专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明专利

JPH06295296A METHOD AND DEVICE FOR ORBIT ANALISIS 失效
公开(公告)号：JPH06295296A
公开(公告)日：1994-10-21
申请号：JP8156993
申请日：1993-04-08
申请人： MATSUSHITA ELECTRIC IND CO LTD
发明人： HATTA SHINICHIRO
IPC分类号： G06F17/13 , G06F19/00 , G06F15/20 , G06F15/328
摘要： PURPOSE:To provide the orbit analysis device capable of easily analyzing a complicated physical system, for example, the orbit of even the motion, the non-harmonic oscillator, or the like for a multipariticle system where particles interact with each other by two-body potential. CONSTITUTION:The device which simulates and analyzes the orbet being the solution of an equation of motion of particles by a computer is provided with means 32 and 33 which use data of preliminarily determined start point and end point to produce simultaneous differential equations with coordinates of the orbit or the time as an unknown quantity, a means 34 which obtains a convergence solution of differential equations so as to minimize the action of the physical system, and a means 35 which easily outputs this solution to illustrate it. That is, the theorem of minimum action and the converging method of simultaneous differential equations are applied to obtain the solution for the purpose of analyzing the orbit.

2. 发明专利

JPH06119369A DEVICE AND METHOD FOR ANALYZING BOLTZMANN EQUATION OF ANALYZED OBJECT SYSTEM 失效
公开(公告)号：JPH06119369A
公开(公告)日：1994-04-28
申请号：JP26766492
申请日：1992-10-06
申请人： MATSUSHITA ELECTRIC IND CO LTD
发明人： HATTA SHINICHIRO
IPC分类号： G06F17/13 , G06F15/328
摘要： PURPOSE:To provide an analyzing device capable of obtaining easily the distribution function of a Boltzmann equation. CONSTITUTION:An input part 21 for inputting various kinds of parameters such as an electric field value, the thickness of a system and temperature, etc., in respect of the system of the analyzed object like an electron and a gas molecule, etc., an element equation generating part 22 which receives the various kinds of the parameters inputted by the input part 21, and generates simultaneous equations in each element of a finite element method by the analytical Boltzmann equation, a total element equation generating part 23 which generates total element simultaneous equations by summing the simultaneous equations in each element generated by the element equation generating part 22 extending over all the elements, an equation solving part 24 which obtains the distribution function (f) by substituting (m+2)-pieces of (f) of a known number to (m+1) -tuple linear simultaneous equations in respect of the equation generated by the total element equation generating part 23, and solving the linear simultaneous equations about (m +m-1)-pieces of (f) of an unknown number, and an output part 25 which outputs an obtained result are provided.

3. 实用新型

JPS61204256U 失效
公开(公告)号：JPS61204256U
公开(公告)日：1986-12-23
申请号：JP7283286
申请日：1986-05-15
IPC分类号： G06F7/544 , G06F15/328

4. 发明专利

JPS5720863A ARITHMETIC METHOD OF DIGITAL DIFFERENTIAL ANALYTICAL ENGINE 失效
公开(公告)号：JPS5720863A
公开(公告)日：1982-02-03
申请号：JP9576680
申请日：1980-07-15
申请人： HITACHI LTD , HITACHI ELECTRONICS
发明人： YABUUCHI SHIGERU , ENDOU TAKESHI
IPC分类号： G06F7/64 , G06F17/13 , G06F15/328
摘要： PURPOSE:To redouble the number of inputs per operation, by controlling the timing of production according to the number of inputs for the reset signal of a register that stores the total sum of the microincrements of the input signals of >=2 interlations. CONSTITUTION:A computer 201 transmits the initial value to a Y memory 203 which stores the result of operation of equation (1), an R memory 204 which stores the result of operation of equation (2) and an increment memory 205 which stores an overflow component DELTAZ1 of the operation of equation (2) respectively via a common bus 202 plus multiplexer 2021, 2022 and 2023, and at the same time transmits an arithmetic control instruction of DDA to a control memory 206. The computer 201 also transmits the desired value of a program counter 207 that designates the address of the memory 206 and furthermore starts and stops a controller 208 which produces a series of timing signals necessary for execution of an operation of DDA.

5. 发明专利

JPH06290204A METHOD OF ITERATIVE SOLUTION WITH PREPROCESSING 失效
公开(公告)号：JPH06290204A
公开(公告)日：1994-10-18
申请号：JP7840493
申请日：1993-04-06
申请人： NEC CORP
发明人： WASHIO TAKUMI
IPC分类号： G06F17/16 , G06F17/13 , G06F15/328 , G06F15/347
摘要： PURPOSE:To prevent deterioration in convergence due to a decrease in vector length, an overhead due to the utilization of a list vector, and paralleling as to the iterative solution of large-scale rule space simultaneous linear equations. CONSTITUTION:Memory access at the time of forward/backward substitution in inverse approximating operation for the iterative solution is facilitated by rearranging an array to improve the operation efficiency of a pipeline processor for vector arithmetic. Further, the forward/backward substitution is performed twice for some part in an area to prevent the deterioration in convergence due to paralleling and the inverse approximating operation which is converged at nearly the same frequency of iteration with a conventional method which is small in parallelism.

