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

滑坡泥石流信息遥感获取与空间预测

On the issue of geological hazard evaluation（GHE）, taking remote sensing and GIS systems as experimental environment, assisting with some programming development, this thesis combines multi-knowledges of geo-hazard mechanism, statistic learning, remote sensing (RS), high-spectral recognition, spatial analysis, digital photogrammetry as well as mineralogy, and selects geo-hazard samples from Hong Kong and Three Parallel River region as experimental data, to study two kinds of core questions of GHE, geo-hazard information acquiring and evaluation model. In the aspect of landslide information acquiring by RS, three detailed topics are presented, image enhance for visual interpretation, automatic recognition of landslide as well as quantitative mineral mapping. As to the evaluation model, the latest and powerful data mining method, support vector machine (SVM), is introduced to GHE field, and a serious of comparing experiments are carried out to verify its feasibility and efficiency. Furthermore, this paper proposes a method to forecast the distribution of landslides if rainfall in future is known baseing on historical rainfall and corresponding landslide susceptibility map. The details are as following: (a) Remote sensing image enhancing methods for geo-hazard visual interpretation. The effect of visual interpretation is determined by RS data and image enhancing method, for which the most effective and regular technique is image merge between high-spatial image and multi-spectral image, but there are few researches concerning the merging methods of geo-hazard recognition. By the comparing experimental of six mainstream merging methods and combination of different remote sensing data source, this thesis presents merits of each method ,and qualitatively analyzes the effect of spatial resolution, spectral resolution and time phase on merging image. (b) Automatic recognition of shallow landslide by RS image. The inventory of landslide is the base of landslide forecast and landslide study. If persistent collecting of landslide events, updating the geo-hazard inventory in time, and promoting prediction model incessantly, the accuracy of forecast would be boosted step by step. RS technique is a feasible method to obtain landslide information, which is determined by the feature of geo-hazard distribution. An automatic hierarchical approach is proposed to identify shallow landslides in vegetable region by the combination of multi-spectral RS imagery and DEM derivatives, and the experiment is also drilled to inspect its efficiency. (c) Hazard-causing factors obtaining. Accurate environmental factors are the key to analyze and predict the risk of regional geological hazard. As to predict huge debris flow, the main challenge is still to determine the startup material and its volume in debris flow source region. Exerting the merits of various RS technique, this thesis presents the methods to obtain two important hazard-causing factors, DEM and alteration mineral, and through spatial analysis, finds the relationship between hydrothermal clay alteration minerals and geo-hazards in the arid-hot valleys of Three Parallel Rivers region. (d) Applying support vector machine (SVM) to landslide susceptibility mapping. Introduce the latest and powerful statistical learning theory, SVM, to RGHE. SVM that proved an efficient statistic learning method can deal with two-class and one-class samples, with feature avoiding produce ‘pseudo’ samples. 55 years historical samples in a natural terrain of Hong Kong are used to assess this method, whose susceptibility maps obtained by one-class SVM and two-class SVM are compared to that obtained by logistic regression method. It can conclude that two-class SVM possesses better prediction efficiency than logistic regression and one-class SVM. However, one-class SVM, only requires failed cases, has an advantage over the other two methods as only "failed" case information is usually available in landslide susceptibility mapping. (e) Predicting the distribution of rainfall-induced landslides by time-series analysis. Rainfall is the most dominating factor to bring in landslides. More than 90% losing and casualty by landslides is introduced by rainfall, so predicting landslide sites under certain rainfall is an important geological evaluating issue. With full considering the contribution of stable factors (landslide susceptibility map) and dynamic factors (rainfall), the time-series linear regression analysis between rainfall and landslide risk mapis presented, and experiments based on true samples prove that this method is perfect in natural region of Hong Kong. The following 4 practicable or original findings are obtained: 1) The RS ways to enhance geo-hazards image, automatic recognize shallow landslides, obtain DEM and mineral are studied, and the detailed operating steps are given through examples. The conclusion is practical strongly. 2) The explorative researching about relationship between geo-hazards and alteration mineral in arid-hot valley of Jinshajiang river is presented. Based on standard USGS mineral spectrum, the distribution of hydrothermal alteration mineral is mapped by SAM method. Through statistic analysis between debris flows and hazard-causing factors, the strong correlation between debris flows and clay minerals is found and validated. 3) Applying SVM theory (especially one-class SVM theory) to the landslide susceptibility mapping and system evaluation for its performance is also carried out, which proves that advantages of SVM in this field. 4) Establishing time-serial prediction method for rainfall induced landslide distribution. In a natural study area, the distribution of landslides induced by a storm is predicted successfully under a real maximum 24h rainfall based on the regression between 4 historical storms and corresponding landslides.

