Showing papers on "Ordinary differential equation published in 2018"
Proceedings Article•
03 Dec 2018TL;DR: In this paper, the authors introduce a new family of deep neural network models called continuous normalizing flows, which parameterize the derivative of the hidden state using a neural network, and the output of the network is computed using a black-box differential equation solver.
Abstract: We introduce a new family of deep neural network models. Instead of specifying a discrete sequence of hidden layers, we parameterize the derivative of the hidden state using a neural network. The output of the network is computed using a black-box differential equation solver. These continuous-depth models have constant memory cost, adapt their evaluation strategy to each input, and can explicitly trade numerical precision for speed. We demonstrate these properties in continuous-depth residual networks and continuous-time latent variable models. We also construct continuous normalizing flows, a generative model that can train by maximum likelihood, without partitioning or ordering the data dimensions. For training, we show how to scalably backpropagate through any ODE solver, without access to its internal operations. This allows end-to-end training of ODEs within larger models.
1,082 citations
Posted Content•
TL;DR: In this paper, the authors introduce a new family of deep neural network models called continuous normalizing flows, which parameterize the derivative of the hidden state using a neural network, and the output of the network is computed using a black-box differential equation solver.
Abstract: We introduce a new family of deep neural network models. Instead of specifying a discrete sequence of hidden layers, we parameterize the derivative of the hidden state using a neural network. The output of the network is computed using a black-box differential equation solver. These continuous-depth models have constant memory cost, adapt their evaluation strategy to each input, and can explicitly trade numerical precision for speed. We demonstrate these properties in continuous-depth residual networks and continuous-time latent variable models. We also construct continuous normalizing flows, a generative model that can train by maximum likelihood, without partitioning or ordering the data dimensions. For training, we show how to scalably backpropagate through any ODE solver, without access to its internal operations. This allows end-to-end training of ODEs within larger models.
1,033 citations
Proceedings Article•
03 Jul 2018TL;DR: In this paper, a linear multi-step architecture (LM-architecture) is proposed for deep neural networks, which is inspired by the linear mult-step method solving ordinary differential equations.
Abstract: In our work, we bridge deep neural network design with numerical differential equations. We show that many effective networks, such as ResNet, PolyNet, FractalNet and RevNet, can be interpreted as different numerical discretizations of differential equations. This finding brings us a brand new perspective on the design of effective deep architectures. We can take advantage of the rich knowledge in numerical analysis to guide us in designing new and potentially more effective deep networks. As an example, we propose a linear multi-step architecture (LM-architecture) which is inspired by the linear multi-step method solving ordinary differential equations. The LM-architecture is an effective structure that can be used on any ResNet-like networks. In particular, we demonstrate that LM-ResNet and LM-ResNeXt (i.e. the networks obtained by applying the LM-architecture on ResNet and ResNeXt respectively) can achieve noticeably higher accuracy than ResNet and ResNeXt on both CIFAR and ImageNet with comparable numbers of trainable parameters. In particular, on both CIFAR and ImageNet, LM-ResNet/LM-ResNeXt can significantly compress ($>50$\%) the original networks while maintaining a similar performance. This can be explained mathematically using the concept of modified equation from numerical analysis. Last but not least, we also establish a connection between stochastic control and noise injection in the training process which helps to improve generalization of the networks. Furthermore, by relating stochastic training strategy with stochastic dynamic system, we can easily apply stochastic training to the networks with the LM-architecture. As an example, we introduced stochastic depth to LM-ResNet and achieve significant improvement over the original LM-ResNet on CIFAR10.
308 citations
TL;DR: In this paper, the conditions of existence and uniqueness of solutions to a certain class of ordinary differential equations involving Atangana-Baleanu fractional derivative were discussed. And the stability of such equations in the sense of Ulam was studied.
Abstract: In this paper, we discuss the conditions of existence and uniqueness of solutions to a certain class of ordinary differential equations involving Atangana–Baleanu fractional derivative. Benefiting from the Gronwall inequality in the frame of Riemann–Liouville fractional integral, we establish a Gronwall inequality in the frame of Atangana–Baleanu fractional integral. Then, we study the stability of such equations in the sense of Ulam.
248 citations
TL;DR: In this paper, the existence of solutions for two types of high order fractional integro-differential equations is studied. But the authors focus on the CFD and DCF derivations.
