贝叶斯决策网在智能决策系统中的应用研究

面向复杂的决策问题，提出基于贝叶斯决策网在智能决策支持系统中的应用，通过决策网可以直接计算不同行动方案下的效用，直接为决策者服务。文章深入阐述了贝叶斯决策网的语义、语法、效用函数的构造和详细探讨了在求解复杂决策问题中决策模型构建和推理算法实现的方法和途径。最后建立了供应链管理中零售商预定冰激凌的实例，实际运行表明把基于贝叶斯决策网的智能决策系统用于复杂决策问题是有效的，尤其适用于多角色参与的决策问题。     

知识化制造环境下航空发动机装配车间自进化

针对产品动态到达的航空发动机装配车间, 对知识化制造系统的自进化问题进行研究. 将自进化的思想应用于该装配车间, 提出了知识化制造环境下该装配车间自进化问题的求解算法. 根据双层规划理论, 建立了系统在每个决策时刻静态决策问题的一般数学模型, 并设计了一种基于可行域搜索的双层遗传算法(FR-BiGA) 对模型进行求解. 仿真结果验证了该模型与算法的有效性和可行性, 且实验数据表明, 自进化的系统具有相对较优的生产性能.

     

基于智能体模型的路径规划与行为模拟仿真研究

智能体模型是当前在模拟仿真领域内能够实现的最为接近人类的方式。是实现复杂行为模拟仿真的重要方法,智能体能够通过逻辑驱动或自学习模式进行决策。多智能体系统在复杂战场环境的模拟仿真方面有着较大的优势,其分布式交互原理能够模拟不同兵种或者不同个体的行为与路径规划决策。文章针对当前智能体模型的研究现状以及路径规划领域的各种方法进行了简要介绍,并指出优点和不足,最后对其发展趋势进行了展望。     

一种基于可传递置信模型的分布智能体决策融合方法*

在分析与研究对抗性多机器人系统决策问题的基础上,提出了一种基于可传递置信模型的多智能体决策融合方法;构建了决策融合体系架构,分别设计了基于证据推理的观测智能体模型,基于TBM的决策智能体模型以及决策融合中心模型,给出了相应的算法。通过在机器人足球中的应用及仿真实验,体现了本方法在对抗性多机器人系统中决策制定的良好性能及效果。     

基于Q-学习和行为树的CGF空战行为决策

 空战行为决策的智能性是学术界关注的重要问题之一。提出一种基于Q-学习和行为树的CGF空战行为决策方法。通过构建CGF空战行为树模型,实现CGF智能行为;通过在行为树上的Q-学习,使CGF具有不断进化的能力。仿真结果表明,该算法在与传统算法对抗中,性能优势明显且学习能力较强。     

基于微粒群算法的城市公交线网模型研究

微粒群算法是求解组合优化问题的一种新的群体智能进化算法,从城市公交乘客选择出行路径的决策因素出发,以微粒群算法进化机理为核心,结合微粒群进化算法中的局部搜索与全局搜索同时进行的优点和运筹学旅行商组合优化理论,系统地建立了规划城市智能交通公交线网最短路径的数学模型进化算法,并通过MATLAB 7.0进行了实例仿真,得到了城市公交线网出行选择模型中总运输里程权重最短的优化目标。仿真结果也表明,该进化算法模型是解决城市公交线网规划的有效方法。     

基于Markov对策的强化学习及其在RoboCup中的应用

论文研究了Markov对策模型作为学习框架的强化学习,提出了针对RoboCup仿真球队决策问题这一类复杂问题的学习模型和具体算法。在实验中,成功实现了守门员决策,并取得了良好的效果,证明了算法的可行性和有效性。     

数字孪生驱动下微电网多智能体协调控制优化

传统微电网系统设备分时控制能力差,存在协同控制不足的问题。对此,提出采用思维进化算法优化多智能体控制系统。首先基于“源-网-荷-储”概念提出微电网多智能体模型,并优化目标环保成本与运维成本;然后在数据分析的基础上,通过历史光伏发电功率数据与当日气象数据,构建分布式“源”功率预测模型;最后采用思维进化算法对智能体种群调度策略的适应值进行趋同异化优化,迭代出最优种群调度策略。功率预测仿真结果表明,在类簇为3时,模型具有最高的预测精确性,较传统预测方法精度提升了5.6%;控制策略仿真结果表明,MEA算法的微电网协调控制决策优化后,提高多智能体协同控制能力,降低了环保成本与运维成本。     

基于智能合约和CPSO$_$DNN的流程工艺参数可信自决策模型

针对流程制造过程中工艺关系复杂、优化效率难保证和数据安全问题,提出一种基于智能合约和改进混沌粒子群算法的工艺参数可信自决策模型PPO-TS.首先,基于区块链技术设计适合流程制造工艺特性的数据集成与存储机制,通过数据上链技术实现数据的可信存储;然后,设计工艺参数自决策智能合约机制,利用智能合约搭建基于区块链广播式通信协议的工艺参数优化网络,启动网络并编译质量指标访问、优化自决策和决策自执行智能合约,通过自动触发工艺参数优化事务完成自决策和自执行过程;在此基础上构建基于改进混沌粒子群算法CPSO和深度神经网络DNN的优化算法CPSO_DNN,实现流程制造工艺参数优化;最后,以现场采集的某流程生产线数据为例,验证了PPO-TS模型的实用性和有效性,为流程制造工艺参数优化提供了一种新思路.     

