一种基于复形调优遗传算法的多学科优化方法

建立良好的优化方法,是多学科设计优化(multidisciplinary design optimization,MDO)求解的关键和难点。结合具有全局搜索能力的遗传算法和局部收敛特性的复形调优算法的优势,建立了全局优化算法,并将全局优化算法应用于多学科可行(multidisciplinary feasible,MDF)方法的多学科分析模型求解,建立了两级优化MDF(double optimization MDF,DO-MDF)方法和单级优化MDF(single optimization MDF,SO-MDF)方法两种计算构架。以CASCADE系统产生的多学科问题求解为例,对比5种计算构架对MDO求解的优化结果,结果表明,DO-MDF、SO-MDF两种计算构架通过有效关联学科分析的前馈信息流表现出更好的计算能力。

正交与调优运算法对硝酸毛果芸香碱滴眼液稳定性的研究

目的 :运用正交试验与调优运算法对硝酸毛果芸香碱滴眼液最佳配方进行优选 ,并对其稳定性进行预测。方法 :选取影响硝酸毛果芸香碱滴眼液稳定性3个主要因素 :pH值、温度和贮存时间 ,每个因素选3个水平 ,按L9(33)正交表安排试验 ,以3个指标评定硝酸毛果芸香碱滴眼液稳定性 ,筛选最稳定因素进行温度加速试验 ,并采用t0 9 法对其稳定性进行预测。结果 :pH为4、温度为8℃、贮存时间为12h为最佳优选点 ;温度加速试验预测其贮存期为76 9d。结论 :用正交与调优运算法对硝酸毛果芸香碱滴眼液稳定性进行预测 ,方法简便、快速。

基于非概率区间模型的可靠性分析与优化

根据影响目标零件结构参数变化因素以及材料性能参数的区间特性,采用可靠性分析技术与结构优化方法,对目标零件结构的控制参数、材料强度及载荷分布等参量的不确定性进行分析,通过对非概率区间可靠性进行分析,构造出结构失效概率度量的可靠性指标,结合区间约束的n维复形调优算法,获得了结构参数的最优结果。以钢坯吊具钳板为例,验证了该方法的实用性和有效性。该方法为基于可靠性的产品设计提供了新的途径。

AccSMBO:一种基于超参梯度和元学习的SMBO加速算法

为了利用最佳超参高概率范围和超参梯度,提出了加速的序列模型优化算法(sequential model-based optimization algorithms,SMBO)——AccSMBO算法.AccSMBO使用了具有良好抗噪能力的基于梯度的多核高斯过程回归方法,利用元学习数据集的meta-acquisition函数.AccSMBO自然对应的并行算法则使用了基于元学习数据集的并行算法资源调度方案.基于梯度的多核高斯过程回归可以避免超参梯度噪音对拟合高斯过程的影响,加快构建较好超参效果模型的速度.meta-acquisition函数通过读取元学习数据集,总结最佳超参高概率范围,加快最优超参搜索.在AccSMBO自然对应的并行算法中,并行资源调度方法使更多的并行计算资源用于计算最佳超参高概率范围中的超参,更快探索最佳超参高概率范围.上述3个技术充分利用超参梯度和最佳超参高概率范围加速SMBO算法.在实验中,相比于基于传统的SMBO算法实现的SMAC(sequential model-based algorithm configuration)算法、基于梯度下降的HOAG(hyperparameter optimization with approximate gradient)算法和常用的随机搜索算法,AccSMBO使用最少的资源找到了效果最好的超参.

灰度发布在智能问答模型训练中的应用研究

随着人工智能的广泛应用,智能模型训练成为研发周期的关键路径。灰度发布的引入,将模型训练实现多"进程",及时抓取用户反馈信息,提高模型的成型效率。

INTEGRATED OPERATOR GENETIC ALGORITHM FOR SOLVING MULTI-OBJECTIVE FLEXIBLE JOB-SHOP SCHEDULING

In the flexible job-shop scheduling problem （FJSP）, each operation has to be assigned to a machine from a set of capable machines before alocating the assigned operations on all machines. To solve the multi-objective FJSP, the Grantt graph oriented string representation （GOSR） and the basic manipulation of the genetic algorithm operator are presented. An integrated operator genetic algorithm （IOGA） and its process are described. Comparison between computational results and the latest research shows that the proposed algorithm is effective in reducing the total workload of all machines, the makespan and the critical machine workload.

