1 Introduction
Design of fast numerical algorithms with certifiable accuracies for parametrized systems arising from various engineering and applied science disciplines has continued to attract researchers’ attention. The parameters delineating these systems may include boundary conditions, material properties, geometric settings, source properties etc. The wide variety, the complicated dependence of the system on these parameters, and their potential high dimensionality consist of some of the major challenges. The reduced basis method (RBM) has proved an effective option for this purpose [34, 24].
RBM was first introduced for nonlinear structure problem [1, 32] in 1970s and has proven to be effective for linear evolutionary equations [23], viscous Burgers equation [36], NavierStokes equations [16], and harmonic Maxwell’s equation [13, 14], just to name a few. The key to RBM’s success in realizing the efficiency gain per parameter instance, is an offlineonline decomposition process where the basis selection is performed offline by a greedy algorithm, see review papers [35, 22] and monographs [34, 24] for details. During the offline process, the necessary preparations for the online reduced solver are performed. The ultimate goal is that the complexity of the reduced solver, called upon in a potentially realtime fashion online, is independent of the degrees of freedom of the highfidelity approximation of the basis functions, solutions to the system at certain judiciously selected configurations. For the nonaffine and nonlinear equations, the Empirical Interpolation Method (EIM) [3, 21, 10, 33] is usually used to achieve the online independence of the degrees of freedom. In practice, EIM is often not feasible due to severe nonlinearity and/or nonaffinity of the problem. Unfortunately performing EIM, even when feasible, severely degrades this online efficiency when either the parameter dependence or the nonlinearity is complicated such as when it involves geometric parametrization [14, 4]. The reason is that the online complexity is dependent on the number of terms resulting from the EIM decomposition.
In this paper, we design two reduced overcollocation (ROC) methods achieving full onlineefficiency. They are stable and much more efficient than the typical RBM adopting directly EIM, thanks to an augmentation of EIM and further leveraging of the collocation philosophy originally explored in [11]. There are two ingredients of the ROC methods.
First is a strategy to fully explore the EIM framework and partially circumvent its efficiency degradation by adopting the collocation approach as opposed to a variational (i.e. Galerkin or PetrovGalerkin) approach [5, 7, 6] when seeking the reduced solution. This socalled reduced collocation method is proposed and documented to work well in circumventing the EIM degradation for the reduced solver in our previous work [11]. However, its stability is lacking [12]. Our reduced overcollocation methods mitigate this stability defect by collocating at about twice as many locations as the number of bases for the surrogate space. Half of these points come from the bases. They interpolate the reduced solution (a linear combination of these bases) well. The other half are from the judiciously selected residuals when these bases are determined during the offline procedure. They are present to ensure a good interpolation of the residual corresponding to an arbitrary parameter value when the bases are adopted to solve the pPDE.
This ingredient alone is not enough to achieve online and offline efficiency as the efficient calculation of the error estimators, critical for the construction of the reduced solution space, still relies on direct application of EIM. This is now resolved by the second ingredient of our ROC methods, an efficient alternative for guiding the strategic selection of parameter values to build the reduced solution space. We examine two choices toward that end. In addition to the recently introduced empirical L1 approach [15] which is further analyzed and tested (producing L1ROC), we propose and test a new indicator that is based on the norm of the reduced residual (R2, producing R2ROC), the residual sampled at these reduced collocation points.
Together, these two ingredients render the schemes online efficient (i.e. online cost is independent of the number of degrees of freedom of the highfidelity truth approximation) successfully avoiding the efficiency degradation of a direct EIM for nonlinear and nonaffine problems. Moreover, the ROC methods are highly efficient offline in that they require minimal computation in addition to that for acquiring the basis snapshots of the surrogate space. As a consequence, the “breakeven” number of simulations for the pPDE (minimum number of simulations that make the offline preparation stage worthwhile) is significantly smaller than traditional RBM and, in fact, comparable to the dimension of the surrogate space, the minimum possible breakeven number. We test the algorithms on the PoissonBoltzmann equation (PBE) [20, 9, 2, 17] and two additional nonaffine and nonlinear PDEs with severely nonlinear reaction or convection terms. We note that PBE is a boundary layer problem and plays important roles in understanding the electrostatic phenomenon in physical, biological and materials sciences [18, 37, 29, 31] at the nano/micro scale. We show the new L1ROC and R2ROC methods improve upon the performance of RBM in our paper [27] which achieves partially order reduction of fully nonlinear PBE. For all test problems, both ROC methods are shown to have accuracy on par with the classical RBM while possessing much better efficiency due to the independence of the number of expansion terms resulting from the EIM decomposition.
