On-line Monitoring System For Scale Accumulation Of Boiler Pipelines In Thermal Power Plants

With the increase of the installed capacity of thermal power plants,the problem of scale peeling caused by the corrosion of high-temperature heating surface pipelines increasingly threatens the safe production of power plants.The subject is based on the problem of scale delamination and accumulation in boiler systems in thermal power plants.Combined with the high-temperature and high-pressure steam scale separation technology,the scale particle magnetic detection system designed under the new scale separation and accumulation scenario can complete the oxidation in the boiler system in the operating state.Skin particle online monitoring.First,the on-line separation and sampling of scale particles in the high-temperature and high-pressure steam pipeline were realized,and how to condense the high-temperature and high-pressure steam mixed with scale particles was analyzed,and the large-scale scale particles were separated and deposited.The success of scale particles Separation is a prerequisite for online monitoring of scale particles.Secondly,according to the magnetic characteristics difference between the scale particles and the deposition pipeline,a magn etic induction intensity hardware detection device is designed,and the design of the multi-point acquisition system is completed according to the requirements of the magnetic induction intensity information vector characteristics.Experiments show that the designed multi-channel magnetic sensitive matrix detection system can Complete the sensing of magnetic induction intensity information around the deposition pipeline.Then,combining deep self-encoding and shrinking self-encoding,a deep shrinking self-encoding(SCAE)data dimensionality reduction algorithm is proposed to extract features from multi-channel data information.In order to verify the dimensionality reduction effect of deep shrinkage self-encoding,the support vector regression machine(SVR)was used to predict the scale accumulation height,and the parameters of the deep-shrinkage self-encoding network were selected after comparing several groups of experiments.Comparing the model of the scale accumulation predicted by deep shrinkage self-encoding dimensionality reduction data and non-dimensionality reduction magnetic induction intensity gradient data training,it is found that deep shrinkage self-encoding dimensionality reduction data can improve the prediction accuracy of support vector regression machine.In this paper,through the research on the separation and sampling of scale particles in high temperature and high pressure steam,a set of scale particles accumulation monitoring system is designed,which can realize the online monitoring of scale particles accumulation in the non-stop state of the boiler.The system has high detection accuracy and has been applied to the actual site of the power plant.It has an important reference value for the evaluation of the generation of scale particles in the boiler system of the power plant.