Combustion Control Strategy Of Marine Natural Gas Engine Based On Data Drive

On the basis of stable operation of the engine,working in the best combustion state is an important technical means to further improve the performance of the engine.The existing Marine natural gas engine modeling is difficult,difficult to design the combustion process control strategy based on the model,unable to effectively control the state of combustion in the cylinder.In this paper,combustion characteristic parameter CA50 is selected as the controlled parameter,and a data-driven combustion process control strategy is designed to realize closed-loop combustion control and improve the efficiency of heat and power conversion in the cylinder,so as to improve the power and economy of the engine,alleviate the after-combustion phenomenon and reduce emissions.Based on the idea of modular modeling and control requirements,the engine mathematical model is established according to the structure and working principle of natural gas engine.At the same time,a combustion model was established using two Wiebe functions to characterize the combustion process in the cylinder.In order to verify the accuracy of the model,the simulation data and experimental data were compared.The maximum error of the model was less than 10%,which met the design requirements of the controller.The control strategy is designed according to the engine operating conditions and requirements,which mainly includes the engine state management,starting open-loop strategy and closed-loop control strategy design.Closed-loop control includes speed control,air-fuel ratio control and CA50 closed-loop control.Considering the influence of air-fuel ratio on CA50,the air-fuel ratio control is combined with CA50 closed-loop control,and the CA50closed-loop control is controlled by adjusting the ignition timing and the target value of air-fuel ratio at the same time,and a model-free adaptive control algorithm is selected in the data-driven category.The design idea is to first give the corresponding dynamic linearized data model for the input and output data at each operating point of the combustion closed-loop system,and then design the controller according to this data model.The design process does not depend on the combustion system model.In order to realize the overshoot free control of CA50,a fast convergence term is introduced on the basis of the original model-free adaptive control algorithm,and the convergence analysis is carried out.The final off-line simulation results show that the CA50 closed-loop is more rapid and stable with no overshoot under the improved model-free adaptive control algorithm.In order to further verify the effectiveness and control performance of the controller,a semi-physical simulation platform was built based on NI simulation system and rapid prototyping system.It is compared with the traditional PID control.In the verification process,different speed,different steady-state load,sudden load and other operating conditions are designed.The results show that compared with PID control,the improved model-free adaptive control algorithm has shorter response time and no overshoot under various working conditions,and less fluctuation under steady-state conditions.Moreover,the modified model-free adaptive control algorithm has little influence on CA50 control when the same set of controller parameters are adopted under different working conditions,indicating that the proposed modified model-free adaptive combustion process controller has strong adaptability.