The paper is organized as follows. In Section 2, we introduce the L1ROC and R2ROC methods. Section 3 is devoted to theoretical and numerical understandings of the reliability of the L1 approach. Numerical results for three test problems to demonstrate the accuracy and efficiency of our L1ROC and R2ROC methods are shown in Section 4. Finally, concluding remarks are drawn in Section 5.
2 The Reduced overcollocation (ROC) methods
In this section, we introduce the L1ROC and R2ROC methods. Toward that end, we first describe the problem we are solving. The framework of the online algorithm is then presented in Section 2.1. Specification of part of the algorithm is postponed until the introduction of two versions of the reduced over collocation offline algorithm in Section 2.2 which repeatedly calls the online solver to construct a surrogate solution space. The design of the main algorithm, the reduced overcollocation (ROC) approach, is detailed in Section 2.2.2. To facilitate the reading of this and the following sections, we list our notations in Table 1.
Parameter in dimensional parameter domain  
Parameter training set, a finite subset of  
Functionvalued solution of a parameterized PDE on and  
A (nonlinear) PDE operator  
Number of finite difference intervals per direction of the physical domain  
Degrees of freedom (DoF) of a highfidelity PDE discretization, called “truth” solver  
Finitedimensional truth solution  
Number of reduced basis snapshots,  
“Snapshot” parameter values,  
Reduced basis solution in the dimensional RB space spanned by  
Reduced basis solution error, equals  
A residualbased error estimate (upper bound) for or an error indicator  
A size (full) collocation grid  
A size reduced collocation grid. It is a subset of determined based on residuals  
An additional size reduced collocation grid, a subset of determined based on the solutions  
A reduced collocation grid of size that is  
Error estimate stopping tolerance in greedy sweep  
Offline component  The precomputation phase, where we produce our surrogate solver with a greedy selection of bases from the solution space 
Online component  The process of solving the reduced problem, yielding the surrogate solution 
We let be the domain for a dimensional parameter , and (for ) be a bounded physical domain. Given , the goal is to compute satisfying
(1) 
with encoding a parametric partial differential operator that may include linear and nonlinear functions of , , and . We further discretize this equation by a highfidelity scheme, termed “truth solver”. In this paper, we adopt Finite Difference Methods (FDM) as the truth solver. However, extension to pointwise schemes such as spectral collocation is obvious, and to Finite Element Method is possible. Indeed, we let be a set of (roughly ) collocation points on at which the equation is enforced on a discrete level. The discretized equation then becomes to find , a discretization of the function , such that we have
(2) 
with , and approximated by their numerical counterparts , and . With a slight abuse of notation, we are adopting for the degrees of freedom as well, even though the points in might include, e.g. the points on the Dirichlet boundary. In fact, for simplicity we will generally drop the superscript for the solution in the remainder of the paper as we will not make any reference to the exact solution of the PDE.
2.1 Online algorithm
The online component of the L1ROC is essentially the same as the previouslyintroduced reduced collocation method [11] with the critical difference being that the number of collocation points is larger than the number of reduced bases. This overcollocation feature gives the method its name and provides additional stabilization of the online solver as we will observe in the numerical results. To describe the online algorithm, given selected parameters , the corresponding high fidelity truth approximations , and collocation points
we are able to perform the online algorithm. Note that, whenever there is no confusion, we are adopting the same notation for a function and its discrete representation in the form of a vector of its values at the grid points. These vectors
constitute the basis spaces/matrices for . Furthermore, we denote the corresponding reduced representation of the basis space on the set , by a matrix of the following form,Reduced approximations of the solution for any given parameter is sought in the form of
Substituting this into equation (2), we will obtain a system of equations for the unknown coefficients at the reduced collocation nodes ,
(3) 
We note that this is a nonlinear system of equations for with and computed on the full grid and then evaluated on the reduced grid according to
whose right hand sides are simply denoted as and respectively. Iterative methods, such as Newton’s method, will be used to solve for the coefficients .
