Analysis of stability and chaotic trajectories in nonlinear fluid wave interactions under forcing effects Authors: Tomas Kozubek, Muhammad Iqbal, Muhammad Bilal Riaz, Syeda Sarwat Kazmi, Muhammad Aziz ur Rehman Published: 2026 ABSTRACT AbstractThe Ablowitz–Kaup–Newell–Segur (AKNS) water wave equation is a fundamental model for describing nonlinear wave propagation in various physical systems, including fluid dynamics, plasma physics, and nonlinear optics. Its higher-dimensional form provides a more realistic framework for understanding complex wave interactions in dispersive media. In this work, the (2+1)-dimensional AKNS equation is studied from both analytical and dynamical perspectives. Analytical solutions of the governing equation are obtained using the modified Khater method, constructing different types of wave structures such as stationary and traveling profiles. These solutions illustrate the influence of nonlinearity and dispersion on the formation and evolution of coherent wave patterns. To further explore the system behavior, a detailed dynamical analysis is performed. The model is reduced to a system of ordinary differential equations, allowing the investigation of phase space dynamics through phase portraits, time series, Poincaré sections, bifurcation diagrams, and Lyapunov exponents. The results demonstrate a wide range of behaviors, including periodic, quasi-periodic, and chaotic motions, depending on the choice of system parameters. In particular, the system shows strong sensitivity to initial conditions and parameter variations, indicating the presence of complex and chaotic dynamics.Keywords: Exact solutions, Ablowitz-Kaup-Newell Segur model, Modified Khater method, Dynamical analysisSubject terms: Mathematics and computing, Physics Introduction Nonlinear partial differential equations (NLPDEs) form the basis for mathematical description and investigation of a large number of complex problems in mathematical physics, ocean engineering, fluid mechanics, and other related fields of science. They play a fundamental role in understanding the propagation of waves, instability, and pattern formation in a wide range of physical systems due to their ability to model nonlinear interactions1–3. These equations are especially interesting because they can describe complicated dynamic processes in both natural and man-made systems4. Furthermore, they also govern other physical processes like mass transport, energy propagation, and wave evolution and have been applied in many ways in different fields, including epidemiology, neural networks, economics, population dynamics, and solid-state physics. Analytical solutions are critical to comprehending the structure and physics of nonlinear wave systems, especially in fluid and water-wave systems. Motivated by these studies, much progress has been made in obtaining analytical solutions of NLPDEs, which has led to a deeper understanding of the behavior of nonlinear wave phenomena in various applications, such as oceanography and engineering. There are also numerous strong methods that have been proposed over the years, and they include Lie symmetry analysis5, the bilinear residual network method6, the modified auxiliary equation method7, the unified solver method8, the extended Sinh-Gordon expansion method9, the Jacobi elliptic function expansion method10, and the modified extended tanh-function method11, the bilinear neural network technique12,13, and the Hirota bilinear method14. Such methods have facilitated the creation of a broad category of nonlinear wave phenomena, including solitons, rogue waves, multi-wave interactions, lump solutions, and other localized excitations. Most of these solutions directly apply to water-wave dynamics, in which nonlinear dispersion and interaction between waves are important in both pattern formation and energy transport. Following these advances, there has been a growing interest in nonlinear models with rich dynamical behaviors in higher dimensions. Examples of these are the long Kadomtsev-Petviashvili equation15, the Landau-Ginzburg-Higgs equations16, the Fokas-Lenells model17, the Schrodinger equation18, and the Kaup-Boussinesq equations19. The models are especially important in the description of the wave propagation phenomena in the multidimensional systems of fluid and water waves, where the dispersion, nonlinearity, and external perturbation interactions contribute to the appearance of complex wave patterns. The development of soliton theory could be traced back to the discovery of solitary water waves by John Scott Russell20, who published the “Great Translation Wave” in 1834. He was initially skeptical, but this observation of shape preservation of the wave was later confirmed by Rayleigh and Boussinesq21,22, respectively. This discovery led to the development of the Korteweg-de