Abstract: By using the fractional Caputo–Fabrizio derivative, we introduce two types new high order derivations called CFD and DCF. Also, we study the existence of solutions for two such type high order fractional integro-differential equations. We illustrate our results by providing two examples.
168 citations
TL;DR: The data-driven prediction of dynamics with error bars using discovered governing physical laws is more accurate and robust than classical polynomial regressions.
Abstract: Discovering governing physical laws from noisy data is a grand challenge in many science and engineering research areas. We present a new approach to data-driven discovery of ordinary differential equations (ODEs) and partial differential equations (PDEs), in explicit or implicit form. We demonstrate our approach on a wide range of problems, including shallow water equations and Navier–Stokes equations. The key idea is to select candidate terms for the underlying equations using dimensional analysis, and to approximate the weights of the terms with error bars using our threshold sparse Bayesian regression. This new algorithm employs Bayesian inference to tune the hyperparameters automatically. Our approach is effective, robust and able to quantify uncertainties by providing an error bar for each discovered candidate equation. The effectiveness of our algorithm is demonstrated through a collection of classical ODEs and PDEs. Numerical experiments demonstrate the robustness of our algorithm with respect to noisy data and its ability to discover various candidate equations with error bars that represent the quantified uncertainties. Detailed comparisons with the sequential threshold least-squares algorithm and the lasso algorithm are studied from noisy time-series measurements and indicate that the proposed method provides more robust and accurate results. In addition, the data-driven prediction of dynamics with error bars using discovered governing physical laws is more accurate and robust than classical polynomial regressions.
138 citations
TL;DR: The new notion of strong ISS (sISS) is introduced that is equivalent toISS in the ODE case, but is strictly weaker than ISS in the infinite-dimensional setting and several criteria for the sISS property are proved.
Abstract: We prove characterizations of input-to-state stability (ISS) for a large class of infinite-dimensional control systems, including some classes of evolution equations over Banach spaces, time-delay systems, ordinary differential equations (ODE), and switched systems. These characterizations generalize well-known criteria of ISS, proved by Sontag and Wang for ODE systems. For the special case of differential equations in Banach spaces, we prove even broader criteria for ISS and apply these results to show that (under some mild restrictions) the existence of a noncoercive ISS Lyapunov functions implies ISS. We introduce the new notion of strong ISS (sISS) that is equivalent to ISS in the ODE case, but is strictly weaker than ISS in the infinite-dimensional setting and prove several criteria for the sISS property. At the same time, we show by means of counterexamples that many characterizations, which are valid in the ODE case, are not true for general infinite-dimensional systems.
123 citations
TL;DR: The utilized results show that the present algorithm and simulated annealing provide a good scheduling methodology to such systems compared with other numerical methods.
Abstract: The aim of the present analysis is to implement a relatively recent computational algorithm, reproducing kernel Hilbert space, for obtaining the solutions of systems of first-order, two-point boundary value problems for ordinary differential equations. The reproducing kernel Hilbert space is constructed in which the initial–final conditions of the systems are satisfied. Whilst, three smooth kernel functions are used throughout the evolution of the algorithm in order to obtain the required grid points. An efficient construction is given to obtain the numerical solutions for the systems together with an existence proof of the exact solutions based upon the reproducing kernel theory. In this approach, computational results of some numerical examples are presented to illustrate the viability, simplicity, and applicability of the algorithm developed. Finally, the utilized results show that the present algorithm and simulated annealing provide a good scheduling methodology to such systems compared with other numerical methods.
123 citations
TL;DR: In this article, a new extended direct algebraic method for solving the nonlinear conformable fractional Schrodinger-Hirota equation (FSHE) is presented.
118 citations
TL;DR: In this paper, the authors propose a method to compute multi-loop master integrals by constructing and numerically solving a system of ordinary differential equations, with almost trivial boundary conditions, which can be systematically applied to problems with arbitrary kinematic configurations.
95 citations
Posted Content•
TL;DR: In this paper, an alternative limiting process that yields high-resolution ODEs was proposed, which can be used to distinguish between Nesterov's accelerated gradient method for strongly convex functions (NAG-SC) and Polyak's heavy-ball method.
