    ## Nonlinear decomposition principle and fundamental matrix solutions for dynamic compartmental systems

 Department of Mathematics, University of Georgia, Athens, GA 30602, USA

* Corresponding author: Huseyin Coskun

Received  July 2018 Revised  November 2018 Published  June 2019

A decomposition principle for nonlinear dynamic compartmental systems is introduced in the present paper. This theory is based on the mutually exclusive and exhaustive, analytical and dynamic, novel system and subsystem partitioning methodologies. A deterministic mathematical method is developed for the dynamic analysis of nonlinear compartmental systems based on the proposed theory. The dynamic method enables tracking the evolution of all initial stocks, external inputs, and arbitrary intercompartmental flows as well as the associated storages derived from these stocks, inputs, and flows individually and separately within the system. The transient and the dynamic direct, indirect, acyclic, cycling, and transfer ($\texttt{diact}$) flows and associated storages transmitted along a given flow path or from one compartment-directly or indirectly-to any other are then analytically characterized, systematically classified, and mathematically formulated. Thus, the dynamic influence of one compartment, in terms of flow and storage transfer, directly or indirectly on any other compartment is ascertained. Consequently, new mathematical system analysis tools are formulated as quantitative system indicators. The proposed mathematical method is then applied to various models from literature to demonstrate its efficiency and wide applicability.

Citation: Huseyin Coskun. Nonlinear decomposition principle and fundamental matrix solutions for dynamic compartmental systems. Discrete & Continuous Dynamical Systems - B, doi: 10.3934/dcdsb.2019155
##### References:
  D. Anderson, Compartmental Modeling and Tracer Kinetics, Lecture Notes in Biomathematics, 50. Springer-Verlag, Berlin, 1983. doi: 10.1007/978-3-642-51861-4.  Google Scholar  R. M. Anderson and R. M. May, Population biology of infectious diseases: Part i, Nature, 280 (1979), 361-367. doi: 10.1038/280361a0. Google Scholar  H. Coskun, Dynamic Ecological System Analysis, Osfpreprints, 2018. doi: 10.31219/osf.io/35xkb. Google Scholar  H. Coskun, Dynamic Ecological System Measures, Osfpreprints, 2018. doi: 10.31219/osf.io/j2pd3. Google Scholar  H. Coskun, Static ecological system analysis, Theoretical Ecology, 2019, 1–36, https://doi.org/10.31219/osf.io/zqxc5. doi: 10.1007/s12080-019-0421-8. Google Scholar  H. Coskun, Static ecological system measures, Theoretical Ecology, 2019, 1–26, https://doi.org/10.31219/osf.io/g4xzt. doi: 10.1007/s12080-019-0422-7. Google Scholar  L. Edelstein-Keshet, Mathematical Models in Biology, Classics in Applied Mathematics, 46. Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA, 2005. doi: 10.1137/1.9780898719147.  Google Scholar  B. Hannon, The structure of ecosystems, Journal of Theoretical Biology, 41 (1973), 535-546. Google Scholar  P. W. Hippe, Environ analysis of linear compartmental systems: The dynamic, time-invariant case, Ecological Modelling, 19 (1983), 1-26. Google Scholar  J. Jacquez and C. Simon, Qualitative theory of compartmental systems, SIAM Review, 35 (1993), 43-79. doi: 10.1137/1035003.  