Prateek Mittal

Kriging Surrogate Based Multi-objective Optimization of Bulk Vinyl Acetate Polymerization with Branching

Despite the established superiority in finding the global as well as well-spread Pareto optimal (PO) points, the need of more numbers of function evaluations for population based evolutionary optimization techniques leads to a computationally demanding proposal. The case becomes more miserable if the function evaluations are carried out using a first principle based computationally expensive model, making the proposal not fit for online usage of the application. In this work, a Kriging based surrogate model has been proposed to replace a computationally expensive model to save execution time while performing an optimization task. A multi-objective optimization study has been carried out for the bulk vinyl acetate polymerization with long-chain branching using these surrogate as well as expensive models and Kriging PO solutions similar to those found by the first principle models are obtained with a close to 85% savings in function evaluations.

A fast and effective algorithm to optimize the total number and placement of wind turbines

Wind turbine placement in wind farms is a nontrivial optimization problem primarily due to aerodynamic wake interaction between turbines. Moreover, in reality, there could be other constraints related to farm topology, capacity factors etc. that can potentially make the problem more constrained and complex. Existing problem formulations typically solve for optimal turbine locations directly while assuming that the total number of turbines in the farm is known. However, wind farm developers often do not know the best number of turbines to be placed in a farm. This paper proposes a novel map-based heuristic search procedure to simultaneously identify the optimal total number of turbines to be placed in the farm along with their placement.

Multi-objective optimization of energy generation and noise propagation: A hybrid approach

Various methodologies have been proposed to optimally place wind turbines inside a wind farm to extract maximum energy. However it is highly likely, that these layouts come in the proximity of human habitation leading to a negative impact on their health by creating noise, visual impact, electromagnetic interference etc. Compared to others, noise has become an important point of concern for wind farm owners, which limits the number of turbines to be erected in a wind farm satisfying its mandatory noise limits. In this study, wind farm layout optimization (WFLO) is performed by considering a trade-off between energy and noise generated in a wind farm. A novel hybrid methodology is proposed to carry out a multi-objective optimization between maximized energy and minimized noise level value. Proposed hybrid methodology is a combination of multi-objective evolutionary algorithm (NSGA-II) followed by a single objective gradient approach to solve a series of integer and continuous problem formulations. Results of a generated Pareto Optimal (PO) front provide an alternative solution of Energy - Noise trade-off along with an additional information on corresponding optimum number of turbines and their optimal location coordinates, which leaves a designer with ample of choices to make in between optimal turbine number as well as two objectives.

Energy-noise trade-off to optimize the total number and the placement of wind turbines on wind farms: A hybrid approach

Wind energy is gaining importance as one the most progressive renewable energies due to rapid depletion of conventional energy resources. Micro-siting is the optimal way of placing turbines inside a wind farm to convert wind power into electrical energy avoiding constraints related to wake loss. Though a significant progress has been made towards proposing efficient methodologies for micro-siting, limited availability of land area has resulted in the construction of wind farms near to the human habitats causing a negative impact on the human health. Compared to the other effects, the effect of noise is a matter of immense concern for the wind farm designers, as it needs to be constrained within the mandatory limits. Using a well-established wake model and ISO-9613-2 noise calculation, this study performs a wind farm layout optimization (WFLO) based on multi-objective trade-off between minimization of the noise propagation and maximization of the energy generation. A novel hybrid methodology is proposed as a combination of probabilistic multi-objective evolutionary algorithm (NSGA-II) and a deterministic gradient based Normalized normal constraint (NNC) method. In contrast to previous studies, the generated Pareto Optimal (PO) front provides several options for a decision maker, where optimal number of turbines and their optimal layouts are obtained at the same time along with the alternative solutions.