--- title: 3D Fractional-Order Toda Lattice url: https://www.emergentmind.com/topics/3-dimensional-fractional-order-toda-lattice type: topic --- # 3D Fractional-Order Toda Lattice Searching arXiv for the specified Toda-lattice papers to ground the article in current records. A 3-dimensional fractional-order Toda lattice is a Toda-type dynamical system in which the underlying three-dimensional structure is combined with non-integer-order evolution, typically through a Caputo derivative. In the current arXiv literature, the most direct formulation is a Caputo fractional extension of the \(3\)-dimensional Toda lattice with two embedded control parameters, analyzed for well-posedness, equilibrium structure, local stability, feedback stabilization, and numerical integration [2507.20788]. The surrounding literature makes clear, however, that the phrase joins two largely distinct research lines: one on higher-dimensional or 3D Toda-type lattices in discrete or \((2+1)\)-dimensional settings [1803.06774], [2409.07017], and another on generalized or fractional Toda hierarchies in which “fractional” refers to fractional powers of a Lax operator rather than to fractional derivatives in time or space [1210.5038]. The topic therefore sits at an intersection of fractional dynamical systems, integrable-lattice theory, and controlled nonlinear evolution. ## 1. Definition and scope The classical \(n\)-dimensional Toda lattice recalled in the recent fractional study is posed in \(\mathbb{R}^{2n-1}\) with coordinates \[ \{x^1,\dots,x^n,y^1,\dots,y^{n-1}\}, \] and governed by \[ \dot{x}^{i}(t) = 2\big[(y^{i})^{2}(t)-(y^{i-1})^{2}(t)\big], \qquad \dot{y}^{j}(t) = y^{j}(t)\big( x^{j+1}(t)-x^{j}(t)\big), \] with \[ y^0(t)=y^n(t)=0. \] For \(n=3\), this yields a five-dimensional first-order system in the variables \((x^1,x^2,x^3,y^1,y^2)\) [2507.20788]. The fractional analogue replaces the first derivative by a Caputo derivative of order \(q\in(0,1)\): \[ D_t^q x^i(t) = 2\big[(y^i)^2(t)-(y^{i-1})^2(t)\big], \qquad D_t^q y^j(t) = y^j(t)\big(x^{j+1}(t)-x^j(t)\big). \] The principal model currently identified on arXiv as a “3-dimensional fractional-order Toda lattice” is the \(n=3\) case with two control terms: \[ \left\{ \begin{array}{lll} D_t^q x^1(t) &=& 2 (y^1)^2(t) + a x^1(t),\\[0.1cm] D_t^q x^2(t) &=& 2\big[(y^2)^2(t)-(y^1)^2(t)\big],\\[0.1cm] D_t^q x^3(t) &=& - 2 (y^2)^2(t),\\[0.1cm] D_t^q y^1(t) &=& y^1(t)( x^2(t)-x^1(t)),\\[0.1cm] D_t^q y^2(t) &=& y^2(t) ( x^3(t)- x^2(t))+b\, y^2(t), \end{array} \right. \] where \(a,b\in\mathbb{R}^\ast\) and \(q\in(0,1)\) [2507.20788]. After the relabeling \[ y^1=x^4,\qquad y^2=x^5, \] the system becomes \[ \left\{ \begin{array}{lll} D_t^q x^1(t) &=& 2 (x^4)^2(t) + a x^1(t),\\[0.1cm] D_t^q x^2(t) &=& 2\big[(x^5)^2(t)-(x^4)^2(t)\big],\\[0.1cm] D_t^q x^3(t) &=& - 2 (x^5)^2(t),\\[0.1cm] D_t^q x^4(t) &=& x^4(t)( x^2(t)-x^1(t)),\\[0.1cm] D_t^q x^5(t) &=& x^5(t) ( x^3(t)- x^2(t))+b x^5(t), \end{array} \right. \] and is explicitly termed the “3-dimensional fractional-order Toda lattice with two controls around axes \(Ox^1\) and \(Ox^5\)” [2507.20788]. This usage of “fractional-order” is specific. It denotes fractional differentiation in the Caputo sense, not merely fractional powers of a Lax operator. That distinction is essential because in the bigraded Toda hierarchy literature the “fractional” aspect is instead encoded by operators such as \(L^{1/N}\) and \(L^{1/M}\), without introducing fractional derivatives in the evolution equations [1210.5038]. ## 2. Fractional structure and governing equations The derivative used in the Caputo-Toda formulation is \[ D_t^q f(t)= I^{m-q} f^{(m)}(t), \qquad q>0, \] with \(m\in\mathbb{N}^\ast\) such that \(m-1\le q\le m\), and the associated Riemann–Liouville fractional integral is \[ I_t^q f(t)=\frac{1}{\Gamma(q)}\int_0^t (t-s)^{q-1}f(s)\,ds. \] For the Toda models in question, the order is restricted to \(q\in(0,1)\) [2507.20788]. The initial condition is simply \[ x(0)=x_0\in\mathbb{R}^5, \] or componentwise \[ x^i(0)=x_0^i,\qquad i=1,\dots,5. \] No additional compatibility conditions are imposed beyond finite initial data [2507.20788]. The same paper rewrites the system in the compact matrix form \[ D_t^\alpha x(t)=x^4(t)A_1x(t)+x^5(t)A_2x(t)+A_3x(t),\qquad x(0)=x_0, \] with \[ x(t)=\big(x^1(t),x^2(t),x^3(t),x^4(t),x^5(t)\big)^T, \] and explicit matrices \[ A_{1} = \begin{pmatrix} 0 & 0 & 0 & 2 & 0\\ 0 & 0 & 0 & -2 & 0\\ 0 & 0 & 0 & 0 & 0\\ -1 & 1 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 0 \end{pmatrix}, \quad A_{2} = \begin{pmatrix} 0 & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 2\\ 0 & 0 & 0 & 0 & -2\\ 0 & 0 & 0 & 0 & 0\\ 0 & -1 & 1 & 0 & 0 \end{pmatrix}, \] \[ A_{3} = \begin{pmatrix} a & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & b \end{pmatrix}. \] The paper notes a notational mix between \(q\) and \(\alpha\), but the operative fractional order remains \(q\in(0,1)\) [2507.20788]. A key clarification in the literature is that this Caputo-based system should not be conflated with other “3D Toda” models. The higher-dimensional discrete Toda-type equation on \(\mathbb{Z}^{a+b}\), \[ T_{t+1,n}T_{t-1,n} = \prod_{i=1}^{a}T_{t,n+e_i}^{k_i}T_{t,n-e_i}^{l_i} + \prod_{i=a+1}^{a+b}T_{t,n+e_i}^{k_i}T_{t,n-e_i}^{l_i}, \] does furnish a genuine spatially three-dimensional discrete specialization when \(a+b=3\), but it contains no fractional derivatives, no fractional differences, and no non-integer-order operators [1803.06774]. Likewise, the \((2+1)\)-dimensional Toda-type lattices with one discrete index \(n\) and two continuous variables \(x,y\) develop a notion of “3D” based on three independent variables, yet remain entirely integer-order [2409.07017]. ## 3. Well-posedness and integral formulation The initial value problem for the controlled fractional system is treated through a standard Caputo-IVP framework. The nonlinear vector field is defined by \[ f(x(t))=x^4(t)A_1x(t)+x^5(t)A_2x(t)+A_3x(t). \] The main existence-and-uniqueness statement is explicit: > **Proposition 2.1.