首页> 外文期刊>電子情報通信学会技術研究報告. コンカレント工学. Concurrent System Technology >A new algorithm to derive generators for both invariants and particular solutions of state equations for P/T Petri Nets - a method via Z-Basis for augmented incidence matrix
【24h】

A new algorithm to derive generators for both invariants and particular solutions of state equations for P/T Petri Nets - a method via Z-Basis for augmented incidence matrix

机译:推导P / T Petri网的不变量和状态方程的特定解的生成器的新算法-基于Z-Basis的增广入射矩阵方法

获取原文
获取原文并翻译 | 示例
           

摘要

Recently; it has been shown that an arbitrary nonnegative integer inhomogeneous solution x = x{sub}(H6) + x{sub}(P6) ∈(Z{sub}(NN)){sup}(n×l) for state (or matrix) equation Ax=b(A∈Z{sup}(m×n),b∈Z{sup}(m×l)) of P/T Petri nets is represented at level 6 as the simplest form such that the homogeneous solution x{sub}(H6) ∈(Z{sub}(NN)){sup}(n×l) is a linear combination of the l{sub}6 nonnegative integer generators U{sub}6 = {u{sub}I ∈(Z{sub}(NN)){sup}(n×l)|Au{sub}i = 0{sup}(m×l), i∈I(l{sub}6)}, with nonnegative integer weights, and the particular solution x{sub}(P6) ∈(Z{sub}(NN)){sup}(n×l) is a single term v{sub}j ∈(Z{sub}(NN)){sup}(n×l), i.e., just one term out of the k{sub}6 nonnegative integer particular solutions V{sub}6 ={v{sub}j∈(Z{sub}(NN)){sup}(n×l)|Av{sub}i =b, j ∈I(k{sub}6)}, with also a nonnegative integer and unity weight. Therefore it is strongly hoped to give an algorithm for finding effectively the level-6 generators (U{sub}6, V{sub}6) from the given matrices A∈Z{sup}(m×n) and b∈Z{sup}(m×l) for a P/T Petri net. In this paper a new and useful algorithm for (U{sub}6, V{sub}6) via Z-basis (i.e., the level-2 generator for (A(top)~)(x{top}~)= 0{sup}(m×l)) (U{top}~){sub}2 ={(x{top}~){sub}i∈Z{sup}((n250L羖)×l)|(A{top}~)(x{top}~){sub}i = 0{sup}(m×l), i∈I(n+l-r)} is proposed, where (A{top}~)=[A, b] ∈Z{sup}(m×(n×l)) and r = rank(A) = rank(A{top}~). First, U{sub}2 is obtained from (A{top}~)=[A, -b] by the extended Kannan and Bachen algorithm which is with polynomial time and space complexity. Secondly, from (U{top}~){sub}2, find the maximal set of minimal T-invariants (U{top}~){sub}6 = {(u{top}~){sub}i ∈(Z{sub}(NN)){sup}((n×l)×l)|(A{top}~)(u{top}~){sub}i =0{sup}(m×l), i∈I((l{top}~){sub}6)} for (A{top}~)(x{top}~) = 0{sup}(m×l). Thirdly from (U{top}~){sub}6, obtain simultaneously both U{sub}6 and V{sub}6 i.e., the level-6 generators for Ax = b. Fourthly reduce (U{sub}6, V{sub}6) to the level-5 and level-4 generators, U{sub}4 and V{sub}4 = V{sub}5, by using support(x) properties for x ∈(Z{sub}(NN)){sup}(n×l), because