冷凝器污垢清洗的智能测量与控制方法研究

《冷凝器污垢清洗的智能测量与控制方法研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、湖南大学博士学位论文冷凝器污垢清洗的智能测量与控制方法研究姓名：樊绍胜申请学位级别：博士专业：控制理论与控制工程指导教师：王耀南20060519冷凝器污垢清洗的智能测量与控制方法研究神经网络预测控制算法控制清洗剂投入量，实现复杂清洗过程中最优清洗液浓度的动态稳定；根据清洗控制的效果实现清洗液浓度的滚动优化，使得既能快速除垢，又能使换热管的腐蚀降至最小。试验结果证明了该方法的有效性。本文的研究成果将为冷凝器的合理、高效运行提供理论基础，对促进电力、化工、机械等行业节能降耗，提高经济效益具有重要意义。本文的研究成果同样可适用于其他行业的大型换热设备，具有良好的应用前景。关键词：冷凝器污

2、垢清洗智能测量智能控制与优化IIAbstractForthepurposeofimprovingtheheattransferperformanceofcondenserandincreasingtheefficiencyofthermalcycleofturbo—generatorunit，thisdissertationinvestigatestheapproachesofintelligentmeasurementandcontrolforfoulingcleaningincondenser．Themaintopicsofthedissertationaredescribed

3、asfollows．Firstly，thedissertationexplainsthenecessityoffoulingmonitoringandfoulingcleaningbyinvestigatingthecauseoffoulingaccumulationandtheinfluenceonheattransferperformanceofcondenser，andanalyzestheadvantageanddisadvantageoftheordinaryfoulingmeasurementandcleaningmethods．Then，basicideaandmai

4、nmethodsofintelligenttechnologyareintroducedandaconclusionismade，whichdemonstratestheapplicationofintelligenttechnologyinfoulingmeasurementandcontrolisaneffectivewayforimprovingtheheattransferperformanceofcondenser．Anovelstrategywhichcombineshighpressurewatercleaningwithchemicalcleaningispropo

5、sedinthisdissertation．Basedonthestrategy，acleaningsystemisdevelopedandthedetailofthecontrolsystemisdescribed．Anovelapproachbasedonfeatureselectionandintelligentmodelingforprecisemeasurementoffoulingisproposed．Intheapproach，sensitivitycalculationbasedonBPneuralnetworkiSintroducedtodeterminethem

6、ostsuitablefeaturevariablewhichreflectsthefoulingstate．T-Smodel，multi-RBFNeuralnetworkanddiagonalrecurrentneuralnetworkareemployedtoseparatetheinfluencesofbothfoulingandotherfactors，suchasoff-designconditiononfeaturevariable．Basedonit，anexperimentalsystemisdevelopedandexperimentonallactualcond

7、enseriscarriedout．Theresultsshowtheapproachmeasuresthefoulingcorrectly，andismoreeffectivethanthermalresistancemethodorheattransfertoe侬cientmethodundertheconditionofblockedtubesorexcessiveamountofairincondenser．Thepredictionoffoulinginco