基于自适应遗传算法的模拟电路自动设计方法

针对电路进化设计的速度和规模瓶颈,提出并讨论一种新的自适应遗传算法,其特点包括:支持结构自动生成和元件参数标准化的编解码方案,兼顾功能设计和结构化简要求的多目标适应度评估,考虑基因位影响力并跟踪进化进程的遗传参数调整策略等.实验证明,该方法可自动生成电路结构、优化元件参数和化简电路,并显著地减小运算量和提高优化程度.     

基本电子电路

TN7 2005030489 基于均匀设计的多目标自适应遗传算法及应用/赵曙光,焦李成,王宇平,杨万海(西安电子科技大学)//电子学报.-2004,32(10).- 1723-1725,1729 提出一种多目标遗传算法,将均匀设计技术应用于适应度函数合成和交叉算子构造,以提高遗传算法的空间搜索均匀性、子代质量和运算效率. 分析和实验结果表明,该方法可缩短算法运行时间和得到分布较均匀的Pareto有效解集;配合基于元件标称值的网表级高效编码方案和考虑基因位差异的遗传概率调整策略,可实现模拟电路自动设计,通过单次运行即获得对应不同偏好的多种实用化设计结果.图4表2参10(李)     

基于函数级进化方法的乘法器设计

介绍了可进化硬件的基本思想，阐述了基于遗传算法和现场可编程门阵列的函数级进化方法及其特点。采用函数级进化的结构，用遗传算法实现了乘法器的设计；讨论了进化设计中影响进化速度和成功率的因素，通过设计恰当的染色体编码提高了进化速度和评估效率。     

可逆逻辑门进化设计方法的CUDA实现

可逆逻辑门优化程度将直接影响可逆逻辑电路的整体优化,目前已有的优化方法难以实现全局最优。文中基于NCV门库,对遗传算法编码方案和适应度进行了改进,并将进化设计方法改造为CUDA架构下的并行算法应用到可逆逻辑门的优化。其不仅发挥了电路进化设计的全局优化能力,且在不增加硬件规模的前提下,明显提高了电路的搜索速度。     

基于均匀设计的多目标进化算法

现实生活中的很多决策问题都要考虑同时优化若干个目标,多目标优化算法就是要从所有可能的方案中找到最合理、最可靠的解决方案。如何在Pareto界面稀疏区域求得更多非劣解,则使所求出的解的分布更加均匀。如何求出距Pareto界面更近的非劣解以使所求出的解的质量更高。论文基于加权平均法和均匀设计方法设计了一种解决多目标优化问题的新算法。首先,为了找到在Pareto界面上尽可能多、且均匀分布的点,利用均匀设计方法设计了一个交叉算子,该算子让稀疏部分的相邻点进行均匀交叉,以使算法在稀疏部分能找到更多的非劣解,从而使其所求解分布更加均匀。其次,为了克服加权平均法不能找到Pareto界面非凸部分解的缺点,考虑到非劣解界面上相邻距离较远的一对点之间有可能是非劣解界面上非凸部分之一的情况,分别将此两点与距其最近的非劣解集外的点进行交叉,以期在该两点之间找到新的非劣解,这样可能在非劣解界面的非凸部分找到更多的解。最后对两个测试问题进行了数值试验,并和著名的NSGA-Ⅱ算法用算法性能评价的三种度量进行了比较,结果表明了本文算法是有效的。     

解无约束非线性全局优化的一种新进化算法及其收敛性

进化算法是解复杂非线性规划问题的一种新型有效方法,但现有方法的计算量通常较大.为减小计算量,提高算法的效率,本文利用均匀设计来构造新的高效进化算法,新的进化算法本身具有类似于传统优化技术中的局部搜索功能,因此它能非常有效地搜索解空间,保持种群的多样性,减小计算量.文中还证明了新算法的全局收敛性.最后的模拟结果表明,新算法计算量小且收敛速度快.     

进化计算在Matlab中的实现方法

进化计算是一类广泛应用且不断发展的智能计算方法,构成进化计算的主要有遗传算法、进化策略、进化规划,这3种算法侧重点和实现方法各不相同。在分析这3种算法关键要素的基础上,依据Matlab的M文件要求,实现了进化计算程序的设计,给出了Matlab中进化计算关键要素的实现代码。     

遗传算法作曲系统研究

对遗传算法作曲系统进行了一定的分析.通过讨论几种遗传算法作曲系统,给出了各种系统的不同特点及适用范围.     

