Zhenhua Shi

Mengshan Xie, Sheng Xu, Cheng-yue Su et al.

With the characteristics of vertical take-off and landing and long endurance, tiltrotor has attracted considerable attention in recent decades for its potential applications in civil and scientific research. However, the problems of strong couplings, nonlinear characteristics and mismatched disturbances inevitably exist in the tiltrotor, which bring great challenges to the controller design in transition mode. In this paper, we combined a super-twisting extended state observer (STESO) with an adaptive recursive sliding mode control (ARSMC) together to design a tiltrotor aircraft attitude system controller in transition mode using STESO-ARSMC (SAC). Firstly, the six degrees of freedom (DOF) nonlinear mathematical model of tiltrotor is established. Secondly, the states and disturbances are estimated by the STES observer. Thirdly, ARSM controller is designed to achieve finite time convergence. The Lyapunov function is used to testify the convergence of the tiltrotor UAV system. The new aspect is that the assessments of the states are incorporated into the control rules to adjust for disruptions. When compared to prior techniques, the control system proposed in this work can considerably enhance anti-disturbance performance. Finally, simulation tests are used to demonstrate the efficacy of the suggested technique.

Changming Zhao, Dongrui Wu, Jian Huang et al.

Bootstrap aggregating (Bagging) and boosting are two popular ensemble learning approaches, which combine multiple base learners to generate a composite model for more accurate and more reliable performance. They have been widely used in biology, engineering, healthcare, etc. This paper proposes BoostForest, which is an ensemble learning approach using BoostTree as base learners and can be used for both classification and regression. BoostTree constructs a tree model by gradient boosting. It increases the randomness (diversity) by drawing the cut-points randomly at node splitting. BoostForest further increases the randomness by bootstrapping the training data in constructing different BoostTrees. BoostForest generally outperformed four classical ensemble learning approaches (Random Forest, Extra-Trees, XGBoost and LightGBM) on 35 classification and regression datasets. Remarkably, BoostForest tunes its parameters by simply sampling them randomly from a parameter pool, which can be easily specified, and its ensemble learning framework can also be used to combine many other base learners.