| The dredging project plays a major role in the construction of the port and waterway,the dredging and remediation of the lake and the maintenance of the Yangtze River channel.The dredging ship as core equipment of the dredging project is the basis for ensuring the efficient,economic and green implementation of the dredging project.The suction system consisting of pump and pipeline is one of the main ways for dredging vessels to realize mud transportation.In order to achieve efficient pump-tube matching,key parameters including pump speed,pipeline flow rate and pressure,slurry concentration and so forth.must be accurately collected.Concentration measurement is the core and difficult.At present,dredging ships use traditional instruments and gamma ray concentration meters for measurement.However,gamma ray concentration meters have problems in human body damage and safety management.The degree of approval for gamma ray of dredging enterprises is not high,and it is hoped that there will be new safe and accurate concentration testing methods.As a non-intrusive detection technology,electrical capacitance tomography(ECT)has the advantages of simple structure,no radiation and it can also reflect the distribution of medium in the pipeline.It has been used in petrochemical pipeline transportation and other fields.Based on the dredging pipeline test platform in laboratory and the ECT technology,this paper carries out the simulation analysis of gas/liquid and solid/liquid flow types.The main research is as follows:(1)Calculate and analyze the ECT capacitive sensor using COMSOL.A commercial code COMSOL was used to carry out ECT capacitive sensor modeling and simulation,and the sensitivity distribution map reflecting the soft field characteristics of ECT capacitive sensor is obtained.(2)Parameter optimization and analysis of ECT capacitive sensor.Combined with the pipeline in laboratory,the ECT capacitive sensor was simulated and designed,and the polar plate coverage,plate length,shielding layer and plate spacing parameters were optimized.At the same time,the influence of variable thickness of the plate and the incorrect install of plate was analyzed.The simulation of the impact on sensor performance provides an auxiliary basis for the manufacture and installation of sensors.(3)Classification of flow patterns and pipeline blockage.The flow regime identification of gas/liquid and solid/liquid two-phase flow is carried out by using Support Vector Machine(SVM)optimized by Genetic Algorithm(GA).Four common flow pattern models of gas/liquid two-phase flow were constructed.The above method was used for flow pattern identification,and the recognition accuracy reached 93.75%.At the same time,the pipeline blockage was divided into “normal”,“blocked” and “severe blockage” according to the amount of sediment deposition.The same method was used to classify the pipeline blockage,and the classification accuracy reached 97.91%. |
