基于多级专家系统的矿井设计智能优化方法

针对现有矿井设计优化中直接的数值优化难以直接获得最优结果的问题,本文以缓倾斜煤层为应用条件,提出一种基于多级专家系统的矿井设计智能优化方法,该方法具有启发与联想的智能特点,经实例应用表明,比较好地解决上述长期困扰着矿井设计优化工作者的一大难题。     

“Unfolding” Natural Resource-Use Information Systems: Fieldwork in Botswana

My research examines the performance of natural resource-use information systems. I question why such systems, despite receiving substantial financial and human investment, appear to have a weak impact on projects, programs, and policy intended for rural poverty alleviation in developing countries. Drawing on my understanding of the process of unfolding introduced by C. West Churchman, and its particular relationship to Habermasian constitutive interests, I reflect on my experiences of using the concept during fieldwork undertaken in Botswana. The concept is found useful on two fronts: first, it provides a purposeful guide for gathering and processing information/knowledge—what I have termed an epistemological intent; second, it provides a useful template for evaluating other information systems; in particular, the role of expertise—what I have termed an ontological intent. The process of unfolding, as I understand it, also provides an invitation for constructive (rather than self-indulgent) personal reflection: what might be termed a constitutive reflexive intent. By making information gathering and knowledge generation less mystical and a more transparent social activity, the conceptual and practical application of the process of unfolding can help toward retrieving inquiry as being a purposeful, openly political, and thereby less deceptive engagement: features which I believe are found particularly wanting in the business of rural development information gathering.     

基于动态贝叶斯网的时变样本空间无监督学习

提出了针对概率分布参数时变的样本空间的参数估计问题,由于一般的聚类和EM算法等参数估计方法是针对概率分布时不变的样本空间的问题所采用的。因此,对于概率分布时变的样本空间而言,这些方法均不能有效、准确地估计实时变化的样本参数。通过构建动态贝叶斯网,利用先验和后验的知识进行预测和滤波,结合贝叶斯增量式学习方法并充分利用了其学习所获得的样本空间概率模型变化演进的规律,这样可以较准确、平滑地学习实时概率模型及其分布参数。     

基于MCMC方法的虚拟试验贝叶斯校准研究

虚拟试验利用计算模型研究复杂物理过程,并预测其性能.计算模型输入参数中通常包含部分固定但未知参数,可利用计算模型结果和少量有限的试验数据校准未知参数,并研究存在建模不确定性时虚拟试验的预测问题.提出了一种贝叶斯统计方法,采用高斯过程为仿真计算模型以及模型不确定性建模,利用Markov chain蒙特卡罗抽样方法计算校准参数和仿真模型预测后验分布.设计测试算例演示所提出方法的高效性.     

基于Bayes近似限法的串联系统置信区间的求解

运用基于Bayes的方法对红外空空导弹进行抗干扰性能评估时,只能分别得到导引头和导弹控制系统的抗干扰概率P<,g>、P<,c>及其置信区间,而难以求出导弹复杂系统的抗干扰概率在一定的置信度下的置信区间.针对该问题,提出用Bayes近似限的方法拟合导弹导引头、控制系统的整个串联系统的抗干扰概率的分布,得出该复杂系统的置信度为y的近似置信区间的求解公式,并给出导弹串联系统数字仿真中抗干扰概率的置信区间的求解实例,对该求解方法的正确性进行了验证.     

Research on the self-defence electronic jamming decision-making based on the discrete dynamic Bayesian network

The manner and conditions of running the decision-making system with self-defense electronic jamming are given.After proposing the scenario of applying discrete dynamic Bayesian network to the decision making with self-defense electronic jamming,a decision-making model with self-defense electronic jamming based on the discrete dynamic Bayesian network is established.Then jamming decision inferences by the aid of the algorithm of discrete dynamic Bayesian network are carried on.The simulating result shows that this method is able to synthesize different targets which are not predominant.In this way,various features at the same time,as well as the same feature appearing at different time complement mutually;in addition,the accuracy and reliability of electronic jamming decision making are enhanced significantly.     

基于MCMC模拟的相关系数平稳序列模型及其应用

提出了基于MCMC方法来估计相关系数平稳序列模型的参数;给出基于贝叶斯分布的相关系数平稳序列模型参数的算法;在无信息先验分布下,模拟证明了用此方法估计相关系数平稳序列模型参数的优良效果。最后对实际的广西电网-月负荷数据,分别用基于相关系数平稳序列模型的MCMC方法和极大似然估计法以及基于经典的ARMA模型建模,结果表明采用MCMC方法得到的模型给出的预测是最好的。     

Coordinated Bayesian optimal approach for the integrated decision between electronic countermeasure and firepower attack

The coordinated Bayesian optimization algorithm (CBOA) is proposed according to the characteristics of the function independence, conformity and supplementary between the electronic countermeasure (ECM) and the firepower attack systems. The selection criteria are combinations of probabilities of individual fitness and coordinated degree and can select choiceness individual to construct Bayesian network that manifest population evolution by producing the new chromosome. Thus the CBOA cannot only guarantee the effective pattern coordinated decision-making mechanism between the populations, but also maintain the population multiplicity, and enhance the algorithm performance. The simulation result confirms the algorithm validity.     

基于民生视角的政府绩效评价灰色贝叶斯网络模型

建立了一种基于民生视角的政府绩效评价贝叶斯网络模型。研究了运用极大熵模型,求解网络的参数分布问题;结合灰色系统的思想,给出了政府绩效评价中的不确定信息处理的方法;通过对网络结构的分析,研究了评价信息在传递过程中的衰减性;最后,通过贝叶斯网络推理,给出网络中各节点的定量评价,为基于民生视角的政府绩效评价提供了一种新的理论方法和工具。     

基于MPI的主从式并行MCMC

马尔可夫链蒙特卡洛(MCMC)是贝叶斯网络(BN)的一种近似推理算法,为保证其时间性能,在生成马尔可夫链的组成序列时,通常只使用较少的样本,其推理精度较低.以Gibbs抽样为基本框架,提出一种BN的并行MCMC推理方法PMCMC,在生成马尔可夫链的组成序列时,通过增加对BN中每个节点的抽样频率来提高推理精度,并在消息传递接口MPJ的支持下,利用主从式并行机制来实现其推理过程,保证了推理的时间性能.在3个不同BN上的推理实验结果表明,PMCMC在提高推理精度的同时有效保证了推理的时间生能.