of U{sub}6 =U{sub}5 {contains} U{sub}4 and V{sub}6 {contains} V{sub}5 = V{sub}4, where U{sub}L = {u{sub}i ∈(Z{sub}(NN)){sup}(n×l)|Au{sub}i = 0{sup}(m×l), i ∈ I(l{sub}L)} and V{sub}L ={v{sub}j ∈(Z{sub}(NN)){sup}(n×l)|Av{sub}j =b, j ∈ I(k{sub}L)} for L=4, 5. Note that l{sub}6 = l{sub}5 and l{sub}4 means the total number of all minimal T- invariants and the that of all minimal support (or all elementary) T-invariants for Ax =0{sup}(m×l), respectively. Note also that k{sub}6 and k{sub}5 = k{sub}4 means the total number of all nonnegative integer particular solutions and that of all nonnegative integer basic particular solutions for Ax =b, respectively, where a basic particular solution is a particular solution, but the converse is not always true.
机译:最近;已经证明,对于状态为(或)的任意非负整数不均匀解x = x {sub}(H6)+ x {sub}(P6)∈(Z {sub}(NN)){sup}(n×l) P / T Petri网的方程Ax = b(A∈Z{sup}(m×n),b∈Z{sup}(m×l))在第6级表示为最简单形式,使得齐次解x {sub}(H6)∈(Z {sub}(NN)){sup}(n×l)是l {sub} 6个非负整数生成器U {sub} 6 = {u {sub } I∈(Z {sub}(NN)){sup}(n×l)| Au {sub} i = 0 {sup}(m×l),i∈I(l {sub} 6)},其中非负整数权重,并且特定解x {sub}(P6)∈(Z {sub}(NN)){sup}(n×l)是单项v {sub} j∈(Z {sub}(NN )){sup}(n×l),即k {sub} 6个非负整数特解V {sub} 6 = {v {sub}j∈(Z {sub}(NN))中仅一项{sup}(n×l)| Av {sub} i = b,j∈I(k {sub} 6)},并且具有非负整数和单位权重。因此,强烈希望提供一种从给定矩阵A∈Z{sup}(m×n)和b∈Z{来有效地找到6级生成器(U {sub} 6,V {sub} 6)的算法。 sup}(m×l)表示P / T Petri网。本文提出了一种基于Z的(U {sub} 6,V {sub} 6)的新实用算法(即(A(top)〜)(x {top}〜)的2级生成器= 0 {sup}(m×l))(U {top}〜){sub} 2 = {(x {top}〜){sub}i∈Z{sup}((n250L羖)×l)|提出(A {top}〜)(x {top}〜){sub} i = 0 {sup}(m×l),i∈I(n + lr)},其中(A {top}〜)= [A,b]∈Z{sup}(m×(n×l)),r =等级(A)=等级(A {top}〜)。首先,通过扩展的Kannan和Bachen算法从(A {top}〜)= [A,-b]中获得U {sub} 2,该算法具有多项式的时间和空间复杂度。其次,从(U {top}〜){sub} 2中,找到最小T不变量的最大集合(U {top}〜){sub} 6 = {(u {top}〜){sub} i∈( Z {sub}(NN)){sup}((n×l)×l)|(A {top}〜)(u {top}〜){sub} i = 0 {sup}(m×l),对于(A {top}〜)(x {top}〜)= 0 {sup}(m×l),i∈I((l {top}〜){sub} 6)}。第三,从(U {top}〜){sub} 6中,同时获得U {sub} 6和V {sub} 6,即Ax = b的6级生成器。第四,通过使用support(x)将(U {sub} 6,V {sub} 6)减少到5级和4级生成器U {sub} 4和V {sub} 4 = V {sub} 5 x∈(Z {sub}(NN)){sup}(n×l)的性质,因为U {sub} 6 = U {sub} 5 {包含} U {sub} 4和V {sub} 6 {包含} V {sub} 5 = V {sub} 4,其中U {sub} L = {u {sub} i∈(Z {sub}(NN)){sup}(n×l)| Au {sub} i = 0 {sup}(m×l),i∈I(l {sub} L)}和V {sub} L = {v {sub} j∈(Z {sub}(NN)){sup}( n×l)| Av {sub} j = b,对于L = 4、5,j∈I(k {sub} L)}。请注意,l {sub} 6 = l {sub} 5和l {sub} 4表示分别对于Ax = 0 {sup}(m×l)的所有最小T不变量的总数和所有最小支持(或所有基本)T不变量的总数。还要注意,k {sub} 6和k {sub} 5 = k {sub} 4分别表示Ax = b的所有非负整数特定解的总数和所有非负整数基本特定解的总数,其中解决方案是一种特殊的解决方案,但是反过来并不总是正确的。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号