N-particles intermediate coordinate-momentum representation and its application

This paper constructs the new common eigenvectors of n intermediate coordinate-momentum operators which are complete and orthonormal. The intermediate coordinate-momentum representation of a multi-particles system is proposed and applied to a generaln-mode quantum harmonic oscillators system with coordinate-momentum coupling.

Operator Fredholm Integration Equation in Intermediate Coordinate-Momentum Representation and Its Analogue to Thermo Conduction Equation

Using the technique of integration within an ordered product （IWOP） of operators we construct intermediate coordinate-momentum representation, with which we build a type of operator Fredholm integration equation that is an operator generalization of the solution of thermo conduction equation. Then we seach for the solution of operator Fredholm integration equations, which provides us with a new approach for deriving some operator identities.

Optical scaled Fresnel-Fourier transform obtained via intermediate coordinate-momentum representation

Using the intermediate coordinate-momentum representation {x}s,r, we introduce a new Hadamard transform. It is found that the operator U corresponding to this transform can be considered as a combination of the Fresnel operator F （r, s） and the Fourier transform operator F- by decomposing U. We also find that the matrix element s,r 〈x｜U｜f） just corresponds to an optical scaled Presnel Fourier transform.

New two-mode intermediate momentum-coordinate representation with quantum entanglement and its application

We construct four linear composite operators for a two-particle system and give common eigenvectors of those operators. The technique of integration within an ordered product （IWOP） of operators is employed to prove that those common eigenvectors are complete and orthonormal. Therefore, a new two-mode intermediate momentum-coordinate representation which involves quantum entanglement for a two-particle system is proposed and applied to some twobody dynamic problems. Moreover, the pure-state density matrix ｜ξ1,ξ2｜ C,D C,D（ξ1, ξ2｜ is a Radon transform of Wigner operator.

Application of Coordinate-Momentum Intermediate Representation in Solving Eigenstate Problem of a Coupled Oscillator System

We study the eigenstate problem of a kind of coupled oscillators in the new quantum mechanical representation ｜q,μ,υ〉, which is defined as the eigenvector of the operator （μQ ＋ υP）, whereμ and υ are two real parameters. We also use the U operator transformation method to deal with the same problem. We obtain the normally ordered product expressions of U operator and eigenvector. It is shown that the ground state of system Hamiltonian is a squeezed state.

Coordinate-Momentum Intermediate Representation and Marginal Distributions of Quantum Mechanical Bivariate Normal Distribution

We introduce bivariate normal distribution operator for state vector [ψ） and find that its marginal distribution leads to one-dimensional normal distribution corresponding to the measurement probability ｜λ,v〈x｜.ψ〉｜^2, where ｜x〉λ,v is the coordinate-momentum intermediate representation. As a by-product, the one-dimensional normal distribution in statistics can be explained as a Radon transform of two-dimensional Gaussian function.

Intermediate Representation Using Graph Visualization Software

In this paper, a method to initiate, develop and visualize an abstract syntax tree (AST) in C++ source code is presented. The approach is in chronological order starting with collection of program codes as a string and split into individual characters using regular expression. This will be followed by separating the token grammar using best first search (BFS) algorithm to determine node having lowest value, lastly followed by graph presentation of intermediate representation achieved with the help of graph visualization software (GraphViz) while former is implemented using python programming language version 3. The efficacy of our approach is used in analyzing C++ code and yielded a satisfactory result.

BWDSP SIMD指令的WHIRL中间表示生成研究

WHIRL是Open64采用的一种通用的中间表示,支持多种高级语言和多种目标机。Open64的后端优化工作都是基于WHIRL设计的。BWDSP是一款支持SIMD指令的数字信号处理器,SIMD指令能够提高该处理器的指令并行性。SIMD优化也需要产生WHIRL中间表示,本文对如何产生BWDSP SIMD指令的WHIRL中间表示进行了研究,并在BWDSP的编译器OCC上得以实现。

基于二幂阶矩阵的量子中间表示与翻译

在两能级量子计算系统中,所有量子门、量子态和测量算子都可以表示为2的幂次方阶矩阵(简称二幂阶矩阵)的形式,而现有量子编程语言未考虑该特性。因此,提出一种二幂阶矩阵类型系统,并设计相应的量子中间表示。首先,在定理证明器Coq中利用递归对偶结构实现二幂阶矩阵系统,可以精确描述量子门、量子态和测量算子;其次,设计一套量子中间表示作为编程工具,可以自动将量子程序翻译为二幂阶矩阵表达式;最后,展示量子傅里叶变换的编写和翻译过程。二幂阶矩阵系统为基于定理证明器的量子编程语言提供了更精确、更简洁的类型系统,量子中间表示实现了从二幂阶矩阵到程序语言的过渡,提供了在二幂阶矩阵系统中编写量子程序的有效手段。

