Aimin An

A robust adaptive hybrid genetic simulated annealing algorithm for the global optimization of multimodal functions

In this paper we presented a novel hybrid genetic algorithm for solving NLP problems based on combining the Genetic algorithm and Simulated annealing, together with a local search strategy. The proposed hybrid approach combines the merits of genetic algorithm (GA) with simulated annealing (SA) to construct a more efficient genetic simulated annealing (GSA) algorithm for global search, which could well maintain the population diversity in GA evolution without becoming easily trapped in local optimum. The iterative hill climbing (IHC) method as a local search technique is incorporated into GSA loop to speed up the convergence of the algorithm. In addition, a self-adaptive hybrid mechanism is developed to maintain a tradeoff between the global and local optimizer searching then to efficiently locate quality solution to complex optimization problem. The computational results indicate that the global searching ability and the convergence speed of this hybrid algorithm are significantly improved. Some well-known benchmark functions are utilized to test the applicability of the proposed algorithm.

PMSG maximum wind power tracking control based on T-S fuzzy method

A novel control strategy for capturing maximum wind energy under multi-operating condition based on T-S fuzzy model method is proposed in this paper. Firstly, the T-S fuzzy model of the permanent magnet synchronous wind power generator (PMSG) is constructed under multi-operating condition. The state observer is designed for estimating the states accurately with considering the random wind disturbances existence. Meanwhile, the Lyapunov stability theory and the parallel distributed compensation(PDC) strategy are used to track maximum wind power steadily in real time. The gain of the proposed controller and state observer are solved through linear matrix inequality(LMI) method. The simulation results demonstrated that the performance of the proposed control strategy is significantly better than PI method, and the maximum wind power tracking can be reached effectively.

Benefit analysis and evaluation of distributed generation in distribution network under active management

The renewable distributed generation such as solar energy and wind energy, which is connected to the distribution network, can enhance the reliability of the power system and produce more economic benefits. This paper analyzes the impact of distributed power access network under active management (AM) at first. And mainly establishes the voltage distribution index and compares the voltage distribution in different ways of reactive power control strategies. Meanwhile, a reasonable economic evaluation model is proposed according to the general model of line loss and pollutant discharge standards. Finally, the evaluation index and model of three kinds of DGs under different accesses are verified by computing and analyzing simulation case with IEEE30 node model. Results of numerical analysis show that reasonable planning for the access of distributed power can improve voltage profiles reduce line losses and produce greater economic benefits under active management.

A hybrid differential evolution algorithm for nonlinear parameter estimation of kinetic systems

The determination of the optimal model parameters for kinetic systems development of kinetic models is a time consuming, iterative process [1]. In this paper, we presented a novel hybrid Differential Evolution (DE) algorithm for solving kinetic parameter estimation problems based on the Differential Evolution technique together with a local search strategy. By combining the merits of DE with Gauss-Newton method, the proposed hybrid approach employs a DE algorithm for identifying promising regions of the solution space followed by use of Gauss-Newton method to determine the optimum in the identified regions. The computational results indicate that the global searching ability and the convergence speed of this hybrid algorithm are significantly improved. Additionally, study of kinetic model parameters for an irreversible, first-order reaction system was carried out to test the applicability of the proposed algorithm. The suggested method can be used to estimate suitable values for the model parameters of a complex mathematical model.

A novel hybrid Genetic Algorithm for HEN synthesis and its industrial application

In this paper we look at a new hybrid Genetic Algorithm (HGA) based on genetic simulated annealing (GSA) algorithm for solving heat exchanger network synthesis (HENS) problems with Mixed Integer Nonlinear Programming (MINLP) model. In order to efficiently locate quality solution to complex optimization problem, a self-adaptive mechanism is developed to maintain a tradeoff between the global and local search. The computational results indicate that the global searching ability and the convergence speed of this hybrid algorithm are significantly improved. Further, the proposed algorithm is tailored to find optimum solution to industrial HENS problem, The results show that the proposed approach could provide designers with a least-cost HEN with less computational cost comparing with other optimization methods.

Data-driven LQG benchmaking for economic performance assessment of advanced process control systems

In this paper, a data-driven subspace approach for economic performance assessment of the advanced process control (APC) systems is presented. The method introduces LQG tradeoff curve to estimate potential of reduction in variance, which is directly obtained from subspace matrices using closed loop data. To exploit feasible economic performance of the APC systems, the proposed approach considers the uncertainties induced by process variability and evaluates the economic performance through solving stochastic optimization problem. Results of the performance evaluation provide a guideline for the control system tuning to realize the potential improvement in profitability of process. The application of the proposed method is illustrated by its benefits evaluation on a simulated example.