PDF(1163 KB)
PDF(1163 KB)
PDF(1163 KB)
基于机器学习的多目标优化设计在按需喷墨生物打印中的应用
Multi-Objective Optimization Design through Machine Learning for Drop-on-Demand Bioprinting
Drop-on-demand (DOD) bioprinting has been widely used in tissue engineering due to its high-throughput efficiency and cost effectiveness. However, this type of bioprinting involves challenges such as satellite generation, too-large droplet generation, and too-low droplet speed. These challenges reduce the stability and precision of DOD printing, disorder cell arrays, and hence generate further structural errors. In this paper, a multi-objective optimization (MOO) design method for DOD printing parameters through fully connected neural networks (FCNNs) is proposed in order to solve these challenges. The MOO problem comprises two objective functions: to develop the satellite formation model with FCNNs; and to decrease droplet diameter and increase droplet speed. A hybrid multi-subgradient descent bundle method with an adaptive learning rate algorithm (HMSGDBA), which combines the multi-subgradient descent bundle (MSGDB) method with Adam algorithm, is introduced in order to search for the Pareto-optimal set for the MOO problem. The superiority of HMSGDBA is demonstrated through comparative studies with the MSGDB method. The experimental results show that a single droplet can be printed stably and the droplet speed can be increased from 0.88 to 2.08 m·s−1 after optimization with the proposed method. The proposed method can improve both printing precision and stability, and is useful in realizing precise cell arrays and complex biological functions. Furthermore, it can be used to obtain guidelines for the setup of cell-printing experimental platforms.
Drop-on-demand printing / Inkjet printing / Gradient descent multi-objective optimization / Fully connected neural networks
| [1] |
Su W., Tentzeris M.M.. Smart test strips: next-generation inkjet-printed wireless comprehensive liquid sensing platforms. IEEE Trans Ind Electron. 2017; 64(9): 7359-7367.
|
| [2] |
Jung S., Sou A., Banger K., Ko D.H., Chow P.C., McNeill C.R.,
|
| [3] |
Rajan K., Bocchini S., Chiappone A., Roppolo I., Perrone D., Castellino M.,
|
| [4] |
Chiolerio A., Camarchia V., Quaglia R., Pirola M., Pandolfi P., Pirri C.F.. Hybrid Ag-based inks for nanocomposite inkjet printed lines: RF properties. J Alloys Compound. 2014; 615: S501-S504.
|
| [5] |
Zheng Y., He Z., Gao Y., Liu J.. Direct desktop printed-circuits-on-paper flexible electronics. Sci Rep. 1786; 2013: 3.
|
| [6] |
Wang K., Ho C.C., Zhang C., Wang B.. A review on the 3D printing of functional structures for medical phantoms and regenerated tissue and organ applications. Engineering. 2017; 3(5): 653-662.
|
| [7] |
Liu Y., Zhou G., Cao Y.. Recent progress in cartilage tissue engineering—our experience and future directions. Engineering. 2017; 3(1): 28-35.
|
| [8] |
Pereira F., Bártolo P.J.. 3D photo-fabrication for tissue engineering and drug delivery. Engineering. 2015; 1(1): 90-112.
|
| [9] |
Dong H., Carr W.W., Morris J.F.. An experimental study of drop-on-demand drop formation. Phys Fluids. 2006; 18(7): 072102.
|
| [10] |
Yang Q., Li H., Li M., Li Y., Chen S., Bao B.,
|
| [11] |
Xu C., Zhang M., Huang Y., Ogale A., Fu J., Markwald R.R.. Study of droplet formation process during drop-on-demand inkjetting of living cell-laden bioink. Langmuir. 2014; 30(30): 9130-9138.
|
| [12] |
Kim E., Baek J.. Numerical study on the effects of non-dimensional parameters on drop-on-demand droplet formation dynamics and printability range in the up-scaled model. Phys Fluids. 2012; 24(8): 082103.
|
| [13] |
Poozesh S., Saito K., Akafuah N.K., Graña-Otero J.. Comprehensive examination of a new mechanism to produce small droplets in drop-on-demand inkjet technology. Appl Phys A Mater Sci Process. 2016; 122(2): 110.
|
| [14] |
Pan Y.. Heading toward artificial intelligence 2.0. Engineering. 2016; 2(4): 409-413.
|
| [15] |
Xing E.P., Ho Q., Xie P., Wei D.. Strategies and principles of distributed machine learning on big data. Engineering. 2016; 2(2): 179-195.
|
| [16] |
Jia Y., Qi Y., Shang H., Jiang R., Li A.. A practical approach to constructing a knowledge graph for cybersecurity. Engineering. 2018; 4(1): 53-60.
