Homotopy-Perturbation Method

Updated 16 April 2026

Homotopy-Perturbation Method is a semi-analytical technique that constructs a one-parameter family (homotopy) to deform nonlinear equations into solvable linear subproblems.
It systematically applies power series expansion and recursive linearization, enabling closed-form or approximate solutions in fields like astrophysics, fluid mechanics, and cosmology.
Though effective for weakly nonlinear problems, HPM lacks adaptive convergence control which can limit its performance for strongly nonlinear or stiff systems.

The Homotopy-Perturbation Method (HPM) is a semi-analytical technique for obtaining approximate or sometimes exact solutions of nonlinear ordinary and partial differential equations, including boundary-value, initial-value, and integro-differential problems. Combining homotopy theory from topology with the regular perturbation expansion, HPM provides a systematic procedure for reducing nonlinear problems to a sequence of linear subproblems, each typically solvable in closed form. This method, originally developed by J. H. He, has found widespread application across mathematical physics, engineering, astrophysics, and cosmology.

1. Structural Foundation and Methodology

HPM constructs a one-parameter family of deformations—termed a homotopy—between a simple linear problem and the original, typically nonlinear, target equation. Let the problem be written in operator form as

$A(u) = f,$

where $A$ is a (possibly nonlinear) differential operator. The operator is split as $A = L + N$ with $L$ linear and $N$ nonlinear. Introducing the homotopy (embedding) parameter $p\in[0,1]$ , a continuous deformation is constructed:

$H(u,p) = (1-p)[L(u) - L(u_0)] + p[L(u) + N(u) - f] = 0,$

with $u_0$ an initial approximation, typically satisfying the boundary/initial conditions.

Assuming an analytic dependence on $p$ , the solution is formally written as a power series,

$u(x,p) = u_0(x) + p\,u_1(x) + p^2\,u_2(x) + \ldots,$

and substituting into $A$ 0, one generates a recursive hierarchy of linear subproblems for the coefficients $A$ 1. The approximate or exact solution is given by the sum at $A$ 2,

$A$ 3

This structural approach does not require the existence of a small physical parameter; rather, the formal $A$ 4 organizes the analytic expansion around the linear reference point.

2. Theoretical Role within Homotopy-Based Methods

Recent theoretical advances demonstrate that HPM is a special, fixed-parameter case of the broader Homotopy Analysis Method (HAM), which provides additional convergence-control flexibility via choice of auxiliary linear operator and parameter(s) (Xu, 19 Mar 2026). In HAM, the deformation equation includes tunable convergence-control ( $A$ 5) and auxiliary function ( $A$ 6) degrees of freedom:

$A$ 7

Fixing $A$ 8, $A$ 9, and $A = L + N$ 0 recovers HPM exactly (Xu, 19 Mar 2026). This embedding positions HPM as an analytically justified, but non-adaptive, degenerate case of HAM; while HPM provides simplicity and ease of use, it sacrifices tunable convergence control and robustness for strongly nonlinear systems.

3. Recursive Expansion and Implementation Details

The recursive structure of HPM is central to its implementation. After constructing the homotopy, substituting the power series into $A = L + N$ 1 and equating terms at each order in $A = L + N$ 2 produces a hierarchy of linear problems:

$A = L + N$ 3: typically yields $A = L + N$ 4, vanishing identically or yielding $A = L + N$ 5 as the initial approximation.
$A = L + N$ 6: $A = L + N$ 7.
$A = L + N$ 8: $A = L + N$ 9 (with $L$ 0 the Fréchet derivative), and so forth.

Each subproblem is linear (in $L$ 1), and higher-order corrections have homogeneous boundary conditions if $L$ 2 is constructed to satisfy all (possibly multi-point) conditions (Iftikhar, 2013). The process is practically algorithmic, allowing direct algebraic or symbolic computation of each term.

For strongly nonlinear $L$ 3, He’s polynomials or equivalent recursive expansions for products/powers are used (Maitama, 2018), while for systems involving memory or fractional derivatives, auxiliary transforms (Laplace, Natural Transform) may be combined with HPM, yielding hybrid methods (e.g., HPTM, HNTHPM) (Singh et al., 2016, Maitama, 2018). In all cases, the principle remains: nonlinearity appears only as explicit driving terms at each order, with convergence determined by the analyticity radius in $L$ 4.

4. Convergence, Parameter Sensitivity, and Error Analysis

Convergence of the HPM series is guaranteed under analyticity assumptions and smallness of the nonlinear operator (in a Banach space), often implemented via the Implicit Function Theorem (Xu, 19 Mar 2026, Iftikhar, 2013). For weakly nonlinear or analytic problems, the convergence radius in $L$ 5 (i.e., the interval over which the $L$ 6-series converges at $L$ 7) is typically sufficient to ensure rapid convergence in practical computations. The lack of a convergence-control parameter in standard HPM, however, can result in divergence or slow convergence for strongly nonlinear or stiff problems, where an adaptive variant (e.g., HAM) is preferable (Xu, 19 Mar 2026).