Vries (KdV) equation by Korteweg and de Vries (1895)23, which provides a well-developed mathematical description of wave propagation in shallow water and soliton formation. Subsequent other findings in nonlinear science came with the Fermi-Pasta-Ulam (FPU) experiment (1951), which revealed unexpected recurrence phenomena in nonlinear lattices and the intricacies of energy transfer in nonlinear wave systems24. More recently, Zabusky and Kruskal25 showed that solitary waves collide elastically and retain their shapes and velocities, thus introducing the modern notion of solitons, which is of great importance for water-wave theory and nonlinear fluid dynamics. In particular, the AKNS formulation, which is based on the inverse scattering transform (IST)26, is fundamental in the construction of exact solutions to nonlinear evolution equations, many of which are related to water waves in dispersive and stratified fluids. Here, we take into account the following fourth-order nonlinear AKNS-type water wave equation27 with a small perturbation B: This equation governs the evolution of the wave field Z(x, y, t) in a (2+1)-dimensional water-wave framework. It models the interaction between nonlinear effects and dispersive mechanisms essential in shallow water waves, internal waves, and other fluid systems. The term represents mixed spatiotemporal evolution, while introduces higher-order dispersion responsible for wave spreading. The nonlinear coupling terms and describe energy transfer between spatial directions, capturing multidimensional wave interaction effects. The factor B is a physical modulator of wave amplitude and wave restoration. In general, the equation represents a combination of nonlinearity and dispersion, which is responsible for the formation of coherent structures (solitons, kink waves, and periodic solutions) in water-wave systems. This family of equations is relevant to fluid mechanics and water-wave physics; therefore, it has been widely researched in the fields of nonlinear optics, plasma physics, and hydrodynamics. It has been used to describe various physical phenomena in dispersive systems, such as wave interactions, energy transport, and stability. Dusunceli28 applied the improved Bernoulli sub-equation method to obtain new exact solutions of the AKNS water wave equation, while Aktürk et al.29 used the modified exponential function method to derive traveling wave solutions of the fourth-order AKNS equation with graphical representations. Similarly, Ahmad et al.30 employed the Power index method to obtain exponential, trigonometric, and logarithmic solutions and showed that the AKNS equation can be reduced to a linear ordinary differential equation under suitable transformations. Duru et al.31 applied the (1/G')-expansion method to obtain hyperbolic and singular soliton solutions of the (2+1)-dimensional AKNS equation, while Gao et al.32 derived dark–bright soliton solutions using the sine-Gordon expansion method. Rani et al.33 further obtained various exact solutions using an improved expansion technique, supported by graphical analysis. Inspired by these advances, the present work is dedicated to the analytical construction of exact soliton solutions and a detailed examination of the corresponding dynamical behavior in a water-wave form. We use the modified Khater method (MKM)34, which has not been previously used to solve this equation, to derive a broad range of exact solutions, such as kink, singular, bright, dark, and periodic solitons. The analytical solution of this method is also illustrated graphically in order to provide a physical interpretation of the wave propagation in fluid mediums. From the dynamical systems perspective, bifurcation explains the sudden change in the nature of waves in response to some physical parameter, often accompanied by a change in stability, instability or chaos. Such changes are particularly important in water-wave systems where the influence of the external forcing, as well as the nonlinear effects, are predominant in determining the nature of the wave propagation. In recent years, periodic waves, solitons, and shockwaves have been shown to be commonly observed in plasma systems as well as fluid systems35. In particular, we can see the same nonlinear effects in water-wave systems when exposed to external conditions. The work of Saha et al.36, Abdikian et al.37, and El-Taibany et al.38 has observed the complex behavior of waves in different physical systems, such as periodic structures and nonlinear stability transitions. Unlike previous independent investigations, in the present study the approach is to consider the dynamics from a coherent dynamical systems and chaos-theoretical perspective on the nonlinear behavior of water waves. The inclusion of an external forcing term gives rise to highly nonlinear responses