Abstract:This paper gives an exact solution of a rectangular plate with complex edges. A general solution of differential equation for solving elastic thin plates in bending is suggested.

Abstract:Through data dependence analysis，this paper discusses the method of multitasking partition at the loop level，gives the condition of parallel execution of Fortran loops，and puts forward some questions needed to be solved. Finally，there will be the possibility of macro multitasking partition.

Abstract:An implicit method and a semi-implicit method applied to first and second-order systems are proposed in this paper. Being 2-Order pricise，these methods have been rigorously jutisfied to be unconditional stable in energy norm. In the struction of formula the coefficeint matrices remain positive definite and symmetry; so the operation of their inverse matrices have been simplified in an approximate procedure.

Abstract:This paper presents a fast and efficient algorithm for generation and processing of 3-D dynamic graphics. It is appropriate to the purpose for the real time graphics simulation of robot manipulators. A procedure is described by which combinative geometry bodies are determined，the geometry transformation and hidden lines elimination are also introduced. The program is written in Turbo-Pascal and implemented on IBM-PC.

Abstract:In this paper，the theory and method are described in detail for solving two-dimensional neutron transport equation in cartesian geometry by using Flat Leakage Discrete Nodal Transport Method (FLDNTM). Based on this，a Linear Leakage Discrete Nodal Transport Method (LLDNTM)，more accurate than FLDNTM，is developed. Two computer codes-FLDNTM & LLDNTM，are created and applied to the calculations of one-and two-dimensional test problems，the excellent numerical results are obtained

Abstract:In this paper a double exposure method with a simpler experimental set up is reported for measurement of small displacements of the speckle pattern of multimode optical fiber which were caused by changing the radial pressure on an optica1 fiber. The measurment of displacement ranges from 10 to 1000μm.

Abstract:This paper gives a detai1ed account of the measuring techniques and the results of the dynamic welding parameters and the pressure field in bonding region，respectively used and obtained in the tube-to-tubeplate explosive welding. In the experiments，the dynamic welding parameters，such as the impact velocity V_p，the collision point velocity V_c，and the dynamic angle of collision β，are measured with slanting wire method，for verifying the slanting wire method，the electrocontact probe technique is used to measure the detonation velocity V_d; while the pressure field P(L)in the bonding region is measured with manganin stress gauge.

Abstract:In this paper，the method of estimating the retrograde orbit 1aunching energy and the to-and-fro orbits parameters by using lunar gravitation to change the orbit moving direction is discussed. The result showes that，for 1aunching a retrograde orbital sate1lite in the moon-earth p1ane，more energy is saved by using 1unar gravitation to change the orbit moving direction and than shows that by using Hohmann transfer direct1y on1y in the case，when the terminal retrograde orbita1 is higher than thirty-thousand ki1ometers. The higher the terminal orbit is，the more energy is saved. This paper a1so points out that the energy needed in lunar gravitation orbital-plane change (i.e.，the to-and-fro orbits are not in the same orbit plane) is less than that in the moon-earth plane change.

Abstract:In this paper the principle of the minimax robust detection of signals is reviewed. Certain minimax robust defectors are analyzed and the design methods and Monte-Carlo Simulation results for them are provided.

Abstract:In this paper，the readers can see a new method of plotting standard electron diffraction patterns for closepacked hexagonal crystals (c/a=1.633). This method is named "Rectangular Method" by the author. It is different from the conventional method and the rectangular method plotting cubic system electron diffraction patterns. In this paper，we have used three mothematics transform factors "a、b、c"，greatly simplified the plotting processes. This method is almost as simple as plotting cubic system standard electron diffraction patterns.

Abstract:The studies of curing kinetics of epoxy resins using isothermal and dynamic differential scaning calorimetry were reviewed and compared to each other. The features of two methods were described. The possible reasons for different resu1ts obtained by these two methods were analysed.

Abstract:In this paper，we discuss necessary and sufficient condition under which the one-dimensional general transforms equivalent to two-dimensional general transforms Theorem：Let [X]，[Y] be the matrix representations of vectors X，Y respectively，and let [R]，[Q]，[P] be three matrices of order N M x N M，N x N，M x M respectively，[X]，[Y] are of order N x M，X ，Y are dimension N M. (1) If and only if [R]=[P]^T [Q]，Y=[R]X and [Y]=[Q] [X] [Q] for any vector X，where denotes Kronecker Product of matrices; (2) If Y=[R]X and [Y]=[Q] [X] [P] for any vector X，then [R] is symmetric，if and only if [Q]，[P] are symmetric or anti-symmetric matrices，simultaneously，and R_i+jN,k+lN=±R_i+lN,k+jN R_i+jN,k+ln=±R_k+jN,i+lN (i，k=0，l，…，N-I;j,l=0，l，…，M-1) Nextly，we have discussed the equivalence of Walsh-Hadamard transforms and that of any general transforms having the cyclic convolution property (CCP)(DFT，NTT，PT). Lastly，we have introduced a fast algorithm of 2D-DFT-Vector Algorithm by using above equivalence theorem，which is more efficient than the row-column method of 2D-DFTs.

Abstract:This paper presents the reasoning method applied to KD Computer-Based Medical Diognosis System. Upon this reasoning method，the auther successfully developed the diognosis system for the uper-digesting system，the blood-circulating system and the lower-digesting system diseases (achieved on the combination system of CROMECO-Ⅲ to the TQ-100 reading end).

Abstract:The purpose of this paper is investigating various solutions of continuous convex games with contably many players. We showed that for such games,the core is the unique stable set and coincides with the bargaininng set (for the grand coalition)，and the kernel (for the grand coalition)is a nonempty subset of the core. These can be viewed as a generalization of similar results obtained by Shapley，Maschler and Peleg. The values of the games are also considered and it is proved that the Banach space acM，introduced by Artstein，is the closed linear subspace generated by all these games. Moreover，the value of every such game is in the core，as is the case for finite convex games.

Abstract:As shown in previous studies[3],the piecewise-recursive Kalman filter is one of the most efficient and promising approach to decrease the computation requirements of the Kalman filter. In this paper the approach and its influence on the accurancy of estimation and the problem of how to choose the ⊿t(the sampling interval) and ⊿τ = N ?⊿t (the interval between gain-matrix calculations)are considered. In the final portion of the paper，a example is given to show the applicability of the approach.