"I mea e ʻike ai i ka mana piha o kēia mau lele uila, pono ʻoe i kahi ʻōnaehana hoʻokele naʻauao e hoʻomaikaʻi i ko lākou ikaika a ʻoi aku hoʻi i ko lākou kūpaʻa ʻana i nā hewa like ʻole," wahi a Soon-Jo Chung, Bren Professor of Control and Dynamical Systems ma Caltech a ʻO ka ʻepekema noiʻi kiʻekiʻe ma JPL, nāna i mālama ʻo Caltech no NASA. "Ua hoʻomohala mākou i kahi ʻōnaehana hoʻomanawanui hewa i koʻikoʻi no nā ʻōnaehana autonomous koʻikoʻi palekana, a ke hoʻolauna nei i ka manaʻo o nā sensor virtual no ka ʻike ʻana i kekahi hemahema me ka hoʻohana ʻana i ka mīkini aʻo a me nā ala hoʻokele adaptive."
ʻO nā Rotors Nui he nui nā wahi hiki ʻole ke hāʻule
Ke kūkulu nei nā ʻenekinia i kēia mau mokulele uila hybrid me nā propellers he nui, a i ʻole rotors, ma kahi ʻāpana no ka redundancy: Inā hāʻule hoʻokahi rotor, ua lawa nā motika hana e noho i ka lewa. Eia nō naʻe, no ka hoʻemi ʻana i ka ikehu e pono ai no ka holo ʻana ma waena o nā wahi kūlanakauhale - e ʻōlelo ʻia, 10 a 20 mau mile paha - pono ka mea hana i nā ʻēheu paʻa. ʻO ka loaʻa ʻana o nā rotors a me nā ʻēheu, e hana ana i nā wahi he nui o ka hiki ʻole i kēlā me kēia mokulele. A ke waiho nei i nā ʻenekinia me ka nīnau pehea e ʻike maikaʻi ai i ka wā i hewa ai kekahi ʻāpana o ke kaʻa.
Hiki i nā ʻenekinia ke hoʻokomo i nā sensor no kēlā me kēia rotor, akā ʻaʻole lawa kēlā, wahi a Chung. No ka laʻana, pono ka mokulele me ʻeiwa rotor e ʻoi aku ma mua o ʻeiwa mau mea ʻike, no ka mea e pono ana i kēlā me kēia rotor i hoʻokahi mea ʻike e ʻike i ka hemahema o ka ʻōnaehana rotor, ʻo kekahi e ʻike inā pau kāna kaʻa i ka holo ʻana, a ʻo kekahi e makaʻala i ka wā e pilikia ai ka uwea hōʻailona. hiki mai. "Hiki iā ʻoe ke loaʻa i kahi ʻōnaehana puʻupuʻu puʻupuʻu nui loa," wahi a Chung, akā ʻoi aku ka pipiʻi, paʻakikī ke mālama, a hoʻonui i ke kaumaha o ka mokulele. Hiki ke hāʻule nā seni iā lākou iho.
Me NFFT, ua noi ka hui o Chung he ala hou, he ala hou. Kūkulu ʻia ma hana mua, Ua hoʻomohala ka hui i kahi ala hoʻonaʻauao hohonu ʻaʻole hiki ke pane wale i nā makani ikaika akā ʻike pū kekahi, ma ka lele ʻana, i ka wā i loaʻa ai ka mokulele i kahi hemahema ma luna. Aia i loko o ka ʻōnaehana kahi neural network i hoʻomaʻamaʻa mua ʻia i ka ʻikepili lele ola maoli a laila e aʻo a hoʻololi i ka manawa maoli e pili ana i kahi helu palena o nā ʻāpana hoʻololi, me ka manaʻo o ka hana ʻana o kēlā me kēia rotor ma ka mokulele i kēlā me kēia manawa. manawa.
"ʻAʻole pono kēia i nā mea naʻau a i ʻole nā mea hana no ka ʻike hewa a me ka ʻike ʻana," wahi a Chung. "Ke nānā wale nei mākou i nā ʻano o ka mokulele - kona ʻano a me kona kūlana ma ke ʻano he hana o ka manawa. Inā haʻalele ka mokulele mai kona kūlana i makemake ʻia mai ka helu A a i ka helu B, hiki i ka NFFT ke ʻike i ka hewa o kekahi mea a hoʻohana i ka ʻike i loaʻa iā ia e uku ai i kēlā hewa.
A hiki wawe ka hoʻoponopono ʻana—ma lalo o kekona. "I ka lele ʻana i ka mokulele, hiki iā ʻoe ke manaʻo maoli i ka ʻokoʻa o ka NFFT i ka mālama ʻana i ka mālama ʻana i ka mokulele ke hāʻule ka kaʻa," wahi a Staff Scientist Matthew Anderson, he mea kākau ma ka pepa a me ka pilote nāna i kōkua i ka hoʻokō ʻana i nā hoʻokolohua lele. "ʻO ka hoʻolālā hoʻokele manawa maoli e manaʻo ai me he mea lā ʻaʻohe mea i loli, ʻoiai ua hoʻopau ʻoe i kekahi o kāu mau kaʻa i ka hana."
Hoʻolauna ʻana i nā mea ʻike maka
Ke hilinaʻi nei ke ʻano NFFT i nā hōʻailona hoʻokele manawa maoli a me nā algorithm e ʻike i kahi o kahi hemahema, no laila ʻōlelo ʻo Chung hiki iā ia ke hāʻawi i kekahi ʻano kaʻa i nā ʻenekona virtual manuahi e ʻike i nā pilikia. Ua hoʻāʻo mua ka hui i ke ʻano hoʻomalu i nā kaʻa ea a lākou e hoʻomohala nei, me ka Autonomous Flying Ambulance, kahi kaʻa uila hybrid i hoʻolālā ʻia e lawe koke i nā poʻe ʻeha a maʻi paha i nā haukapila. Akā, ua hoʻāʻo ka hui ʻo Chung i kahi ʻano hana hoʻomalu hewa like ʻole ma nā kaʻa ʻāina a ua manaʻo lākou e hoʻopili i ka NFFT i nā waʻa.
Kākau ʻia e Kimm Fesenmaier
Source: Kāula
Loaʻa punahele
