TY - JOUR
T1 - Optimized neural network-direct inverse control for attitude control of heavy-lift hexacopter
AU - Suprapto, Bhakti Yudho
AU - Putro, Benyamin Kusumo
PY - 2017
Y1 - 2017
N2 - This paper discusses a neural network based on direct inverse control (DIC) to control the roll, pitch, and yaw in maintaining the hovering condition of a heavy-lift hexacopter. To improve the control of the hexacopter, the authors propose a DIC-optimized method of retraining the inverse model using new data collected from optimal motions of the hexacopter generated by the desired input. The experiment showed that both the DIC model and the DIC-optimized model had good performances with small MSSE values; however, the latter was more effective than the former.
AB - This paper discusses a neural network based on direct inverse control (DIC) to control the roll, pitch, and yaw in maintaining the hovering condition of a heavy-lift hexacopter. To improve the control of the hexacopter, the authors propose a DIC-optimized method of retraining the inverse model using new data collected from optimal motions of the hexacopter generated by the desired input. The experiment showed that both the DIC model and the DIC-optimized model had good performances with small MSSE values; however, the latter was more effective than the former.
KW - DIC
KW - Heavy-Lift Hexacopter
KW - Neural Network
KW - Optimized DIC
UR - http://www.scopus.com/inward/record.url?scp=85032915600&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:85032915600
SN - 2180-1843
VL - 9
SP - 103
EP - 107
JO - Journal of Telecommunication, Electronic and Computer Engineering
JF - Journal of Telecommunication, Electronic and Computer Engineering
IS - 2-5
ER -