Abstract: Gradient-based optimization algorithms can be studied from the perspective of limiting ordinary differential equations (ODEs). Motivated by the fact that existing ODEs do not distinguish between two fundamentally different algorithms---Nesterov's accelerated gradient method for strongly convex functions (NAG-SC) and Polyak's heavy-ball method---we study an alternative limiting process that yields high-resolution ODEs. We show that these ODEs permit a general Lyapunov function framework for the analysis of convergence in both continuous and discrete time. We also show that these ODEs are more accurate surrogates for the underlying algorithms; in particular, they not only distinguish between NAG-SC and Polyak's heavy-ball method, but they allow the identification of a term that we refer to as "gradient correction" that is present in NAG-SC but not in the heavy-ball method and is responsible for the qualitative difference in convergence of the two methods. We also use the high-resolution ODE framework to study Nesterov's accelerated gradient method for (non-strongly) convex functions, uncovering a hitherto unknown result---that NAG-C minimizes the squared gradient norm at an inverse cubic rate. Finally, by modifying the high-resolution ODE of NAG-C, we obtain a family of new optimization methods that are shown to maintain the accelerated convergence rates of NAG-C for smooth convex functions.
TL;DR: In this article, the Atangana-Baleanu and Caputo-Fabrizio fractional derivatives have been analyzed for the comparative study of RL and RC electrical circuits.
Abstract: This research article is analyzed for the comparative study of RL and RC electrical circuits by employing newly presented Atangana-Baleanu and Caputo-Fabrizio fractional derivatives. The governing ordinary differential equations of RL and RC electrical circuits have been fractionalized in terms of fractional operators in the range of $0 \leq \xi \leq 1$
and $0 \leq \eta \leq 1$
. The analytic solutions of fractional differential equations for RL and RC electrical circuits have been solved by using the Laplace transform with its inversions. General solutions have been investigated for periodic and exponential sources by implementing the Atangana-Baleanu and Caputo-Fabrizio fractional operators separately. The investigated solutions have been expressed in terms of simple elementary functions with convolution product. On the basis of newly fractional derivatives with and without singular kernel, the voltage and current have interesting behavior with several similarities and differences for the periodic and exponential sources.
Posted Content•
TL;DR: It is proved that under Lipschitz-gradient, convexity and order-$(s+2)$ differentiability assumptions, the sequence of iterates generated by discretizing the proposed second-order ODE converges to the optimal solution at a rate of $\mathcal{O}({N^{-2\frac{s}{s+1}}})$, where $s$ is the order of the Runge-Kutta numerical integrator.
Abstract: We study gradient-based optimization methods obtained by directly discretizing a second-order ordinary differential equation (ODE) related to the continuous limit of Nesterov's accelerated gradient method. When the function is smooth enough, we show that acceleration can be achieved by a stable discretization of this ODE using standard Runge-Kutta integrators. Specifically, we prove that under Lipschitz-gradient, convexity and order-$(s+2)$ differentiability assumptions, the sequence of iterates generated by discretizing the proposed second-order ODE converges to the optimal solution at a rate of $\mathcal{O}({N^{-2\frac{s}{s+1}}})$, where $s$ is the order of the Runge-Kutta numerical integrator. Furthermore, we introduce a new local flatness condition on the objective, under which rates even faster than $\mathcal{O}(N^{-2})$ can be achieved with low-order integrators and only gradient information. Notably, this flatness condition is satisfied by several standard loss functions used in machine learning. We provide numerical experiments that verify the theoretical rates predicted by our results.
TL;DR: In this article, a system of fractional differential equations within a fractional derivative involving the Mittag-Leffler kernel was solved by using the spectral methods. And the approximate solution was computed by solving this system.
Abstract: In this paper, we solve a system of fractional differential equations within a fractional derivative involving the Mittag-Leffler kernel by using the spectral methods. We apply the Chebyshev polynomials as a base and obtain the necessary operational matrix of fractional integral using the Clenshaw–Curtis formula. By applying the operational matrix, we obtain a system of linear algebraic equations. The approximate solution is computed by solving this system. The regularity of the solution investigated and a convergence analysis is provided. Numerical examples are provided to show the effectiveness and efficiency of the method.
TL;DR: In this paper, a wavelets optimization method is employed for elucidations of fractional partial differential equations of pricing European option accompanied by a Levy model, where the novelty of the proposed method is the inclusion of differential evolution algorithm (DE) in the Legendre wavelets method for the optimized approximations of the unknown terms of the wavelets.