** The initial value problem of the 3-dimensional fractional-order Toda lattice with two controls (2.5) has a unique solution [2507.20788]. The proof strategy relies on continuity and Lipschitz estimates on a closed box \[ D=\{x\in\mathbb{R}^5:\ x^i\in[x_0^i-\delta,x_0^i+\delta],\ i=1,\dots,5\}, \] and on the decomposition \[ f(x(t))-f(y(t))=g_1(t)+g_2(t)+g_3(t), \] where \[ g_1(t)=x^4(t)A_1x(t)-y^4(t)A_1y(t),\qquad g_2(t)=x^5(t)A_2x(t)-y^5(t)A_2y(t), \] \[ g_3(t)=A_3x(t)-A_3y(t). \] The estimate obtained is \[ |f(x(t))-f(y(t))| \le \big(\|A_1\|+\|A_2\|+\|A_3\|+|y^4(t)|+|y^5(t)|\big)\,|x(t)-y(t)|, \] and with \(\|A_1\|=\|A_2\|=\sqrt{10}\), the paper concludes a global Lipschitz bound \[ |f(x(t))-f(y(t))|\le L|x(t)-y(t)| \] for some \(L>0\), after which Diethelm’s existence-uniqueness theorems are invoked [2507.20788]. For a general Caputo IVP \[ D_t^q y(t)=f(t,y(t)),\qquad y(0)=y_0, \] the equivalent Volterra equation is stated as \[ y(t)=y(0)+I_t^q f(t,y(t)). \] Applied to the Toda system, this becomes componentwise \[ x^i(t)=x^i(0)+I_t^q F_i(x^1(t),x^2(t),x^3(t),x^4(t),x^5(t)), \qquad i=\overline{1,5}, \] or, in vector form, \[ x(t)=x_0+\frac{1}{\Gamma(q)}\int_0^t (t-s)^{q-1}f(x(s))\,ds. \] This integral representation is the main analytical manifestation of the fractional memory structure in the model [2507.20788]. A plausible implication is that the fractionalization moves the Toda system from a purely local ODE phase portrait to a hereditary dynamical system in which the state depends on the entire prior trajectory. The paper itself emphasizes this only indirectly through the Caputo and Volterra formulations [2507.20788]. ## 4. Equilibria and stability theory The equilibrium analysis begins from the component functions \[ \begin{cases} f_1(x)=2(x^4)^2+a x^1,\quad f_2(x)=2\big[(x^5)^2-(x^4)^2\big],\\ f_3(x)=-2(x^5)^2,\quad f_4(x)=x^4(x^2-x^1),\quad f_5(x)=x^5(x^3-x^2)+b x^5. \end{cases} \] Solving \(f_i(x)=0\), the paper derives the equilibrium set \[ E=\{e_0=(0,0,0,0,0)\}\cup E_1, \] where \[ E_1=\left\{e_{23}^{k,m}=(0,k,m,0,0)\in\mathbb{R}^5:\ k,m\in\mathbb{R},\ k\neq 0\ \text{or}\ m\neq 0\right\}. \] Hence all equilibria lie on the plane \[ x^1=x^4=x^5=0. \] Three subfamilies are singled out: \[ E_1^1=\{e_2^k=(0,k,0,0,0):k\in\mathbb{R}^\ast\}, \] \[ E_1^2=\{e_3^m=(0,0,m,0,0):m\in\mathbb{R}^\ast\}, \] \[ E_1^3=\{e_{23}^{m,m}=(0,m,m,0,0):m\in\mathbb{R}^\ast\} \] [2507.20788]. The Jacobian of the uncontrolled fractional system is \[ J(x,a,b)= \begin{pmatrix} a & 0 & 0 & 4x^4 & 0\\ 0 & 0 & 0 & -4x^4 & 4x^5\\ 0 & 0 & 0 & 0 & -4x^5\\ -x^4 & x^4 & 0 & x^2-x^1 & 0\\ 0 & -x^5 & x^5 & 0 & x^3-x^2+b \end{pmatrix}. \] The stability test used is the Matignon criterion: > An equilibrium \(x_e\) is locally asymptotically stable iff every eigenvalue \(\lambda\) of \(J(x_e,a,b)\) satisfies > \[ > |\arg(\lambda)|>\frac{q\pi}{2}. > \] > It is locally stable iff either it is asymptotically stable, or every critical eigenvalue with > \[ > |\arg(\lambda)|=\frac{q\pi}{2} > \] > has geometric multiplicity one [2507.20788]. At an equilibrium \(e_{23}^{k,m}=(0,k,m,0,0)\), \[ J(e_{23}^{k,m},a,b)= \begin{pmatrix} a & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & k & 0\\ 0 & 0 & 0 & 0 & m-k+b \end{pmatrix}, \] with characteristic polynomial \[ p_{J(e_{23}^{k,m},a,b)}(\lambda) = -\lambda^2(\lambda-a)(\lambda-k)(\lambda-b+k-m). \] Since \(0\) is always an eigenvalue, indeed with multiplicity at least \(2\), the paper concludes that every equilibrium of the uncontrolled system is unstable for all \(q\in(0,1)\) [2507.20788]. This instability result is one of the sharpest features of the current model: the uncontrolled Caputo fractional 3D Toda lattice possesses a continuum of equilibria, but none is locally asymptotically stable. This should not be confused with integrability-theoretic stability notions in the integer-order discrete Toda literature. For instance, the multidimensional discrete Toda-type recursions over \(\mathbb{Z}^{a+b}\) are studied instead through Laurentness, irreducibility, and coprimeness, while their iterates exhibit exponential degree growth and are thus non-integrable in the algebraic-entropy sense except in the standard low-dimensional Toda case [1803.06774]. ## 5. Control and stabilization The stabilization problem is treated by augmenting the plant with an external control vector \(u=(u_1,\dots,u_5)\): \[ \left\{ \begin{array}{lll} D_t^q x^1(t) &=& 2 (x^4)^2(t) + a x^1(t)+ u_1(t),\\[0.1cm] D_t^q x^2(t) &=& 2[(x^5)^2(t)-(x^4)^2(t)] + u_2(t),\\[0.1cm] D_t^q x^3(t) &=& - 2 (x^5)^2(t)+ u_3(t),\\[0.1cm] D_t^q x^4(t) &=& x^4(t)( x^2(t)-x^1(t))+ u_4(t),\\[0.1cm] D_t^q x^5(t) &=& x^5(t) ( x^3(t)- x^2(t))+b x^5(t) + u_5(t). \end{array} \right. \] The chosen feedback law is \[ u_1=0,\qquad u_2=c_1(x^2-x_e^2),\qquad u_3=c_2(x^3-x_e^3),\qquad u_4=c_3(x^4-x_e^4),\qquad u_5=0, \] where \[ c_1,c_2,c_3\in\mathbb{R}^\ast. \] This yields the closed-loop system \[ \left\{ \begin{array}{lll} D_t^q x^1(t) &=& 2 (x^4)^2(t) + a x^1(t),\\[0.1cm] D_t^q x^2(t) &=& 2[(x^5)^2(t)-(x^4)^2(t)] + c_1 (x^2(t)- x_e^2),\\[0.1cm] D_t^q x^3(t) &=& - 2 (x^5)^2(t)+ c_2 (x^3(t)- x_e^3),\\[0.1cm] D_t^q x^4(t) &=& x^4(t)( x^2(t)-x^1(t))+c_3 (x^4(t)- x_e^4),\\[0.1cm] D_t^q x^5(t) &=& x^5(t) ( x^3(t)- x^2(t))+b x^5(t). \end{array} \right. \] The corresponding Jacobian is \[ J(x,a,b,c_1,c_2,c_3)= \begin{pmatrix} a & 0 & 0 & 4x^4 & 0\\ 0 & c_1 & 0 & -4x^4 & 4x^5\\ 0 & 0 & c_2 & 0 & -4x^5\\ -x^4 & x^4 & 0 & x^2-x^1+c_3 & 0\\ 0 & -x^5 & x^5 & 0 & x^3-x^2+b \end{pmatrix} \] [2507.20788]. At the origin \(e_0=(0,0,0,0,0)\), the Jacobian is diagonal with characteristic polynomial \[ p_0(\lambda)=(\lambda-a)(\lambda-b)(\lambda-c_1)(\lambda-c_2)(\lambda-c_3). \] The paper states: - if \[ a<0,\qquad b<0,\qquad c_1<0,\qquad c_2<0,\qquad c_3<0, \] then \(e_0\) is asymptotically stable for every \(q\in(0,1)\); - if \(a>0\), then \(e_0\) is unstable for all parameter choices [2507.20788]. At a nonzero equilibrium \(e_{23}^{k,m}=(0,k,m,0,0)\), \[ J_1= \begin{pmatrix} a & 0 & 0 & 0 & 0\\ 0 & c_1 & 0 & 0 & 0\\ 0 & 0 & c_2 & 0 & 0\\ 0 & 0 & 0 & k+c_3 & 0\\ 0 & 0 & 0 & 0 & m-k+b \end{pmatrix}, \] with eigenvalues \[ \lambda_1=a,\quad \lambda_2=c_1,\quad \lambda_3=c_2,\quad \lambda_4=c_3+k,\quad \lambda_5=b-k+m. \] Hence, if \[ a<0,\qquad c_1<0,\qquad c_2<0,\qquad k< -c_3,\qquad m