Prediction of the Height of Fractured Water-Conducting Zone Based on the Improved Cuckoo Search Algorithm–Extreme Learning Machine Model

The research aims to improve prediction accuracy for heights of fractured water-conducting zones (FWCZs) and effectively prevent control roof water disasters, ensure safe coal mining. For this purpose, the method that integrates improved cuckoo search (ICS) algorithm extreme learning machine (ELM) is used predict FWCZs. Based on an analysis factors influencing FWCZs, ICS employed optimize two key parameters ELM model, input weight ѡ bias b hidden elements, thus establishing ICS–ELM model predicting height FWCZ. trained using 42 measured samples, remaining six sample data points. obtained results show a relative error only 3.97% are more consistent with actual situation. To verify effectiveness compared those adaptive particle swarm optimization based least squares support vector (APSO–LSSVM) (PSO) backpropagation (PSO–BP) models. average errors models 8.21 9.75%, respectively, which further proves improves FWCZs predicted accurate reliable, meets requirements engineering practice.