Understanding the Impact of Constraints: A Rank Based Fitness Function for Evolutionary Methods

Fresa: A Plant Propagation Algorithm for Black-Box Single and Multiple Objective Optimization

“…One of these is the non-dominated sorting algorithm, exemplified by the popular NSGA-II genetic algorithm implementation [9]. In Fresa, an alternative method is the assignment of fitness values based on the Hadamard product of the rankings with respect to the individual criteria [10]. A variation on this latter method is an alternative based on the Borda sum of those individual rankings.…”

“…A variation on this latter method is an alternative based on the Borda sum of those individual rankings. These latter two fitness assignment methods tend to emphasise solutions found towards the ends of the Pareto frontier whereas the non-dominated sorting algorithm may lead to the ends of the frontier having less representation in the population [10]. The advantage of the Hadamard and Borda based fitness assignment methods is that they scale linearly with both the population size and the number of objectives whereas a non-dominance sorting approach scales as with .…”

On the application of a nature-inspired stochastic evolutionary algorithm to constrained multi-objective beer fermentation optimisation

“…al., 2015). Recently, using a new fitness function for multi-objective problems, the algorithm has been applied to integrated energy systems design for off-grid mining operations (Fraga and Amusat, 2016), also a dynamic optimisation problem. The algorithm can be summarised as follows: Algorithm 1: The strawberry plant propagation algorithm, adapted from Fraga and Amusat, 2016.…”

“…The weights assigned to the various process targets to produce a single objective function may be considered arbitrary in many cases, with decision-makers (brewers) not necessarily able to quantify a priori the relative importance of competing objectives. A number of multi-objective optimisation algorithms have been successfully applied to a wide range of engineering problems, where visualisation of the trade-offs can provide decision makers with valuable insight (Li et al, 2014;Gujarathi et al, 2015;Zhang et al, 2015;Fraga and Amusat, 2016;Che at al., 2017;Maria and Crişan, 2017;Kessler et al, 2017). Systematically exploring the trade-off and visualising Pareto optimal temperature manipulations for efficient fermentation is desirable to gain insight and assist brewers with the selection of the most preferable operation strategy.…”

“…al., 2015). Recently, using a new fitness function for multi-objective problems, the algorithm has been applied to integrated energy systems design for offgrid mining operations (Fraga and Amusat, 2016). Given the demonstrated success of the algorithm for gaining insight into a bi-criteria objective trade-off, we propose using the Strawberry Algorithm evaluate our solution representations and compute Pareto optimal solution sets for the industrial beer fermentation process.…”