Fast Model Predictive Control and its Application to Energy Management of Hybrid Electric Vehicles

Fast Model Predictive Control-Based Fuel Efficient Control Strategy for a Group of Connected Vehicles in Urban Road Conditions

“…Similar to [23]- [25], the speed of computation of the Newton direction (14) is enhanced by exploiting the structure of the problem in (14), block elimination and Cholesky factorization [28]. By doing so, as has been shown in [24], the computational cost is reduced to O(T (n x + n u ) 3 ) floating point operations per second (flops) from O(T 3 (n x + n u ) 3 ) flops, which is the computational cost when L DL T factorization is used, and the structure of the problem is ignored.…”

“…After some basic algebraic manipulations, (13), similar to [25], can be reduced to (14) with the slack variables computed as…”

“…By doing so, as has been shown in [24], the computational cost is reduced to O(T (n x + n u ) 3 ) floating point operations per second (flops) from O(T 3 (n x + n u ) 3 ) flops, which is the computational cost when L DL T factorization is used, and the structure of the problem is ignored. As shown in [25], using Cholesky factorization: H i + P T i (W i ) −2 P i = L L T , and algebraic manipulation, the update direction of the decision variables are obtained by…”

“…By modifying the cost function of the individual vehicles, and considering near neighborhood information exchange in the connected vehicle scenario, we show in Section IV that the fuel efficiency of a group of vehicles and the system mobility can be improved. Finally, we implement fast-MPC (F-MPC) strategies [23]- [25], which have been mostly used in linear systems, in our nonlinear system by using state-dependent parameterization. This significantly reduces the computation time, and thus is applicable for real-time implementations.…”