as the system is sensitive to parameters such as the ratio of two frequencies and strength of the perturbation. This results in the transition of regular wave dynamics, periodic and chaotic dynamics, which give a better understanding of the nonlinear dynamics of water waves. To systematically analyze these behaviors, we investigate several key dynamical features, including: Phase portraits (PPs) structures such as centers, saddles, and cusp points, Phase portraits (PPs) structures such as centers, saddles, and cusp points, Bifurcation diagrams (BDs), Bifurcation diagrams (BDs), Chaotic behavior (CB), Chaotic behavior (CB), Time series (TS), Time series (TS), Poincaré maps (PMs), Poincaré maps (PMs), Lyapunov exponents(LEs), Lyapunov exponents(LEs), Multistability phenomena(MS), Multistability phenomena(MS), Sensitivity analysis (SA) to initial conditions(ICs). Sensitivity analysis (SA) to initial conditions(ICs). This paper is structured as follows. Section 9c Graphical representation of system’s trajectories of system (22). Implementation of analytical methods Analytical methods for obtaining exact solutions of NPDEs play a crucial role in applied mathematics and theoretical physics. In the present study, the MKM is utilized to derive exact solutions, including soliton structures, periodic waveforms, and rational-type solutions. The general procedure of this method begins with the application of a traveling wave transformation, which reduces the original nonlinear partial differential equation into an ordinary differential equation. Execution of Modified Khater Method In this section, we describe the algorithm of the MKM34,39 to obtain the solitary wave solutions of NLPDEs. The following main steps are involved: Step (1): Consider the NLPDE in the form. where represents a polynomial function of Z(x, y, t) , and its partial derivatives in which the highest order derivatives and nonlinear terms are involved. Step (2): Consider traveling wave transformation where w denotes the wave speed, while x and y correspond to the spatial variables. Substituting Eq. (3) into Eq. (2) reduces the NLPDE to the following NODE: Step (3): Assume a general solution of the form: where are constant coefficients and satisfies an auxiliary differential equation. Specifically, is assumed to satisfy with being real constants. Step (4): By applying the homogeneous balance principle between the highest-order linear derivative and the highest-order nonlinear terms in Eq. (4), the positive integer N is determined to be N=1. Consequently, Eq. (5) simplifies to Step (5): After that substitute the Eq. (7) with Eq. (6) into Eq. (4) and setting the polynomial coefficients equal to zero, we have Step(6): The constant parameters are determined through symbolic computation by solving the resulting algebraic system. All symbolic calculations and algebraic manipulations are performed using Maple 2025.2. Based on these computations, the analytical solutions are consequently obtained as follows: Substituting the above solutions of parameters into Eq. (7) and applying the transformation given in Eq. (3), the resulting solutions of Eq. (2) are obtained as follows: Case 1: If and Case 2: If and Case 3: If and Case 4: If and , Case 5: If and , Case 6: If and , Case 7: If , Case 8: If , and , Case 9: If Case 10: If Case 11: If and , Case 12: If and Wave solution plots and discussion This section illustrates various graphical patterns of Eq. (1) obtained using the computational software MATLAB (2018). The derived solutions are expressed in terms of trigonometric, hyperbolic, and Jacobi elliptic functions. Based on their physical and structural characteristics, these solutions are categorized into four main types: solitary waves, bright solitons, dark solitons, and periodic wave structures. Solutions in terms of Jacobi elliptic functions are a special type of solution that is important in solving mathematical physics problems due to their ability to represent periodic and soliton solutions. These include, to name a few, bright soliton solutions, which are highly excited and localized waves with amplifications. These are stable and preserve their shape as they propagate; they are useful in the dynamics of strong signals and high-intensity waves in applications such as plasma physics and nonlinear optics. Dark soliton solutions are associated with the dip in amplitude in a continuous wave train background. These solutions are generally low-energy states and can be employed to study low-intensity waves and their interactions, as well as weak signals. Their dip property makes them significant in studying nonlinear optical systems and other processes, particularly in understanding how these low-energy states can influence the propagation of light and the behavior of weak signals in various applications. On the other hand, periodic