柴油机高压共轨系统轨压控制仿真及实验研究-机械电子工程专业论文.docxVIP

摘要 摘要 万方数据 万方数据 摘要 柴油机是现代应用最为广泛的原动机之一，随着 21 世纪全球燃油与环保问题日 趋严重，人们在油耗与减排方面对柴油机设计提出了更高的要求。高压共轨燃油喷射 技术能明显提高柴油机工作效率，目前已成为改善整机性能的重要途径。 本文通过研究高压共轨燃油喷射系统的开发历程，深入剖析 CR 系统结构及工作 原理，在全面了解国内外同类研究的基础上，以 BOCSH 公司高压共轨燃油喷射系统 产品为研究对象，利用 AMEsim 软件建立了 CR 系统的仿真模型；将 BP 神经网络控 制与传统 PID 控制相结合，提出了 BP 神经网络整定 PID 控制算法对轨压进行闭环控 制，在 MATLAB 环境下编写 BP 神经网络整定 PID 函数；利用 AMEsim 软件与 MATLAB 的接口，建立 CR 系统轨压控制的联合仿真环境；对 CR 仿真模型进行轨压 控制联合仿真，采集和处理相关参数，完善控制策略。 对联合仿真结果的分析表明，本研究设计的 BP 神经网络整定 PID 控制方式与模 糊 PID 控制相比，有更好的鲁棒性和动态响应特性，明显提高了系统轨压的控制效果 。 关键词：柴油机；共轨系统；神经网络；轨压控制；联合仿真 I II II 万方数据 Abstract Diesel is one of the most widely used prime motors in modern society. With the 21st century fuel and environmental problems becoming more serious, people are setting higher requirements for the design of diesel engine in terms of fuel consumption and emission reduction. The technique of high pressure common rail fuel injection can improve the efficiency of diesel engine significantly, it has, therefore, become an important way to improve the machine performance. This paper studies the developing process of high pressure common-rail fuel injection system, in-depth analysis of CR system structure and working principle. Based on similar studies at home and abroad, it takes the high pressure common-rail fuel injection system products developed by BOCSH as its research object and employes AMEsim software to establish a simulation model of CR system. The neural network control combined with the traditional PID control is proposed that to employ neural network adjusting PID control algorithm for close-loop control rail pressure, to write functions of neural network adjusting PID in MATLAB environment; the use of the interface between AMEism soft ware and MATLAB established an co-simulation environment of CR system rail pressure control to conduct an co-simulation of rail pressure controlling over CR simulation model, then collect and process parameters and improve controlling strategies. Analyses on the results of the co-simulat