| 基于模糊Markov博弈控制求解伙伴选择问题 |
| |
| 引用本文： | 王惠,符策,陈燕. 基于模糊Markov博弈控制求解伙伴选择问题[J]. 辽宁工程技术大学学报(自然科学版), 2006, 0(Z2) |
| |
| 作者姓名： | 王惠 符策 陈燕 |
| |
| 作者单位： | 大连海事大学经济与管理学院，大连海事大学经济与管理学院，大连海事大学经济与管理学院 辽宁大连 116026，辽宁大连 116026，辽宁大连 116026 |
| |
| 基金项目： | 国家自然科学基金资助项目(70540005) |
| |
| 摘 要： | 针对不确定条件下的伙伴选择决策问题,引入基于ANFIS和神经网络的模糊神经网络,实现了一种全新的进行值函数逼近的梯度下降Q学习的算法,对多影响因素进行FNN学习,将输出量作为标准Markov博弈模型的输入量,得到影响的策略,最后研究了一个应用实例,利用具体历史数据对建模方法和模型进行了验证和分析。
|
| 关 键 词： | 伙伴选择 多智能体 自适应模糊控制系统 神经网络 Markov博弈 Q学习 |
Solving partner selection problem based on fuzzy Markov game controller |
| |
| WANG Hui,FU Ce,CHEN Yan. Solving partner selection problem based on fuzzy Markov game controller[J]. Journal of Liaoning Technical University (Natural Science Edition), 2006, 0(Z2) |
| |
| Authors: | WANG Hui FU Ce CHEN Yan |
| |
| Abstract: | According to partner selection under uncertain conditions, a multi-agent fuzzy Markov game controller is proposed based on adaptive neuron-fuzzy inference system (ANFIS), neural network and Markov game. In this model, FNN is used to train the factors which influence the partner selection and the results of FNN are taken as the input for the standard Markov game while the finial policy is taken as the output. A case is studied and the simulation model is validated by historical data. |
| |
| Keywords: | partner selection multi-agent anfis neural network markov game q learning |
| 本文献已被 CNKI 等数据库收录！ |
|