solutions are associated with periodic, or repeating, waves, which are important in modeling naturally periodic processes, such as ocean waves and oscillating systems. In addition, the singular solutions relate to the points where it diverges. These solutions might not always be directly physically meaningful but can provide valuable information on important transitions, instabilities, and constraints of the mathematical model, such as how they can reveal critical points in the evolution of nonlinear waves. Over all, the different solution structures provide a comprehensive framework for studying nonlinear waves and their dynamics, in which strong and weak signals, periodic solutions, and even the limitations of the mathematical model associated with the AKNS equation can be studied. Figure 1 shows the 3D surface, 2D line, and contour plots of the solution from Eq. (9). These plots are made using the parameter values , , , v=3, w=1, B=2, and y=0. These plots show that the soliton solution is periodic. Figure 2 shows the 3D, 2D, and contour views of the solution from Eq. (10) in the same way. These graphs show the kink-type behavior of the soliton solution when the parameters are set at , , , v = 3, w = 1, B = 2, and y = 0. Figure 3 shows three-dimensional, two-dimensional, and contour plots of the solution from Eq. (11). These graphs clearly show that the soliton solution is periodic in the Jacobi sense when the parameters are set at , , , v=3, w=1, B=2, and y=0. Figure 4 also shows the 3D, 2D, and contour plots of the solution from Eq. (11). These visualizations show the multiple singular behaviors that come with the soliton solution when the same parameters are used. Figure 5 shows how the solution from Eq. (12) behaves as a kink in three-dimensional, two-dimensional, and contour plots. The parameter values , , , v=3, w=1, B=2, and y=0 are used to make these plots. Figure 6 shows the 3D, 2D, and contour plots of the solution from Eq. (13) in the same way. The graphs indicate that the soliton solution has periodic properties when you set , , , v=3, w=1, B=2, and y=0. Figure 7 shows the contour plot and the three-dimensional and two-dimensional views of the solution from Eq. (14). These results are produced with the parameter values , , , v=3, w=1, B=2, and y=0, demonstrating the periodic characteristics of the soliton solution. Figure 8, on the other hand, displays the singular features of the solution from Eq. 15. These three-dimensional, two-dimensional, and contour plots are constructed by using the values , , , v=3, w=1, B=2, and y=0. Shows the periodic wave solution of in 3D, 2D, and contour graphs. Reveals the kink behavior of the dynamic wave solution . Discloses the Jacobi periodic characteristics of the soliton solution in 3D, 2D, and contour plots. Shows the multiple singular behaviors of the dynamic wave solution in 3D, 2D, and contour graphs. Indicates the kink-type depiction of through 3D, 2D, and contour graphs. Reveals the periodic wave structure of in 3D, 2D, and contour graphs. Displays the bright soliton properties of with 3D, 2D, and contour graphs. Indicates the singular wave behavior of with 3D, 2D, and contour graphs. Dynamical analysis of the AKNS water wave dynamical equation Dynamical analysis is the study of the time-dependent behavior of nonlinear systems and assists in the study of their stability and complexity. It entails PPs, BDs, CBs, LEs, TS, PMs, MS, and SA. These tools show the stability, periodicity, quasi-periodicity, and chaos of solutions at various parameter values. Phase portraits analysis To analyze the dynamical behavior of the system, PPs are employed to illustrate the evolution of the state variables in phase space. For this purpose, the third-order NODE given in Eq. (4) is reduced to a lower-order system by introducing the transformation . As a result, the original equation is converted into a second-order system, which can be written as follows: Applying the Galilean transformation, Eq. (21) is reformulated as a system of coupled first-order differential equations by setting S' = V and S'' = V' as: Where and . The above system exhibits Hamiltonian characteristics and admits the following first integral: where g is the Hamiltonian constant. The fact that the Hamiltonian function provides a clear and detailed explanation of the overall energy dynamics of a physical phenomenon illustrates its prominence. Both classical and quantum physics are based on it, and it is possible to develop equations of motion and predictions of the system’s long-term behavior. Furthermore, the development of both analytical and computational methods of studying and modeling physical structures relies on the Hamiltonian. The planar dynamical system has a vector field that restricts its phase trajectories. We are therefore forced to investigate