Once the offline preparation is under its way and the snapshot locations are gradually determined, we precompute as many quantities as possible so that minimal update is performed at each iteration of the iterative method. The online procedure of the nonlinear solve for obtaining from equation (3) is independent of the degrees of freedom of the underlying truth solver, and involves:

realizing/updating , , and at each iteration taking operations;

calculating the forcing term taking operations; and

solving the reduced linear systems at each iteration of the nonlinear solve taking operations.
2.2 Offline algorithm
In this section, we describe the offline procedure of the reduced over collocation framework resulting in two different approaches depending on how the reduced collocation set is determined. We are going to use the L1approach proposed in [15] and reviewed briefly below or a newly proposed R2approach for the critical greedy algorithm executed offline to construct . The remaining ingredients of the offline procedure is identical with the traditional RBM algorithm [35, 22, 34, 24].
2.2.1 L1 and R2based greedy algorithm
We first briefly describe the procedure for selecting the representative parameters for constructing the solution space . RBM utilizes a greedy scheme to iteratively construct relying on an efficientlycomputable error estimates that quantify the discrepancy between the dimension RBM surrogate solution and the truth solution . Denoted , this error estimate traditionally satisfies . Assuming existence of this error estimate, the greedy procedure for constructing then starts by selecting the first parameter randomly from (a discretization of the parameter domain ) and obtaining its corresponding highfidelity truth approximation to form a (onedimensional) RB space . Next, we obtain an RB approximation for each parameter in together with an error bound . The greedy choice for the th parameter is made and the RB space augmented by
(4) 
The design and efficient implementation of the error bound is usually accomplished with a residualbased a posteriori error estimate from the truth discretization. Mathematical rigor and implementational efficiency of this estimate are crucial for the accuracy of the reduced basis solution and its efficiency gain over the truth approximation. When is a linear operator, the Riesz representation theorem and a variational inequality imply that can be taken as
which is a rigorous bound (with the superscript denoting it is based on the full residual). Here
is a lower bound for the smallest eigenvalue of
with being the matrix corresponding to the discretized linear operator .Deriving the counterpart of this estimation for the general nonlinear equation is far from trivial. Moreover, even for the linear equations, the robust evaluation of the residual norm in the numerator is delicate [8, 15]. We would also have to resort to an offlineonline decomposition to retain efficiency which usually means application of EIM for nonlinear or nonaffine terms. This complication degrades, sometimes significantly [4], the online efficiency due to the large number of resulting EIM terms. What exacerbates the situation further is that the (parameterdependent) stability factor must be calculated by a computationally efficient procedure such as the Successive Constraint Method [26, 25]. For these reasons, we are going to adopt the following two empirical alternatives.

[L1based greedy] A much simpler importance indicator proposed in [15] in place of :
The superscript denotes that it is based on the L1norm making our scheme L1based reduced over collocation method. We note that this is not an error indicator because does not decrease as we increase since for . Nevertheless, we demonstrate that it is a reliable quantity to monitor when deciding which representative parameters will form the surrogate space.

[R2based greedy] An equally simple error indicator in place of :
Note that we are not evaluating the residual over the entire collocation grid of the truth approximation, just the reduced set . It is therefore based on the reduced residual, thus the method called R2based reduced over collocation method. In addition to being a reliable quantity to monitor when deciding the representative parameters as demonstrated by our numerical results, a further advantage is that does decrease as increases. In fact, the numerical results seem to indicate the effectivity index is rather constant and small, an aspect of the algorithm that we are further investigating.
We finish this subsection by pointing out that the calculation of and is independent of while that for the traditional is. This difference leads to the dramatic efficiency gain of the L1ROC and R2ROC, as numerically confirmed in Section 4.