程序自动并行化系统中IR的面向对象设计

从构造一个高性能的程序自动并行化系统的需求出发，介绍了程序自动并行化系统ＡＧＡＳＳＩＺ中ＩＲ（ＩｎｔｅｒｍｅｄｉａｔｅＲｅｐｒｅｓｅｎｔａｔｉｏｎ）的设计原则与设计方法，阐明了此种ＩＲ的设计给整个程序自动并行化系统的设计所带来的便利。

Normal ordering and antinormal ordering of the operator （fQ＋gP）n and some of their applications

In this paper by virtue of the technique of integration within an ordered product （IWOP） of operators and the intermediate coordinate-momentum representation in quantum optics, we derive the normal ordering and antinormal ordering products of the operator （fQ＋gP）n when n is an arbitrary integer. These products are very useful in calculating their matrix elements and expectation values and obtaining some useful mathematical formulae. Finally, the applications of some new identities are given.

New Approach for Calculating Tomograms of Quantum States by Virtue of the IWOP Technique

We show that for tomographic approach there exist two mutual conjugate quantum states [p,σ, τ） and [x,μ, ν） （named the intermediate coordinate-momentum representation）, and the two Radon transforms of the Wigner operator are just the pure-state density matrices [p）σ1τσ1τ （p｜ and （p）λ,ν,λ,ν,（x｜ respectively. As a result, the tomogram of quantum states can be considered as the module-square of the states＇ wave function in these two representations. Throughout the paper we fully employ the technique of integration within an ordered product of operators. In this way we establish a new convenient formalism of quantum tomogram.

O2-Oxidation of Cyanomethylene Radical:Infrared Identification of Criegee Intermediates syn-and anti-NCC(H)OO

Cyanomethylene radical(HCCN)is an important intermediate in the nitrile chemistry in both the earth’s and the Titan’s atmosphere.Despite that the mechanism for the oxidation of HCCN has been already computationally explored,the key Criegee intermediate,NCC(H)OO,remains unobserved yet.By photolyzing mixtures(1:50:1000)of either HC(N2)CN/O2/N2(266 nm)or HCCNCO/O2/N2(193 nm)at 15.0 K,the elusive carbonyl oxides NCC(H)OO,in syn-and anti-conformations,have been generated and characterized with IR spectroscopy.The spectroscopic identification is supported by ^18O-labeling experiments and the quantum chemical calculations at the BP86/6-311++G(3df,3pd)level.Upon subsequent UV-light irradiation,both conformers of NCC(H)OO further react with O2 and yield NCC(O)H and O3,whereas,the dioxirane isomer HC(O2)CN,which is lower than syn-NCC(H)OO by 23.7 kcal/mol at the CCSD(T)-F12a/aug-cc-pVTZ//BP86/6-311++G(3df,3pd)level,was not observed experimentally.

A New Kind of Bipartite Entangled State Representation and its Application

We construct four linear composite operators for a two-particle system and give common eigenvectors of those operators. The technique of integration within an ordered product of operators is employed to prove that those common eigenvectors are complete and orthonormal. Therefore, a new intermediate coordinate-momentum representation for a two-particle system is proposed and applied to some two-body dynamic problems.

Recast combination functions of coordinate and momentum operators into their ordered product forms

By using the parameter differential method of operators,we recast the combination function of coordinate and momentum operators into its normal and anti-normal orderings,which is more ecumenical,simpler,and neater than the existing ways.These products are very useful in obtaining some new differential relations and useful mathematical integral formulas.Further,we derive the normally ordered form of the operator(fQ+gP)^-n with n being an arbitrary positive integer by using the parameter tracing method of operators together with the intermediate coordinate-momentum representation.In addition,general mutual transformation rules of the normal and anti-normal orderings,which have good universality,are derived and hence the anti-normal ordering of(fQ+gP)^-n is also obtained.Finally,the application of some new identities is given.

