【作者】 裴海灵；

【导师】 周乃君；

【作者基本信息】 中南大学 ， 热能工程， 2009， 博士

【摘要】 电控共轨柴油机作为清洁节能环保型柴油机,在节能和环保意识日益增强的今天,已成为热力发动机的必然发展趋势。柴油机高压共轨喷油控制策略及共轨管的优化研究,对增强共轨系统运行稳定性、提高油量控制精确度、优化发动机整体性能、实现节能减排目标具有重要的现实意义。本文以“新一代轿车用节能环保高效内燃机研发”(国家“863”项目)和“清洁节能乘用车柴油动力系统的关键技术研发与应用”(国家科技支撑计划项目)等科研项目为依托,在广泛查阅相关文献的基础上,对高压共轨喷油控制策略及共轨管的多学科优化设计方面进行了较系统的研究,主要工作包括:(1)通过对现有的多次喷射协调控制进行分析,采用层次化结构分模块设计方法,将多次喷射协调控制分为环境层、限制层和设定层进行设计,基于自适应神经网络技术建立了多次喷射动态组合模型,并分起动、正常和关机三种工况对多次喷射动态组合模型进行了检验。(2)提出了两次喷射间最小时间间隔的定义和计算方法,并基于高压共轨燃油喷射实验台、开放式电控单元及高速摄影技术,研究了两次喷射间时间间隔对燃油喷射雾化状况的影响。(3)提出了基于神经网络的部分微分PID共轨压力控制方法,直接将发动机工况和所处环境状况与PID参数相关联,使共轨压力始终处于与发动机工况相对应的最佳状态,从而实现了对共轨压力的全工况实时动态控制,并有效降低了轨压波动。(4)根据共轨管内压力波动情况,结合喷油器和高压油泵的工作原理,运用上次喷射结束至下次喷射开始间时间间隔和每次喷射电控与液压之间的时间延迟两个物理量,建立了基于前次喷射的喷油量压力波动修正算法,实现了每次喷射油量的平衡与优化。(5)应用多学科设计优化方法,对共轨管多学科特性进行了分析,通过挖掘各种参数与最佳共轨容积间存在的理论关系,对共轨管进行了全局寻优,从而提高了共轨管运行的稳定性和经济性。本文的主要研究结论有:●建立的多次喷射动态组合模型具有较好的适用性,可满足ECU控制的需要。●两次喷射间时间间隔对燃油喷射雾化有重要影响,在本文的实验条件下,喷射频率为20Hz时燃油喷射雾化状况最佳。●提出的基于神经网络部分微分PID的共轨压力控制方法,具有必要的精度;该方法不仅提供了共轨压力控制的新手段,同时也为PID参数整定提供了新的思路,因而具有一定的理论价值。●基于前次喷射的压力波动修正算法,提供了喷油量补偿的一种新方法,对喷油量的调整具有较强的指导作用,可减少由于共轨压力波动导致的油量偏差,为控制参数的优化提供了依据。●采用多学科设计优化方法,使共轨管容积减小了1.07%,压力波动下降了16.7%,总质量减少了6.8%,有利于提高高压共轨燃油喷射系统的整体性能。本文的研究成果已在GS-1000型高压共轨燃油喷射实验台上进行检验。结果表明,所开发的控制策略可以有效降低共轨压力波动和使各次喷射油量得到补偿,可使高压共轨燃油喷射系统的整体功能得到提升。总之,本文的研究为高压共轨燃油喷射系统研发探索了新思路,其成果不但具有一定的理论价值,而且具有较好的应用价值。更多还原

【Abstract】 As a type of clean diesel engine, electronic control common rail diesel engine is one of the best choices in save energy and environments protect conception. The research of fuel injection control strategy and optimization of the rail for High Pressure Common Rail (HPCR) has a very important significance to strengthen the operation stability, to improve the fuel calculation precision, to optimize the engine performance, to save energy and decrease wastage. In this dissertation, the fuel injection control strategy and Multidisciplinary Design Optimization (MDO) for rail had been systematically studied by the support of "Development on New Generation Energy-saving and Environment-protect Efficient Internal Combustion Engine for Car" (the National High Technology Research and Development Program ("863"Program) of China) and "Development and Application on Key Technologies for Clean and Energy-save Diesel Power System Used on Passenger Car"(National Key Technology R&D Program).The main work in this dissertation is as follows:1. By fully demonstrating and comparing the multiply injection coordinate control strategies in existence, the control strategy had been designed as environment layer, limitation layer and setup layer, using hierarchical and module structure. By adapting fuzzy neural network technology, multiply injection dynamic combination model had been built. The model had been tested in starting, normal and after run working conditions.2. Put up the definition and calculate method for internals between two injections. Based on the high pressure common rail test pump bench, open ECU and high speed photography, the effect of fuel spraying and atomizing by internals between two injections had been researched.3. Based on neural network and portion differential PID, a new control strategy for rail pressure had been brought forward. This method associated engine working condition and the PID parameters together, achieved control the rail pressure in all the working conditions, obtained certain precision, and decreased rail pressure oscillation effectively.4. According to the fact of pressure oscillation in rail tube and the working principle for high pressure oil pump and electrical control injectors, with the intervals from the end of last injection to the start of next injection and the hysteresis between electronic control and hydraulic together, the backoff algorithms of pressure oscillation for injection quantity had been put forward. This method can make the injection quantity balancer and more homogenous.5. By MDO method, the multidisciplinary characteristics of rail tube had been analyzed. Optimization design for rail in whole situation had been achieved through winkling the relationship between kinds of parameters and the optimum rail volume. By doing this, the economy and stability for rail had been improved further.There are main conclusions in the dissertation:The applicability of multiply injection dynamic combination model is good enough to satisfy the demand of ECU.The internals between two injections have important effects on fuel spraying and atomizing. When the injection frequency is 20Hz, the fuel spraying and atomizing performance is the best on the condition of the equipments in this dissertation.The pressure control method with neural network and portion differential PID put forward by the author not only provides the new means to control common rail pressure, but also provides new thoughts to set PID parameters, therefore it has higher theoretic values.The backoff algorithms of pressure oscillation for injection quantity based on last injection inaugurate a new field to compensate injection quantity, and can instruct scientifically to adjust the injection quantity and decrease the deviation caused by pressure oscillation so as to optimize the operation parameters.By MDO method, the rail tube volume is decreased by 1.07%, pressure oscillation is decreased by 16.7%, and the mass of rail tube is decreased by 6.8%. The whole performance of HPCR system has been improved largely.The research work developed by author had been successfully used on HPCR test pump bench. The running practice showed the control strategy and optimization rail can decrease the pressure oscillation and compensate the injection quantity effectively, and can make the operation of HPCR stabilization and optimization so that can improve the running efficiency of HPCR.It can be said that the research work offers a new way to upgrade the HPCR. Thus the research results have higher theoretic and utility value, and can been extended to apply in the same HPCR in our country.更多还原