Mechanism Analysis Of Dredging And Mud Water Transportation Based On Artificial Intelligence

The suction dredger is the main dredging vessel in the dredging industry.Under the background of big data and intelligent era,the emergence of various emerging technologies,dredgers have been moving towards intelligent development.The traditional construction of suction dredgers relies mainly on the experience of the operators.Due to the uneven experience of the operators,the actual dredging efficiency is not high.The intelligent dredging system uses mechanism analysis to construct mathematical models,collects and analyzes construction data,predicts dredging yield,conducts independent learning,and optimizes construction strategies.Therefore,it is of great significance to carry out intelligent dredging research to improve the dredging efficiency of dredgers.This thesis relies on the foundation project of CCCC National Engineering Research Center of Dredging Technology and Equipment.This paper relies is aimed at modeling the mining and transportation modeling of the suction dredger,the modeling of the mud and water transport mechanism,the sensor data filtering,the missing value processing,Pipeline delay calibration,steam head production forecasting and soil particle size classifiers were studied.The research results provide a solution for the dredging optimization of the suction dredger.main tasks as follows:(1)Research on the mechanism analysis and modeling of taro excavation and mud water transportation.The traditional model of Shantou excavation is analyzed,and the cutting force model and yield model are sorted out.The mud transport process is divided into two models: mud pump and pipeline.(2)Data preprocessing.Aiming at the noise and data loss caused by the sensor's data acquisition,the data filtering algorithm and the missing value processing method are studied to solve the data noise and data loss problems.The sensor for the density and flow of the mud-water mixture has a transmission delay problem at the mud pump and the data from the hoe,and a data delay calculation method is proposed to solve the problem of data delay in the hoe model.(3)Research on production forecasting models.Aiming at the timeliness of the Shantou production data and the traditional dredger production model,the density estimation of the black box model is poor.The extreme learning machine and its two improved algorithms are used to study the incremental extreme learning machine and regularization extreme learning machine,using extreme learning machine to build models,learning network parameters,and predicting the suction density of steamed bread,experiments show that the algorithm can predict Shantou well.Inhalation density,good real-time.(4)Soil particle size classifier.In view of the problem that the sensor does not sense the soil particle size on the suction dredger,combined with the analysis of the construction conditions,the soil particle size parameters of different construction conditions are different,and the principal component analysis method is used to obtain the construction of different working conditions.The parameters have different distributions,and the support vector machine is used to construct the classifier to distinguish the construction data from different working conditions.The test results show that the classifier has a good classification effect.