基于能量模型的水稻生长模型

该文考虑生命活动与物质交换和能量流动的关系,从热力学出发,引入一维能量模型。在一维能量模型的基础上,将水稻对外界能量的摄取能力与水稻生长季中的气温变化规律联系起来,建立在气温变化条件下水稻的生长模型。结合气温增长率、水稻最适气温和不同阶段的生物量,以徐士良单形调优算法为基础,利用Forcal二维方程参数拟合,得到关于水稻自然增长率、气温波幅和初始值等参数的最优拟合。将最优拟合数据代人水稻的生长模型,拟合得到水稻的生长曲线和水稻生长随气温的变化趋势。以水稻的生长模型为基础,定性分析水稻的临界气温和最大生长率出现时间。通过分析发现,水稻生长季的临界气温与气温增长率无关;水稻的最大生长率出现时间为水稻分蘖后期、孕穗期。为考虑异常气温对水稻生长的影响,利用Matlab对水稻分蘖期低温和生殖生长期高温的水稻生长进行模拟,结果表明:水稻后期高温对产量影响比前期受低温影响严重。

基于云计算的油气管道SCADA系统设计

为了有效提高现有油气管道SCADA系统的可靠性和资源利用率,提出基于云计算的油气管道SCADA系统和最优资源调度算法。采用最优资源调度算法的基于云计算的油气管道SCADA系统通过整合数据中心服务器内存资源和处理器资源提高SCADA系统资源利用率,通过实现历史数据多机备份并保障实时数据主副版本并行执行从而有效提高SCADA系统可靠性。实验证明最优资源调度算法具有良好的可靠性和可调度性。

Optimization of total harmonic current distortion and torque pulsation reduction in high-power induction motors using genetic algorithms

This paper presents a powerful application of genetic algorithm （GA） for the minimization of the total harmonic current distortion （THCD） in high-power induction motors fed by voltage source inverters, based on an approximate harmonic model. That is, having defined a desired fundamental output voltage, optimal pulse patterns （switching angles） are determined to produce the fundamental output voltage while minimizing the THCD. The complete results for the two cases of three and five switching instants in the first quarter period of pulse width modulation （PWM） waveform are presented. Presence of harmonics in the stator excitation leads to a pulsing-torque component. Considering the fact that if the pulsing-torques are at low frequencies, they can cause troublesome speed fluctuations, shaft fatigue, and unsatisfactory performance in the feedback control system, the 5th, 7th, 1 lth, and 13th current harmonics （in the case of five switching angles） are constrained at some pre-specified values, to mitigate the detrimental effects of low-frequency harmonics. At the same time, the THCD is optimized while the required fundamental output voltage is maintained.

人工智能在网络运维中的应用

本文简要介绍了人工智能在网络运维中的应用优势,针对智能化网络运维方案的构建展开探讨,并根据当前网络运维中的常见故障分析了人工智能在网络运维中的实际应用策略。

Peak shaving operation optimization of high proportion new energypower generation considering wind-solar complementationand source-load coupling

To optimize peaking operation when high proportion new energy accesses to power grid,evaluation indexes are proposed which simultaneously consider wind-solar complementation and source-load coupling.A typical wind-solar power output scene model based on peaking demand is established which has anti-peaking characteristic.This model uses balancing scenes and key scenes with probability distribution based on improved Latin hypercube sampling(LHS)algorithm and scene reduction technology to illustrate the influence of wind-solar on peaking demand.Based on this,a peak shaving operation optimization model of high proportion new energy power generation is established.The various operating indexes after optimization in multi-scene peaking are calculated,and the ability of power grid peaking operation is compared whth that considering wind-solar complementation and source-load coupling.Finally,a case of high proportion new energy verifies the feasibility and validity of the proposed operation strategy.

An optimizing algorithm of static task scheduling problem based on hybrid genetic algorithm

To reduce resources consumption of parallel computation system, a static task scheduling opti- mization method based on hybrid genetic algorithm is proposed and validated, which can shorten the scheduling length of parallel tasks with precedence constraints. Firstly, the global optimal model and constraints are created to demonstrate the static task scheduling problem in heterogeneous distributed computing systems（HeDCSs）. Secondly, the genetic population is coded with matrix and used to search the total available time span of the processors, and then the simulated annealing algorithm is introduced to improve the convergence speed and overcome the problem of easily falling into local minimum point, which exists in the traditional genetic algorithm. Finally, compared to other existed scheduling algorithms such as dynamic level scheduling （ DLS）, heterogeneous earliest finish time （HEFr）, and longest dynamic critical path（ LDCP）, the proposed approach does not merely de- crease tasks schedule length, but also achieves the maximal resource utilization of parallel computa- tion system by extensive experiments.

A hybrid differential evolution algorithm for meta-task scheduling in grids

Task scheduling is one of the core steps to effectively exploit the capabilities of heterogeneous re-sources in the grid.This paper presents a new hybrid differential evolution(HDE)algorithm for findingan optimal or near-optimal schedule within reasonable time.The encoding scheme and the adaptation ofclassical differential evolution algorithm for dealing with discrete variables are discussed.A simple but ef-fective local search is incorporated into differential evolution to stress exploitation.The performance of theproposed HDE algorithm is showed by being compared with a genetic algorithm(GA)on a known staticbenchmark for the problem.Experimental results indicate that the proposed algorithm has better perfor-mance than GA in terms of both solution quality and computational time,and thus it can be used to de-sign efficient dynamic schedulers in batch mode for real grid systems.