人事甄选智能决策支持系统分析与设计

在人事招聘中,对大量候选人进行选择是一个复杂的多因素决策问题.这个决策过程涉及到多种相关因素,主观性强,干扰因素多,科学决策难度大.构建相应的企业人事甄选智能决策支持系统可以提供决策支持.本文首先分析了相应的心理测评模型,然后根据心理模型特点对适用神经网络模型进行了对比,最后提出了综合应用心理学和人工智能技术的系统详细功能模块设计,实现了企业招聘过程的客观化、人工智能化和科学决策支持.     

自组织柔性ERP研究

制造企业结成自组织制造网络,以提高自身的柔性和快速反应能力,应对复杂、多变的混沌环境的挑战。面向事务处理的经典ERP系统,不能有效支持企业在混沌环境下不断变化业务模型及业务模型的持续改进。提出了基于Mul-ti-agent的自组织柔性ERP(SOF-ERP),SOF-ERP不仅是可执行代码的集合,而且是模型与软件组件的集合。基于Multi-agent技术和CORBA规范的组件化SOF-ERP具有分散化、智能化决策与预测能力,实现了软件系统的自适应性修改,提高了系统对环境的适应性和演化能力,解决了异构企业信息集成的兼容性问题,降低ERP实施的复杂性,促进企业向自组织柔性企业的演化。实施案例证明了其可行性。     

Intelligent control systems. II. Design of self-organized robust knowledge bases in contingency control situations

Informational technology for design of robust knowledge bases based on discovery synergetic effect of their self-organization in contingencies is developed. The principle of minimum of information entropy used in quantum algorithm guarantees the necessary condition of self-organization, the minimum of required initial information in learning signals; thermodynamic minimum criterion of the new measure of generalized entropy production provides sufficient condition of self-organization, robustness of control processes. In risky conditions and contingencies, the optimization of knowledge bases according to information-thermodynamic criteria, using the quantum algorithm of self-organization, provides invariant real-time achieving of the control objective with required robustness level of an intelligent control system. Example of efficient simulation of self-organization of robust knowledge bases in an intelligent control system of dynamically unstable essentially nonlinear object in contingencies is given.     

基于动态蚁群劳动分工模型的多AUV任务分配方法

多水下自主航行器(autonomous underwater vehicle,AUV)的动态任务分配问题具有高度非线性、动态不确定性以及多模态的特征,对多AUV任务分配方法的自组织性、鲁棒性以及快速性提出了更高的要求.动态蚁群劳动分工(dynamic ant colony''s labor division,DACLD)模型是一种采用分布式框架的群智能算法,众多行为简单的个体相互作用过程中涌现产生的整体智能行为能很好地适应复杂多变的环境,在解决任务分配问题上具有很好的柔性.引入动态蚁群劳动分工中的刺激-响应原理,建立动态蚁群劳动分工与多AUV任务分配问题之间的映射关系,将任务的状态预测纳入响应阈值,研究基于动态蚁群劳动分工模型的多AUV任务分配方法.同时,针对任务分配过程中可能出现的任务冲突现象,提出新的循环竞争方案以实现最大限度地利用AUV资源.仿真结果表明,所提出的方法能高效地完成任务分配过程,具有很好的自组织性、鲁棒性及快速性.     

CIMOSA: enterprise engineering and integration

Enterprises are rather complex systems which have to be managed for their internal affairs, but more importantly for the many relations to the different environments in which they are operating. Today, these environments are changing much more rapidly and the need for relevant information becomes of paramount importance in the decision making processes at all levels of enterprise management. Fluctuations in market demands, technology evolution and changing regulations require very flexible enterprise operations, capable of reacting to those changes. These reactions must be based on relevant and up-to-date information which must be supported by new decision support technology. The challenges in decision support concern the identification of relevant information, easy access and intelligent use of this information. Building and maintaining the enterprise knowledge base and enabling its efficient use for decision support are major tasks of enterprise engineering. Enterprise integration and its subsequent operation in the global environment of customers, suppliers and regulatory bodies will heavily depend on the availability and the continuous extension of this knowledge base. Enterprise modelling will play an important role in creating the knowledge base and in using it for enterprise integration and operational decision support. The paper discusses enterprise engineering as an enterprise life-cycle oriented discipline for identification, design, and implementation of enterprises and their continuous evolution. Current problems in the field are identified and initiatives are presented.     

