of 2×10¹⁵ and 5×10¹⁵cm^-2 after annealing diffiuse into SiO₂ region of this gate. Anomalous migration of fluorine atoms is observed in both poly-silicon and SiO₂. This is due to the collection of fluorine atoms in the regions of residual damage and bond defects.

The low temperature frequency characteristics of polysilicon emitter bipolar transistors are investigated, considered the carrier freezingout effect and trapping effect of shallow energy impurities in semiconductors. It shows that the base sheet resistance R_db increases quasi-exponentially at low temperature due to carrier freezingout, which results in the degradation of frequency characteristics of the transistors at low temperature. The base and emitter transit time increase and the cutoff frequency decreases at low temperature due to the trapping effect of shallow energy level impurities. Those should be considered carefully in the design of devices for low temperature operation.

This paper introduces the designing thought of modular constant-current circuits and new-made current regulators with wide current range. Some data about the new device read as follows: current range is 0.1A-8A, current temperature coefficient is 10^-4-10^-5/℃, limiting-voltage is lower than 0.4V. It is a kind of excellent device with both high stability and wide current range

Basic theory of designing SCPs with less capactor spread is studied. Two topologial methods for reducing capacitor spread of SCFs are presented. Realizations of those methods are also proposed.