a large number of system phase profiles (22) with different parameter settings. Two equilibrium points and exist for the system (22) about the non zero components and . By satisfying the following conditions. the system (22) regarded as Hamiltonian with the potential function. The stability and classification of the equilibrium points depend on the sign of the Jacobian determinant, as summarized in Table 1. A positive determinant typically indicates center behavior, a negative determinant corresponds to saddle behavior, while a zero determinant represents the parabolic case, marking a transition between stability types. The corresponding Jacobian matrix of the system takes the form: whose determinant is given by Table 1Classification of equilibrium points based on the Jacobian determinant.DeterminantCenterSaddleCuspid\det (J(S,V))> 0< 0= 0 Classification of equilibrium points based on the Jacobian determinant. By assigning different values to the parameters involved, a variety of qualitative behaviors can be observed in the system. These results are summarized in Table 2. Table 2Phase portrait classification for different sign combinations of and .CaseFigures(1) CenterSaddleFig. 9a(2) CenterSaddleFig. 9b(3) SaddleCenterFig. 9c(4) SaddleCenterFig. 9d Phase portrait classification for different sign combinations of and . Case 1: For and , Fig. 9a takes the parameters B= 2 and w= 2, system 22 reveals two critical points: and . In this particular configuration, behaves as a center point while acts as a saddle. These critical points are represented in Fig. 9a. and are stable and unstable points, respectively. Case 2: For and , Fig. 9b reveals two critical points: and by taking the parameters B= 5 and w= -1. In this particular configuration, behaves as a center point while behaves as a saddle. These critical points are represented in Fig. 9b. and are stable and unstable points, respectively. Case 3: For and , Fig. 9c with the parameters B= 2 and w= 2, system 22 also reveals two critical points: and . In this particular case, behaves as a saddle point while acts as a center point. These critical points are represented in Fig. 9c. and are unstable and stable points, respectively. Case 4: For and , Fig. 9d chose the values w= -1 and B=-16, in system (22). In this particular configuration, behaves as a saddle point while acts as a center. Once again presents two critical points. and , represents the saddle point, whereas indicates the center point in Fig. 9d. In these portraits the central points equilibrium correspond to a stable oscillatory state, where the wave system remains bounded and exhibits closed orbits in its PP. From a physical perspective, this is a situation where water waves are confined to a steady profile and grow and decay around that profile (for example, small-amplitude water waves in a tank). The saddle points are an unstable equilibrium. Any small deviations from this condition will amplify and result in different dynamical behaviors. For a physical water wave, the saddle point may represent a critical wave setup that is very unstable to perturbations, such as the initiation of wave breaking or the change in the system from regular wave dynamics to more chaotic or disordered wave dynamics. Quasi-periodic and chaotic patterns Chaos in dynamical systems is a type of motion that is very intricate and unpredictable, even when the system is governed by strict and deterministic laws. One of the most essential characteristics of chaotic dynamics is how sensitive they are to the ICs. This means that even small changes in the IC can lead to significantly different results over time. The scientific community agrees that weather is chaotic, with a finite predictability, and that the source of chaos is nonlinearity. People typically name this attribute the butterfly effect40–44. Therefore, even small changes grow bigger and bigger, making it impossible to make long-term predictions. To investigate these CBs, an external driving factor is included in system (22). The value of indicates the intensity, whereas the value of Λ signifies the frequency of the perturbed system (24). Therefore, this section analyzes the chaotic and quasi-periodic patterns of the understudy model via adding a perturbation term to the planar dynamical system. This system is autonomous and can be expressed in the form of dynamical equations as . To investigate its qualitative behavior, both two- and three-dimensional PPs, along with TS analysis, are employed. These graphical tools provide valuable insight into the nature of the system, allowing the identification of periodic, quasi-periodic, and chaotic regimes. The main features of the perturbed system dynamics are captured by varying the physical parameters. In particular, the parameters and Λ are important in controlling the system dynamics, while