基于BPPID的疏浚管道稳定流速控制

BPPID-based steady flow control for dredged pipelines

摘要:

针对疏浚船管道输送过程中泥浆管道流速难以控制,容易造成较大的功耗、磨损甚至会堵管、爆管等风险,本文以疏浚泥泵管道输送实验台为研究对象,提出一种基于BP神经网络的PID控制器,对管道输送流速进行稳定控制。在采用系统辨识方法对实验台进行建模的基础上,将BPPID与传统PID控制器进行仿真对比分析并利用模型实验台分别进行了流速阶跃变化和流速跟踪实验。实验结果表明BPPID控制对突变工况具有自适应自学习能力,为实际挖泥船输泥管道的稳定流速提供借鉴。

Abstract:

The paper proposes a PID controller based on BP neural network for stable control of the pipeline conveying flow rate, taking the dredging mud pump pipeline conveying experimental bench as the research object. Based on the modeling of the experimental bench using the system identification method, the BPPID is compared with the conventional PID controller and the flow rate step change and flow rate tracking experiments are carried out using the model experimental bench. The experimental results show that the BPPID control has the ability of adaptive self-learning for abrupt changes in operating conditions, which can be used as a reference for the stabilisation of the flow rate in a real dredger mud transfer pipeline .