Google Scholar  W. O. Kermack and A. G. McKendrick, A contribution to the mathematical theory of epidemics, Proceedings of the Royal Society of London A: Mathematical, Physical and Engineering Sciences, 115 (1927), 700–721, http://rspa.royalsocietypublishing.org/content/115/772/700.Google Scholar  W. W. Leontief, Quantitative input and output relations in the economic systems of the united states, The Review of Economic Statistics, 18 (1936), 105-125. Google Scholar  W. W. Leontief, Input-output Economics, Oxford University Press on Demand, New York, 1986.Google Scholar  B. C. Patten, Systems approach to the concept of environment, Ohio Journal of Science, 78 (1978), 206-222. Google Scholar  J. Szyrmer and R. E. Ulanowicz, Total flows in ecosystems, Ecological Modelling, 35 (1987), 123-136. Google Scholar  L. J. Tilly, The structure and dynamics of Cone Spring, Ecological Monographs, 38 (1968), 169-197. Google Scholar

show all references

##### References:
  D. Anderson, Compartmental Modeling and Tracer Kinetics, Lecture Notes in Biomathematics, 50. Springer-Verlag, Berlin, 1983. doi: 10.1007/978-3-642-51861-4.  Google Scholar  R. M. Anderson and R. M. May, Population biology of infectious diseases: Part i, Nature, 280 (1979), 361-367. doi: 10.1038/280361a0. Google Scholar  H. Coskun, Dynamic Ecological System Analysis, Osfpreprints, 2018. doi: 10.31219/osf.io/35xkb. Google Scholar  H. Coskun, Dynamic Ecological System Measures, Osfpreprints, 2018. doi: 10.31219/osf.io/j2pd3. Google Scholar  H. Coskun, Static ecological system analysis, Theoretical Ecology, 2019, 1–36, https://doi.org/10.31219/osf.io/zqxc5. doi: 10.1007/s12080-019-0421-8. Google Scholar  H. Coskun, Static ecological system measures, Theoretical Ecology, 2019, 1–26, https://doi.org/10.31219/osf.io/g4xzt. doi: 10.1007/s12080-019-0422-7. Google Scholar  L. Edelstein-Keshet, Mathematical Models in Biology, Classics in Applied Mathematics, 46. Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA, 2005. doi: 10.1137/1.9780898719147.  Google Scholar  B. Hannon, The structure of ecosystems, Journal of Theoretical Biology, 41 (1973), 535-546. Google Scholar  P. W. Hippe, Environ analysis of linear compartmental systems: The dynamic, time-invariant case, Ecological Modelling, 19 (1983), 1-26. Google Scholar  J. Jacquez and C. Simon, Qualitative theory of compartmental systems, SIAM Review, 35 (1993), 43-79. doi: 10.1137/1035003.  Google Scholar  W. O. Kermack and A. G. McKendrick, A contribution to the mathematical theory of epidemics, Proceedings of the Royal Society of London A: Mathematical, Physical and Engineering Sciences, 115 (1927), 700–721, http://rspa.royalsocietypublishing.org/content/115/772/700.Google Scholar  W. W. Leontief, Quantitative input and output relations in the economic systems of the united states, The Review of Economic Statistics, 18 (1936), 105-125. Google Scholar  W. W. Leontief, Input-output Economics, Oxford University Press on Demand, New York, 1986.Google Scholar  B. C. Patten, Systems approach to the concept of environment, Ohio Journal of Science, 78 (1978), 206-222. Google Scholar  J. Szyrmer and R. E. Ulanowicz, Total flows in ecosystems, Ecological Modelling, 35 (1987), 123-136. Google Scholar  L. J. Tilly, The structure and dynamics of Cone Spring, Ecological Monographs, 