Accurate ultra-short-term wind speed prediction is extremely important for the power control of wind farms, the safe dispatch of power systems, and the stable operation of power grids. At present, most wind farms mainly rely on supervisory control and data acquisition systems to obtain operation and maintenance data which includes operating characteristics of wind turbines. In the ultra-short-term wind speed prediction, a long short-term memory network is one of the commonly used deep learning methods. To address the problem that improper selection of long short-term memory network's hyperparameters may affect the prediction results, In the present study, a hybrid prediction model based on the long short-term memory and the modified tuna swarm optimization algorithm was established, and was used to predict after the wind speed sample data had been decomposed by successive variational mode decomposition method. The experimental results reveal that the proposed model effectively improved the accuracy of wind speed prediction for wind farms compared with the support vector regression, deep belief networks, and long short-term memory models optimized by particle swarm optimization algorithm.
Mayfly algorithm is a new intelligent optimization algorithm with unique optimization capabilities recently proposed. It has strong research value, but there are also insufficient explorations, and it is easy to fall into the problem of local optimization. This paper aims to improve the optimization performance of the mayfly algorithm and explore its application value in practical engineering optimization problems. An improved mayfly algorithm based on the median position of the group is proposed. In its velocity update, the median position of the group is introduced with emphasis, and a non-linear gravity coefficient is introduced at the same time. Through the benchmark test function, its superiority in exploitation, convergence speed and accuracy and the improvement of exploration are verified. At the same time, the simulation model of the hydro-turbine governor using MATLAB/Simulink is established, and 10% frequency disturbance experiments of this model are carried out separately in two typical working conditions. The experiments results show that the optimal ITAE index value of the system obtained by the improved mayfly algorithm is smaller, and 16.5 and 18.1 iterations to complete on average. In addition, the experiments results reveal that the PID parameters optimized by the improved mayfly algorithm can make the dynamic performance of the regulation system better than other popular swarm intelligence algorithms, where the overshoot decreased by more than 3.1%, and the adjustment time also decreased in different degrees. The proposal of the median position of the group provides a new idea for the improvement of the swarm intelligence optimization algorithm. Meanwhile, a new effective method for optimizing the PID parameters of the hydro-turbine governor has been found.
The use of failure recognition technology can detect unusualness timely and deal with them properly to ensure safe and stable operation of wind turbine (WT). An effective troubleshooting method can quickly distinguish the type of WT fault and reduce wind farm operation and maintenance costs. At present, the relevant data required for fault diagnosis methods come from the supervisory control and data acquisition (SCADA) system, because the SCADA data contains information associated with the operating characteristics of WT, which can provide a rich source of data for WT fault diagnosis. A deep belief network (DBN) is commonly used the deep learning method. In the present study, an optimized DBN based on the modified tuna swarm optimization (MTSO) algorithm was established to construct a MTSO-DBN WT fault diagnostic model, so as to address the problem that the selection of DBN hyperparameters may affect the classification results. After preprocessing the WT fault data acquired by SCADA, the MTSO-DBN model was used to classify the WT faults. The experimental results reveal that, compared with the support vector machine (SVM), extreme learning machine (ELM), DBN, PSO-DBN, and TSO-DBN classification models, the MTSO-DBN model could effectively improve the accuracy of WT faults for wind farms.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.