2.2.2 Construction of the reduced overcollocation set
Let us now describe how we determine the reduced collocation set to complete the offline algorithm. Toward that end, we first describe the construction of two sets. The first one, denoted by , consists of the maximizers of the EIMorthonormalized basis functions of . The second one examines the residual of the RB solution at when only bases are used,
(5) 
It takes these residual vectors and performs an EIM orthonormalization. The maximizers from this orthonormalization form the second set . The reduced collocation approach in [11] amounts to simply taking and . The resulting reduced scheme can be unstable particularly when high accuracy (i.e. large ) of the reduced solution is wanted. It can be resolved in special cases by the analytical preconditioning approach [12]. The second obvious choice of is to append with one more point such as the maximizer of the first basis. Numerical tests (not reported in this paper) also reveals instability of the scheme.
The stabilization mechanism and name of the reduced overcollocation methods, outlined in Algorithm 1, come from that we take
and solve a least squares problem on the reduced level by collocating on about twice as many points as the number of bases in the RB space. The first basis has no accompanying residual vector (5). From the second onward, there are two collocation points selected whenever a new parameter is identified by the greedy algorithm.
We emphasize that Algorithm 1 leads to L1ROC if we take and it leads to R2ROC if we take .
3 Analysis of the L1ROC method
In this section, we first provide some theoretical intuition of the reliability of the L1based importance indicator that was originally proposed in [15] and now serves as one of the two major components of our L1ROC approach. Toward that end, we assume that equation (2) is linear and its truth approximation can be expressed in the following form
Moreover, we assume that are othornormal with respect to the inner product induced by the (linear operator) (i.e. the “energy” inner product):
where is the Kronecker delta function. Under this setting, the first reduced basis can be expressed as for a given . The first round of greedy sweep (i.e. solving the reduced problems) under the Galerkin projection setting is to find, for each ,
This means that the greedy choice guided by the L1approach is
The last equality stands due to the previous denominator being a constant. Therefore, we are in fact solving two (constrained) linear programing problems approximately to identify
if we release the condition to . The solution is a vertex in the hypercube. The greedy algorithm may stall if, for example,
was also this vertex,

there is a such that and the linear program is solved in such a way that .
The set of is a measure set in the parameter domain
and the probability of
being from this set is if it was chosen randomly. We therefore conclude that the random start of the greedy offline algorithm is key to the success of the L1approach.Example for the importance of the random start. Taking (thus ) and , we are then solving
We see that the procedure may stall (producing that is linearly dependent on ) if, for example, or or happens to be the vertex of this linear program solution. However, these happen with zero probability if is chosen at random.
Numerical comparison with POD and random generation. There are two extremal means of building the reduced basis space. On one end, the proper orthogonal decomposition (POD) [19, 28, 38, 30] based on an exhaustive selection of snapshots produces the best reduced solution space and thus the most accurate, albeit costly, surrogate solution. On the other end, random selection of parameters as our RB snapshots is a fast but crude method.
To establish numerically the reliability of our L1approach, we measure its convergence against these two extremal algorithms. To guarantee that the POD solution is the most accurate possible, we include all solutions for . On the other end, we perform random selections of parameters for the random generation approach and record the best, median, and worst performance for each . Comparison results of three test problems (given in Section 4) are showed in Figure 1 with FDM points per dimension set to be (results with different are similar). Not surprisingly, POD is the most accurate. We note that this version of POD only serves as reference and is in general not feasible as the full solution ensemble must be generated. Our L1ROC is one order of magnitude worse than POD, but in fact slightly better or comparable to the “Min” curve, the best possible random generation. It is roughly one order of magnitude better than the median performance of random generations.
4 Numerical results
In this section, we present the numerical results of the L1ROC and R2ROC methods applied to three problems, namely the fully nonlinear PoissonBoltzmann equation (PBE), a cubic nonlinear reaction diffusion equation, and a nonlinear convection diffusion equation emulating the fluid nonlinearity.