同事关系质量及其作用机制

Although the research into coworker relationship quality has been recognized one of key factors related to organization performance, and has been thought a new trend in organization behavior research with the flatting of organization structure and complication of task assignment, there is relatively little empirical research on the mechanism between coworkers’ interaction, contraring to the fruitful results on member exchange research based on social network theory, say nothing of the influence of cultural differences such as GUANXI. This research developed the scale for the assessment of Coworker Relationship Quality by literature review, deep interview, and questionnaires, compared the predictable ability of Coworker Relationship Quality (CRQ) scale and Coworker Exchange (CWX) scale on employees’ work attitudes and behaviors. Finally, the mediating effect of Coworker Relationship Quality between employees’ similarities on personality and their work attitudes and behaviors was investigated. Following are main results. Firstly, we found that the interpersonal communication, trust, and mutual support are the key factors of coworker relationship quality, which is similar to the result getting from western samples. But Chinese people are more GUANXI ORIENTATION, means they want to build longtime relationship with others, not only when they are coworkers, but also when one of them left the organization. Secondly, though the core meaning of CRQ and CWX are same, their predictable ability on organization outcomes is different. CRQ is more powerful than CWX, especially on turnover intention. The result showed that after controlling the effect of demographic variables and CRQ, CWX cannot predict turnover intention significantly, but CRQ can still predict turnover intention significantly after controlling demographic variables and CWX. Thirdly, the partial mediating effect of CRQ between positive affectivity similarity and organizational citizenship behavior, coworker satisfaction, organizational commitment and turnover intention are validated, but we did not find the mediating effect of CRQ between demographic variable similarity and workers’ attitudes and behaviors. The Similarity Attraction Paradigm, Social Identity Theory, and Self Category Theory were supported.

应用解析水平理论考察时间距离对决策的影响

Construal Level Theory is a kind of social cognitive theory, with the main conclusion of that the more time distance is, the more people use high level construal and the more high level construal affect in judging, decision making and such of these cognitive processes, and the low level construal has the mirror process as well. In these ten years, construal level theory is developing very fast and some of the researches of consumer behavior has concerned this developed trend very much. In 2007’s Journal of Consumer psychology, half of the papers focused on construal topics and ask for more researches about these topics. So our research mainly focused on such topics, the application of construal level theory in consumer behavior research. The first part is contained 3 experiments which mean to discuss the questions from reviewing the prior literatures. The research methods, manipulated definition, construal level factors VS construal level events are the goal of first part of this study. The results showed that construal level’s effect might be a consecutive variable and has a consecutive trend when time goes by. Time distance might not be a semantic cognitive effect to construal level but a internal ones for when there’s no verbal information of time distance the results would have the same one as well. The second part of the study is focus on a simplified model of decision-making, which contained two times decision-making, to detect whether and how would the prior decision making affect the next decision making. The study conducted the time distance in first decision making with the situation for the decision and conducted the second time distance in the time distance between the two decision situations. Although we couldn’t get the detail of the weight changes in construal level factors limited by our research, we can have a logical hypothesis based on “one-down- the other-up”. The results showed that the level of construal factors can affect the memory and the more the factors affected, the bigger the error would be. The first time distance and second time distance could effect the second decision in a interaction-way, and support the hypothesis. The whole research is mean to explore the application of construal level theory in consumer behavior research. The conclusion suggests that the periods of products and the consuming may need different marketing strategies in a long-term perspective and time distance should be control are detected in marketing analysis even there’s no time information on the paper.

弈棋问题解决启发式的元策略模型

Problem solving is one of the basic processes of human cognition and heuristic strategy is the key to human problem solving, hence, the studies on heuristic strategy is of great importance in cognitive psychology. Current studies on heuristics in problem solving may be summarized as follows: nature and structure of heuristics, problem structure and representation, expert knowledge and expert intuition, nature and role of image, social cognition and social learning. The present study deals with the nature and structure of heuristics. The Solitaire problem was used in our the experiments. Both traditional experimental method and computer simulation were used to study the nature and structure of heuristics. Through a series of experiments, the knowledge of Solitaire problem solving was summed up, its metastrategy is worked out, and then the the metastrategy by computer simulation and experimental verification are tested.