智能控制系统仿真技术分析

本文详细地叙述了智能控制系统仿真的实现过程,针对其中涉及的数学原理、数值算法、编程应注意的问题、仿真精度、几种典型干扰的仿真仿真模型、仿真技巧和程序调试等常见问题进行了方法讨论和技术分析,并用Matlab语言给出了一个具体模糊控制系统的仿真例子。     

GENERAL THEORY OF INTELLIGENT SYSTEMS

The basic foundations of the general theory of intelligent systems are constructed in this paper through formalization of the evolution in an open system. Evolution is presented as behavior and self-organization of an open system, and it is assumed that intelligence is given by the laws of evolution. Four laws of self-organization are obtained, depending on the type of open system and its kind of self-organization. The redundancy and entropy of self-organization are obtained as basic evolution characteristics. A pattern recognition and an expert subsystems of an intelligent system are constructed. Discrete and analog intelligent systems have been defined, respectively, as an artificial intelligence system and a system of functional diagnostics.     

Modeling and simulation of supply network evolution based on complex adaptive system and fitness landscape

A supply network (SN) is a complex adaptive system, and its structure and collaboration mechanism evolves over time. However, most literature views SN as a static system and the study on the evolution of SNs is very limited. Based on complex adaptive system and fitness landscape theory, this paper first proposes an evolution model of SNs in order to understand the general principle of SN evolution. Then the paper conducts a multi-agent simulation on the evolution model, and discloses that the SN emerges and evolves from firms’ dynamic interaction under the dynamic environment. Dominated by the environment and firms’ internal mechanism, the evolution is highly sensitive to the initial condition, and it is path-dependent and difficult to predict precisely. Although the dynamics of environments is different, a SN enjoys the stable structure in different environments. Higher level of structure stability and fitness of the SN are achieved when the firms in the SN adopt the long-term collaboration strategy rather than the short-term strategy. Finally, a China case is explored which validates the self-organization evolution of SNs.     

基于鸽群的鲁棒强化学习算法

Reinforcement learning(RL) is an artificial intelligence algorithm with the advantages of clear calculation logic and easy expansion of the model. Through interacting with the environment and maximizing value functions on the premise of obtaining little or no prior information, RL can optimize the performance of strategies and effectively reduce the complexity caused by physical models . The RL algorithm based on strategy gradient has been successfully applied in many fields such as intelligent image recognition, robot control and path planning for automatic driving. However, the highly sampling-dependent characteristics of RL determine that the training process needs a large number of samples to converge, and the accuracy of decision making is easily affected by slight interference that does not match with the simulation environment. Especially when RL is applied to the control field, it is difficult to prove the stability of the algorithm because the convergence of the algorithm cannot be guaranteed. Considering that swarm intelligence algorithm can solve complex problems through group cooperation and has the characteristics of self-organization and strong stability, it is an effective way to be used for improving the stability of RL model. The pigeon-inspired optimization algorithm in swarm intelligence was combined to improve RL based on strategy gradient. A RL algorithm based on pigeon-inspired optimization was proposed to solve the strategy gradient in order to maximize long-term future rewards. Adaptive function of pigeon-inspired optimization algorithm and RL were combined to estimate the advantages and disadvantages of strategies, avoid solving into an infinite loop, and improve the stability of the algorithm. A nonlinear two-wheel inverted pendulum robot control system was selected for simulation verification. The simulation results show that the RL algorithm based on pigeon-inspired optimization can improve the robustness of the system, reduce the computational cost, and reduce the algorithm’s dependence on the sample database. © 2022, Beijing Xintong Media Co., Ltd.. All rights reserved.     

求解模糊相关机会规划模型的混合差分进化算法

针对不确定规划领域中存在的模糊相关机会规划模型,基于群体智能的差分进化算法,设计一种新的求解模糊相关机会规划模型的混合智能算法．该算法基于粒子群优化算法对差分进化算法进行改进,并运用模糊模拟技术对模糊相关机会规划模型进行分析和数值求解,无需像传统的基于遗传算法的混合智能算法需要很长时间并经过复杂的计算才能得到合理的结果．最后,通过实例表明了所提混合智能算法的合理性和有效性．     

Cooperative exploration of distributed worlds in WAVE

This paper modifies the WAVE model for parallel processing in virtual networks to explore and process continuous physical worlds. Expressing distributed activity by cooperative jobs spreading in space and “seeing” each other, rather than as vehicles exchanging messages, the model allows complex missions to be planned in a very flexible manner, with mobile hardware being assigned to the evolving space-conquering programs (waves) dynamically, when required or available. A number of cooperative scenarios in a physical world demonstrate the simplicity and compactness of the wave code. The execution of waves by mobile hardware is discussed, including run-time mapping of waves to vehicles, and supporting multiple distributed jobs in cases of hardware shortages. WAVE can be used efficiently for solving complex problems in space by organized groups of cheap specialized mobile robots, where intelligent behavior is provided by very high level of system organization rather than by the smartness of individual units. In a broader sense, it may also serve as a basic technology for parallel and distributed simulation, and the management of evolution and self-organization of large open systems of different natures.