Research And Realization Of Temperature Estimation Algorithm For High Voltage Cable Joint

Cable joints may be aging due to poor splicing process or long-term effects of electrical,thermal,chemical and other factors,which seriously affect the safety of power supply.The cable is a multi-medium heat transfer model,the cable core will produce a lot of heat during the work,the heat through the multi-layer medium non-uniform transfer to the outside,if the heat generated during the core work can not be exported in time will cause a fire,so the cable joint temperature monitoring is an important means to ensure the safety of power supply.It is a direct and effective method to monitor the temperature of cable by using the directly buried sensor,but the sensor is difficult to install and replace,and the interference is difficult to avoid.In this paper,the core temperature of the cable joint is obtained indirectly through the measurement of the cable external parameters and the estimation method.In order to study the estimation models,the main temperature estimation models were compared from the complex mapping ability,fault tolerance ability,calculation accuracy,calculation speed and extrapolation ability through comparative experiments,and the cable joint temperature estimation model based on the generalized neural network(GRNN)was determined.In order to overcome the slow speed of GRNN in the calculation process,the fruit fly optimization algorithm(FOA)was used to find the smooth value(SPREAD).However,the FOA algorithm would reduce the calculation accuracy of GRNN.Aiming at the problems such as poor computational accuracy,low robustness and slow iteration speed of drosophila optimization algorithm,this paper studied a drosophila optimization algorithm(GCFOA)based on the theory of co-evolution in biology.In order to verify the superiority of the new algorithm,a group of comparative experiments were set up in this paper.The GCFOA algorithm,FOA algorithm,particle swarm optimization(PSO),Bat algorithm(BA),adaptive fruit fly optimization algorithm(IFOA)and three-dimensional fruit fly optimization algorithm(WFOA)are compared.The experiments show that the GCFOA algorithm has obvious advantages in iteration speed,computational accuracy and robustness.Combining the GCFOA algorithm with the generalized neural network,the cable joint temperature estimation algorithm is obtained.Finally,this paper realizes the cable joint temperature estimation algorithm through the hardware system FPGA,and further analyzes the cable joint temperature estimation algorithm in Model Sim software.The results show that the cable joint temperature estimation system realized based on FPGA not only improves the speed greatly,but also provides a complete demonstration for engineering application.This paper studies the estimation algorithm of cable joint temperature,and improves the calculation speed and accuracy of the model by improving the algorithm.Finally,the hardware system FPGA is used to realize the estimation algorithm of cable joint temperature,which has important reference significance for indirect monitoring of cable joint temperature in practical engineering.