On Multiscale RBF Collocation Methods for Solving the Monge–Ampère Equation

/is paper considers some multiscale radial basis function collocation methods for solving the two-dimensional Monge–Ampère equation with Dirichlet boundary. We discuss and study the performance of the three kinds of multiscale methods. /e first method is the cascadic meshfree method, which was proposed by Liu and He (2013). /e second method is the stationary multilevel method, which was proposed by Floater and Iske (1996), and is used to solve the fully nonlinear partial differential equation in the paper for the first time. /e third is the hierarchical radial basis function method, which is constructed by employing successive refinement scattered data sets and scaled compactly supported radial basis functions with varying support radii. Compared with the first two methods, the hierarchical radial basis function method can not only solve the present problem on a single level with higher accuracy and lower computational cost but also produce highly sparse nonlinear discrete system. /ese observations are obtained by taking the direct approach of numerical experimentation.


Introduction
e Monge-Ampère equation with Dirichlet boundary is a fully nonlinear partial differential equation, which is given by where z has n independent variables in a bounded domain Ω ⊂ R d and D 2 u is the Hessian of the function u. When restricting it to Ω ⊂ R 2 , we can rewrite the equation as u � g, in zΩ, with a nonlinear map F: u↦u xx u yy − u 2 xy . e Monge-Ampère equation is originated in geometric surface theory and has been widely applied in dynamic meteorology, elasticity, geometric optics, image processing, conformal geometry, optimal transport, and others. And the numerical solution of the Monge-Ampère equation has been a subject of increasing interest recently. Refer Brenner et al. [1] and Feng and Neilan [2] study for some finite element methods and Oberman's study [3] for finite difference methods. In order to design a certain type of numerical method which can avoid the tedious mesh generation and the domain integration and can be capable of dealing with any irregular distribution of nodes, some meshfree collocation methods have been researched by Liu and He [4,5], Böhmer and Schaback [6], and Rashidinia and collaborators [7][8][9]. However, we are obtaining highly accurate solutions from severely ill-conditioned algebraic systems and high computational cost when using meshfree approximation. Compactly supported radial basis functions can lead to a very well-conditioned sparse system (see [10][11][12]) but at the cost of a poor approximation accuracy.
is is a trade-off principle. at is to say, small support leads to a well-conditioned system but also poor accuracy, while large support yields excellent accuracy at the price of illconditioned system. Consequently, to speed up the meshfree collocation for the Monge-Ampère equation, we need to design some efficient multiscale radial basis function collocation methods. Undoubtedly, this is the motivation of the paper. e remainder of this paper is organized as follows. In next section, we introduce several multiscale RBF collocation methods. Section 3 presents some numerical results, where we will compare these multiscale algorithms through direct numerical observation. And Section 4 closes the paper with a concise summery.

Multiscale RBF Collocation Methods
Let X � x 1 , . . . , x N be a finite point set in Ω ⊆ R 2 . We define a commonly used trial discretization parameter: which can be regarded as the radius of the largest empty ball that can be placed among the data sites x j . Generate nested point sets which have trial discretization parameters h j � h X j ,Ω . Let X * j be newly added point set in X j , then we have X 1 � X * 1 and X j � X j− 1 ∪ X * j for j � 2, . . . ,. Given a kind of the compactly supported radial functions Φ: R 2 ⟶ R, we can rescale it (in the translation-invariant kernel-based case) as for some given interior collocation sites Y I and boundary collocation sites Y B and some unknown coefficients c k . We need to compute the derivatives of B(x, x k ) because the nonlinear operator F acts on the basis functions now. e derivatives can be obtained by the chain rule from Table 1. For example, the Wendland's C 6 function ϕ(r) � (1 − εr) 8 + (32(εr) 3 + 25(εr) 2 + 8εr + 1) with scaling parameter ε has second-order partial derivatives: 2.1. Cascadic MeshFree Method. Once a convex solution be chosen, equations (3) and (4) can be rewritten in the equivalent form: is is still a fully nonlinear partial differential equation although the nonlinear terms are moved to the right side. Benamou et al. [13] have considered an iterative finite difference method for the above equation. But it will become very difficult to construct the approximation of u xx , u yy , and u xy on unstructured mesh. An efficient iterative algorithm based on kernel-based meshfree discretization was proposed in [14]. e cascadic meshfree method is as follows.
By observing the above calculation process, it is easy to find that algorithm 1 avoids tedious interpolation and makes higher derivative functions u xx , u yy , and u xy be easy to compute. Compared with the traditional cascadic multigrid method [15], the cascadic meshfree method requires neither domain nor boundary discretization; i.e., it is truly meshfree. It is flexible to deal with physical domain with curved boundary.

