一种新的模糊Petri网推理机制

摘要

针对模糊Petri网(FPN)建立过程中模糊产生式规则各项参数的确定问题,通过引入一种新的FPN推理机制,利用虚库所和虚变迁构建分层FPN模型.该方法的实现不依赖经验数据,对初始输入无严格要求.仿真实例结果表明,利用该推理机制对非训练样本中的输入数据进行模糊推理,所得的FPN模型具有较强的泛化和自适应能力.%Aiming at the problem of determining all parameters of fuzzy production rules in building a Fuzzy Petri Net(FPN), by introducing a new FPN reasoning mechanism, this paper uses virtual places and virtual transitions to construct layered FPN model. Its realization does not depend on experiential data, and the requirements for primary input are not critical. Simulation experimental result shows that for the input data that do not include training samples, the reasoning mechanism possesses strong generalizing capability and self-adjustion.