An improved cellular automaton model considering the effect of traffic lights and driving behaviour

This paper proposes an improved cellular automaton model to describe the urban traffic flow with the consideration of traffic light and driving behaviour effects. Based on the model,the characteristics of the urban traffic flow on a single-lane road are investigated under three different control strategies,i.e.,the synchronized,the green wave and the random strategies. The fundamental diagrams and time-space patterns of the traffic flows are provided for these strategies respectively. It finds that the dynamical transition to the congested flow appears when the vehicle density is higher than a critical level. The saturated flow is less dependent on the cycle time and the strategies of the traffic light control,while the critical vehicle density varies with the cycle time and the strategies. Simulated results indicate that the green wave strategy is proven to be the most effective one among the above three control strategies.     

Optimality of the barrier strategy in de Finetti’s dividend problem for spectrally negative Lévy processes: An alternative approach

The optimal dividend problem proposed in de Finetti [1] is to find the dividend-payment strategy that maximizes the expected discounted value of dividends which are paid to the shareholders until the company is ruined. Avram et al. [9] studied the case when the risk process is modelled by a general spectrally negative Lévy process and Loeffen [10] gave sufficient conditions under which the optimal strategy is of the barrier type. Recently Kyprianou et al. [11] strengthened the result of Loeffen [10] which established a larger class of Lévy processes for which the barrier strategy is optimal among all admissible ones. In this paper we use an analytical argument to re-investigate the optimality of barrier dividend strategies considered in the three recent papers.     

Integration of business function models into an aggregate enterprise systems model

Introduced are methods that combine models of distinct business functions into an aggregate model of an enterprise system to assist management’s strategic decision making. Models of individual business functions are reviewed, and equations quantifying relationships presented. Using methods of system theory, including block diagrams, non-dimensionalization, and state equation methods, these business function sub-models were assembled into a composite enterprise systems model. The formulated aggregate model is illustrated with industry examples for tire companies; nonetheless the aggregate model can be used to assess other industries. Values of parameters for the system model were determined from data obtained from annual reports of publicly owned companies. Simulations closely matched the companies’ published performance over ensuing years. The developed aggregate enterprise model has significant predictive capabilities for modern corporations.     

基于可拓策略生成技术的商品搜索服务改进研究

购买者在使用商品搜索服务时所得搜索结果数量不符合其主观要求是一个典型而普遍的矛盾问题.在分析矛盾问题的基础上,提出以可拓策略生成技术改进商品搜索服务的方法.该方法通过对购买者提出的搜索要求进行发散分析及相关分析,以形成各种可能变换,并依据关联函数寻求问题的解变换,以及利用评价函数对解变换进行优选.给出了关于策略生成的基本算法,并实现于笔记本电脑导购网站平台上.实验及测试结果表明,系统能有效生成矛盾问题的解变换,从而改进了搜索服务的质量.     

海关走私监管的演化博弈分析

利用演化博弈的方法对海关缉私部门和走私商之间相互作用时的策略,进行了分析。结果表明：当企业进行走私活动的收益大于不走私的收益、海关缉私部门对走私企业处罚力度过轻或对缉私检查的成本过高时,走私必然发生。要确保走私的发生率降低,就必须对进行走私的企业进行严惩,降低其的预期收益。同时,要加强缉私队伍建设。     

具有红利边界的Erlang（2）风险模型

给出了具有边界红利策略的Erlang（2）风险模型,在此红利策略下,若保险公司的盈余在红利线以下时不支付红利,否则红利以低于保费率的常速率予以支付．对于该模型,本文推导了Gerber-Shiu折现惩罚函数所满足的两个积分－微分方程和更新方程．     

Neuroevolution strategies for episodic reinforcement learning

Because of their convincing performance, there is a growing interest in using evolutionary algorithms for reinforcement learning. We propose learning of neural network policies by the covariance matrix adaptation evolution strategy (CMA-ES), a randomized variable-metric search algorithm for continuous optimization. We argue that this approach, which we refer to as CMA Neuroevolution Strategy (CMA-NeuroES), is ideally suited for reinforcement learning, in particular because it is based on ranking policies (and therefore robust against noise), efficiently detects correlations between parameters, and infers a search direction from scalar reinforcement signals. We evaluate the CMA-NeuroES on five different (Markovian and non-Markovian) variants of the common pole balancing problem. The results are compared to those described in a recent study covering several RL algorithms, and the CMA-NeuroES shows the overall best performance.     

基于多信号流模型的诊断策略动态生成

针对静态故障树诊断序列长、人工可干预度差和不支持多现象并行推理问题,提出了基于D-矩阵和Rollout信息启发搜索算法的故障诊断策略动态生成方法,介绍了基于TEAMS多信号流模型的D-矩阵数据获取方法和Rollout信息启发搜索算法的DLL独立封装技术、输入输出数据规范和动态交互控制方法;基于此开发的交互式智能诊断系统(IIDS)软件平台经实际现场测试诊断验证,对诊断现场出现的多种客观条件,如系统工作模式或测试资源变化、用户干预和多故障现象并发推理等具有很好的适应性,故障覆盖率和隔离率指标高,对提高装备故障诊断的效率、准确性和灵活性具有显著作用。     

基于路况预测的PHEV能量管理策略

为了提高插电式混合动力汽车(plug-in hybrid electric vehicle, PHEV)的燃油经济性,减少排放,提出了基于路况预测的PHEV能量管理策略;首先,建立PHEV系统结构并在此基础上依据动力电池SOC(State of charge)变化规律定义了3种PHEV基本工作模式;然后,设计路况识别模糊控制器对当前行驶路况进行识别并预测;最后,根据预测的路况类型结合合理规划的动力电池SOC的曲线约束,制定PHEV能量管理策略;仿真结果表明,该能量管理策略能够较好的使动力电池SOC保持在设定的参考轨迹附近,提高燃油经济性,减少排放。     

Hopf bifurcation and chaos control in a new chaotic system via hybrid control strategy

In this paper, we investigate the problem of Hopf bifurcation and chaos control in a new chaotic system. A hybrid control strategy using both state feedback and parameter control is proposed. Theoretical analysis shows that the Hopf bifurcation critical value can be changed via hybrid control. Meanwhile, this control strategy can also control the chaos state. The direction and stability of bifurcating periodic solutions are determined by the normal form theory and the center manifold theorem. Finally, numerical simulations are carried out to illustrate the effectiveness of the main theoretical results.