6. 发明专利

JPH0675986A LARGE SCALE PARALLEL COMPUTING SYSTEM DEVICE 失效
公开(公告)号：JPH0675986A
公开(公告)日：1994-03-18
申请号：JP5060393
申请日：1993-03-11
申请人： IBM
发明人： JIERARUDO JIYOOJI PECHIYANETSU , SUTAMATEISU BUATSUSHIRIADEISU , HOSE GUADARUUPU DERUGAADOOFURI
IPC分类号： G06F15/16 , G06F15/80 , G06F17/13 , G06F15/328
摘要： PURPOSE: To provide the new PE and related composition of a computer system which can be used for a large-scaled parallel computing system. CONSTITUTION: A most adjacent mesh computer is constituted of processor elements(PE) of an N×N rectangular array, each PE is connected only with the north, south, east, and west PE. When a single line interface is assumed among the PE, total 2N pieces of wires are present in the mesh structure. For example, when an SIMD operation for allowing all the PE to transfer data to north is assumed, symmetric processing elements are overlapped, and the north and south wires and the east and west writes are shared, so that the number of the wires can be reduced into half N , and the array can be reconstituted. Thus, a matrix deforming operation can be executed in one cycle only by exchanging data elements in dual symmetric processing elements.

7. 发明专利

JPH0619953A CALCULATING METHOD FOR ORDINARY DIFFERENTIAL EQUATION 失效
公开(公告)号：JPH0619953A
公开(公告)日：1994-01-28
申请号：JP17226092
申请日：1992-06-30
申请人： CANON KK
发明人： YONEDA HIROSHI
IPC分类号： G06F17/13 , G06F15/328
摘要： PURPOSE:To reduce the quantity of calculation and to simplify a program also by evaluating an error of a numeral integration by a specific equation in a method for numerically integrating an ordinary differential equation and discretely calculating it. CONSTITUTION:A ROM 20 stores a control program for an operation controlling CPU 10 and a RAM 30 secures storage areas for a required error value 31 to be the threshold of error evaluation, time counting width DELTAt 32, a position r0 34, speed 35 and external force F0 36 at time t0, and a position r1 38, speed 39 and external force F1 40 at time t1 obtained after time counting width. In the case of discretely calculating an ordinary differential equation by numerical integration, the equation is expressed by d r/dt =F, and in the case where subscripts i, i+1 express values at time points ti, ti+1 (=ti+DELTAt) which are adjacent to each other, error evaluation is executed based upon the shown equation. Consequently, calculation time can be saved, the program can be simplified and the sorts of memories to be used can be reduced, attaining easy management.

8. 发明专利

JPH06348746A FINITE ELEMENT METHOD OSCILLATION ANALYSIS SYSTEM FOR MULTIPROCESSOR 失效
公开(公告)号：JPH06348746A
公开(公告)日：1994-12-22
申请号：JP14030993
申请日：1993-06-11
申请人： HITACHI LTD
发明人： TAKAYANAGI MASAAKI
IPC分类号： G06F15/16 , G06F17/13 , G06F17/14 , G06F19/00 , G06F15/328 , G06F15/20
摘要： PURPOSE:To shorten arithmetic time by parallelly calculating the characteristic matrix of elements and setting a processor to the hierarchy of a Gaussian point or a Fourier harmonic. CONSTITUTION:Individual data for preparing respective elements are prepared from the coordinate data, material data, physical constant and node coupling data of an entire system by a main processor 1, those data are transferred to processors 2 corresponding to the respective elements, and arithmetic for preparing the matrix of elements is executed by parallel processing. On the other hand, a processor 3 is arranged under the hierarchy of an element level for each Gaussian numeral integration point at the element, arithmetic is executed by parallel processing at the element levels of plural points and after the arithmetic is completed, the results are transferred to the processors of element levels and added. Further, processors 4 of hierarchies corresponding to the wave number parameters of Fourier series to express the peripheral changes of node parameters are arranged in the low order of Gaussian numeral integration points, and parallel processing is performed at the respective hierarchies in a certain order from the least significant Fourier series through the Gaussian numeral integration point to the element level. Thus, arithmetic is enabled by the characteristic matrix on the element stage.

9. 实用新型

JPH0675045U 失效
公开(公告)号：JPH0675045U
公开(公告)日：1994-10-21
申请号：JP2276693
申请日：1993-03-23
IPC分类号： G06F17/13 , G06F15/328

10. 发明专利

JPH06167926A ENERGY TRANSPORT TYPE DEVICE SIMULATING METHOD 失效
公开(公告)号：JPH06167926A
公开(公告)日：1994-06-14
申请号：JP32183692
申请日：1992-12-01
申请人： SONY CORP
发明人： FUIRITSUPU ORUDEIJISU , TOOMASU BEEMU
IPC分类号： G09B9/00 , G06F17/13 , G06F17/50 , G06F19/00 , H01L29/00 , G06F15/20 , G06F15/328 , G06F15/60
摘要： PURPOSE:To surely perform the convergence of simulation by multiplying difference between an updated value in a computing process and a value before update by a relaxation coefficient. CONSTITUTION:It is assumed that computation by repeat of a k-th case is performed, and the updated value at that time shows a value in which a subscript (k+1) is attached on the upper right of each of variables (phi), (n), and T, and the value before update shows a value in which a subscript (k) is attached on the upper right of the variables. In such a case, the Poisson equation is solved by introducing values phi , n , and T for potential, electron density, and carrier temperature, and the updated value phi for the potential can be obtained. Thence, equation of continuity is solved by introducing the update value phi , an initial values n and T , and the updated value n for the electron density can be obtained similarly. Furthermore, the updated values phi , n , and the value T are introduced to an energy conservation equation, however, in this computation, a value DELTAT*R in which DELTAT is multiplied by R is employed instead of DELTAT. Such relaxation coefficient R is selected as a value capable of evading the oscillation of the carrier temperature and current density i.e., 0.01-0.1.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式