Search

With increasing wind farm, solar farm, and pump storage plant integrations, intense frequency fluctuation of the pumped storage hydro unit (PSHU) under the no-load running condition, which is caused by its operation along the S-shaped curve, has been noted and researched. So, parameter identification of the PSHU regulation system (PSHURS) is crucial in precise modeling of the PSHU and can provide support for the optimized control and stability analysis of the power system. In this paper, a parameter adaptive identification method together with an improved gravitational search algorithm (IGSA) is proposed and applied to solve the identification problem for a PSHURS under the no-load condition. The IGSA, which is based on the standard gravitational search algorithm, accelerates convergence speed with a combination of the Pbest-Gbest-guided search strategy and the adaptive elastic-ball method and improves the local optimal with the added chemotaxis operator of the bacterial foraging algorithm. Furthermore, for the parameter adaptive identification method, the parameter performance evaluator is employed to devise the moving step of the agent of the chemotaxis operator. The illustrative experiment for parameter identification of the PSHURS is used to verify the feasibility and effectiveness of the proposed method. Comparison with other methods clearly shows that the adaptive parameter identification method along with the IGSA perform best for all identification indicators. [ABSTRACT FROM AUTHOR]

Extreme Learning Machine (ELM) is a novel single-hidden-layer feed forward neural network with fast learning speed and better generalization performance compared with the traditional gradient-based learning algorithms. However, ELM has two issues: the hidden node number of ELM needs to be predefined and the random determination of the input weights and hidden biases lead to ill-condition problem. In this paper, a two-stage evolutionary extreme learning machine (TSE-ELM) algorithm was proposed to overcome the drawbacks of original ELM, which used an improved artificial bee colony (ABC) algorithm to optimize the input weights and hidden biases. The proposed TSE-ELM algorithm was applied on the UCI benchmark datasets and rolling bearing fault diagnosis. The numerical experimental results demonstrated that TSE-ELM had an improved generalization performance than traditional ELM and other evolutionary ELMs. [ABSTRACT FROM AUTHOR]