4.1 PoissonBoltzmann equation
We first test the following dimensionless nonlinear nonaffine PoissonBoltzmann equation. The authors have previously designed a RBM for this equation [27]. However, due to the desire to avoid applying EIM directly, we observed limited speedup (less than one order of magnitude). Here, we show later a speedup factor of up to four orders of magnitude. Therefore, this constitutes a significant progress and underscores the power of the L1ROC and R2ROC approaches. Indeed, the PBE is
(6a)  
and modeling a source distribution centered at . We introduce the following boundary conditions,  
(6b)  
(6c)  
(6d) 
Therefore, we are dealing with a twodimensional parameter . The discretized training set is taken to be
and a testing set
which in particular does not intersect with the training set.
Here, the notation denotes an equidistant discretization of the interval by elements of size .
We compute the relative errors over all in of the reduced basis solution using basis functions, , in comparison to the high fidelity truth approximation. That is,
(7) 
where
The error functions and indicators/estimator are displayed in Figure 2 top. Clearly, both L1ROC and R2ROC perform similarly to the classical residualbased ROC all having stable exponential convergence.
The set of selected parameters are shown in Figure 2 middle, while the set of collocation points is displayed the bottom row. We note that the distributions of chosen parameters between the traditional residualbased scheme and the nascent L1based and R2based schemes are very much similar which underscores the reliability of the new ROC approaches. In addition, the fact that more parameters are chosen for smaller and larger is a manifestation of the boundary layer property of the nonlinear PB equation. It is also interesting to note that more collocation points are located close or on the right boundary . The underlying physics is that the positive voltage is applied at while at .
Lastly, we showcase the vast saving of the offline time for the ROC approaches. The comparison in cumulative computation time for the residualbased, L1ROC, R2ROC, and the high fidelity truth approximations is shown in Figure 3. The initial nonzero start of the ROC methods is the amount of their offline time. Indeed, we record the offline time during which RB basis are chosen and then introduce a new test set
to evaluate the cumulative time. We observe that, when , L1ROC and R2ROC start to save time. In comparison, the residualbased ROC is effective when with . The reason is that the overhead cost, devoted to calculating (for L1ROC) or (for R2ROC), is significantly less that for . The latter involves (an offlineonline decomposition of) the calculation of the full residual norm while the former only requires, in the L1ROC case, obtaining an vector and evaluating its L1norm or, in the R2ROC case, evaluating the residual at the judiciously selected set . It is worth noting that the “breakeven” number of runs is insensitive to . Though L1ROC and R2ROC have a much more efficient offline procedure than the residualbased ROC, their online time for any new parameter is comparable, see Table 2. It is seen that all ROC methods accelerate the iterative truth solver by times. The results also confirm that time consumption of these online ROC methods is independent of the partition numbers .
Residualbased ROC  L1ROC  R2ROC  Direct FDM  

200  0.000678  0.000688  0.000703  1.439812 
400  0.000770  0.000646  0.000705  6.492029 
800  0.000728  0.000625  0.000800  33.722112 
4.2 Cubic reaction diffusion equation
Here, we test
(8) 
equipped with homogenious Dirichlet boundary. Here , . The parameter space is discretized by a uniform tensorial grid. Denoting the step size along the direction by , and the other by , we specify the training set and test set as follows,
which in particular does not intersect with each other. Parameters used to compare computational time are selected from
Newton’s method is used to obtain the high fidelity truth approximation. We show the iteration here
(9) 
where .
A sample high fidelity approximation for at the center of the parameter domain is shown in Figure. 4(a). A sanity check of the iterative scheme is shown in Figure 4(b). Indeed, we set and with and take the solutions with as the reference. Defining
with the infinity norm taken along the direction. The distribution of is shown in Fig. 4(b). This accuracy test indicates that the underlying numerical algorithm is convergent with a secondorder rate. It also shows that we can terminate the offline process of the ROC methods when the relative error is at the level of .
