Prediction based Multiobjective Solution of Economic Emission and Load Dispatch for Solar Integrated Power Systems

In present day power systems, the conventional thermal generating stations are being interconnected with solar photovoltaic sources to reduce the running cost along with environmental emission. For proper load dispatch, short term forecasting of electric load is essential to avoid overloads, surges and instability because of varying demand. In this work, Improved Multi-Objective Teaching Learning-Based Optimization (IMOTLBO) algorithm has been developed for effective Economic Emission and Load Dispatch (EELD). Here, the predicted load of a real time load center on a test Solar Integrated System (SIS) has been utilized to obtain Pareto solution, considering cost and emission as two objectives. The performance of the proposed IMOTLBO algorithm is compared with four other MOEAs, namely, Non-dominated Sorting Genetic Algorithm-II (NSGA-II), Multi-Objective Particle Swarm Optimization (MOPSO), Modified Multiobjective Cat Swarm Optimization (MMOCSO) and Multi-Objective Differential Evolution with Recursive Distributed Constraint Handling (MODE-RDC). For efficient prediction of the expected electrical load demand, an efficient single layer low complexity neural network i.e. Functional Link Artificial Neural Network (FLANN) model is considered. The weights of FLANN model are optimized by utilizing four different algorithms; one derivative based i.e. Least Mean Squares (LMS), and three others heuristic algorithms, namely Particle Swarm Optimization (PSO), Jaya and TLBO. To compare the performance of the proposed TLBO based FLANN models with the other three models, the Root Mean Square Error (RMSE) has been considered as the performance index. The dominance of the proposed FLANN-TLBO models over others is investigated by conducting non-parametric statistical testing.