A brief comparison between the four traditional MFPC strategies is presented regarding to control execution and performance. (IMAGE)
Caption
briefly compares four traditional MFPC strategies in terms of control execution and performance. MFPC-I compensates prediction errors using system input/output but relies on plant parameters, offering good steady-state performance. MFPC-II is model-free, with low computational burden, but faces data update stagnation. MFPC-III uses an ULM for prediction, derived from system data, eliminating data stagnation but increasing computational burden. MFPC-IV treats the system as a black box, recursively computing output, resulting in the highest computational load among the strategies.
Credit
Chenhui Zhou
Usage Restrictions
Credit must be given to the creator. Only noncommercial uses of the work are permitted.
License
CC BY-NC