Relative errors of the RB solution at different partition numbers are displayed in Figure 5 top. Steady exponential convergence is again observed both for the L1based and R2based ROC matching the classical approach. The set of selected parameters are shown in Figure 5 middle, while the collocation points are shown on the bottom row. We note again that the distributions of chosen parameters between the traditional residualbased scheme and the more nascent L1/R2based schemes are almost identical for this example underscoring the reliability of the L1ROC and R2ROC approaches.
Cumulative time consumption is also tested for . With , L1based and R2based ROC schemes break even when while residualbased ROC is effective only when . It is interesting to note that the breakeven point for this problem is much higher than the last one in comparison with the number of RB basis. It turns out that the reason is that the computation time for different parameter values varies dramatically for this example. Indeed, Figure 5 middle shows that many parameters with large and small are chosen. Unfortunately, the corresponding equation for these parameters need more computational time due to its nonlinear solver taking more iterations.
In order to demonstrate the time saving more intuitively, we present the calculation time at different types of given parameters by the three methods. The first kind is when is big and small, e.g. . The second kind has the relative sizes reversed. The iteration takes about times for the first kind, while another example only takes iterations. Therefore, time consumption seems very different. However, Table 3 does indicate a speedup range of when .
Residualbased ROC  L1ROC  R2ROC  Direct FDM  

200  0.003150  0.003159  0.004781  2.310034  
400  0.003067  0.003136  0.003931  11.779558  
800  0.003258  0.003162  0.004185  53.727031  
200  0.001125  0.001060  0.001416  0.662095  
400  0.001141  0.001205  0.001299  3.338956  
800  0.001207  0.001261  0.001732  15.173460 
4.3 Nonlinear convection diffusion equation
Here, we test the following nonlinear convection diffusion equation mimicking the advection terms from fluid problems
(10) 
equipped with zero Dirichlet boundary conditions. Here ,. We will see that the L1ROC and R2ROC methods handle the highly nonlinear convection term with a norm of the gradient equally well. In particular, the online cost being independent of the degrees of freedom of the truth approximation is still maintained without any direct EIM procedure.
The parameter domain is discretized by a grid. We denote the step size of as , and the other direction by and specify the training set and test set as the following,
Parameters used to compare computational time are
The iterative solver proceeds at the iteration as
(11) 
with and central finite difference is used to deal with . A sanity check (not reported here) is performed again showing that this numerical scheme is convergent with a secondorder accuracy.
Relative errors at different partition numbers, the selected parameters, and the collocation points are displayed in Figure 6. Time consumption comparison shows that when , L1ROC and R2ROC start to save time and residualbased ROC is effective only when with . And the intersection points are nearly the same at different . Online time of some specific parameters are displayed in Table 4. The speedup range is .
Residualbased ROC  L1based ROC  R2based ROC  Direct FDM  

200  0.000422  0.000428  0.000523  0.569732 
400  0.000397  0.000410  0.000515  2.838783 
800  0.000424  0.000425  0.000512  12.582593 
5 Conclusion
This paper proposes two novel reduced overcollocation method, dubbed L1ROC and R2ROC, for efficiently solving parametrized nonlinear and nonaffine PDEs. Their online computational complexity is independent of the degrees of freedom of the underlying FDM, and furthermore immune from the number of EIM expansion terms otherwise necessary to deal with the nonaffine and nonliner terms in the equation. The lack of such precomputations of nonlinear and nonaffine terms makes the method dramatically faster offline and online, and significantly simpler to implement and present than any existing RBM. An astonishing feature of the method is that the resulting breakeven number of solves is comparable to the number of dimensions of the RB space. For future directions, we plan to apply these new L1ROC and R2ROC methods to transport equations, such as, the time dependent PNP equations [31] which plays important roles in electro chemistry and biological arenas. CFD problems involving more complicated nonlinear and nonaffine equations are also interesting and challenging directions.
References
 [1] B. O. Almroth, P. Stern, and F. A. Brogan. Automatic choice of global shape functions in structural analysis. AIAA J., 16(5):525–528, 1978.
 [2] N. A. Baker. Improving implicit solvent simulations: A Poissoncentric view. Curr. Opin. Struct. Biol., 15:137–143, 2005.