38 (1968), 169-197. Google Scholar Schematic representation of the dynamic subcompartmentalization in a three-compartment model system. Each subsystem is colored differently; the second subsystem ($k = 2$) is blue, for example. Only the subcompartments in the same subsystem ($x_{1_2}(t)$, $x_{2_2}(t)$, and $x_{3_2}(t)$ in the second subsystem, for example) interact with each other. Subsystem $k$ receives external input only at subcompartment ${k_k}$. The initial subsystem (gray) has no external input. The dynamic flow decomposition is not represented in this figure. Compare this figure with Fig. 2, in which the subcompartmentalization and corresponding flow decomposition are illustrated for $x_1(t)$ only Schematic representation of the dynamic flow decomposition in a three-compartment model system. The figure illustrates the subcompartmentalization of compartment $1$ and the corresponding flow decomposition of $f_{j1}(t,\mathit{\boldsymbol{ x}})$. The figure also illustrates further decomposition of initial subcompartment $1_0$ and the corresponding initial subflow function, $f_{j_0 1_0}(t,\bf{x})$ (both dark gray) Schematic representation of the dynamic subsystem decomposition. The transient inflow and outflow rate functions, $f^w_{\ell_k j_k i_k}(t)$ and $f^w_{n_k \ell_k j_k}(t)$, at and associated transient substorage, $x^w_{n_k \ell_k j_k}(t)$, in subcompartment ${\ell_k}$ along subflow path $p^w_{n_k j_k} = i_k \mapsto j_k \to \ell_k \to n_k$ Schematic representation of the simple and composite $\texttt{diact}$ flows. Solid arrows represent direct flows, and dashed arrows represent indirect flows through other compartments (not shown). The composite $\texttt{diact}$ flows (black) generated by outward throughflow $\hat{\tau}_{j}(t,\mathit{\boldsymbol{x}}) - \hat{\tau}_{j_0}(t,{\bf x})$ (i.e. derived from all external inputs): direct flow, $\tau^\texttt{d}_{i j}(t)$, indirect flow, $\tau^\texttt{i}_{ij}(t)$, acyclic flow, $\tau^\texttt{a}_{ij}(t) = \tau^\texttt{t}_{ij}(t) - \tau^\texttt{c}_{ij}(t)$, cycling flow, $\tau^\texttt{c}_{ij}(t)$, and transfer flow, $\tau^\texttt{t}_{ij}(t)$. The simple $\texttt{diact}$ flows (blue) generated by outward subthroughflow $\hat{\tau}_{i_i}(t,{\bf x})$ (i.e. derived from single external input $z_{i}(t)$): direct flow, ${\tau}^\texttt{d}_{j_i}(t) = {\tau}^\texttt{d}_{j_i i_i}(t)$, indirect flow, ${\tau}^\texttt{i}_{j_i}(t) = \tau^\texttt{i}_{j_i i_i}(t)$, acyclic flow, ${\tau}^\texttt{a}_{j_i}(t) = \tau^\texttt{a}_{j_i i_i}(t) = {\tau}^\texttt{t}_{j_i}(t) - {\tau}^\texttt{c}_{j_i}(t)$, cycling flow, ${\tau}^\texttt{c}_{j_i}(t) = \tau^\texttt{c}_{j_i i_i}(t)$, and transfer flow, $\tau^{\texttt{t}}_{j_i}(t) = \tilde{\tau}_{j_i}(t,{\bf x}) = \check{\tau}_{j_i}(t,{\bf x}) - z_{j_i}(t)$. Note that the cycling flows at the terminal (sub)compartment may include the segments of the direct and/or indirect flows at that (sub)compartment, if the cycling flows indirectly pass through the corresponding initial (sub)compartment (see Fig. 10). Consequently, the acyclic flows are composed of the segments of the direct and/or indirect flows Schematic representation of the model network. Arrows are labeled by the corresponding rate constants. Subflow paths $p^2_{0_1 1_1}$ and $p^3_{0_1 1_1}$ along which the transient external outputs are computed