会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 82. 发明专利
    • METHOD FOR SEARCHING MINIMUM VALUE
    • JPH04205049A
    • 1992-07-27
    • JP32581690
    • 1990-11-29
    • OKI ELECTRIC IND CO LTD
    • MIYAMAE TORU
    • G06F17/11G06F19/00
    • PURPOSE:To move search operation to accurate search while searching the minimum value by finding the average inclination of a function in a search range and determining the next search position that can be predicted to take a smaller value to achieve the minimum value of the function. CONSTITUTION:The output from an error function arithmetic section 3 that receives input pattern 1 and standard pattern 2 is wired so as to be inputted to buffer memory 4, and the content of memory 4 is searched for the minimum value. To the minimum value search section 5, a general search parameter register 6 is connected, and this content of memory 4 can be sequentially switched in accordance with various conditions. Further, the output from minimum value search section 5 is inputted to the mininum value register 7 of the function, where vibration detection and determination of the minimum value are carried out. Furthermore, when obtaining the minimum value of the function, factor output 8 is obtained. Arithmetic section 3 and search section 5 consists of microprocessor, and so on, and in memory 4 an error function is stored being made to correspond to its factor. When the minimum value of the function is obtained, the corresponding output 8 is used to recognize characters, and so on with the aid of a circuit not shown in the figure.
    • 83. 发明专利
    • FINITE ELEMENT MESH GENERATING DEVICE
    • JPH03175580A
    • 1991-07-30
    • JP31497189
    • 1989-12-04
    • NEC CORPNEC SOFTWARE LTD
    • NAKAYAMA MASAKIKUMAZAWA KAZUYAICHINOSE TOSHIO
    • G06F17/50G06F17/11
    • PURPOSE:To easily generate finite element meshes having a complicated form based on already prepared finite element meshes by designating an expansion/ reduction rate by a user to generate a coordinate transform matrix and moving already prepared finite element meshes along a guide line and generating new finite element meshes based on the locus of this movement. CONSTITUTION:The expansion/reduction rate with which the size of already prepared finite element meshes is expanded/reduced at the time of moving these meshes along the guide line is inputted from an input device 3 and is stored in an expansion/reduction rate storage means 23, and a coordinate transform matrix generating means 12 generates the coordinate transform matrix based on the stored expansion/reduction rate and information related to the guide line, and a new finite element meash generating means 13 uses the generated coordinate transform matrix to perform coordinate transformation. Thus, existing finite element meshes are moved along the guide line while being expanded/reduced with the designated expansion/reduction rate, and new finite element meshes are generated based on the locus of this movement.
    • 84. 发明专利
    • CUBIC EXPRESSION ARITHMETIC UNIT
    • JPH02184970A
    • 1990-07-19
    • JP569789
    • 1989-01-11
    • RICOH KK
    • MATSUOKA SHIGEKI
    • G06F17/10G06F17/11G06T11/20
    • PURPOSE:To shorten computation time and to reduce a required hardware quantity by replacing a multiplication to be necessary in an arithmetic process with a summing operation. CONSTITUTION:The values of f(T) are obtained with a cubic expression f(T)= AT +BT +CT+D in the sequence of the values of T from T=0 with a 1 increment. Namely, the values of f(T) are obtained in the sequence of f(0), f(1), f(2) - f(n). At this time, when the values of f(T) are to be obtained for T>=1, in the case of T=1, which is the first value of T, expressions A1=A, B1=B, C1=C, D1=D are set up, An, Bn, Cn, and Dn in the case of T=n, which is the n-th value of T, are obtained according to expressions An=An-1, Bn=3 An-1+2Bn-1+Cn-1, and Dn=An-1+Bn-1+Cn-1+Dn-1 based on An-1, Bn-1, and Cn-1 in the case of T=n-1, which is the (n-1)th value of T, and the n-th value of f(T), namely, f(n), is obtained by obtaining the value of an expression An+Bn+Cn+Dn. Thus, the computation time can be shortened without increasing the whole hardware quantity.
    • 86. 发明专利
    • FORMING DEVICE FOR TRIANGULAR MESH FOR DIGITIZING
    • JPS63238665A
    • 1988-10-04
    • JP7212087
    • 1987-03-26
    • TOSHIBA CORP
    • SHIGYO NAOYUKISATO KOICHI
    • G06F17/11G06T17/20
    • PURPOSE:To reduce the number of obtuse angle triangles by using an average interval of dots of a dot string as the reference to control lengths of sides of a triangle mesh to be generated and moving coordinates of a nodal point to be noticed. CONSTITUTION:An average value of lengths of two sides of a part to be processed and an average value of lengths of all sides are compared with each other and a correction ratio (r) is calculated based on the comparison result in accordance with a formula I, and the average value of lengths of two sides is multiplied by the correction ratio (r) to generate triangle elements and sides of triangle elements are averaged. Coordinates Pi of a nodal point (i) and indicated by a formula II are moved based on distances from nodal points adjacent to the nodal point (i) and coordinates of respective nodal points to form a mesh. In these formulas, (n), hi, Ni, Pj, and Hj are the total number of dots of the dot string indicating an area where triangular areas should be generated, the distance between points (i) and i-1, the number of nodal points adjacent to the nodal point (i), coordinates of a nodal point (j), and the distance between nodal point (i) and (j) respectively. Thus, the occurrence of an obtuse angle triangle in the triangular mesh is suppressed.
    • 90. 发明专利
    • OPTIMIZATION PROCESSING SYSTEM FOR LINEAR PROGRAMMING
    • JPS6235965A
    • 1987-02-16
    • JP17529585
    • 1985-08-09
    • FUJITSU LTD
    • KANEZASHI TETSUYA
    • G06F17/11
    • PURPOSE:To make optimization processing efficiently by processing optimization of cutout part matrix regarding it as original linear programming of the cutout matrix and determining pivot successively from rows of less non-zero elements and making sweeping out calculation. CONSTITUTION:A cutout matrix extracting and processing section 3 takes out the cutout matrix as a program of separate linear programming basing on information of non-cutout matrix. A cutout matrix row index table preparing section 4 arranges rows constituting the matrix in order of larger number of non-zero elements, gives them row index and prepares a table, and further prepares column index vector that indicates correspondence of number of non- zero elements and names of variables looked in the direction of column. A cutout matrix optimization processing section 5 makes sweeping out calculation of specified rows in order of smaller number of non-zero elements basing on the content of the row index table, and makes optimization processing of cutout matrix.