基于控制流和数据流分析的内存拷贝类函数识别技术

内存错误漏洞仍是当前网络攻击中造成危害最严重的漏洞之一.内存错误漏洞的产生往往与对内存拷贝类函数的误用有关.目前针对内存拷贝类函数的识别主要借助于符号表和代码特征模式匹配,具有较高的误报率和漏报率,并且适用性较差.提出了一种内存拷贝类函数识别技术CPYFinder(copy function finder).该技术在内存拷贝类函数控制流特征的基础上,将二进制代码转换为中间语言表示VEX IR(intermediate representation)进行数据流的构建和分析,根据内存拷贝类函数在数据流上的特征进行识别.该技术能够在较低的运行时间下对多种指令集架构(x86,ARM,MIPS,PowerPC)的二进制程序中的内存拷贝类函数进行识别.实验结果表明,相比于最新的工作BootStomp和SaTC,CPYFinder在对内存拷贝类函数识别上具有更好的表现,在精准率和召回率上得到更好的平衡,并且运行时间与SaTC几乎相等,仅相当于BootStomp耗时的19%.此外,CPYFinder在漏洞函数识别上也具有更好的表现.

基于轴向空间注意力和中间融合表示的单图像三维重建模型

提出一种基于注意力和中间融合表示的三维重建模型,旨在重建具有精细化结构的三维模型。该方法利用轴向空间注意力机制学习不同方向的信息,将其嵌入编码器中以捕获局部结构特征;并基于双流网络推测深度图和三维平均形状以设计中间融合表示模块,该模块能够有效地融合可见表面细节信息,从而更好地描绘对象的三维空间结构。实验结果表明:所提出的轴向空间注意力机制和中间融合表示模块增强了特征提取的能力,IoU和F-score比PixVox++分别提升了1.3%和0.4%,三维重建效果更优。

基于混合代码表示的源代码脆弱性检测

软件脆弱性对网络与信息安全产生了极大的威胁,而脆弱性的根源在于软件源代码。因为现有的传统静态检测工具和基于深度学习的检测方法没有完整地表示代码特征,并且简单地使用词嵌入方法转换代码表示,所以检测结果准确率低,误报率高或漏报率高。因此,提出了一种基于混合代码表示的源代码脆弱性检测方法来解决代码表示不完整的问题,并提升检测性能。首先将源代码编译为中间表示(IR),并提取程序依赖图;然后基于数据流和控制流分析进行程序切片来得到结构化的特征,同时使用doc2vec嵌入节点语句得到非结构化的特征;接着使用图神经网络(GNN)对混合特征进行学习;最后使用训练好的GNN进行预测和分类。为了验证所提方法的有效性,在软件保证参考数据集(SARD)和真实世界数据集上进行了实验评估,检测结果的F1值分别达到了95.3%和89.6%。实验结果表明,所提方法有较好的脆弱性检测能力。

面向多源异构数据库的SQL解析与转换方法研究

传统的单一数据库模式难以适应如今多样化的数据管理需求。如何将多个异构独立的数据库进行集成,对数据库系统进行整体控制和协同操作成为研究热点。针对此问题进行面向多源异构数据库的SQL解析与转换方法研究,通过建立通用的中间表示模型,对异构数据库请求进行语法树解析、语义分析与模型转换,实现了不同数据库之间的互操作。在基于TPC-H基准测试数据集的功能测试中,测试系统对数据类型和语法操作的支持度达到100%。在性能测试中,测试系统在跨平台的增删改查操作时间上,较官方工具分别快了13.1 ms、8.8 ms、22.5 ms与2.3ms。实验验证了该方法的正确性与可行性。

面向深度学习算子的循环不变式外提算法

TVM是一个深度学习编译器,支持将TVM的领域专用语言即张量表达式定义的算子编译生成目标平台的代码,并在高级中间表示TVM IR上进行一系列优化。张量表达式对算子执行循环变换,产生与循环迭代变量相关的复杂表达式的计算,在多层嵌套循环内这些计算包含了大量的循环不变式。然而,传统的循环不变量外提技术不能判断不变量外提是否能带来额外收益,无法发现操作数顺序不同的循环不变表达式,不能处理嵌套的条件表达式,并且与目标平台编译器优化存在冲突等。由于这些挑战,传统的循环不变量外提算法无法直接用于深度学习编译器的优化,提出了一种融合深度学习代价函数和启发式策略的循环不变量外提算法。该算法基于深度学习编译器的高层中间表示,通过调整操作数顺序和简化嵌套条件表达式等方法规范化表达式。为了衡量优化的收益,在结合TVM IR和目标平台的特点的基础上,提出了一个新的面向深度学习的不变式外提代价指标函数。在开源编译器TVM 0.7版本上,通过新增优化遍的形式,具体实现了所介绍的算法以及代价函数。为评测算法的有效性,在Tesla P4的图形处理器(GPU)平台上对TVM TOPI的测试算子集中27个典型算子不同输入规模的511个测例进行了测试。实验结果表明47.6%的算子性能得到提升,最大加速比大于40.0%。