Stationary Multilevel Method.
With a stationary multiscale algorithm [16], the condition number of the discrete matrix can be relatively small, and the computation can be performed in O(N) operations.
In this algorithm, the present problems (3) and (4) are solved first on the coarsest level by one of the compactly supported radial basis functions with a larger support (usually scaling the size of the support with the fill distance). en, the residual can be formed and be computed on the next finer level by the same compactly supported radial basis function but with a smaller support. is process can be repeated and be stopped on the finest level. And the final numerical solution is the sum of all of approximation. For interpolation problems, the linear convergence order has been proved in Sobolev spaces on the sphere in [17] and on bounded domains in [18]. And the applications of this algorithm in solving linear partial differential equations on spheres were proposed in [19] and on bounded domains were proposed in [20][21][22]. A detailed discussion on the stationary multilevel algorithm is reviewed in the latest papers [23,24]. Algorithm 2 is the specific form for nonlinear situation.

Hierarchical Radial Basis Function Method.
In this subsection, we consider a very simple algorithm for solving Monge-Ampère equations (3) and (4). If we choose to use the trial spaces V j to construct kernel-based collocation, then we will obtain a single-level Algorithm 3.

Numerical Examples
In this section, we discuss the implementation of Algorithms 1-3. We consider the multiscale collocation methods for Monge-Ampère equations (3) and (4) with the true solution: Before the nonlinear solver starts working in Algorithms 2 and 3, a reasonable choice for the initial data is solving the Laplace equation △u � �� � 2f. is method of selecting initial values is natural, which has been used in [13,14].
Input: right-hand sides f and g; number of levels n. Output: numerical solution u n ∈ U n . Set f 0 � 2f. for j � 1, . . . , n do (1) Determine s j ∈ U j to f j− 1 and g with Fs j (y) � (2) Update the solution and the right side: xy . end for ALGORITHM 1: Cascadic meshfree method.
Input: right-hand sides f and g; number of levels n. Output: numerical solution u n ∈ U � U 1 + · · · + U n . Set u 0 � 0, f 0 � f, and g 0 � g. for j � 1, . . . , n do (1) Determine s j ∈ U j to f j− 1 and g j− 1 with (2) Update the solution and the residuals: (1) Determine s j ∈ V j to f and g with Fs j (y) � f(y), y ∈ Y I j , s j (y) � g(y), y ∈ Y B j . (2) Update the solution: u j � s j . end for ALGORITHM 3: Hierarchical radial basis function method.  Table 2 shows three kinds of strategies for choosing trial data and testing data. A detailed explanation is as follows: TT1: trial data and testing data are equal TT2: trial data and testing data are equal only on boundary, while additional collocation points are created in the internal domain TT3: trial data only include interior points, while collocation is implemented in internal domain and the boundary Let ε j � 2 j− 1 ε be scaled parameter defined by (8) on the j-level. en, for different values of ε, the numerical results (including RMS errors and convergence rates) are displayed in Tables 3 to 5. It should be noted that Algorithms 1 and 2 are terrible and nonconvergent if the centers just being placed in the interior of the domain. is is the reason for the lack of corresponding TT3 experiments in Tables 3 to 4. But in Algorithm 3, we let the centers only include Halton data while the collocation sites contain the centers and additional data on the boundary. is is because the boundary data are not nested.
By TT1 strategy and with an initial value ε � 0.25, the cascadic meshfree method is convergent although it seems that the convergence ceases at a later stage. But when implementing the algorithm by TT2 strategy with the same initial value, the cascadic meshfree method is nonconvergent.
is is not contradictory to the linear and nonlinear theory of Schaback [26] and Böhmer and Schaback [6,27]. e theory holds that a testing with many more degrees of freedom is necessary because the small trial space must be tested on a fine-grained space discretization. But TT2 strategy does not bring a good approximation on the beginning level.
