【作者】 张乐；

【导师】 季天晶；

【作者基本信息】 哈尔滨工业大学 ， 流体机械及工程， 2011， 硕士

【摘要】 目前,随着世界各国对石油需求不断增长,使得油田油井深度不断加深而增加产量。为了保持地层能量,各大油田普遍采用高压注水的方法。近年来,高压欠注问题日益突出,但是油田应用的高压注水泵无论是在工艺结构上还是对复杂工况的适应性上已经越来越难以满足实际要求。针对以上情况,本文首次提出一种传动方式新颖、受力结构简单、可输出高压的新型往复式高压注水泵以及与之相匹配的精确变量装置。这种新型泵不仅结构简单,制造容易,安装工艺上要优于油田目前常用的三缸往复泵,而且效率、可靠性高,同时能够满不同地质条件下油井的高压注水需要并具有很强适应性。本文首先根据油田给定的目标参数进行进行了运动学分析和动力学分析,研究了各关键零部件运动规律及受力,利用VB6.0编制了的受力分析程序,计算出了注水泵的危险工况,并以此作为了泵体结构参数设计的依据;然后利用UG5.0绘制了泵体结构的三维模型,检验了所有零部件之间无干涉,验证了设计的结构的可行性。其次,设计了与新型泵相匹配的变量机构,使之成为由伺服阀和液压缸构成的电液伺服控制系统,能够实现在油田注水控制中心对注水泵进行远程自动控制。同时建立控制系统的仿真模型,并在simulink中对模型的稳定性等指标进行了仿真分析。最后,设计了BP神经网络PID控制器和常规PID控制器,对控制系统进行了校正,并在不同负载的条件下进行了仿真分析、对比,可知相对于常规PID控制器,BP神经网络PID控制器对于新型往复泵具有更强的有效性、实用性。从目前油田高压注水应用前景来看,新型往复式高压注水泵完全可以取代现有高压注水泵。更多还原

【Abstract】 With the continuous improvement of the oil production, the oil well depth deepened ceaselessly and working pressure keeps increasing. In order to maintain and ensure the crude output and the formation energy, each big oilfield widely use the method of high pressure water injection. Recently years, the problem of the injection shortcoming of High-pressure is getting worse, so the normally used High-pressure injection pump is no longer suitable for actual engineering demands.In view of the above situation , this article put forward a new kind of High-pressure injection pump which is more suitable for Engineering actual demand and a Variable device to match the pump.First of all, according to objective parameter and the body structure of the pump, we make dynamics analysis to ensure the force of key parts. Then we calculate the ultimate working condition of the pump with Program Composition of VB6.0, and use the results as the basis of parameter design. At last , use UG5.0 to draw the 3D model of the pump and make sure that the structure is correct.Design the Variable device which is made of servo valve and hydraulic cylinder, so that the pump can be controlled in waterinjection center. Make the nonlinear model of theVariable device, and analyse the model’s stability and its response of typical signals in simulink.Design the BP neural network PID controller and the conventional PID controller to correct the control system, and analyse it in condition of non-load and 5000N load. Compare the results of the two controller, we find that the BP neural network PID controller is more suitable for actual demands.更多还原