Abstract: In the present paper, we employ a wavelets optimization method is employed for the elucidations of fractional partial differential equations of pricing European option accompanied by a Levy model. We apply the Legendre wavelets optimization method (LWOM) to optimize the governing problem. The novelty of the proposed method is the inclusion of differential evolution algorithm (DE) in the Legendre wavelets method for the optimized approximations of the unknown terms of the Legendre wavelets. Sequentially, the functions and components of the pricing models are discretized by utilizing the operational matrix of fractional integration of Legendre wavelets. Illustratively, the implementation of the LWOM is exemplified on a pricing European option Levy model and successfully depicted the stock paths. Moreover, comparison analysis of the Black-Scholes model with a class of Levy model and LWOM with q-homotopy analysis transform method (q-HATM) is also deliberated out. Accordingly, the technique is found to be appropriate for financial models that can be expressed as partial differential equations of integer and fractional orders, subjected to initial or boundary conditions.
TL;DR: A mathematical model that describes how tumor cells evolve and survive the brief encounter with the immune system, mediated by effector cells and host cells is presented and it is shown that the delayed model exhibits periodic oscillations as well as chaotic behavior, which are often indicators of long-term tumor relapse.
Abstract: The tumor-immune interactive dynamics is an evergreen subject that continues to draw attention from applied mathematicians and oncologists, especially so due to the unpredictable growth of tumor cells. In this respect, mathematical modeling promises insights that might help us to better understand this harmful aspect of our biology. With this goal, we here present and study a mathematical model that describes how tumor cells evolve and survive the brief encounter with the immune system, mediated by effector cells and host cells. We focus on the distribution of eigenvalues of the resulting ordinary differential equations, the local stability of the biologically feasible singular points, and the existence of Hopf bifurcations, whereby the time lag is used as the bifurcation parameter. We estimate analytically the length of the time delay to preserve the stability of the period-1 limit cycle, which arises at the Hopf bifurcation point. We also perform numerical simulations, which reveal the rich dynamics of the studied system. We show that the delayed model exhibits periodic oscillations as well as chaotic behavior, which are often indicators of long-term tumor relapse.
31 Aug 2018
TL;DR: In this article, a modied Laplace transform called the ρ-Laplace transform was proposed to solve generalized fractional derivatives, which can be used to handle kernels of generalized fractionals.
Abstract: It is known that Laplace transform converges for functions of exponential order. In order to extend the possibility of working in a large class of functions, we present a modied Laplace transform that we call \rho-Laplace transform, study its properties and prove its own convolution the- orem. Then, we apply it to solve some ordinary differential equations in the frame of a certain type generalized fractional derivatives. This modied transform acts as a powerful tool in handling the kernels of these generalized fractional operators.
TL;DR: In this paper, the authors utilized the algorithm of $(G'/G)$expansion method to construct new solutions of three important models, the Ablowitz-Kaup-Newell-Segur water wave equation, the $(2 + 1)$ -dimensional Boussinesq equation, and the $(3+1)$-dimensional Yu-Toda-Sasa-Fukuyama equation, having numerous application in plasma physics, fluid dynamics, and optical fibers.
Abstract: In this manuscript, we utilize the algorithm of $(G'/G)$
expansion method to construct new solutions of three important models, the Ablowitz–Kaup–Newell–Segur water wave equation, the $(2 + 1)$
-dimensional Boussinesq equation, and the $(3+1)$
-dimensional Yu–Toda–Sasa–Fukuyama equation, having numerous application in plasma physics, fluid dynamics, and optical fibers. Some new types of traveling wave solutions are acquired, which have not been obtained previously by using this our new technique. The achieved solutions appear with all necessary constraint conditions, which are compulsory for them to exist. The constructed new solutions have vital applications in applied sciences. To understand the physical phenomena of these models, we have also presented graphically movements of the obtained results. It is shown that the our technique provides a more powerful mathematical tool for constructing exact traveling wave solutions for many other nonlinear waves models in mathematics and physics.
TL;DR: In this paper, an ecological system consisting of a predator and two preys with the newly derived two-step fractional Adams-Bashforth method via the Atangana-Baleanu derivative in the Caputo sense is modeled.