Gravitational search algorithm for coordinated design of PSS and TCSC as damping controller

A newly developed heuristic global optimization algorithm, called gravitational search algorithm （GSA）, was introduced and applied for simultaneously coordinated designing of power system stabilizer （PSS） and thyristor controlled series capacitor （TCSC） as a damping controller in the multi-machine power system. The coordinated design problem of PSS and TCSC controllers over a wide range of loading conditions is formulated as a multi-objective optimization problem which is the aggregation of two objectives related to damping ratio and damping factor. By minimizing the objective function with oscillation, the characteristics between areas are contained and hence the interactions among the PSS and TCSC controller under transient conditions are modified. For evaluation of effectiveness and robustness of proposed controllers, the performance was tested on a weakly connected power system subjected to different disturbances, loading conditions and system parameter variations. The cigenvalues analysis and nonlinear simulation results demonstrate the high performance of proposed controllers which is able to provide efficient damping of low frequency oscillations.

Peer priority based data scheduling algorithm in P2P streaming system

In order to solve the problem that the existing data scheduling algorithm cannot make full use of neighbors' bandwidth resources when allocating data request among several senders in the multisender based P2P streaming system,a peer priority based scheduling algorithm is proposed.The algorithm calculates neighbors' priority based on peers' historical service evaluation as well as how many wanted data that the neighbor has.The data request allocated to each neighbor is adjusted dynamically according to the priority when scheduling.Peers with high priority are preferred to allocate more data request.Experiment shows the algorithm can make full use of neighbors' bandwidth resources to transmit data to reduce server pressure effectively and improve system scalability.

Rescheduling Problems with Bicriteria

In the rescheduling on a single machine,a set of original jobs has already been scheduled to minimize some cost objective,when a new set of jobs arrives and creates a disruption.The decision maker needs to insert the new jobs into the existing schedule without excessively disrupting it.In this paper,we consider hierarchical optimization between the scheduling cost of all the jobs and the degree of this disruption.For every problem,we provide either a polynomial time algorithm or an intractable result.

A real-time fault-tolerant scheduling algorithm with low dependability cost in on-board computer system

To make the on-board computer system more dependable and real-time in a satellite, an algorithm of the fault-tolerant scheduling in the on-board computer system with high priority recovery is proposed in this paper. This algorithm can schedule the on-board fault-tolerant tasks in real time. Due to the use of dependability cost, the overhead of scheduling the fault-tolerant tasks can be reduced. The mechanism of the high priority recovery will improve the response to recovery tasks. The fault-tolerant scheduling model is presented simulation results validate the correctness and feasibility of the proposed algorithm.

Modeling and Optimisation of Precedence-Constrained Production Sequencing and Scheduling for Multiple Production Lines Using Genetic Algorithms

This paper presents an integrated methodology for the modelling and optimisation of precedence-constrained production sequencing and scheduling for multiple production lines based on Genetic Algorithms （GA）. The problems in this class are NP-hard combinatorial problems, requiring a triple optimisation at the same time： allocation of resources to each line, production sequencing and production scheduling within each production line. They are ubiquitous to production and manufacturing environments. Due to nature of constraints, the length of solutions for the problem can be variable. To cope with this variability, new strategies for encoding chromosomes, crossover and mutation operations have been developed. Robustness of the proposed GA is demonstrated by a complex and realistic case study.

Genetic Optimization Algorithm of PID Decoupling Control for VAV Air-Conditioning System

Variable-air-volume (VAV) air-conditioning system is a multi-variable system and has multi coupling control loops. While all of the control loops are working together, they interfere and influence each other. A multivariable decoupling PID controller is designed for VAV air-conditioning system. Diagonal matrix decoupling method is employed to eliminate the coupling between the loop of supply air temperature and that of thermal-space air temperature. The PID controller parameters are optimized by means of an improved genetic algorithm in floating point representations to obtain better performance. The population in the improved genetic algorithm mutates before crossover, which is helpful for the convergence. Additionally the micro mutation algorithm is proposed and applied to improve the convergence during the later evolution. To search the best parameters, the optimized parameters ranges should be amplified 10 times the initial ideal parameters. The simulation and experiment results show that the decoupling control system is effective and feasible. The method can overcome the strong coupling feature of the system and has shorter governing time and less over-shoot than non-optimization PID control.

Strategy of changing cracking furnace feedstock based on improved group search optimization

The scheduling process of cracking furnace feedstock is important in an ethylene plant. In this paper it is described as a constraint optimization problem. The constraints consist of the cycle of operation, maximum tube metal temperature, process time of each feedstock, and flow rate. A modified group search optimizer is proposed to deal with the optimization problem. Double fitness values are defined for every group. First, the factor of penalty function should be changed adaptively by the ratio of feasible and general solutions. Second, the "excellent" infeasible solution should be retained to guide the search. Some benchmark functions are used to evaluate the new algorithm. Finally, the proposed algorithm is used to optimize the scheduling process of cracking furnace feedstock. And the optimizing result is obtained.