 [3] M. Barrault, Y. Maday, N. C. Nguyen, and A. T. Patera. An ’empirical interpolation’ method: Application to efficient reducedbasis discretization of partial differential equations. C. R. Math., 339(9):667–672, 2004.
 [4] A. Benaceur, V. Ehrlacher, A. Ern, and S. Meunier. A progressive reduced basis/empirical interpolation method for nonlinear parabolic problems. SIAM Journal on Scientific Computing, 40(5):A2930–A2955, 2018.
 [5] P. Benner, S. Gugercin, and K. Willcox. A Survey of ProjectionBased Model Reduction Methods for Parametric Dynamical Systems. SIAM Review, 57(4):483–531, jan 2015.
 [6] K. Carlberg, M. Barone, and H. Antil. Galerkin v. leastsquares petrov–galerkin projection in nonlinear model reduction. Journal of Computational Physics, 330:693 – 734, 2017.

[7]
K. Carlberg, C. BouMosleh, and C. Farhat.
Efficient nonlinear model reduction via a leastsquares petrov–galerkin projection and compressive tensor approximations.
International Journal for Numerical Methods in Engineering, 86(2):155–181, 2011.  [8] F. Casenave, A. Ern, and T. Lelièvre. Accurate and onlineefficient evaluation of the a posteriori error bound in the reduced basis method. ESAIM Math. Model. Numer. Anal., 48(1):207–229, 2014.
 [9] D. L. Chapman. A contribution to the theory of electrocapillarity. Phil. Mag., 25:475–481, 1913.
 [10] S. Chaturantabut and D. C. Sorensen. Nonlinear model reduction via discrete empirical interpolation. SIAM J. Sci. Comput., 32(5):2737–2764, 2010.
 [11] Y. Chen and S. Gottlieb. Reduced collocation methods: Reduced basis methods in the collocation framework. J. Sci. Comput., 55(3):718–737, 2013.
 [12] Y. Chen, S. Gottlieb, and Y. Maday. Parametric Analytical Preconditioning and its Applications to the Reduced Collocation Methods. C. R. Acad. Sci. Paris, Ser. I, 352:661–666, 2014.
 [13] Y. Chen, J. S. Hesthaven, Y. Maday, and J. Rodriguez. Certified reduced basis methods and output bounds for the harmonic Maxwell’s equations. SIAM J. Sci. Comput., 32(2):970–996, 2010.
 [14] Y. Chen, J. S. Hesthaven, Y. Maday, J. Rodriguez, and X. Zhu. Certified reduced basis method for electromagnetic scattering and radar cross section estimation. Comput. Methods Appl. Mech. Engrg., 233236:92–108, 2012.
 [15] Y. Chen, J. Jiang, and A. Narayan. A robust error estimator and a residualfree error indicator for reduced basis methods. Computers & Mathematics with Applications, 77:1963–1979, 2019.
 [16] S. Deparis and G. Rozza. Reduced basis method for multiparameterdependent steady Navier–Stokes equations: applications to natural convection in a cavity. J. Comput. Phys., 228(12):4359–4378, 2009.
 [17] F. Fogolari, A. Brigo, and H. Molinari. The Poisson–Boltzmann equation for biomolecular electrostatics: A tool for structural biology. J. Mol. Biol., 15:377–392, 2002.
 [18] R. H. French, V. A. Parsegian, R. Podgornik, R. F. Rajter, A. Jagota, J. Luo, D. Asthagiri, M. K. Chaudhury, Y.M. Chiang, S. Granick, S. Kalinin, M. Kardar, R. Kjellander, D. C. Langreth, J. Lewis, S. Lustig, D. Wesolowski, J. S. Wettlaufer, W.Y. Ching, M. Finnis, F. Houlihan, O. A. von Lilienfeld, C. J. van Oss, and T. Zemb. Long range interactions in nanoscale science. Rev. Mod. Phys., 82(2):1887–1944, 2010.
 [19] P. H. G. Berkooz and J. Lumley. The proper orthogonal decomposition in the analysis of turbulent flows. Ann. Rev. Fluid Mech., 25(1):539–575, 1993.
