基于模糊专家规则的离散事件辨识及其在交通诱导中的应用

Identification of Discrete Events Based on Fuzzy Expert Rules and Application in Traffic Induction

传统离散事件动态系统中的离散事件状态的转换具有不确定性,其不确定性主要来自状态转向发生时刻的不确定性,而所转向的状态一般具有确定性。本文对具有状态转向时刻和转向状态的二重不确定性的离散事件动态系统进行了讨论,用模糊专家系统来对未来状态和状态迁移时刻二重不确定性的离散事件动态系统进行评价,并以智能交通系统中车辆诱导技术为例,说明了此类模糊专家规则的应用价值。本文提出了一种模拟淬火算法,通过模拟物质加温后急剧冷却的过程来求得目标函数的局部极值,以模拟淬火算法的转向概率作为模糊专家系统中的规则选择概率。该方法有效地保证了事件转换的实时性,提高了交通疏导的效率。

State transition in traditional discrete event dynamic systems （DEDS） features uncertainty. In general, the transition time is uncertain and the corresponding state is determined. The discrete event dynamic system in which the states and state transition are all undetermined is discussed in this paper. A fuzzy expert system is proposed to evaluate the system. Taking the traffic-induction technology in intelligent transportation systems as an example, the application value of the fuzzy expert rules is illustrated. A simulated quenching algorithm is proposed in the paper. The local extreme value of the object function can be gained by increasing the object temperature and cooling sharply. The state transition probability in the simulated quenching algorithm can be seen as the rule selection probability in fuzzy expert system. The state transition is real-time and increases the efficiency of traffic induction.