Swarm and Evolutionary Computation

Data File

Energy-Efficient Hybrid Flowshop Scheduling

Source code and set of instances.

1. User load and wind predicted data for 'Multi-scenario microgrid optimization using an evolutionary multi-objective algorithm'. 2. Run data

This repository contains data extracted from the oil spill images used in this manuscript. In all, there are ten image folders, one for each oil spill image. The extracted data is captured in the LAB color space (see https://goo.gl/WtXnBz), accounting for three sub-folders within each image folder. Each image subfolder represents a LAB color space channel, of which there are three. One channel represents Luminance (lightness) (labeled ‘L’) while the other two (labeled ‘a’ and ‘b’), represent color channels. Each channel is a MATLAB mat file of size 1000 x 1000, based on the 1 km^2 scaling used for each image.

The direct program output. This data was used in the non-parametric Wilcoxon one-sided signed-rank test. MUT1_I_out.txt - evolution algorithm uses the mutation operator (Scramble Mutation) based on the conditional expectation value for the QAP problem MUT_S_out.txt - evolution algorithm uses the standard mutation operator (Scramble Mutation) OXI_out.txt - evolution algorithm uses the OX operator (order crossover) and the conditional expectation value for the QAP problem (intensification version) OX_out.txt - evolution algorithm uses the standard OX operator (order crossover) PMX2E_out.txt - evolution algorithm uses the PMX operator (partially matched crossover) and the conditional expectation value for the QAP problem (2nd exploration version) PMX_out.txt - evolution algorithm uses the standard PMX operator (partially matched crossover) POX2E_out.txt - evolution algorithm uses the POX operator (position based crossover) and the conditional expectation value for the QAP problem (2nd exploration version) POX_out.txt - evolution algorithm uses the standard POX operator (position based crossover)

The archive contains raw data of the various DE mutation strategies, tested with different selection methods, applied to mutation, as well as the jupyter notebook, containing the code for data processing, graphs generating, statistical tests and so on.

This is supplementary material to the article ``Comparison of Mutation Strategies in Differential Evolution -- a Probabilistic Perspective''. It provides more detailed benchmark results than those reported in the paper. The design of the experiment, which serves also as the ``table of contents'', is provided in Table 1.

The Matlab source code of DE/GM.

Vanishing Point Detection in an Autonomous Car using Harmony Search and RANSAC algorithms