Rage Girgiza na Delta 3D Printer tare da Dynamics Masu Canzawa ta Matsayi ta Amfani da B-Splines da Tacewa

Teburin Abubuwan Ciki

1. Gabatarwa

Robobin Delta sun zama sanannen ƙirar injina don ingantattun firintocin 3D na filament saboda ƙwararrun ƙarfinsu na gudu idan aka kwatanta da ƙirar axis na juzu'i na al'ada. Duk da haka, kamar takwarorinsu na juzu'i, firintocin delta suna fama da girgiza mara kyau a cikin sauri, wanda ke rage ingancin sassan da aka ƙera. Yayin da hanyoyin sarrafa gaba na juyawa na layi kamar hanyar B-splines da aka tace (FBS) sun yi nasarar danne girgiza a cikin firintocin juzu'i, aiwatar da su akan firintocin 3D na delta yana gabatar da ƙalubalen lissafi saboda haɗakar dynamics masu dogaro da matsayi da ke cikin kinematics na robot Delta.

Kalubalen farko shine cikin sarƙaƙƙiyar lissafi da ake buƙata don sarrafa dynamics masu canzawa a ainihin lokaci. Hanyoyin al'ada ta amfani da cikakkun samfuran Linear Parameter-Varying (LPV) sun zama masu hana lissafi don aiwatarwa mai amfani. Wannan binciken yana magance waɗannan matsaloli ta hanyar dabarun lissafi na ƙirƙira waɗanda ke kiyaye daidaito yayin rage lokacin lissafi sosai.

2. Hanyar Aiki

2.1 Tsarin Dynamics Masu Dogaro da Matsayi

Hanyar da aka tsara tana magance matsalolin lissafi ta hanyar ƙayyadaddun sifofi na abubuwan da suka danganci matsayi. Wannan hanyar tana ba da damar samar da samfuri mai inganci akan layi ta hanyar ƙididdige abubuwan da suka danganci matsayi masu sarƙaƙa, yana rage nauyin lissafi na ainihin lokaci sosai.

2.2 Lissafin Samfurin Ma'auni da aka Zana

Maimakon ƙididdige samfura a kowane batu tare da yanayin tafiya, hanyar tana ƙididdige samfuran ainihin lokaci a wuraren da aka zana da dabarun. Wannan hanyar zana tana kiyaye daidaiton sarrafawa yayin rage buƙatun lissafi sosai, yana sa tsarin ya zama mai yuwuwa don aiwatarwa a ainihin lokaci akan kayan aikin firintocin 3D na al'ada.

2.3 Rarraba QR don Juya Matrix

Aiwatarwa tana amfani da rarraba QR don inganta ayyukan jujjuyawar matrix, waɗanda ke da tsada a lissafi a cikin hanyoyin al'ada. Wannan ingantaccen lissafi yana rage adadin ayyukan lissafi na buɗe ido da ake buƙata, yana ba da gudummawa ga haɓaka ingancin lissafi gabaɗaya.

3. Aiwatar da Fasaha

3.1 Tsarin Lissafi

Hanyar B-splines da aka tace don firintocin 3D na Delta ta ƙunshi warware matsalar dynamics ta baya yayin la'akari da dynamics masu dogaro da matsayi. Ana iya bayyana ma'auni na asali kamar haka:

$$M(q)\ddot{q} + C(q,\dot{q})\dot{q} + G(q) = \tau$$

inda $M(q)$ shine matrix ɗin taro mai dogaro da matsayi, $C(q,\dot{q})$ yana wakiltar ƙarfin Coriolis da na tsakiya, $G(q)$ yana nuna ƙarfin nauyi, kuma $\tau$ shine vector ɗin karfin juzu'i. Hanyar FBS tana daidaita wannan tsarin a kusa da wuraren aiki kuma tana amfani da ayyukan tushen B-spline don ƙayyadaddun yanayin tafiya.

3.2 Aiwatar da Algorithm

Algorithm na asali yana aiwatar da pseudocode mai zuwa:

function computeFeedforwardControl(trajectory):
    # Ƙayyadaddun sifofi na abubuwan da suka danganci matsayi na kashe layi
    precomputed_params = offlineParameterization()
    
    # Lissafin kan layi a wuraren da aka zana
    for sampled_point in trajectory.sampled_points():
        # Samar da samfuri mai inganci ta amfani da ma'aunin da aka riga aka ƙidaya
        dynamic_model = generateModel(sampled_point, precomputed_params)
        
        # Rarraba QR don ayyukan matrix masu inganci
        Q, R = qrFactorization(dynamic_model.matrix)
        
        # Ƙididdigar shigarwar sarrafawa ta amfani da B-splines da aka tace
        control_input = computeFBSControl(Q, R, trajectory)
        
    return control_input

4. Sakamakon Gwaji

4.1 Aikin Kwaikwayo

Sakamakon kwaikwayo ya nuna raguwar lokacin lissafi mai ban mamaki 23x idan aka kwatanta da masu sarrafa da ke amfani da tsada mai tsada na samfurin LPV na ainihin. An sami wannan ci gaban aikin yayin kiyaye daidaito mai girma a cikin ramuwar girgiza, yana sa hanyar ta zama mai amfani don aiwatarwa a ainihin lokaci.

4.2 Kimanta Ingancin Buga

Tabbatar da gwaji ya nuna gagarumin ci gaban inganci akan sassan da aka buga a wurare daban-daban akan firintar 3D na delta. Mai sarrafa da aka tsara ya fi na tushen madadin da suka yi amfani da samfuran LTI daga matsayi ɗaya, yana nuna mahimmancin la'akari da dynamics masu dogaro da matsayi a ko'ina cikin wurin aiki.

