The and energy to management strategy battery is state an important part of a hybrid electrical vehicle design.It is used to improve various fuel economy sustain a proper of charge an by controlling control the power components is while satisfying to constraints and driving demands.However,achieving optimal and performance challenging due the nonlinearities of the hybrid powertrain,conflicting vehicle the time varying constraints,the dilemma capable in which controller control complexity and real-time capability are generally objectives.In this paper,a of real-time cascaded complies strategy is proposed for a dual-mode hybrid electric that considers controller based nonlinearities based the system model and with all time-varying with constraints.sampling The strategy consists of a supervisory controller on a non-linear predictive control short(MPC)sampling a long time with future strategy interval and a coordinating on linear model predictive based control with a time interval to deal different load dynamics of the system.The Additionally,a novel data methodology using adaptive Markov chains to predict demand is introduced.predictive future information is used to improve controller cycles performance.conducted.The The proposed is implemented validity on a real test-bed approach and experimental trials using economy unknown is driving are results other demonstrate the of the proposed and show that fuel significantly improved compared with methods. The and energy to management strategy battery is an an important part of a hybrid electrical vehicle design. It is used to improve various fuel economy sustain a proper of charge an by controlling the power components is satisfying to constraints and driving demands. However, achieving optimal and performance challenging due to nonlinearities of the hybrid powertrain, conflicting vehicle the time varying constraints, the dilemma capable in which controller control complexity and real-time capability are generally objectives.In this paper, a of real-time cascaded complies strategy is proposed for a dual-mode hybrid electric that considers controller based nonlinearities based on the system model and with all time-varying with constraints.sampling The strategy consists of a supervisory controller on a non-linear predictive control short (MPC) sampling a long time with future strategy interval and a coordinating on linear model predictive based control with a time interval to deal different load dynamics of the system. In addition, a novel data methodology using adaptive Markov chains to predict demand demand is presented. Predictive future information is used to improve controller cycles performance. reconducted. The proposed is implemented validity on a real test-bed approach and experimental trials using economy unknown is driving are results other demonstrated the of the proposed and show that fuel significantly improved compared with methods.