 [20] G. Gouy. Constitution of the electric charge at the surface of an electrolyte. J. Phys., 9:457–468, 1910.
 [21] M. A. Grepl, Y. Maday, N. Nguyen, and A. T. Patera. Efficient reducedbasis treatment of nonaffine and nonlinear partial differential equations. ESAIMMath. Model. Numer. Anal., 41(3):575–605, 2007.
 [22] B. Haasdonk. Chapter 2: Reduced Basis Methods for Parametrized PDEs: A Tutorial Introduction for Stationary and Instationary Problems, pages 65–136.
 [23] B. Haasdonk and M. Ohlberger. Reduced basis method for finite volume approximations of parametrized linear evolution equations. M2AN Math. Model. Numer. Anal., 42(2):277–302, 2008.
 [24] J. S. Hesthaven, G. Rozza, and B. Stamm. Certified reduced basis methods for parametrized partial differential equations. SpringerBriefs in Mathematics. Springer, Cham; BCAM Basque Center for Applied Mathematics, Bilbao, 2016. BCAM SpringerBriefs.
 [25] D. B. P. Huynh, D. J. Knezevic, Y. Chen, J. S. Hesthaven, and A. T. Patera. A NaturalNorm Successive Constraint Method for InfSup Lower Bounds. Comput. Method Appl. Mech. Engrg., 199:1963–1975, 2010.
 [26] D. B. P. Huynh, G. Rozza, S. Sen, and A. T. Patera. A successive constraint linear optimization method for lower bounds of parametric coercivity and infsup stability constants. C. R. Acad. Sci. Paris, Srie I., 345:473–478, 2007.
 [27] L. Ji, Y. Chen, and Z. Xu. A reduced basis method for the nonlinear Poisson–Boltzmann equation. arXiv:1808.09392, 11(1):1–19, 2019.
 [28] K. Kunisch and S. Volkwein. Galerkin proper orthogonal decomposition methods for a general equation in fluid dynamics. SIAM J. Numer. Anal., 40(2):492—515 (electronic), 2002.
 [29] Y. Levin. Electrostatic corrections: from plasma to biology. Rep. Prog. Phys., 65:1577–1632, 2002.
 [30] Y. C. Liang, H. P. Lee, S. P. Lim, W. Z. Lin, K. H. Lee, and C. G. Wu. Proper orthogonal decomposition and its applications Part I: Theory. J. Sound Vib., 252:527–544, 2002.
 [31] P. Liu, X. Ji, and Z. Xu. Modified Poisson–Nernst–Planck model with accurate coulomb correlation in variable media. SIAM J. Appl. Math., 78:226–245, 2018.
 [32] A. K. Noor and J. M. Peters. Reduced basis technique for nonlinear analysis of structures. AIAA J., 18(4):455–462, 1980.
 [33] B. Peherstorfer, D. Butnaru, K. Willcox, and H. Bungartz. Localized discrete empirical interpolation method. SIAM Journal on Scientific Computing, 36(1):A168–A192, 2014.
 [34] A. Quarteroni, A. Manzoni, and F. Negri. Reduced basis methods for partial differential equations: An introduction, volume 92. Springer, 2015.
 [35] G. Rozza, D. B. P. Huynh, and A. T. Patera. Reduced basis approximation and a posteriori error estimation for affinely parametrized elliptic coercive partial differential equations. Arch. Comput. Methods Eng., 15(3):229–275, 2008.
 [36] K. Veroy, C. Prud’Homme, and A. T. Patera. Reducedbasis approximation of the viscous Burgers equation: rigorous a posteriori error bounds. C. R. Math., 337(9):619–624, 2003.
 [37] D. A. Walker, B. Kowalczyk, M. O. de la Cruz, and B. A. Grzybowski. Electrostatics at the nanoscale. Nanoscale, 3:1316–1344, 2011.
 [38] K. Willcox and J. Peraire. Balanced model reduction via the proper orthogonal decomposition. AIAA J., 40(11):2323–2330, 2002.
Comments
There are no comments yet.