the parameters and are fixed. The PPs corresponding to and Λ = 0.3 are presented in Fig. 10. For these small amplitudes and frequencies of the forcing, the system has periodic motion. The shape of the trajectories is smooth, closed, and periodic, suggesting a stable motion. For water waves, this type of motion is related to the propagation of regular and harmonic surface waves, where wave profiles are stable and the water wave field has a uniform oscillatory structure (e.g., shallow water waves with a weak disturbance). The two- and three-dimensional PPs in Fig. 11 are for and Λ = 0.9. Here, the system is quasi-periodic as a result of a stronger forcing. The solution stays periodic, but it is no longer strictly periodic, as various wave frequencies start to interact. This phenomenon can be physically interpreted as the interaction of water wave modes with different, but close, frequencies giving rise to complex but structured wave formations like a modulated wave train, which is frequently observed in the ocean and along coasts. Figure 12 shows the phase portraits for the values of and . In this case, the system remains quasi-periodic, but the phase portraits are distorted and irregular. This suggests the system is approaching a dynamical instability. In the context of water waves, this refers to an increase in nonlinear interactions between the wave modes, where wave packets distort and energy starts to transfer between the modes, leading to irregular oscillations of the water surface while still being bounded. In Fig. 13, we show the system dynamics for and , where the system is chaotic. This motion is very erratic, non-periodic, and highly dependent on the ICs. This leads to very different wave evolution despite the small change of the initial wave profile. Physically, it is connected to the chaotic water-wave dynamics, such as strongly disturbed sea states, turbulent wave interactions, and chaotic surface wave dynamics in a highly forced external environment (wind or pressure variations). Surprisingly, this chaotic phenomenon is not only limited to theoretical water-wave systems but can also be often observed in real water-wave systems. The chaotic behavior of water surface waves in stormy seas, the nonlinear wave interactions, and the energy redistribution in turbulent sea states are prevalent. These observations have verified that water wave systems can change from periodic to chaotic behavior as a result of forcing and nonlinear interactions. From a physical perspective, this periodic-chaotic transition of the wave is a consequence of the universal nature of nonlinear systems. It underlines that water wave propagation is a very sensitive parameter, and the analysis of these transitions is crucial in modeling complex ocean and coastal waves. The variation of parameters and their corresponding dynamical regimes are summarized in Table 3. Exposure of periodic dynamical behavior of the perturbed system (24) and model parameters , , and Λ = 0.3. Identification of the perturbed system’s (24) quasi-periodic dynamical behavior and model parameters , , and Λ = 0.9. Identification of the perturbed system’s (24) quasi-periodic dynamical behavior and model parameters , , and . Detection of chaotic dynamical behavior of the perturbed system (24) and model parameters , , and . Table 3Summary of dynamical behaviors of the system for different forcing intensity and frequency Λ values. The results illustrate the transition from periodic to quasi-periodic and finally chaotic dynamics, as confirmed by the PPs in Figs. 10, 11, 12 and 13.CaseParameter Values Observed DynamicsCorresponding figure1Periodic motion: smooth, closed, and repeating oscillationsFig. 102Quasi-periodic motion: bounded, non-repetitive oscillationsFig. 113Quasi-periodic motion with irregular trajectories (near chaos)Fig. 124Chaotic motion: irregular, non-repeating, highly sensitive dynamicsFig. 13 Summary of dynamical behaviors of the system for different forcing intensity and frequency Λ values. The results illustrate the transition from periodic to quasi-periodic and finally chaotic dynamics, as confirmed by the PPs in Figs. 10, 11, 12 and 13. Time series TS is a method employed in the dynamical system to determine recurrent systems and distinguish the behavior such as periodic, quasi-periodic, and chaotic. The perturbed system (24) is investigated by changing the intensities and frequencies as in Fig. 14. The outputs of the systems show that, an increase or decrease in the intensity causes substantial changes in the behavior of the system and irregular appearances of chaotic motion. Also Fig. 14 indicates that the specified system is highly susceptible to the ICs; this is a description from periodic to chaotic dynamics. Chaotic reaction of system (24) via time series at optimal parameters , , and diverse values