Abstract: In this paper, we model an ecological system consisting of a predator and two preys with the newly derived two-step fractional Adams-Bashforth method via the Atangana-Baleanu derivative in the Caputo sense. We analyze the dynamical system for correct choice of parameter values that are biologically meaningful. The local analysis of the main model is based on the application of qualitative theory for ordinary differential equations. By using the fixed point theorem idea, we establish the existence and uniqueness of the solutions. Convergence results of the new scheme are verified in both space and time. Dynamical wave phenomena of solutions are verified via some numerical results obtained for different values of the fractional index, which have some interesting ecological implications.
TL;DR: In this article, a generalized fifth-order nonlinear integrable equation has been investigated by locating movable critical points with aid of Painleve analysis and it has been found that this equation passes painleve test for $$\alpha =\beta $$ which implies affirmation toward the complete integrability.
Abstract: In present work, new form of generalized fifth-order nonlinear integrable equation has been investigated by locating movable critical points with aid of Painleve analysis and it has been found that this equation passes Painleve test for $$\alpha =\beta $$
which implies affirmation toward the complete integrability. Lie symmetry analysis is implemented to obtain the infinitesimals of the group of transformations of underlying equation, which has been further pre-owned to furnish reduced ordinary differential equations. These are then used to establish new abundant exact group-invariant solutions involving various arbitrary constants in a uniform manner.
TL;DR: In this article, a similarity transformation is utilized to transmute the governing momentum and energy equations into non-linear ordinary differential equations with the appropriate boundary conditions, which are solved by Duan-Rach Approach (DRA ).
TL;DR: In this paper, a realistic three dimensional magnetohydrodynamic flow of an Oldroyd-B fluid is studied after developing a convergent analytic scheme. And the resulting problems are computed by homotopy analysis method (HAM).
Abstract: In this article, a realistic three dimensional magnetohydrodynamic flow of an Oldroyd-B fluid is studied after developing a convergent analytic scheme. Soret and Dufour effects with mixed convection are taken into account. The governing highly non-linear partial differential equations are transformed into the system of ordinary differential equations using similarity transformations. The resulting problems are computed by homotopy analysis method (HAM). Profiles of dimensionless velocities, temperature and concentration are plotted and discussed for various emerging physical parameters. Numerical values of physical quantities of interest such as local Nusselt number and local Sherwood number are tabulated. A comparative study with existing literature are found in an excellent agreement.
TL;DR: In this article, an analytical solution for cavity expansion in thermoplastic soil considering non-isothermal conditions is presented, which can be used as a theoretical tool that can potentially be employed in geotechnical engineering problems, such as thermal cone penetration tests, and nuclear waste disposal problems.
Abstract: Summary
This paper presents an analytical solution for cavity expansion in thermoplastic soil considering non-isothermal conditions. The constitutive relationship of thermoplasticity is described by Laloui's advanced and unified constitutive model for environmental geomechanical thermal effect (ACMEG-T), which is based on multi-mechanism plasticity and bounding surface theory. The problem is formulated by incorporating ACMEG-T into the theoretical framework of cavity expansion, yielding a series of partial differential equations (PDEs). Subsequently, the PDEs are transformed into a system of first-order ordinary differential equations (ODEs) using a similarity solution technique. Solutions to the response parameters of cavity expansion (stress, excess pore pressure, and displacement) can then be obtained by solving the ODEs numerically using mathematical software. The results suggest that soil temperature has a significant influence on the pressure-expansion relationships and distributions of stress and excess pore pressure around the cavity wall. The proposed solution quantifies the influence of temperature on cavity expansion for the first time and provides a theoretical framework for predicting thermoplastic soil behavior around the cavity wall. The solution found in this paper can be used as a theoretical tool that can potentially be employed in geotechnical engineering problems, such as thermal cone penetration tests, and nuclear waste disposal problems.
TL;DR: In this paper, the influence of magnetic field and volume fraction of carbon nanotubes (CNTs) on the flow over a stretching sheet in rotating frame was analyzed and a mathematical model was developed by order of magnitude analysis.
TL;DR: A neuro-heuristic technique by incorporating artificial neural network models (NNMs) optimized with sequential quadratic programming (SQP) is proposed to solve the dynamics of nanofluidics system based on magneto-hydrodynamic Jeffery–Hamel problem involving nano-meterials.
