Multi-objective optimization of a recuperative gas turbine cycle using non-dominated sorting genetic algorithm

2011 ◽  
Vol 225 (8) ◽  
pp. 1041-1051 ◽  
Author(s):  
H Sayyaadi ◽  
H R Aminian
Keyword(s):  
Genetic Algorithm ◽  
Gas Turbine ◽  
Heat Exchangers ◽  
Mixed Integer ◽  
Design Stage ◽  
Tube Length ◽  
Pareto Optimal ◽  
Multi Objective Optimization ◽  
Multi Objective ◽  
Gas Turbine Cycle

A regenerative gas turbine cycle with two particular tubular recuperative heat exchangers in parallel is considered for multi-objective optimization. It is assumed that tubular recuperative heat exchangers and its corresponding gas cycle are in design stage simultaneously. Three objective functions including the purchased equipment cost of recuperators, the unit cost rate of the generated power, and the exergetic efficiency of the gas cycle are considered simultaneously. Geometric specifications of the recuperator including tube length, tube outside/inside diameters, tube pitch, inside shell diameter, outer and inner tube limits of the tube bundle and the total number of disc and doughnut baffles, and main operating parameters of the gas cycle including the compressor pressure ratio, exhaust temperature of the combustion chamber and the air mass flowrate are considered as decision variables. Combination of these objectives anddecision variables with suitable engineering and physical constraints (including NO x and CO emission limitations) comprises a set of mixed integer non-linear problems. Optimization programming in MATLAB is performed using one of the most powerful and robust multi-objective optimization algorithms, namely non-dominated sorting genetic algorithm. This approach is applied to find a set of Pareto optimal solutions. Pareto optimal frontier is obtained, and a final optimal solution is selected in a decision-making process.

Micro-Turbine Design Point Definition Using Genetic Algorithm With Single and Multi-Objective Optimization

2012 ◽  
Author(s):  
Diogo F. Cavalca ◽  
Cleverson Bringhenti
Keyword(s):  
Genetic Algorithm ◽  
Gas Turbine ◽  
Pressure Ratio ◽  
Inlet Temperature ◽  
Specific Work ◽  
Multi Objective Optimization ◽  
Total Cost ◽  
Design Point ◽  
Multi Objective ◽  
Engine Design

During a gas turbine development phase an important engineer task is to find the appropriate engine design point that meet the required specifications. This task can be very arduous because all possible operating points in the gas turbine operational envelope need to be analyzed, for the sake of verification of whether or not the established performance might be achieved. In order to support engineers to best define the engine design point that meet required performance a methodology was developed in this work. To accomplish that a computer program was written in Matlab®. In this program was incorporated the thermoeconomic and thermodynamic optimization. The thermodynamic calculation process was done based in enthalpy and entropy function and then validated using a commercial program. The methodology uses genetic algorithm with single and multi-objective optimization. The micro gas turbine cycle chosen to study was the recuperated. The cycle efficiency, total cost and specific work were chosen as objective functions, while the pressure ratio, compressor and turbine polytropic efficiencies, turbine inlet temperature and heat exchange effectiveness were chosen as decision variables. For total cost were considered the fixed costs (equipment, installation, taxes, etc.) and variable costs (fuel, environmental and O&M). For emissions were taken into account the NOx, CO and UHC. An economic analysis was done for a recuperated cycle showing the costs behavior for different optimized design points. The optimization process was made for: single-objective, where each objective was optimized separately; two-objectives, where they were optimized in pairs; three-objectives, where it was optimized in trio. After, the results were compared each other showing the possible design points.

Mixed-integer non-linear programming (MINLP) multi-period multi-objective optimization of advanced power plant through gasification of municipal solid waste (MSW)

2020 ◽  
Vol 0 (0) ◽  
Author(s):  
Ahmad Syauqi ◽  
Widodo Wahyu Purwanto
Keyword(s):  
Power Generation ◽  
Linear Programming ◽  
Power Plant ◽  
Municipal Solid Waste ◽  
Solid Waste ◽  
Gas Turbine ◽  
Mixed Integer ◽  
Multi Objective Optimization ◽  
Non Linear Programming ◽  
Multi Objective

AbstractMulti-objective optimization is one of the most effective tools for the decision support system. This study aims to optimize the gasification of municipal solid waste (MSW) for advanced power plant. MSW gasifier is simulated using Aspen Plus v11 to produce syngas, to be fed into power generation technologies. Four power generation technologies are selected, solid oxide fuel cell, gas turbine, gas engine, and steam turbine. Mixed-integer non-linear programming (MINLP) multi-objective optimization is developed to provide an optimal solution for minimum levelized cost of electricity (LCOE) and minimum CO2eq emissions. The optimization is conducted with a ε-constraint method using GAMS through time periods of 2020–2050. Decision variables include gasifier temperature, steam to carbon ratio, and power generation technologies. The optimization result demonstrates that the lower steam to carbon ratio gives lower LCOE and higher CO2eq emissions, and temperature variation gives no significant impact on LCOE and as it increases, CO2eq emission is reduced. It demonstrates that a gas turbine is the best option for generating electricity from 2020 to 2040 and beyond 2040 SOFC is the best option.

Multi-objective optimization of a new cogeneration system driven by gas turbine cycle for power and freshwater production

2021 ◽  
Vol 288 ◽  
pp. 125639
Author(s):  
Pan Ding ◽  
Xiaojuan Liu ◽  
Hongling Qi ◽  
Hongtao Shen ◽  
Xiaochan Liu ◽  
...  
Keyword(s):  
Gas Turbine ◽  
Multi Objective Optimization ◽  
Multi Objective ◽  
Cogeneration System ◽  
Gas Turbine Cycle

Improved Memetic NSGA-II Using a Deep Neighborhood Search

2021 ◽  
Vol 12 (4) ◽  
pp. 138-154
Author(s):  
Samir Mahdi ◽  
Brahim Nini
Keyword(s):  
Genetic Algorithm ◽  
Genetic Algorithms ◽  
Pareto Front ◽  
Neighborhood Search ◽  
Pareto Optimal ◽  
Multi Objective Optimization ◽  
Nsga Ii ◽  
Pareto Optimal Front ◽  
Slow Convergence ◽  
Multi Objective

Elitist non-sorted genetic algorithms as part of Pareto-based multi-objective evolutionary algorithms seems to be one of the most efficient algorithms for multi-objective optimization. However, it has some shortcomings, such as low convergence accuracy, uneven Pareto front distribution, and slow convergence. A number of review papers using memetic technique to improve NSGA-II have been published. Hence, it is imperative to improve memetic NSGA-II by increasing its solving accuracy. In this paper, an improved memetic NSGA-II, called deep memetic non-sorted genetic algorithm (DM-NSGA-II), is proposed, aiming to obtain more non-dominated solutions uniformly distributed and better converged near the true Pareto-optimal front. The proposed algorithm combines the advantages of both exact and heuristic approaches. The effectiveness of DM-NSGA-II is validated using well-known instances taken from the standard literature on multi-objective knapsack problem. As will be shown, the performance of the proposed algorithm is demonstrated by comparing it with M-NSGA-II using hypervolume metric.

Sign in / Sign up

Export Citation Format

Share Document