of and Λ. Analysis of time series plots Figure 14 demonstrates time series responses of system (24) at different parameter values, and the development of periodic behavior into chaotic behavior. Analysis of periodic behavior In Fig. 14a of the waveform at the values of 0.001 of and 0.3 of Λ indicates that the waveform is regular and repetitive with constant amplitude and frequency, which is a purely periodic motion. The periodic regimes are identical in cycles and the signal repeats after a fixed duration of time and this is typical of a stable, periodic regime. This act can be compared to the continuous movements of a simple pendulum or an undamped harmonic oscillator of which the movement is completely predictable with time. Analysis of quasi periodic behavior The parameter of the wave form is changed to a larger value of is 0.2 and the parameter of the waveform is changed to the value of the period of the waveform Λ, which is π, the waveform begins to lose its strict periodicity as shown in Fig. 14b,c. The amplitude and frequency start to vary slightly and the vibrations get more chaotic and this is a quasi-periodic response. The system is now two or more incommensurate frequencies interacting to give a complex and yet bounded oscillation. In this regime, the signal is not repeated as such, but instead it is structured, as in the example of the quasi-periodic oscillations of a coupled mechanical system or a multi-frequency driven electrical circuit. Analysis of chaotic behavior In Fig. 14d, the parameters and \varLambda of the waveform are highly jagged and aperiodic with haphazard changes in amplitude and frequency. This is a completely chaotic time series, where the system is sensitive to initial conditions and long-term unpredictable. The movement appears to be random, but it remains within certain limits, which is characteristic of deterministic chaos. Such chaotic behavior is common to both natural and artificial systems, including weather dynamics, turbulent fluid flows, electrical oscillators, and population models in ecology, where even minor changes in parameters lead to unbelievably different long-term behavior. Poincaré map The PMs creates a geometrical visualization of the trajectory of the system by recording its state at discrete intervals, allowing deeper understanding of the underlying dynamics. The presence of a closed orbit like shape of dots on a map, suggest quasi-periodic activity. However, when the scattering of dots is not uniform or when the pattern is non-cohesive then this is a sign of chaotic dynamics, which are highly sensitive to ICs and unpredictable. Figure 15a–d is a representation of PMs of the two-dimensional plots of the perturbed system, which are applied to describe the behavior of dynamical systems. In this instance, every point defines the situation of the system at a certain stage of the cycle and abnormalities and peaks that can be articulated are able to lead to chaos. In Fig. 15a, if , , and Λ =0.7, the plot is relatively normal, which suggests that the system behaves nearly periodically. In Fig. 15b, when , , and varying and Λ =0.9, the plot looks more and more crowded and seems to be periodic in nature. In Fig. 15c, for , , and , the structure is intricate, also suggesting signs of quasi-chaotic behavior. In Fig. 15d, for , , and , the structure is much more intricate, also suggesting signs of quasi-chaotic behavior. These maps are indicative of how the system behaves depending on the ICs and chaos that ensue when parameters are adjusted. Sometimes when a dynamical system is exposed to disturbances, it responds with multistability. Analysis of Poincaré maps Figure 15 demonstrates the validation of chaotic behavior in system (24) using PMs for different values of , and Λ, with parameters , , as specified. The PM provides a cross-sectional view of the system’s trajectories by recording their intersections with a specific phase plane, allowing the complex temporal evolution to be analyzed in a reduced-dimensional space. Chaotic behavior in system (24) is identified via the Poincaré map for , , and varying and Λ. Figure 15a: Periodic motion Figure 15a shows that the points in Figure are on a single closed curve in case and Λ = 0.5, indicating the existence of a periodic orbit. This periodicity is a sign that the system will revert to its motion after some time such as the one of a simple harmonic oscillator or a perfect pendulum with constant energy. Figure 15b,c: quasi periodic motion When the control parameter, , is further increased to 0.8 and the value of Λ = 1.5, the single curve is further divided into a series of nested loops, indicating quasi-periodic behavior as in Fig. 15b,c. The path is no longer going to the same point but rather it follows a toroidal shape in the complete phase space. This regime has two or more