Demand response management (DRM) aims to curb the energy consumption by reducing the peak load and thereby improve the performance of distributed generation (DG). From the utility perspective, the cost of generation should be kept minimum, whereas from the customer’s view, the social welfare should be maximum. Therefore this research focuses on a new method based on intelligent algorithm to optimally operate the demand side management in the presence of DG units and demand response. Initially, the location, size and rating of DG units are selected to achieve the optimization of real, reactive power and voltage deviation Index. Finally comprehensive case studies are taken into consideration with DRM strategy to vindicate the technical and economical benefits. Reduction in congestion and power loss with maximized customer benefit are taken as an objective function in this stage. Probability-Based Incremental Learning is implemented and it is tested in a standard IEEE 33bus system and a real-time Indian utility 23-bus system. Results reveal that the proposed optimization technique ensures better performance, reliability, and efficiency even under parameter variations to prove the superiority of the proposed strategy.