1887

Abstract

In the recent years the performance of photo-voltaic (PV) cells has been significantly improved by using multi-cell devices. In such systems, few layers of different cells are combined to maximize the efficiency by exploiting the use of different wave lengths of light. Using this approach efficiency of over 40% has been achieved. The two most promising methods are splits spectrum and multi-junction photo-voltaic cells. With the development of new technologies it has become possible to create more complex PV sells consisting of a higher number of layers. For future development of such PV cells it is of significant importance to have bounds for the optimal possible efficiency. While it is relatively simple to find them in case of two or three layers it becomes significantly more complex in case of a higher number of layers. This is due to the fact that it is necessary to find the of a multi-parameter function, which is computationally expensive. The problem becomes even more complex because it is hard to find the corresponding gradient that could simplify the calculation. There is a wide range of non-gradient based methods like simulated annealing, genetic algorithms, particle swarm optimization, Nelder-Mead Simplex method[1] that are generally used solve this type of problems. The performance of such methods is highly dependent on the function that we wish to minimize. In the case of the problem of interest, initial test have shown that Nelder-Mead Simplex algorithm manages to out preform mentioned more complex population based methods. One of the reasons for this is the fact that due to physical properties of the problem we have a good initial guess of the solution. Our research has focused on improving the performance of this algorithm by incorporating some type of swarm intelligence. Previously, similar hybridization of Nelder-Mead simplex algorithm using genetic algorithms[2], ant colony optimization[3] and particle swarm optimization[4] have proven to be very efficient. In the recent years the Cuckoo Search[5] algorithm has been gaining on popularity as an optimization method due to its good performance, robustness and simplicity of implementation. One of the main problems of hybridized methods is that although they achieve better results they often become very complex for implementation. In our work we introduce a cuckoo search inspired hybridization of the Nelder-Mead simplex algorithm that manages to avoid this drawback but still achieves significantly better results than the original method. In our tests we show that the proposed method also achieves good results on standard benchmark functions. The cuckoo search is often a competing method to the particle swarm optimization; because of this we also give a comparison to previously published results of hybridization of Nelder-Mead simplex using this method.

Loading

Article metrics loading...

/content/papers/10.5339/qfarf.2013.EEP-037
2013-11-20
2019-11-15
Loading full text...

Full text loading...

http://instance.metastore.ingenta.com/content/papers/10.5339/qfarf.2013.EEP-037
Loading
This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error