are red (subsystems are not shown) (Case study 3.1) Numerical results for the evolution of the initial populations, $\underline {\mathit{\boldsymbol{x}}} \left( t \right)$, and state variables, $\mathit{\boldsymbol{x}}(t)$. The populations generated by external inputs alone, $\bar{\mathit{\boldsymbol{x}}}(t)$, are presented in Fig. 7 (Case study 3.1) The graphical representations of the initial substate and substate functions ${\underline x _{{i_k}}}\left( t \right)$ and ${x}_{i_k}(t) = \bar{x}_{i}(t)$ for all $i,k$. The substates that are equal to zero are not labeled. (Case study 3.1) The graphical representations of the transient substorage functions at subcompartment $2_1$, $x^{1,m}_{3_1 2_1 1_1}(t)$, along path $p^1_{2_1 1_1}$, and the transient output rates $f^{2}_{0_1 2_1 1_1}(t)$ and $f^{3}_{0_1 2_1 1_1}(t)$ along paths $p^2_{0_1 1_1}$ and $p^3_{0_1 1_1}$, respectively (Case study 3.1) The graphical representation for the indirect storages from compartment $1$ to $3$ through $2$ within both the subsystems and initial subsystems, $x^\texttt{i}_{31}(t)$ and $x^\texttt{i}_{31}(t)$, respectively, and the residence times, $r_i(t)$ (Case study 3.1) Schematic representation for the complementary nature of the indirect (or acyclic) and cycling subflows within the $k^{th}$ subsystem. The direct subflow, $f_{k_k i_k}(t,{\bf x})$, is represented by solid arrow. This subflow also contributes to the simple cycling flow at subcompartment $k_k$. The indirect subflow, ${\tau}^\texttt{i}_{i_k k_k}(t)$, (or the acyclic subflow, ${\tau}^\texttt{a}_{i_k k_k}(t)$) through other compartments (not shown) is represented by dashed arrow Schematic representation of the model network. Subflow path $p^1_{1_1}$, along which the cycling flow and associated substorage function are computed, is red (subsystems are not shown) (Case study E.1) The graphical representation of the substorage and inward subthroughflow matrices, $X(t)$ and $\check{T}(t)$, for time dependent input $z(t) = \left[ 3+\operatorname{sin}(t), 3+\operatorname{sin}(2 \, t) \right]^T$, and cycling flows and storages, $\tau^c_{i_0}(t)+\tau^c_i(t)$ and $x^c_{i_0}(t)+x^c_i(t)$ (Case study E.1) Schematic representation of the model network. (subsystems are not shown) (Case study E.2)
The dynamic $\texttt{diact}$ flow distribution and the simple and composite $\texttt{diact}$ (sub)flow matrices. The superscript ($^\texttt{*}$) in each equation represents any of the $\texttt{diact}$ symbols. For the sake of readability, the function arguments are dropped
 $\texttt{diact}$ flow distribution matrix flows $\texttt{d}$ $N^\texttt{d} = F \, \mathcal{T}^{-1}$ $\texttt{i}$ $N^\texttt{i}= {N}^\texttt{*} \, (\mathcal{T} - \mathcal{\hat T}_0)$ ${T^*} = {N^*}\left( {{\cal T} - {{\widehat {\cal T}}_0}} \right)$ $\texttt{a}$ ${N}^\texttt{a} = \tilde{T} \, \hat{\mathsf{T}}^{-1} - \tilde{\mathsf{T}} \, \hat{\mathsf{T}}^{-1} \, \hat{T} \, \hat{\mathsf{T}}^{-1}$ ${T}^\texttt{*}_\ell= {N}^\texttt{*} \, \mathcal{\hat T}_\ell$ $\texttt{c}$ ${N}^\texttt{c} = \tilde{\mathsf{T}} \, \hat{\mathsf{T}}^{-1} \, \hat{T} \, \hat{\mathsf{T}}^{-1}$ ${\tilde T^*} = {N^*}{\mkern 1mu} \mathsf{\hat T}$ $\texttt{t}$ $N^\texttt{t} = \tilde{T} \, \hat{\mathsf{T}}^{-1}$
