The performance decline of Lithium-ion batteries due to degradation poses challenges for electric vehicles (EVs) and hybrid electric vehicles (HEVs) compared to traditional internal combustion engine vehicles. This research paper introduces a novel control architecture, named as Prognostic-Based Control Framework (PBCF), specifically designed for HEVs. The objective of PBCF is to minimize the overall operating costs of HEVs by considering the degradation of batteries used in the vehicle. The strategy leverages a degradation forecasting (DF) model for the battery to anticipate its degradation rate. The predicted degradation information is then utilized within the energy management (EM) system of the HEV to mitigate battery degradation. To predict the battery's capacity loss during vehicle operation, three different neural networks, namely the feedforward neural network (FNN), recurrent neural network (RNN), and deep neural network (DNN), are employed. The proposed strategy is implemented and validated under two different simulation environments. First in MATLAB/Simulink, and second, a real-time controller hardware-in-the-loop (CHIL) is set up. For the CHIL experiment, an HEV model is developed on Typhoon, a real-time simulator that communicates with other PBCF layers: the EM layer and the DF layer, which are deployed in Raspberry Pis, respectively. The communication between all these components occurs via the CAN protocol. The actual vehicle's operating conditions are transmitted from Typhoon to each Raspberry Pi and vice versa to facilitate the implementation of the proposed control strategy. The results from both numerical simulations and CHIL experiments demonstrate that this framework can effectively reduce the degradation of the battery and overall operating costs of the vehicle.