化学添加剂优化配方试验设计与分析的研究

针对化学添加剂优化配方试验设计优化方法的研究。阐述了配方试验设计与数学模型优化的常用方法,内容涉及全面试验设计、正交试验设计、均匀试验设计、方差分析法、回归分析、人工神经网络、遗传算法;综合分析了各种方法的优点及存在的问题并对所存在的问题提出了解决的方法。     

基于遗传算法的软硬件协同设计探索

软硬件协同设计的目标结构包括一个CPU和多个ASIC,它们通过一条总线进行通信.本文介绍一种用于多目标、多模式系统综合的协同设计的新方法.各个工作模式具有不同的运行概率.在满足设计约束的条件下,我们应用遗传算法对系统的速度和功耗两个目标进行优化.遗传算法是全局算法,它能避免陷入局部最小.     

Multiobjective programming using uniform design and genetic algorithm

The notion of Pareto-optimality is one of the major approaches to multiobjective programming. While it is desirable to find more Pareto-optimal solutions, it is also desirable to find the ones scattered uniformly over the Pareto frontier in order to provide a variety of compromise solutions to the decision maker. We design a genetic algorithm for this purpose. We compose multiple fitness functions to guide the search, where each fitness function is equal to a weighted sum of the normalized objective functions and we apply an experimental design method called uniform design to select the weights. As a result, the search directions guided by these fitness functions are scattered uniformly toward the Pareto frontier in the objective space. With multiple fitness functions, we design a selection scheme to maintain a good and diverse population. In addition, we apply the uniform design to generate a good initial population and design a new crossover operator for searching the Pareto-optimal solutions. The numerical results demonstrate that the proposed algorithm can find the Pareto-optimal solutions scattered uniformly over the Pareto frontier.     

基于Pareto支配的MPRM电路面积与可靠性优化

针对MPRM（Mixed-Polarity Reed-Muller）电路的面积与可靠性折中优化问题,在逻辑级建立面积估算模型以及电路SER（Soft Error Rate）解析评价模型,并采用Pareto支配概念对MPRM电路进行面积与可靠性多目标优化.通过对MPRM电路的XOR部分进行树形异或门分解,并考虑多个输出之间异或门的共享,建立面积估算模型.采用信号概率和故障传播方法,并考虑电路中的逻辑屏蔽因素以及信号相关性,建立电路SER解析评价模型.根据所提出的面积和SER评价模型,采用极性向量的格雷码序穷举搜索MPRM的极性空间得到MPRM电路面积与可靠性的Pareto最优解集,并使用效率因子技术指标选取最终解.MCNC基准电路的实验结果表明,与面积最小MPRM电路相比,所选取的MPRM电路可以在较小面积开销的前提下获得较高电路可靠性.     

A New Automated Design Method Based on Machine Learning for CMOS Analog Circuits

A new simulation based automated CMOS analog circuit design method which applies a multi-objective non-Darwinian-type evolutionary algorithm based on Learnable Evolution Model (LEM) is proposed in this article. The multi-objective property of this automated design of CMOS analog circuits is governed by a modified Strength Pareto Evolutionary Algorithm (SPEA) incorporated in the LEM algorithm presented here. LEM includes a machine learning method such as the decision trees that makes a distinction between high- and low-fitness areas in the design space. The learning process can detect the right directions of the evolution and lead to high steps in the evolution of the individuals. The learning phase shortens the evolution process and makes remarkable reduction in the number of individual evaluations. The expert designer’s knowledge on circuit is applied in the design process in order to reduce the design space as well as the design time. The circuit evaluation is made by HSPICE simulator. In order to improve the design accuracy, bsim3v3 CMOS transistor model is adopted in this proposed design method. This proposed design method is tested on three different operational amplifier circuits. The performance of this proposed design method is verified by comparing it with the evolutionary strategy algorithm and other similar methods.     

Design of a DTCTGAL circuit and its application

By research on the switch-signal theory for multiple-valued logic circuits, the theory of three essential elements and the principle of adiabatic circuits, a design scheme for a double power clock ternary clocked transmission gate adiabatic logic (DTCTGAL) circuit is presented. The energy injection and recovery can be conducted by the bootstrapped NMOSFET, which makes the circuit maintain the characteristics of energy recovery as well as multiple-valued input and output. An XOR/XNOR circuit based on DTCTGAL is also presented using this design scheme. Finally, using the parameters of a TSMC 0.25 μm CMOS device, PSPICE simulation results indicate that the proposed circuits have correct logic and significant low power characteristics.     

多个体参与交叉的Pareto多目标遗传算法

Pareto多目标遗传算法是利用Pareto最优的概念发展出的一种求解多目标优化问题的向量优化方法,能够得到Pareto最优解集.由于采用常规的两个体参与交叉的遗传算法,使整个算法耗费在小生境(Niche)算子上的时间太多,导致算法的效率较低.本文发展出多个体参与交叉的Pareto多目标遗传算法,群体中的个体采用真实值表示,使该算法的速度大大提高,同时证明了相应的模式定理,并提出用方差和熵来分析该算法对解群多样性的影响.最后用算例说明了采用多个体参与交叉的Pareto多目标遗传算法与常规算法比较的结果,证明了本文提出算法的优越性.     