 $\texttt{diact}$ flow distribution matrix flows $\texttt{d}$ $N^\texttt{d} = F \, \mathcal{T}^{-1}$ $\texttt{i}$ $N^\texttt{i}= {N}^\texttt{*} \, (\mathcal{T} - \mathcal{\hat T}_0)$ ${T^*} = {N^*}\left( {{\cal T} - {{\widehat {\cal T}}_0}} \right)$ $\texttt{a}$ ${N}^\texttt{a} = \tilde{T} \, \hat{\mathsf{T}}^{-1} - \tilde{\mathsf{T}} \, \hat{\mathsf{T}}^{-1} \, \hat{T} \, \hat{\mathsf{T}}^{-1}$ ${T}^\texttt{*}_\ell= {N}^\texttt{*} \, \mathcal{\hat T}_\ell$ $\texttt{c}$ ${N}^\texttt{c} = \tilde{\mathsf{T}} \, \hat{\mathsf{T}}^{-1} \, \hat{T} \, \hat{\mathsf{T}}^{-1}$ ${\tilde T^*} = {N^*}{\mkern 1mu} \mathsf{\hat T}$ $\texttt{t}$ $N^\texttt{t} = \tilde{T} \, \hat{\mathsf{T}}^{-1}$
The $\texttt{diact}$ flow and storage distribution and the simple and composite $\texttt{diact}$ (sub)flow and (sub)storage matrices. The superscript ($^\texttt{*}$) in each equation represents any of the $\texttt{diact}$ symbols
 $\texttt{diact}$ flow and storage distribution matrices flows storages $\texttt{d}$ $N^\texttt{d} = F \, \mathcal{T}^{-1}$ ${S}^\texttt{*} = \mathcal{R} \, {N}^\texttt{*}$ $\texttt{i}$ $N^\texttt{i} = (N-I) \, \mathcal{N}^{-1} - F \,\mathcal{T}^{-1}$ ${T}^\texttt{*} = {N}^\texttt{*} \, \mathcal{T}$ ${X}^\texttt{*} = {S}^\texttt{*} \, \mathcal{T}$ $\texttt{a}$ ${N}^\texttt{a} = (\mathcal{N}^{-1} \, N - I) \, \mathcal{N}^{-1}$ ${T}^\texttt{*}_\ell = {N}^\texttt{*} \, \mathcal{T}_\ell$ ${X}^\texttt{*}_\ell = {S}^\texttt{*} \, \mathcal{T}_\ell$ $\texttt{c}$ ${N}^\texttt{c} = (N - \mathcal{N}^{-1} N) \, \mathcal{N}^{-1}$ $\tilde{T}^\texttt{*} = {N}^\texttt{*} \, \mathsf{T}$ $\tilde{X}^\texttt{*} = {S}^\texttt{*} \, \mathsf{T}$ } $\texttt{t}$ $N^\texttt{t} = (N-I) \, \mathcal{N}^{-1}$
 $\texttt{diact}$ flow and storage distribution matrices flows storages $\texttt{d}$ $N^\texttt{d} = F \, \mathcal{T}^{-1}$ ${S}^\texttt{*} = \mathcal{R} \, {N}^\texttt{*}$ $\texttt{i}$ $N^\texttt{i} = (N-I) \, \mathcal{N}^{-1} - F \,\mathcal{T}^{-1}$ ${T}^\texttt{*} = {N}^\texttt{*} \, \mathcal{T}$ ${X}^\texttt{*} = {S}^\texttt{*} \, \mathcal{T}$ $\texttt{a}$ ${N}^\texttt{a} = (\mathcal{N}^{-1} \, N - I) \, \mathcal{N}^{-1}$ ${T}^\texttt{*}_\ell = {N}^\texttt{*} \, \mathcal{T}_\ell$ ${X}^\texttt{*}_\ell = {S}^\texttt{*} \, \mathcal{T}_\ell$ $\texttt{c}$ ${N}^\texttt{c} = (N - \mathcal{N}^{-1} N) \, \mathcal{N}^{-1}$ $\tilde{T}^\texttt{*} = {N}^\texttt{*} \, \mathsf{T}$ $\tilde{X}^\texttt{*} = {S}^\texttt{*} \, \mathsf{T}$ } $\texttt{t}$ $N^\texttt{t} = (N-I) \, \mathcal{N}^{-1}$
  Jonathan H. Tu, Clarence W. Rowley, Dirk M. Luchtenburg, Steven L. Brunton, J. Nathan Kutz. On dynamic mode decomposition: Theory and applications. Journal of Computational Dynamics, 2014, 1 (2) : 391-421. doi: 10.3934/jcd.2014.1.391  Alessia Marigo, Benedetto Piccoli. A model for biological dynamic networks. Networks & Heterogeneous Media, 2011, 6 (4) : 647-663. doi: 10.3934/nhm.2011.6.647  Xilin Fu, Zhang Chen. New discrete analogue of neural networks with nonlinear amplification function and its periodic dynamic analysis. Conference Publications, 2007, 2007 (Special) : 391-398. doi: 10.3934/proc.2007.2007.391  Andrey Olypher, Jean Vaillant. On the properties of input-to-output transformations in neuronal networks. Mathematical Biosciences & Engineering, 2016, 13 (3) : 579-596. doi: 10.3934/mbe.2016009  Toufik Bakir, Bernard Bonnard, Jérémy Rouot. A case study of optimal input-output system with