Tag Archives: polynomial neural networks

Ideas for doing research in Salinity Gradient Based Power Plants

Power generation from Salinity Gradient is a new topic of research which is steadily attracting interest from the researchers of renewable energy. The scarcity and rising demand of conventional fuels has also aggrevated the importance of alternative sources of energy where power from salinity gradient or Blue Energy can be an useful solution.

As the topic is new and mostly unexplored,doing research in this aspect have a very high potential. Research ideas like :

1) Cost Minimization of Power Extraction from Saline Water Gradient based Power Plants by Nature Based Optimization Techniques

The first objective of the work will be to identify parameter/s which contribute towards the total expenditure for installation of Salinity Gradient based power plants.After all the parameters were identified and the significance of the weights of the parameters are determined the next step will be to formulate the indicator in such a manner that it become directly proportional to the expenditure for operation and maintenance of salinity gradient based power plants.The second objective will be to implement nature based optimization algorithms like Ant Colony,Artificial Bee Hive etc to minimize the cost function.The significance of the parameters can be estimated by the application of Multi Criteria Decision Making methods(MCDM) and the nature based optimization techniques can be used for minimization of the function.The result of the optimization will yield the ratio at which the significant parameters must behave to produce minimum expenditure to run a saline water based power plant.

2)Estimation of Power Potential from Saline Water Gradient in European/Asian/American Coastal regions : A Polynomial Neural Network Approach

Polynomial Neural Networks are an advanced form of neural network which can self-detect the number of inputs and hidden layers, required for prediction of the most accurate output.The Group Method of Data Handling(GMDH) is a technique by which polynomial neural networks(PNN) are trained where it analyse hundreds of different training algorithms and identify the best algorithm automatically with the help of a fitness function.This kind of technologies can be implemented to estimate the power potential from saline water gradient in various locations of different continents.An iso-hyetal map can be produced from the predictions so that the location with highest potential can be easily delineated.

3)Location selection for installation of Pressure Retarded Osmosis (PRO)/Reverse Electrodialysis (RED)/Hydrocratic Ocean Energy based Salinity Gradient Power plants with the help of MCDM-GMDH or MCDM or GMDH.

MCDMs are an objective procedure to identify the better option among the many available alternatives.GMDH is a technique for training PNNs.This two technologies can be applied to identify locations with higher feasibility among multiple locations selected for installation of Saline Water based power plants.At first the parameters which can select the plausible locations can be identified and then with the help of a function the locations with highest potentiality can be selected.The function will be a product function of the selected parameters and its significance in detecting suitable locations.For different kind of saline water based power generation units(such as Pressure Retarded Osmosis (PRO)/Reverse Electrodialysis (RED)/Hydrocratic Ocean Energy) different features will be selected and the significance of the featuress will also change.The calculation of significance will be executed with the help of MCDM methods and function will be predicted with the help of GMDH based PNNs.

…can attract project funds as well as may help to publish research articles in reputed journals.As a pioneer, citations for such research will be high compared to that for other conventional topics of renewable energy.