氢发动机排气污染及NO_x排放优化控制

RESEARCH ON EXHAUST CHARACTERISTICS OF HYDROGEN-FUELED ENGINE AND OPTIMIZING CONTROL OF NO_x

研究了在汽油中加入氢气后混合燃料发动机的排放特性,给出了氢汽油混合燃料可以全面改善汽油机的废气排放,然后进行了高压喷射型氢发动机的两个主要运转参数(点火提前角和喷氢提前角)对NOx排放影响的研究。试验表明点火提前角、喷氢提前角对NOx排放量有很大影响。在建立以点火提前角、喷氢提前角、喷氢量为控制变量,以动力性或经济性为性能指标泛函,且排放不超标等为约束条件的氢发动机最优控制模型的基础上,提出了分别以径向基(RBF)网络、模糊神经网络(FNN)求解最优控制模型的新方法,进行了仿真计算和试验数据的对比研究。研究结果给出了两种网络均可以成功取代传统MAP而满足要求。其中以模糊神经网络所用时间较短。

The Experimental research on exhaust characteristics of a hydrogen-gasoline dual fuel engine was carried out; The experimental results show that the dual fuel engine can improve the exhaust characteristics of gasoline engines. Then, the research on exhaust characteristics of hydrogen-fueled engine under high-pressure injection with different ignition timing and injection timing was also implemented. The results show that both ignition timing and injection timing have great effect on NOx exhaust characteristics. Moreover, a optimizing control model and control system were set up, in which ignition timing, injection timing and injection duration were separately chosen as control variation, and output power and rate of fuel consumption （enacting NOx exhaust performance and so on was as confine condition） were separately selected as performance index function. A radial basis function（RBF） and improved fuzzy-neural network （FNN） system were separately adopted to calculate the optimizing control laws. Finally, the calculating results were compared with experimental results. The contrasting results show that both of the networks can be satisfactorily used in replacing traditional control method with MAP on meeting the requirements of control precision, and the calculation time with FNN system is less under the condition of same calculation precision.