给出了一种新型海水淡化热力学系统的数学模型 ,针对这一类具有不确定因素的复杂系统 ,提出基于模糊集值统计下修改模糊推理模型的关系生成规则方法。在系统运行过程中 ,当外界条件和系统内部结构参数变化时 ,该推理模型具有自适应能力 ,通过输入输出数据对 ,在线修改关系生成规则 ,从而改变系统输出性能。最后 ,经仿真验证 ,该方法对于具有不确定因素的复杂系统 ,可以得到满意的性能指标 A new mathematic model of desalination thermodynamics system is presented. Aiming at this kind of complex system with uncertain factors, a method of generating relation rules based on fuzzy set value statistics is proposed. In the process of system operation, when the external conditions and the internal structural parameters of the system change, the inference model has the ability of self-adaptation. The rules of generating the relationships are modified online through the input and output data pairs to change the output performance of the system. Finally, the simulation results show that this method can get satisfactory performance for complex systems with uncertainties