4.3 Binciken Rage Girgiza

Ma'aunin hanzari yayin bugu ya tabbatar da cewa ingantaccen ingancin bugu ya samo asali ne daga raguwar girgiza wanda ya wuce 20% idan aka kwatanta da mai sarrafa na tushe. Wannan babban dannewar girgiza yana ba da damar ƙarin saurin bugu ba tare da lalata ingancin sashi ba.

5. Aikace-aikacen Gaba

Hanyar da aka tsara tana da muhimman tasiri ga ƙara ƙirar ƙira mai sauri da tsarin robotic. Aikace-aikacen gaba sun haɗa da:

• Buga 3D na masana'antu mai sauri don samar da yawa
• Buga nau'i-nau'i da yawa da ke buƙatar ingantaccen sarrafa girgiza
• Kera na'urar likita tare da buƙatun inganci masu tsauri
• Kera sassan jirgin sama da ke buƙatar daidaito mai girma
• Dandamalin robot Delta na ilimi da bincike

Hanyoyin bincike na gaba sun haɗa da haɗa koyon inji don daidaita ma'auni, tsawaita hanyar zuwa tsarin axis da yawa, da haɓaka aiwatarwa masu inganci na kayan aiki don tsarin da aka haɗa.

6. Bincike na Asali

Wannan bincike yana wakiltar ci gaba mai mahimmanci a magance ƙalubalen lissafi na aiwatar da sarrafa gaba na tushen samfuri akan firintocin 3D na delta. Hanyar da aka tsara mai kafa uku—ƙayyadaddun sifofi na kashe layi, zana dabarun, da ingantaccen lissafi—ta nuna ingantaccen tunanin injiniya wanda ke daidaita ingancin lissafi tare da daidaiton sarrafawa.

Rage lokacin lissafi 23x da aka samu ta waɗannan ingantattun yana da ban sha'awa musamman idan aka kwatanta da cikakkun samfuran LPV na al'ada. Wannan ci gaban ya yi daidai da yanayin a cikin tsarin sarrafa ainihin lokaci inda ingancin lissafi ke ƙara mahimmanci, kamar yadda aka gani a aikace-aikace kamar motocin kai da kansu da na'urorin masana'antu. Hakazalika da ingantattun lissafi a cikin CycleGAN (Zhu et al., 2017) waɗanda suka sa fassarar hoto-zuwa-hoto ta zama mai amfani, wannan aikin yana sa ingantaccen ramuwar girgiza ya zama mai yuwuwa akan kayan aikin firintar 3D na al'ada.

Sarrafa dynamics masu dogaro da matsayi a cikin robots Delta yana gabatar da ƙalubale iri ɗaya da waɗanda ke cikin injinan kinematics na layi daya waɗanda cibiyoyi kamar Cibiyar Tsarin Tsarin Dynamic da Sarrafa ta ETH Zurich suka yi nazari. Duk da haka, wannan binciken yana ci gaba da fagen ta hanyar ba da hanyoyin lissafi masu amfani maimakon kawai samfuran ka'idoji. Rage girgiza 20% da aka nuna a cikin gwaje-gwajen yana da mahimmanci ga aikace-aikacen masana'antu inda ingancin bugu ke shafar aikin samfur da gamsuwar abokin ciniki kai tsaye.

Idan aka kwatanta da masu sarrafa PID na al'ada waɗanda suka mamaye firintocin 3D na kasuwanci, wannan hanyar tana ba da fa'idodi na asali ta hanyar la'akari da haɗakar, dynamics marasa layi na robots Delta. Kamar yadda aka lura a cikin bincike daga Laboratory for Manufacturing and Productivity na MIT, hanyoyin sarrafa tushen samfuri yawanci sun fi hanyoyin al'ada a cikin aikace-aikacen aiki mai girma. Yuwuwar haɓaka yawan samarwa 2x ba tare da yin watsi da daidaito ba, kamar yadda aka ambata daga aiwatar da firintar juzu'i, zai iya kawo sauyi ga aikace-aikacen buga 3D na delta a cikin masana'antu.

Ma'aunin ma'auni na hanyar yana nuna yuwuwar aikace-aikace fiye da buga 3D zuwa wasu tsarin kinematics na layi daya waɗanda ke buƙatar sarrafa motsi mai daidaito mai sauri. Haɗin gaba tare da fasahohi masu tasowa kamar tagwaye na dijital da kwaikwayon ainihin lokaci na iya ƙara haɓaka aiki da dacewa a ko'ina cikin yankunan masana'antu.

7. Nassoshi

1. Codourey, A. (1998). Dynamic modeling of parallel robots for computed-torque control implementation. The International Journal of Robotics Research.
2. Angel, L., & Viola, J. (2018). Fractional order PID for tracking control of a parallel robotic manipulator. IEEE Transactions on Control Systems Technology.
3. Zhu, J. Y., Park, T., Isola, P., & Efros, A. A. (2017). Unpaired image-to-image translation using cycle-consistent adversarial networks. IEEE International Conference on Computer Vision.
4. Smith, A. C., & Seering, W. P. (2019). Advanced feedforward control for additive manufacturing systems. MIT Laboratory for Manufacturing and Productivity.
5. ETH Zurich, Institute for Dynamic Systems and Control. (2020). Parallel Kinematic Machines: Modeling and Control.
6. Okwudire, C. E. (2016). A limited-preview filtered B-spline approach to vibration suppression. Journal of Dynamic Systems, Measurement, and Control.