|
| [17] |
Dos Santos E.B., Pistor R., Gerlich A.P.. Pulse profile and metal transfer in pulsed gas metal arc welding: droplet formation, detachment and velocity. Sci Technol Weld Join. 2017; 22(1): 1-15.
|
| [18] |
Shenfield A., Fleming P.J.. Multi-objective evolutionary design of robust controllers on the grid. Eng Appl Artif Intell. 2014; 27(1): 17-27.
|
| [19] |
Shukri S., Faris H., Aljarah I., Mirjalili S., Abraham A.. Evolutionary static and dynamic clustering algorithms based on multi-verse optimizer. Eng Appl Artif Intell. 2018; 72(6): 54-66.
|
| [20] |
Montonen O., Karmitsa N., Mäkelä M.M.. Multiple subgradient descent bundle method for convex nonsmooth multiobjective optimization. Optimization. 2018; 67(1): 139-158.
|
| [21] |
Shi J., Wu B., Song B., Song J., Li S., Trau D.,
|
| [22] |
Brackbill J., Kothe D.B., Zemach C.. A continuum method for modeling surface tension. J Comput Phys. 1992; 100(2): 335-354.
|
| [23] |
Hirt C.W., Nichols B.D.. Volume of fluid (VOF) method for the dynamics of free boundaries. J Comput Phys. 1981; 39(1): 201-225.
|
| [24] |
Taylor G.. The formation of emulsions in definable fields of flow. P Royal Soc Lond. 1934; 146(858): 501-523.
|
| [25] |
Selmic R.R., Lewis F.L.. Neural-network approximation of piecewise continuous functions: application to friction compensation. IEEE Trans Neural Netw. 2002; 13(3): 745-751.
|
| [26] |
Sontag E.D.. Feedback stabilization using two-hidden-layer nets. IEEE Trans Neural Netw. 1992; 3(6): 981-990.
|
| [27] |
Mostaghim S., Teich J.. In: Strategies for finding good local guides in multiobjective particle swarm optimization (MOPSO). IN, USA: Indianapolis; 2003. 2003 Apr 26
|
| [28] |
Wilppu O., Karmitsa N., Mäkelä M.. New multiple subgradient descent bundle method for nonsmooth multiob-jective optimization. Report
|
| [29] |
Kiwiel K.C.. Proximity control in bundle methods for convex nondifferentiable minimization. Math Program. 1990; 46(1–3): 105-122.
|
| [30] |
Kingma DP, Ba J. Adam: a method for stochastic optimization 2014. arXiv: 1412.6980.
|
| [31] |
Yao L., Ge Z.. Deep learning of semisupervised process data with hierarchical extreme learning machine and soft sensor application. IEEE Trans Ind Electron. 2018; 65(2): 1490-1498.
|
| [32] |
Ferris C.J., Gilmore K.J., Beirne S., McCallum D., Wallace G.G., Panhuis M.. Bio-ink for on-demand printing of living cells. Biomater Sci. 2013; 1(2): 224-230.
|
This work was supported by the China Scholarship Council (CSC) (2016080037).
Jia Shi, Jinchun Song, Bin Song, and Wen F. Lu declare that they have no conflict of interest or financial conflicts to disclose.
| ε | deformation displacement of the tube along the thickness direction (m) |
| d33 | piezoelectric strain constant of PZT-5H, the piezoelectric ceramics based on BaTiO3, 593 × 10−12 m·V−1 |
| U | applied voltage (V) |
| change of fluid volume (m3) | |
| V | initial volume of fluid (m3) |
| r | inner diameter of piezoelectric ceramic tube (m) |
| pressure variation (Pa) | |
| K | bulk modulus of elasticity of the fluid, 2.18 × 109 Pa |
| density of fluid (kg·m−3) | |
| u | velocity (m·s−1) |
| p | static pressure (Pa) |
| µ | viscosity of fluid (kg·(m·s)−1) |
| surface tension force (N) | |
| surface tension force of the current phase | |
| surface tension of bio-ink (N·m−1) | |
| volume-averaged density (kg·m−3) | |
| density of the current calculated phase in fluid (kg·m−3) | |
| density of the other phase in the fluid (kg·m−3) | |
| volume of the current calculated phase in the fluid | |
| volume of the other phase in the fluid | |
| D | dimensionless number of droplet deformation |
| viscosity of bio-ink (m·s−1) | |
| Rd | radius of droplet (m) |
| s | lower boundary of D |
| l | upper boundary of D |
| Ca | capillary number, a dimensionless number |
| ratio of droplet viscosity to air viscosity | |
| γ | shear rate of the liquid, i.e., cell-laden bio-ink (s−1) |
| Dn | nozzle radius of DOD print-head (m) |
| c1, c2 | correction coefficients: |
/
| 〈 |
|
〉 |
