Research on Equivalent Modeling Method of AC-DC Power Networks Integrating with Renewable Energy Generation

Along with the increasing integration of renewable energy generation in AC-DC power networks,investigating the dynamic behaviors of this complex system with a proper equivalent model is significant.This paper presents an equivalent modeling method for the AC-DC power networks with doubly-fed induction generator(DFIG)based wind farms to decrease the simulation scale and computational burden.For the AC-DC power networks,the equivalent modeling strategy in accordance with the physical structure simplification is stated.Regarding the DFIG-based wind farms,the equivalent modeling based on the sequential identification of multi-machine parameters using the improved chaotic cuckoo search algorithm(ICCSA)is conducted.In light of the MATLAB simulation platform,a two-zone four-DC interconnected power grid with wind farms is built to check the efficacy of the proposed equivalentmodelingmethod.Fromthe simulation analyses and comparative validation in different algorithms and cases,the proposed method can precisely reflect the steady and dynamic performance of the demonstrated system under N-1 and N-2 fault scenarios,and it can efficiently achieve the parameter identification of the wind farms and fulfill the equivalent modeling.Consequently,the proposed approach’s effectiveness and suitability are confirmed.

A novel stacking-based ensemble learning model for drilling efficiency prediction in earth-rock excavation

Accurate prediction of drilling efficiency is critical for developing the earth-rock excavation schedule.The single machine learning(ML)prediction models usually suffer from problems including parameter sensitivity and overfitting.In addition,the influence of environmental and operational factors is often ignored.In response,a novel stacking-based ensemble learning method taking into account the combined effects of those factors is proposed.Through multiple comparison tests,four models,e Xtreme gradient boosting(XGBoost),random forest(RF),back propagation neural network(BPNN)as the base learners,and support vector regression(SVR)as the meta-learner,are selected for stacking.Furthermore,an improved cuckoo search optimization(ICSO)algorithm is developed for hyper-parameter optimization of the ensemble model.The application to a real-world project demonstrates that the proposed method outperforms the popular single ML method XGBoost and the ensemble model optimized by particle swarm optimization(PSO),with 16.43%and 4.88%improvements of mean absolute percentage error(MAPE),respectively.