Research On Temperature Prediction And Optimal Layout Of Cable Cluster Based On Multi-physical Field Coupling

With the rapid development of electric power network,in order to solve the conflict between urban land shortage and urban overhead cables,the laying methods of tunnel cable cluster are becoming more common.As a result,the problem of load distribution of power cables leads to cable life reduction and cable resource waste.However,temperature is an important parameter to reflect the normal operation of the power cable,and the calculation of the cable load is also an important basis for cable layout and reasonable load distribution.Therefore,these two characteristic quantities are of great significance to the layout and reasonable load distribution of the actual project.This thesis introduces the basic structure of cable,the solution method of the steady-state cable carrying capacity and the research status of the cable temperature,and uses COMSOL finite element software to establish the three-dimensional electromagnetic-thermal-current coupling steady state and transient models of the tunnel cable cluster.Compared with the two-dimensional model,the built model is more suitable for the actual tunnel laying.The accuracy of the model is verified by analytical and experimental methods,and the layout is optimized according to the tunnel cable cluster carrying capacity and temperature as scalar,and the cable temperature and carrying capacity are predicted by combining with MATLAB.The main research is as follows.(1)the three-dimensional electromagnetic,thermal and flow multi-physical coupling steady state and transient models of the cable is established to calculate the tunnel cable cluster.The calculated results are compared with IEC standard and experimental data respectively.The results show that the transient and steady state results of the cable cluster meet the error requirements,so as to verify the accuracy of the model.(2)Based on the cable transient model,the temperature simulation of tunnel cable cluster is carried out.Based on the model data and CNN-LSTM algorithm,MATLAB is used to predict the temperature of cable core on the fifth day.The results show that the maximum error of the temperature of the A-phase cable core is 0.3℃,1.2℃ for the B-phase,0.2℃ for the C-phase.all of these meet the error requirements,which proves that the CNN-LSTM method based on three-dimensional finite element method is feasible.However,the error of the B-phase is larger than that of A-phase and C-phase.So the attention mechanism is added to optimize the performance of the cable core temperature prediction.The temperature distribution of the cable is calculated according to the transient model of the electromagnetic,thermal and current coupling three-dimensional cable,and the maximum predicted carrying capacity is calculated by the empirical formula.(3)Based on the established three-dimensional cable steady-state model,the layout of the 6-loop cable is optimized,and six representative combination modes are obtained,and the distribution of its carrying capacity and temperature is analyzed.It provides an effective way to realize the transformation from traditional preventive maintenance to proactive predictive management and to improve the load capacity and cable utilization rate of tunnel cable cluster.