多机器人动态编队的强化学习算法研究

在人工智能领域中 ,强化学习理论由于其自学习性和自适应性的优点而得到了广泛关注 随着分布式人工智能中多智能体理论的不断发展 ,分布式强化学习算法逐渐成为研究的重点 首先介绍了强化学习的研究状况 ,然后以多机器人动态编队为研究模型 ,阐述应用分布式强化学习实现多机器人行为控制的方法 应用SOM神经网络对状态空间进行自主划分 ,以加快学习速度 ;应用BP神经网络实现强化学习 ,以增强系统的泛化能力 ;并且采用内、外两个强化信号兼顾机器人的个体利益及整体利益 为了明确控制任务 ,系统使用黑板通信方式进行分层控制 最后由仿真实验证明该方法的有效性

基于多Agent可重构机器人控制方法的研究

根据多 Agent理论中的协商、合作机制和可重构机器人结构的分布性 ,将集中式的机器人控制分配到一组关节 Agent中 ,每个 Agent控制机器人的一个关节 ,使用这种分布式方法 ,得到了一种新的通用机器人控制方法 ,即将关节机器人的复杂控制转换为多个简单子系统的控制 ,该方法可应用于具有不同构型的机器人系统 ,特别适用于可重构模块化机器人的控制。利用微分运动理论提出了一种新的决策方法 ,便于 Agent之间的合作与协商。仿真实验结果表明该方法是一种可行的机器人控制方法

两栖机器人轮桨腿驱动机构多目标优化设计

提出了一种既能够在陆地上爬行,又能够在一定深度的水下浮游和在海底爬行的新概念轮桨腿一体化两栖机器人;多运动模式和复合移动机构是该机器人的突出特点.分析了轮桨腿复合式驱动机构的运动机理,并采用多目标优化设计理论和算法,对驱动机构的爬行性能和浮游特性进行了综合优化,得到了两栖机器人驱动机构的结构优化参数.虚拟样机的仿真结果证明了该轮桨腿一体化两栖机器人驱动机构的综合运动性能良好,对非结构环境具有一定的适应能力。

基于独立分量分析的地震属性分析方法研究

Conventional seismic attribute analysis is not only time consuming, but also has several possible results. Therefore, seismic attribute optimization and multi-attribute analysis are needed. In this paper, Fuyu oil layer in Daqing oil field is our main studying object. And there is much difference between seismic attributes and well logs. So under this condition, Independent Component Analysis (ICA) and Kohonen neural net are introduced to seismic attribute optimization and multi-attribute analysis. The main contents are as follows: (1) Now the method of seismic attribute compression is mainly principal component analysis (PCA). In this article, independent component analysis (ICA), which is superficially related to PCA, but much more powerful, is used to seismic reservoir characterizeation. The fundamental, algorithms and applications of ICA are surveyed. And comparation of ICA with PCA is stydied. On basis of the ne-entropy measurement of independence, the FastICA algorithm is implemented. (2) Two parts of ICA application are included in this article: First, ICA is used directly to identify sedimentary characters. Combined with geology and well data, ICA results can be used to predict sedimentary characters. Second, ICA treats many attributes as multi-dimension random vectors. Through ICA transform, a few good new attributes can be got from a lot of seismic attributes. Attributes got from ICA optimization are independent. (3) In this paper, Kohonen self-organizing neural network is studied. First, the characteristics of neural network’s structure and algorithm is analyzed in detail, and the traditional algorithm is achieved which has been used in seism. From experimental results, we know that the Kohonen self-organizing neural network converges fast and classifies accurately. Second, the self-organizing feature map algorithm needs to be improved because the result of classification is not very exact, the boundary is not quite clear and the velocity is not fast enough, and so on. Here frequency sensitive principle is introduced. Combine it with the self-organizing feature map algorithm, then get frequency sensitive self-organizing feature map algorithm. Experimental results show that it is really better. (4) Kohonen self-organizing neural network is used to classify seismic attributes. And it can be avoided drawing confusing conclusions because the algorithm’s characteristics integrate many kinds of seismic features. The result can be used in the division of sand group’s seismic faces, and so on. And when attributes are extracted from seismic data, some useful information is lost because of difference and deriveative. But multiattributes can make this lost information compensated in a certain degree.

多分量对偶超导理论和口袋模型；Multi-component Dual Superconductor Theory and Bag Model

基于联络新参数化方案研究了多分量对偶超导模型。给出了多分量Ginzburg-Landau模型中的自对偶解,并研究了磁通量子数趋于无穷大时的墙涡旋解,以及与口袋模型之间的联系。

The multi-weight neuron with geometry algorithm and its application

A novel approach for multi-dimension signals processing, that is multi-weight neural network based on high dimensional geometry theory, is proposed. With this theory, the geometry algorithm for building the multi-weight neuron is mentioned. To illustrate the advantage of the novel approach, a Chinese speech emotion recognition experiment has been done. From this experiment, the human emotions are classified into 6 archetypal classes: fear, anger, happiness, sadness, surprise and disgust. And the amplitude, pitch frequency and formant are used as the feature parameters for speech emotion recognition. Compared with traditional GSVM model, the new method has its superiority. It is noted that this method has significant values for researches and applications henceforth.

