Constrained voting extreme learning machine and its application

Extreme learning machine(ELM)has been proved to be an effective pattern classification and regression learning mechanism by researchers.However,its good performance is based on a large number of hidden layer nodes.With the increase of the nodes in the hidden layers,the computation cost is greatly increased.In this paper,we propose a novel algorithm,named constrained voting extreme learning machine(CV-ELM).Compared with the traditional ELM,the CV-ELM determines the input weight and bias based on the differences of between-class samples.At the same time,to improve the accuracy of the proposed method,the voting selection is introduced.The proposed method is evaluated on public benchmark datasets.The experimental results show that the proposed algorithm is superior to the original ELM algorithm.Further,we apply the CV-ELM to the classification of superheat degree(SD)state in the aluminum electrolysis industry,and the recognition accuracy rate reaches87.4%,and the experimental results demonstrate that the proposed method is more robust than the existing state-of-the-art identification methods.

Exergy Analysis and Thermal Optimization of a Double-Turbine Regeneration System in a Ultra-Supercritical Double-Reheat Unit

Improving the primary steam parameters is one of the most direct ways to improve the cycle efficiency of a powergeneration system. In the present study, the typical problem connected to the excessively high superheat degree ofextraction steam in an ultra-supercritical (USC) double-reheat unit is considered. Using a 1000 MW power plantas an example, two systems (case 1 and case 2) are proposed, both working in combination with a regenerativesteam turbine. The thermal performances of these two systems are compared with that of the original systemthrough a heat balance method and an exergy balance strategy. The results reveal that the two coupled systemscan significantly reduce the superheat degree of extraction steam, turbine heat rate, and coal consumption of theunit and improve the energy utilization efficiency. These results will provide useful theoretical guidance to futureinvestigators wishing to address the general problem relating to energy conservation and modelling of the coupledextraction steam regenerative system of USC double-reheat units.

Energy-Saving Optimization Study on 700℃ Double Reheat Advanced Ultra-Supercritical Coal-Fired Power Generation System

700°C double reheat advanced ultra-supercritical power generation technology is one of the most important development directions for the efficient and clean utilization of coal.To solve the great exergy loss problem caused by the high superheat degrees of regenerative steam extractions in 700°C double reheat advanced ultra-supercritical power generation system,two optimization systems are proposed in this paper.System 1 is integrated with the back pressure extraction steam turbine,and system 2 is simultaneously integrated with both the outside steam cooler and back pressure extraction steam turbine.The system performance models are built by the Ebsilon Professional software.The performances of optimized systems are analyzed by the unit consumption method.The off-design performances of optimization systems are analyzed.The results show that:the standard power generation coal consumption rates of optimization systems 1 and 2 are decreased by 1.88 g·(kW·h)^(–1),2.97 g·(kW·h)^(–1)compared with that of the 700°C reference system;the average superheat degrees of regenerative steam extractions of optimized systems 1 and 2 are decreased by 122.2°C,140.7°C(100%turbine heat acceptance condition),respectively.The comparison results also show that the performance of the optimized system 2 is better than those of the optimized system 1 and the 700°C reference system.The power generation standard coal consumption rate and the power generation efficiency of the optimized system 2 are about 232.08 g·(kW·h)^(–1)and 52.96%(100%turbine heat acceptance condition),respectively.