In practical terms, convergence is monitored by the relative size of successive terms or by residual evaluation,

$L$ 8

computed for the truncated sum. Empirical demonstrations in domains such as population balances (Kaur et al., 2017), cosmological models (Shchigolev, 2013), and eigenvalue-boundary-value problems (Iftikhar, 2013) show that only a few terms (3–6) are needed for high-precision approximation. Explicit error estimates for the truncated sum (up to order $L$ 9) are possible when the operator norm admits a contractivity constant $N$ 0 (Singh et al., 2016).

5. Applications Across Mathematical Physics and Engineering

HPM has been successfully deployed in a vast range of nonlinear problems, including:

Astrophysics and General Relativity: Analytically approximating neutron star structures via the TOV equation (yielding mass functions and compactness ranges) (Aziz et al., 2019), calculation of gravitational waveforms (Zare et al., 2012), perihelion precession, and null-geodesic deflections (with exact or rapid-convergent series for orbits and deflection angles) (Shchigolev et al., 2016, Shchigolev, 2017).
Cosmology: Analytical Friedmann equation solutions, including dust and quintessence cosmologies (Shchigolev, 2013), and luminosity distance-redshift relationships in FLRW models (Shchigolev, 2015).
Fluid Mechanics and Applied Mathematics: Blasius boundary layer, Burgers and KdV equations, MHD Jeffery-Hamel flow, heat transfer in nanofluids, where closed-form polynomial approximations for velocity, temperature, and other observables are constructed (Ebenezer, 2023, Pourabdian et al., 2016, Marinca et al., 2015, Joohy, 2019).
Population Balances and Biomathematics: Analytic series solutions for Smoluchowski-type aggregation and fragmentation equations (Kaur et al., 2017).
Control Theory and Optimization: Transformation of nonlinear optimal control problems into recursive linear two-point boundary-value problems, enabling explicit construction of state, costate, and optimal control law as convergent series (Jajarmi et al., 2014).
Fractional PDEs: Integration with Laplace or Natural Transform methods for time-fractional evolution and delay equations (Singh et al., 2016, Maitama, 2018).

In all such applications, HPM provides a framework where nonlinearity is handled recursively via driving terms, sidestepping the necessity for direct numerical iteration or small-parameter expansions.

6. Strengths, Limitations, and Methodological Extensions

Strengths:

General applicability to both linear and nonlinear initial/boundary-value problems.
Absence of any need for physical or artificial smallness parameter.
Reduction of the nonlinear problem to a sequence of linear subproblems, enabling analytical or semi-analytical solutions in closed form.
Rapid convergence demonstrated empirically in diverse contexts.
Straightforward incorporation of multi-point (including nonlocal) boundary conditions (Iftikhar, 2013).

Limitations:

No adaptive mechanism for convergence control: for strongly nonlinear, stiff, or highly nonanalytic systems, the HPM series may lack sufficient convergence radius at $N$ 1, necessitating use of HAM or introduction of auxiliary parameters (Xu, 19 Mar 2026).
Sensitivity to the choice of linear operator $N$ 2 and initial guess $N$ 3; poor choices can impede convergence.
Algebraic complexity increases for high-order terms or for strongly nonlinearities, especially in multi-variable settings.
Rigorous convergence proofs available only in specific norms and under analytic/smallness assumptions (Iftikhar, 2013).

Potential Methodological Enhancements:

Integration with Padé approximants or Adomian decomposition to expand the region of convergence or accelerate series summation (Iftikhar, 2013, Marinca et al., 2015).
Use of hybrid transforms (Laplace, Natural, Sumudu) for memory/nonlocal/fractional differential equations (Singh et al., 2016, Maitama, 2018).
Adoption of "optimal" variations (OHPM), introducing auxiliary functions or parameters to minimize global residuals and further improve convergence (Marinca et al., 2015).
Extension to multi-dimensional and strongly coupled systems via suitable operator and homotopy constructions.

7. Representative Case: Compact Stellar Objects via HPM

A concrete application is the use of HPM to solve the Tolman-Oppenheimer-Volkoff (TOV) equation for isotropic, spherically symmetric perfect fluids under a linear equation of state $N$ 4:

The TOV equation is reduced to an ODE for the mass function $N$ 5 with nonlinear terms in $N$ 6 (Aziz et al., 2019).
The homotopy is constructed by separating the linear terms ( $N$ 7) from the nonlinear remainder.
Power-series expansion with recursive determination of $N$ 8, etc., subject to regularity at $N$ 9.
The final approximate solution is a cubic (or higher) polynomial in $p\in[0,1]$ 0, with parameters determined by physical constraints (e.g., $p\in[0,1]$ 1 for surface matching).
The resulting models yield mass, compactness, redshift, and core-crust structure in agreement with known properties and numerically established maximum mass bounds (Aziz et al., 2019).

This approach generalizes naturally to broader astrophysical, cosmological, and field-theory contexts, where high accuracy and analytic insight are desirable.

HPM thus provides a powerful, general, and conceptually unifying approach for the analysis and construction of analytical approximations to nonlinear differential equations across physics and applied mathematics. Its role as a special case of HAM contextualizes its strengths and boundaries, guiding rational application and motivating further methodological refinement.