Face recognition based on PCA and multi-degree of freedom neurons

An algorithm of PCA face recognition based on Multi-degree of Freedom Neurons theory is proposed, which based on the sample sets' topological character in the feature space which is different from "classification". Compare with the traditional PCA+NN algorithm, experiments prove its efficiency.

Multi-exponentially photoelectric response of bacteriorhodopsin

A thin oriented bacteriorhodopsin (bR) him is deposited on a stainless steel slide by use of the electrophoretic sedimentation method. A junction is made with electrolyte gels having a counterelectrode to construct a bR-based photoelectric detector;. The photoelectric response signal to a 10 ns laser pulse is measured. A theory on the photoelectric kinetics of bR is developed based on the concept of the charge displacement current and the bR photocycle rate equations. Comparison between the theoretical and experimental results proves that the bR photoelectric response to a short laser pulse is a multi-exponential process. The decay time constants and amplitudes of each, exponential component are obtained by data fitting.

Deformation Monitoring and Forecasting Model Study Based on Grey System Theory

The grey system theory studies the uncertainty of small sample size problems. This paper using grey system theory in the deformation monitoring field, based on analysis of present grey forecast models, developed the spatial multi-point model. By using residual modification, the spatial multi-point residual model eras developed in further study. Then, combined with the sedimentation data of Xiaolangdi Multipurpose Dam, the results are compared and analyzed, the conclusion has been made and the advantages of the residual spatial multi-point model has been proved.

Self-assembly of star block copolymers by dynamic density functional theory

The dynamic mean-field density functional method, driven from the generalized time-dependent Ginzburg-Landau equation, was applied to the mesoscopic dynamics of the multi-arms star block copolymer melts in two-dimensional lattice model. The implicit Gaussian density functional expression of a multi-arms star block copolymer chain for the intrinsic chemical potentials was constructed for the first time. Extension of this calculation strategy to more complex systems, such as hyperbranched copolymer or dendrimer, should be straightforward. The original application of this method to 3-arms block copolymer melts in our present works led to some novel ordered microphase patterns, such as hexagonal (HEX) honeycomb lattice, core-shell HEX lattice, knitting pattern, etc. The observed core-shell HEX lattice ordered structure is qualitatively in agreement with the experiment of Thomas [Macromolecules 31, 5272 (1998)].

黄骅坳陷马东地区沙河街组重力流水道储层评价及剩余油研究

In this paper, taking Madong district of Huanghua depression as a case, based on the theory of sequence stratigraphy, sedimentology, reservoir geology and geophysics, according to core analysis, seismic attribute analysis, logging constrained inversion, multi-data correlation of strata, reservoir modeling, etc. the lower and middle first member of Shahejie formation of the study area was forecasted and evaluated. As a result, a number of reservoir prediction and remaining oil distribution methods suitable to oil exploitation of gravity flow channel reservoir are presented. Scientific foundation is provided to the next adjustment of development program and exploitation of the remaining oil. According to high resolution sequence stratigraphy theory, precise stratigraphic framework was founded, the facies types and facies distribution were studied under the control of stratigraphic framework, the technologies of seismic attribute abstraction and logging constrained inversion. Result shows that gravity flow channel, as the main facies, developed in the rising period of base-level cycle, and it was formed during the phase of contemporaneous fault growth. As the channel extends, channel width was gradually widened but thickness thined. The single channels were in possession of a great variety of integrated modes, such as isolated, branching off, merging and paralleling, forming a kind of sand-mud interblending complex sedimentary units. Reservoir quality differs greatly in vertical and horizontal direction, and sedimentary microfacies is main controlling factor of the reservoir quality. In major channel, deposition thickness is great, and petrophysical property is well. While in marginal channel, reservoir is thinner, and petrophysical property is unfavorable. Structure and reservoir quality are main factors which control the oil and gas distribution in the study area. On the basis of the research about the reservoir quality, internal, planar and 3-D reservoir heterogeneities are characterized, and the reservoir quality was sorted rationally. At last, based on the research of reservoir numerical simulation of key well group, combined with reservoir performance analysis and geological analysis above, remaining oil distribution patterns controlled by internal rhythm of gravity flow channel were set up. Through this research, a facies-restrained reservoir prediction method integrating multi-information was presented, and potential orientation of remaining oil distribution in gravity flow channel reservoir is clarified.