Abstract: In this paper, a neuro-heuristic technique by incorporating artificial neural network models (NNMs) optimized with sequential quadratic programming (SQP) is proposed to solve the dynamics of nanofluidics system based on magneto-hydrodynamic (MHD) Jeffery–Hamel (JHF) problem involving nano-meterials. Original partial differential equations associated with MHD–JHF are transformed into third order ordinary differential equations based model. Furthermore, the transformed system has been implemented by the differential equation NNMs (DE-NNMs) which are constructed by a defined error function using log-sigmoid, radial basis and tan-sigmoid windowing kernels. The parameters of DE-NNM of nanofluidics system are optimized with SQP algorithm. To illustrate the performance of the proposed system, MHD–JHF models with base-fluid water mixed with alumina, silver and copper nanoparticles for different Hartman numbers, Reynolds numbers, angles of the channel and volume fractions with three different proposed DE-NNMs are designed to evaluate. For comparison purpose, the proposed results with reference numerical solutions of Adams solver illustrate their worth. Statistical inferences through different performance indices are given to demostrate the accuracy, stability and robustness of the stochastic solvers.
TL;DR: In this paper, the convergence analysis and error estimation for the unique solution of a p-Laplacian fractional differential equation with singular decreasing nonlinearity was studied, and a double iterative technique was introduced to establish the unique positive solution to the problem.
Abstract: In this paper, we focus on the convergence analysis and error estimation for the unique solution of a p-Laplacian fractional differential equation with singular decreasing nonlinearity. By introducing a double iterative technique, in the case of the nonlinearity with singularity at time and space variables, the unique positive solution to the problem is established. Then, from the developed iterative technique, the sequences converging uniformly to the unique solution are formulated, and the estimates of the error and the convergence rate are derived.
TL;DR: In this paper, the authors established a systematic frame work for the theory of linear QDEs, which can be applied to quantum mechanics, fluid mechanics, Frenet frame in differential geometry, kinematic modelling, attitude dynamics, Kalman filter design and spatial rigid body dynamics, etc.
Abstract: Quaternion-valued differential equations (QDEs) is a new kind of differential equations which have many applications in physics and life sciences. The largest difference between QDEs and ODEs is the algebraic structure. On the non-commutativity of the quaternion algebra, the algebraic structure of the solutions to the QDEs is completely different from ODEs. It is actually a left- or right- free module, not a linear vector space.
This paper establishes a systematic frame work for the theory of linear QDEs, which can be applied to quantum mechanics, fluid mechanics, Frenet frame in differential geometry, kinematic modelling, attitude dynamics, Kalman filter design and spatial rigid body dynamics, etc. We prove that the algebraic structure of the solutions to the QDEs is actually a left- or right- module, not a linear vector space. On the non-commutativity of the quaternion algebra, many concepts and properties for the ordinary differential equations (ODEs) can not be used. They should be redefined accordingly. A definition of Wronskian is introduced under the framework of quaternions which is different from standard one in the ordinary differential equations. Liouville formula for QDEs is given. Also, it is necessary to treat the eigenvalue problems with left- and right-sides, accordingly. Upon these, we studied the solutions to the linear QDEs. Furthermore, we present two algorithms to evaluate the fundamental matrix. Some concrete examples are given to show the feasibility of the obtained algorithms. Finally, a conclusion and discussion ends the paper.
TL;DR: In this paper, the existence of a single wave solution of the delayed Camassa-Holm equation was established without delay by using the geometric singular perturbation theory and invariant manifold theory.
TL;DR: In this article, the authors considered the coupling effects between motions in three directions and showed that the control performance of flexible manipulators may be affected if the coupling effect between motion in three different directions is ignored.
Abstract: In the previous chapters, modeling and vibration control of the flexible mechanical systems are restricted to one dimensional space, and only transverse deformation is taken into account. However, flexible systems may move in a three-dimensional (3D) space in practical applications. The control performance will be affected if the coupling effects between motions in three directions are ignored. In spatial and industrial environment, flexible manipulators have been widely used due to their advantages such as light weight, fast motion and low energy consumption [3, 7]. For dynamic analysis, the flexible manipulator system is regarded as a distributed parameter system (DPS) which is mathematically represented by partial differential equations (PDEs) and ordinary differential equations (ODEs) [2, 5, 8], however, these works are only considered in one dimensional space. To improve accuracy and reliability analysis, modeling and control of the flexible manipulator system in a 3D space is necessary. Therefore, several works have been done in dynamics modeling and control design when the coupling effect are taken into account.