Font Size: a A A

Research Of Improving Artificial Bee Colony Algorithm And Its Application In Wind Power Forecasting

Posted on:2018-09-02Degree:MasterType:Thesis
Country:ChinaCandidate:J ChenFull Text:PDF
GTID:2348330518986490Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of social science and technology,the optimization problem has become one of the most common problems in the field of literature and engineering.It is difficult for traditional methods to deal with the optimization problem which is high dimension,multi-peak and non-micro.But artificial intelligent optimization algorithms which are inspired by the behavior of social insect colonies have obvious advantages to solve this kind of complex optimization problem.Recently the research of intelligent optimization algorithms have been widely concerned by many scholars from different countries.Artificial bee colony(ABC)algorithm as a newer artificial intelligent optimization algorithm,has the characteristics of simplicity,high efficiency and good convergence.But ABC algorithm still has the weakness of high convergence precision and is easy to fall into the local optimum.What's more,the research of multi-objective artificial bee colony(MOABC)algorithm is quite a few,and the application of MOABC algorithm needs to be further widened.In this paper,the basic artificial bee colony algorithm is improved in the aspect of convergence accuracy and speed.A new multi-objective artificial bee colony algorithm is proposed and applied in the field of wind power prediction intervals.Therefore,what we research has the positive meaning not only in theoretical research but also in real life.In this paper,the present research of ABC algorithm is analyzed deeply,and the principle and characteristics of basic ABC algorithm are expounded.On the basis of this,we carried out a bit of innovative work:Firstly,an advanced artificial bee colony algorithm based on information feedback and improved fitness evaluation is proposed.The lead search strategy and the probability selection function are improved to change the way of bee colony in searching the optimal solution.The test of standard functions shows that the improved algorithm has faster convergence speed and higher convergence precision.Then a new MOABC algorithm based on evolutionary knowledge fusion is proposed.The evolutionary knowledge is used to guide the evolution of bee colony.The individual dominance relationship and population distribution relationship are combined to improve the probability selection formula.In addition,the new algorithm adopts a more stringent external file maintenance strategy to reduce the costs of algorithm and improve the distribution performance of the solution set.Compared with the other multi-objective optimization algorithms,the results show that the proposed algorithm has good performance both in convergence and distribution.At the same time,the coverage of the solution set becomes wider.Finally,an intelligent multi-objective optimized prediction intervals(PIs)modelfor wind power is proposed and wavelet neural network(WNN)is adopted as the basic prediction model.The selection probability and constraint pruning strategy of basic MOABC algorithm are improved to optimize the scaling factor,the shifting parameter and the weights of WNN.The results show that the proposed model has better performance for wind power interval prediction,which can provide decision-making for electrical power departments.
Keywords/Search Tags:artificial bee colony algorithm, wind power forecast, artificial intelligence, multi-objective optimization, prediction intervals
PDF Full Text Request
Related items