基于fnn的故障诊断报警系统大学论文.doc

硕 士 学 位 论 文 ault Diagnosis Alarm System about Combine Harvester based on FNN algorithm  摘 要 联合收割机作为现代农业必备的设备之一，在农业的机械化生产中起着非常重要的作用，它极大地改善了农民的劳动条件，提高了农产品收获效率。但是农业机械往往都是在高速、高负荷、高压、高温同时伴随着较大的风沙的环境下工作，这使得联合收割机极易发生故障。造成故障的原因有很多，比如湿密的收获物、多余的杂草、过快的速度、过小的脱粒间隙、过低的脱粒速度、欠载的发动机马力或者不均匀的喂入量等因素都有可能引起联合收割机发生故障，从而影响收获质量。，目前许多研究学者针对联合收割机的故障诊断方面做了许多的研究工作，虽然取得了一定的研究成果，但是不能实系统输入与系统输出之间的线性映射关系，所以效果不尽如人意。本文的目的就是设计出一套联合收割机故障诊断报警系统，将FNN（Fuzzy Neural Network，模糊神经网络）算法引入到故障诊断报警系统中，很好的解决了多输入系统训练难、非线性差的问题。 本论文针对联合收割机故障输入非线性性差等问题，采用“太湖之星”TH988联合收割机为试验样机，利用传感器采集割台搅龙 、输粮搅龙、输送槽、切流滚筒、纵轴流滚筒信号作为系统的输入，选择可靠性高EPEC 2023 PLC和高亮度的HD064MV3 彩色按键高性能显示器为硬件平台，设计了联合收割机故障诊断报警系统。将FNN引入联合收割机故障诊断报警系统中来，解决了系统输入非线性、时变和大延时的缺点。本系统的主要工作过程为：PLC通过上述FNN算法对传感器采集到的信号进行分析处理之后，将得到的故障诊断结果通过CAN总线传输给上位机进行显示，并及时的进行声光预警报警。当有故障发生时，步进电机开始自动控制前进速度的变化，将故障危害降低到最低，提高收割机的作业质量和工作效率。wet Crops, overmuch weeds, excessively speed, small Threshing gap, low roller rotation speed, drive belt skid, engine horsepower insufficiency, walkers wheel and walkers slip and uneven feed etc.At present, many research scholars had made a lot of research work on the fault diagnosis of combine harvester.However,it is not satisfactory for the bad ability of linear mapping relationship between input and output.The purpose of this paper is to design a fault diagnosis alarm system about combine harvester.Moreover,the FNN (Fuzzy Neural Network)algorithm is introduced to figure out the nonlinear problem. This thesis is supposed to improve the overmuch faults and fault nonlinearity input by the regard the Star of Taihu Lake TH988 combine harvester as the test prototype. The fault diagnosis alarm system of combine promoted in this paper is consist of signal input system, EPEC 2023 PLC and HD064MV3 color high-performance monitors. The inputs system includes several parts such as cutting table auger, lost grain auger, conveyor trough, cut the flow drum, drum longitudinal axis of the stream and forward speed.Since FNN algorithm has a powerful