sampled-data control: Ding et al. force and fatigue muscular control model. Networks & Heterogeneous Media, 2019, 14 (1) : 79-100. doi: 10.3934/nhm.2019005  Alexei Pokrovskii, Dmitrii Rachinskii. Effect of positive feedback on Devil's staircase input-output relationship. Discrete & Continuous Dynamical Systems - S, 2013, 6 (4) : 1095-1112. doi: 10.3934/dcdss.2013.6.1095  Bogdan Sasu, A. L. Sasu. Input-output conditions for the asymptotic behavior of linear skew-product flows and applications. Communications on Pure & Applied Analysis, 2006, 5 (3) : 551-569. doi: 10.3934/cpaa.2006.5.551  Magdi S. Mahmoud. Output feedback overlapping control design of interconnected systems with input saturation. Numerical Algebra, Control & Optimization, 2016, 6 (2) : 127-151. doi: 10.3934/naco.2016004  James P. Nelson, Mark J. Balas. Direct model reference adaptive control of linear systems with input/output delays. Numerical Algebra, Control & Optimization, 2013, 3 (3) : 445-462. doi: 10.3934/naco.2013.3.445  Angelo B. Mingarelli. Nonlinear functionals in oscillation theory of matrix differential systems. Communications on Pure & Applied Analysis, 2004, 3 (1) : 75-84. doi: 10.3934/cpaa.2004.3.75  Carl. T. Kelley, Liqun Qi, Xiaojiao Tong, Hongxia Yin. Finding a stable solution of a system of nonlinear equations arising from dynamic systems. Journal of Industrial & Management Optimization, 2011, 7 (2) : 497-521. doi: 10.3934/jimo.2011.7.497  Suoqin Jin, Fang-Xiang Wu, Xiufen Zou. Domain control of nonlinear networked systems and applications to complex disease networks. Discrete & Continuous Dynamical Systems - B, 2017, 22 (6) : 2169-2206. doi: 10.3934/dcdsb.2017091  Simone Cacace, Maurizio Falcone. A dynamic domain decomposition for the eikonal-diffusion equation. Discrete & Continuous Dynamical Systems - S, 2016, 9 (1) : 109-123. doi: 10.3934/dcdss.2016.9.109  Steven L. Brunton, Joshua L. Proctor, Jonathan H. Tu, J. Nathan Kutz. Compressed sensing and dynamic mode decomposition. Journal of Computational Dynamics, 2015, 2 (2) : 165-191. doi: 10.3934/jcd.2015002  Xianchao Xiu, Lingchen Kong. Rank-one and sparse matrix decomposition for dynamic MRI. Numerical Algebra, Control & Optimization, 2015, 5 (2) : 127-134. doi: 10.3934/naco.2015.5.127  Rein Luus. Optimal control of oscillatory systems by iterative dynamic programming. Journal of Industrial & Management Optimization, 2008, 4 (1) : 1-15. doi: 10.3934/jimo.2008.4.1  Eduardo Espinosa-Avila, Pablo Padilla Longoria, Francisco Hernández-Quiroz. Game theory and dynamic programming in alternate games. Journal of Dynamics & Games, 2017, 4 (3) : 205-216. doi: 10.3934/jdg.2017013  Joseph D. Skufca, Erik M. Bollt. Communication and Synchronization in Disconnected Networks with Dynamic Topology: Moving Neighborhood Networks. Mathematical Biosciences & Engineering, 2004, 1 (2) : 347-359. doi: 10.3934/mbe.2004.1.347  Jingli Ren, Dandan Zhu, Haiyan Wang. Spreading-vanishing dichotomy in information diffusion in online social networks with intervention. Discrete & Continuous Dynamical Systems - B, 2019, 24 (4) : 1843-1865. doi: 10.3934/dcdsb.2018240  Tai-Ping Liu, Shih-Hsien Yu. Hyperbolic conservation laws and dynamic systems. Discrete & Continuous Dynamical Systems - A, 2000, 6 (1) : 143-145. doi: 10.3934/dcds.2000.6.143

2018 Impact Factor: 1.008