A Large-Scale Nonlinear Mixed-Binary Goal Programming Model to Assess Candidate Locations for Solar Energy Stations: An Improved Real-Binary Differential Evolution Algorithm with a Case Study

Solar energy is considered one of the important sources for new and renewable energy; all countries are seeking to take advantage of these sources in various fields. Saudi Arabia enjoy a privileged position compared to all the countries in the world for the length of the brightness of the sun as well as the degree of emitted radiation throughout the whole year and in all different seasons, particularly in the western region of the Kingdom. The choice of the appropriate sites to build solar power plants is a necessary aspect, the right place should enjoy several properties that differ in its priority such as: the land space currently available and its expansions in the future, the intensity of solar radiation, total costs, distance to the National network of electric power, the distance to the nearest source of water, the distance to the nearest residential city, and the type of access road to the place. To assess the places that are selected and ranked according to the degree of suitability for the establishment of solar power plants with regard to the properties that characterize these places. A Large-Scale Nonlinear Mixed-Binary Goal Programming Model (LSNLMBGPM) will be designed to express the problem and arrange places that are selected according to the degree of suitability. In this paper, an improved real-binary Differential Evolution (DE) algorithm for solving constrained optimization, named IRBDE, is developed to solve the proposed non-linear integer GP problem with binary and real variables. The proposed algorithm introduces a new search mutation to improve both the local search tendency and the global exploration capability, and to accelerate the convergence of DE technique. Moreover, to deal with binary variables, a new binary mutation rule is introduced. Besides, adaptive crossover rate and randomized scale factors will be introduced as uniformly random numbers to enhance the population diversity.