【作者】 张航；

【导师】 余世明；

【作者基本信息】 浙江工业大学 ， 控制科学与工程， 2019， 硕士

【摘要】 各类流程工业过程的高性能自动化控制是工控领域的研究热点。实际的生产过程中常常涉及到流体物料的混合加工,基于不同生产工艺的要求,工业隔膜计量泵逐渐成为重要的生产执行器,主要完成流体的定量运输和比例投加。为了提升计量泵的工作性能,近年来的文献主要关注优化计量泵的机械结构,很少有人考虑将性能更好的控制算法带到泵机控制领域,而事实上,设计合适的控制算法以及控制方案可以大大节约生产成本,提高经济效益。在此背景下,本文以工业隔膜计量泵为研究对象,考虑实际生产中流量远程精确控制的要求,依据模型控制论的思想,设计了一种流量优化控制方案,搭建了一套以计量泵为核心的流体投加系统。本文主要研究工作如下:(1)基于计量泵的内部机械结构分析了转速调节流量的原理,考虑泵机流量和转速之间存在非稳态的动态过渡关系,设计了相应的辨识实验获取该动态模型,包括辨识信号的设计,选取小波分析结合阈值处理的算法对实测流量数据进行预处理,以系统的先验参数模型为基础,分别应用基于神经网络和改进差分进化算法对系统进行辨识,通过仿真实验证明了算法的有效性,实际数据和模型数据的特性拟合验证了模型的可靠性。(2)以辨识的计量泵动态模型作为依据,设计了系统上位机的流量在线优化调节算法,对传统的广义预测控制算法进行改进,引入差分进化算法对实际系统的约束条件进行处理,算法能完成系统最优转速控制率的在线求解。考虑计量泵实际工作过程中受到脉动冲击负载的情况,分析了可能影响本地电机转速的原因,设计了基于模糊控制的转速补偿控制器,减轻了由于转速控制不精确带来的流量控制误差,结合上位机算法的最优转速求解共同完成计量泵流量的在线优化调节。(3)以实验室搭建的流体投加平台为基础,对本文提出的流量控制方案进行系统实测,上位机流量优化算法通过JAVA语言编译集成在自主开发的系统监控软件中,转速补偿算法通过C语言编译整合在本地的计量泵变频控制器中,通过网关控制网络实现控制信息的远程交互,模拟不同工况对本文提出方案进行性能评估,可以满足一般流程工业中的工艺需求。更多还原

【Abstract】 High-performance automation control of various process industry is a research hotspot in the field of industrial control.In the actual production process,the mixing process of fluid materials is often involved.Based on the requirements of different production processes,industrial diaphragm metering pump has gradually become an important production actuator,mainly for quantitative transportation and proportional dosing of fluids.In order to improve the performance of the metering pump,the literature in recent years has focused more on optimizing the mechanical structure of the metering pump.Few people have considered exerting better control algorithms in the field of pump control.In fact,an appropriate control algorithms and control scheme can greatly save production costs and improve economic efficiency.In this context,this paper takes the industrial diaphragm metering pump as the research object,considering the requirements of remote precise control of flow in actual production,designs a flow optimization control scheme based on the idea of model control theory,and builds a remote fluid dosing system based on the metering pump.The main research work of this paper is as follows:(1)Based on the internal mechanical structure of the metering pump,the principle of the flow regulation is analyzed.Considering the unsteady dynamic transition relationship between the pump flow and the motor speed,the corresponding identification experiment is designed to obtain the dynamic model,including the design of identification signal,a data preprocessing method based on wavelet analysis combined with threshold is involved to preprocess the measured flow data.Based on the prior parameter model of the system,the system is identified based on neural network and improved differential evolution algorithm.The simulation results prove the superiority of the algorithm,actual data and the model output signal fits perfectly,which verifies the reliability of the model.(2)According to the identified dynamic model of metering pump,an online flow regulation optimization algorithm is designed.Based on traditional generalized predictive control algorithm,the differential evolution algorithm is introduced to deal with the constraints of the actual system.Optimal speed is calculated online.Considering the pulsating impact load during the actual working process of the metering pump,the reasons that may affect the local motor speed are analyzed.The speed compensation controller based on fuzzy control is designed to reduce the flow error caused by the inaccurate speed.Combining the optimal speed to realize the online optimization of the metering pump flow.(3)The flow control scheme proposed in this paper is tested in the remote fluid dosing platform built by the laboratory.The upper computer flow optimization algorithm is integrated into the self-developed monitoring software via JAVA,and the speed compensation algorithm is compiled by C in the local metering pump variable frequency controller.The Gateway can realize the remote interaction of control information.The performance of the proposed scheme is evaluated by simulating the different working conditions,which can meet the requirements of the general process industry.更多还原