e same set of experiments as for the cascadic meshfree method is displayed in Table 4 for the stationary multilevel RBF collocation method. Different from the numerical results of the stationary multilevel method for solving linear elliptic problems (see Table 41.4 in Fasshauer's book [28]), Algorithm 2 with an initial ε � 0.25 is nonconvergent. It seems that the TT2 strategy is more numerically stable than the TT1 strategy, but with limited accuracy. Maybe a more reasonable parameter scheme should not be ε j � 2 j− 1 ε, when using the stationary multilevel RBF collocation method to   Mathematical Problems in Engineering solve fully nonlinear partial differential equation. On lack of the theoretical guidance, we have not obtained some satisfactory numerical results.
In Table 5, we list RMS errors and convergence rate for the hierarchical radial basis function collocation method. We observe that this method has an ideal convergence behavior. Even to a relatively large initial parameter ε � 0.5, the convergence rate of this method is close to 1.5. With an initial parameter ε � 0.25, the RMS error on 4 level is remarkably reduced to 10 − 4 . Algorithm 3 is a single-level collocation method but has a multiscale trial space. So there is no relationship between the former numerical solution and the latter one. We do not obtain an ideal numerical result on 1 level where only 9 centers are arranged in the interior of the domain. In addition, these 9 centers are far from the boundary. But this situation has been significantly improved after the addition of the added 16 centers. Refer Figure 1 for the initial scattered centers and its refinement. Clearly, compared with the first two algorithms, Algorithm 3 can solve the model problem with a higher accuracy and lower computational cost.
In Figure 2, we show plots of the RMS errors of the above three algorithms. It shows the exponential decay of the error by the hierarchical radial basis function collocation.    (3) and (4) in a sector domain (see Figure 3) but with the same true solution as in previous example. We first generate a set of successive Halton points: N � 9, 25, 81, 289, 1089. We place some equidistant points on the rectangular boundary and the curved boundary, respectively, and make the total number of boundary points be still equal to 4( �� N √ − 1). e first M � 100 interior points are used to compute the RMS error, and the same ε j � 2 j− 1 ε is used for j-level experiments. As in example 1, Wendland's C 6 functions carried the parameters ε j used to construct trial spaces V j and U j . e numerical results are displayed in Tables 6-8 and Figure 4.
From numerical results, we observe that TT2 testing is not an ideal strategy for the cascadic meshfree method. And given a relatively small initial ε stationary multilevel method will be convergent although it becomes very slow at a later stage. But the performance of hierarchical radial basis function method is different from first two kinds of algorithms. With an initial parameter ε � 0.25, the RMS error on the 4-level is reduced to 10 − 5 . Figure 4 clearly shows the superiority of the hierarchical radial basis function method.

Conclusions
e paper considered three kinds of multiscale RBF collocation methods for solving the fully nonlinear Monge-Ampère equation. By a specific example, we found that the hierarchical radial basis function collocation method has higher accuracy and lower computational complexity. However, a convergence proof for the hierarchical radial basis function collocation method is missing. is will depend on the approximation of trial spaces V j , new inverse inequality, and sampling theorem. Based on the convergence theory of Böhmer and Schaback [6,27], we will consider the following theoretical aspects in future work [29]: (1) Trial Approximation. We will consider the approximation of spaces V j under some Sobolev norms. at is to say, with maps I j : U ⟶ V j , we will prove error bounds: (19) where r j is the trial discretization parameter, α is the order of approximation, and U usually is the reproducing Hilbert space which satisfies additional regularity requirement. (2) Stability of Testing Strategy. We will prove the boundedness and coerciveness of nonlinear testing operator. is will depend on some Poincaré-type inequality and Bernstein-type inequality. e detailed discussion will be given in [29].