船用柴油机燃烧闭环控制策略仿真与试验

Simulation and Experimental Investigation of Closed-Loop Combustion Control Strategy for Marine Diesel Engines

以传统柴油机MAP图为数据集,基于人工神经网络建立面向控制的柴油机仿真模型;针对各缸平均指示压力(IMEP)循环间随机波动及其缸间不均匀性的特性,提出基于反馈型卡尔曼滤波算法的前馈与燃烧闭环协同控制策略.结合柴油机神经网络仿真模型,设计了燃烧闭环控制策略仿真架构;在此基础上,通过仿真分析表明所提出的协同控制策略相对于开环控制和常规燃烧闭环控制策略,在稳态和瞬态工况下均表现出了良好性能.最后,在柴油机试验台架上对所提出的协同控制策略进行了配机试验.结果表明:在稳态工况下,所提出的协同控制策略能够明显改善各缸做功均匀性,相比开环控制策略,各缸间IMEP变异系数从5.20%降至0.09%,相对降低了98.27%;在瞬态工况下,响应时间降低至1个工作循环.

The MAP of a traditional diesel engine was used as a data set,and a simulation model of the diesel engine was established based on a neural network.To solve the problem of cycle-to-cycle variations and cylinder-tocylinder inhomogeneity of indicated mean effective pressure(IMEP),a cooperative control strategy,which consists of the feedforward control and the closed-loop combustion control,was proposed based on a modified Kalman filter with a feedback loop.A simulation construction of a closed-loop combustion control strategy was designed in combination with the simulation model.Then,simulation analysis was conducted and indicated that both the steady-state and transient control performance under the proposed cooperative control strategy are better than that of the open-loop control and conventional closed-loop combustion control strategies.Finally,the cooperative control strategy was experimentally validated on a commercial diesel engine test bench.The experimental results show that the cooperative control strategy can effectively reduce cylinder-to-cylinder IMEP variations under steady operations.Compared with the open-loop control,the coefficient of variation of the IMEP decreases from 5.20%to 0.09%,reduced by 98.27%,and the response time is decreased to one engine cycle under transient conditions.