In this paper, the future load condition is predicted by proposing an effective single-layer Black Widow Optimization based Functional Link Artificial Neural Network (BWO-FLANN) model. Also, a novel Non-dominated Sorting Multi-Objective Teaching Learning Based Optimization (NS-MOTLBO) has been proposed for obtaining the solution to Economic Emission Load Dispatch (EELD) in a combined operation of the grid with solar photovoltaic source by considering the future predicted load. The EELD is an inherently multi-objective problem with nonlinear constraints, such as the generator power limits, power balance criteria, transmission line capacity, prohibited operating zones, etc., that make it more complex, causing extra computational burden and convergence issues. An effective constraint-handling methodology is also used in this paper to achieve the lowest possible constraint violation. The proposed approach has been implemented in two test situations, each with distinct solar radiation. The Pareto optimal fronts for various hours were obtained, indicating possible scheduling of the corresponding periods. Non-parametric statistical testing is carried out to investigate the dominance of our proposed approach compared to the other state-of-the-art prediction models.