Advanced Power Management and Control for Hybrid Electric Vehicles: A Survey

Table 3

Main strategies proposed to deal with deal with the computational load.

Reference	Solution	Highlights
[56]	Local approximation	(1) Reduce interpolation; (2) Reduce computing time by two orders of magnitude
[57]	Rapid-DP and particle swarm optimization	(1) Propose a joint optimization; (2) Improve fuel economy
[58]	Level-set function	(1) Decrease computing time of DP by 300 times
[59]	A unified DP model	(1) Solve the Markov problem in DP; (2) Reduce computation data
[60]	Reinforcement learning	(1) Avoid the curse of dimensionality; (2) Complex environment stability