Stochastic-based placement template generator for analog IC layout-aware synthesis

In this paper, a methodology for automatic generation of placement templates for analog integrated circuit design targeted to state-of-the-art optimization-based layout-aware circuit-sizing flows, is proposed. The multi-objective optimization-based placement template generator inputs a Pareto set of sizing solutions and outputs a set of optimal sizing-independent non-slicing B*-tree floorplan representations, i.e., placement templates. Those templates fit the current state of the optimization process and are used within the layout-aware synthesis methodology to generate the floorplan of the following candidate solutions. This innovative methodology combines the advantages of template-based placement approaches, due to its fast packing, with the optimization-based ones, presenting floorplan solutions with improved compactability through the complete evolution of the Pareto set, completely eliminating the template setup effort. Moreover, as the placement template generator runs in parallel with the layout-aware loop, it has no impact on the overall execution time. Experimental results show that the proposed methodology outperforms state-of-the-art multi-template layout-aware synthesis approaches by achieving smaller placement areas for the same performances earlier in the optimization, and further, with a strongly reduced setup effort.     

Elitist nondominated sorting genetic algorithm based RF IC optimizer

An optimization tool for radio frequency integrated circuits (RFICs) based on an elitist nondominated sorting genetic algorithm is introduced. It casts RF circuit synthesis as a multi-objective optimization problem and produces multiple solutions along the Pareto optimal front. Optimization is followed by sensitivity assessment wherein Monte Carlo simulations are performed for the Pareto points with respect to process, voltage, and temperature variations. The tool is validated in the synthesis of a 5.2-GHz direct-conversion receiver front-end that includes a common-gate differential low-noise amplifier, I/Q down-conversion mixers, and a quadrature voltage-controlled oscillator in a 250-nm SiGe BiCMOS process.     

Test Set and Fault Partitioning Techniques for Static Test Sequence Compaction for Sequential Circuits

We propose a new static test set compaction method based on a careful examination of attributes of fault coverage curves. Our method is based on two key ideas: (1) fault-list and test-set partitioning, and (2) vector re-ordering. Typically, the first few vectors of a test set detect a large number of faults. The remaining vectors usually constitute a large fraction of the test set, but these vectors are included to detect relatively few hard faults. We show that significant compaction can still be achieved by partitioning faults into hard and easy faults, and compaction is performed only for the hard faults. This significantly reduces the computational cost for static test set compaction without affecting quality of compaction. The second key idea re-orders vectors in a test set by moving sequences that detect hard faults to the beginning of the test set. Fault simulation of the newly concatenated re-ordered test set results in the omission of several vectors so that the compact test set is smaller than the original test set. Experiments on several ISCAS 89 sequential benchmark circuits and large production circuits show that our compaction procedure yields significant test set reductions in low execution times.     

一种非均匀分布问题分布性维护方法

 几乎所有多目标进化算法(multi-objective optimization evolutionary algorithm,MOEA)都是针对Pareto最优面为均匀分布问题而言.然而现实中很多问题Pareto最优面是非均匀分布的,决策者希望得到一个与Pareto最优面分布类似的解集.现存算法并不能有效解决该问题.对此,提出一种针对于非均匀分布多目标优化问题的维护方法(non-uniformly diversity maintenance method,NUDMM).该方法定义一个反映个体分布"规则"程度的指标——杂乱度,并设计一种降低种群杂乱度的方法,在未知Pareto最优面分布规律情况下有效剔除造成种群混乱的个体.通过与NSGA-II和SPEA2在不同维数下8个非均匀函数上对比实验,表明NUDMM在有效保持问题真实分布的同时,具有良好的收敛性.     

多目标量子编码遗传算法

如何使算法快速收敛到真正的Pareto前沿,并保持解集在前沿分布的均匀性是多目标优化算法重点研究解决的问题。该文提出一种基于量子遗传算法的多目标优化算法,利用量子遗传算法的高效全局搜索能力,在整个解空间内快速搜索多目标函数的Pareto最优解,利用量子遗传算法维持解集多样性的特点,使搜索到的Pareto最优解在前沿均匀分布。通过求解带约束的多目标函数优化问题,对该文算法的多目标优化性能进行了考察,并与NSGAII,PAES,MOPSO和Ray-Tai-Seows算法等知名多目标优化算法进行比较,结果证明了该文算法的有效性和先进性。