incommensurate frequencies and the motion is confined but not repetitive. This kind of quasi-periodic response occurs in coupled oscillatory systems or electrical circuits that are excited by many independent frequencies. Figure 15d: chaotic motion Further increase in the value of the parameter, , of 2.14 and frequency Λ to 2π, Fig 15d leads to an extreme change in the PM structure. The continuous curves which previously were discontinued by a sprinkling of points occupying a large region of the phase plane, indicating the onset of chaos. Such scattered-around fractal-like distribution is a property of chaotic dynamics, the trajectories of which are very sensitive to initial conditions and never recur, yet lie within a finite area of the phase space. This change is a sign of the direction of the system towards irregular behavior as nonlinear effects replace. Lyapunov exponent LE is a significant parameter with regard to the dynamic systems, which quantifies how successive paths within a system evolve naturally. This is a scalar that is a key measure of the inherent chaos of the system. A negative value of LE shows a steady state of the system but a positive LE shows a chaotic state. In our analysis we used Wolf technique to calculate the LE of the system in question including Gram-Schmidt operationalization. A further plot of the change of these exponents with time then gave us clarifications of the functioning of the system. Using the values , , , Λ =0.3, and the initial (0.7, 0.4, 0.7), we find out the modified system behaves chaotically by displaying the LE over time in Fig. 16a. Likewise, in Fig. 16b, the LE is shown with updated parameters as follows: , , , , and the initial condition (0.3, -0.3, 0.3). This illustration demonstrates that the perturbed system (24) is chaotic. To examine the dynamical behavior of the perturbed system, the Lyapunov exponents (LEs) are computed numerically over the time interval t ∈ [1,80]. Tables 4 and 5 present the evolution of the three exponents at different time instances. The results reveal that remains positive, stays negative, and is approximately zero throughout the considered time span. This distribution of the Lyapunov spectrum indicates the presence of a chaotic attractor, where the system exhibits exponential divergence in one direction and contraction in another, while preserving the phase space volume due to the zero exponent. Chaotic response of system (24) through the Lyapunov Exponent with an appropriate choice of parameters , , and different values of and Λ. Table 4Computation of Lyapunov exponent values for perturbed dynamical system using time span [0, 80], , , and Λ =0.3..TimeTime1.00000.618342-0.6183420.00000041.00000.077016-0.0770160.0000002.00000.289928-0.2899280.00000042.00000.073492-0.0734920.0000003.00000.309359-0.3093590.00000043.00000.067316-0.0673160.0000004.00000.289562-0.2895620.00000044.00000.067752-0.0677520.0000005.00000.327875-0.3278750.00000045.00000.076578-0.0765780.0000006.00000.177488-0.1774880.00000046.00000.063489-0.0634890.0000007.00000.189961-0.1899610.00000047.00000.061761-0.0617610.0000008.00000.221979-0.2219790.00000048.00000.061088-0.0610880.0000009.00000.234959-0.2349590.00000049.00000.064599-0.0645990.00000010.00000.256433-0.2564330.00000050.00000.073913-0.0739130.00000011.00000.196005-0.1960050.00000051.00000.063166-0.0631660.00000012.00000.173832-0.1738320.00000052.00000.060815-0.0608150.00000013.00000.149072-0.1490720.00000053.00000.056345-0.0563450.00000014.00000.155744-0.1557440.00000054.00000.055643-0.0556430.00000015.00000.173503-0.1735030.00000055.00000.064121-0.0641210.00000016.00000.133776-0.1337760.00000056.00000.053007-0.0530070.00000017.00000.127369-0.1273690.00000057.00000.051446-0.0514460.00000018.00000.125285-0.1252850.00000058.00000.052002-0.0520020.00000019.00000.138413-0.1384130.00000059.00000.053386-0.0533860.00000020.00000.152352-0.1523520.00000060.00000.061973-0.0619730.00000021.00000.131729-0.1317290.00000061.00000.052227-0.0522270.00000022.00000.120891-0.1208910.00000062.00000.049716-0.0497160.00000023.00000.106161-0.1061610.00000063.00000.046240-0.0462400.00000024.00000.108943-0.1089430.00000064.00000.044645-0.0446450.00000025.00000.120162-0.1201620.00000065.00000.052402-0.0524020.00000026.00000.099423-0.0994230.00000066.00000.043577-0.0435770.00000027.00000.094404-0.0944040.00000067.00000.041357-0.0413570.00000028.00000.090227-0.0902270.00000068.00000.042167-0.0421670.00000029.00000.099308-0.0993080.00000069.00000.042384-0.0423840.00000030.00000.110129-0.1101290.00000070.00000.049278-0.0492780.00000031.00000.097474-0.0974740.00000071.00000.041462-0.0414620.00000032.00000.091344-0.0913440.00000072.00000.038058-0.0380580.00000033.00000.082226-0.0822260.00000073.00000.034711-0.0347110.00000034.00000.083761-0.0837610.00000074.00000.032476-0.0324760.00000035.00000.093140-0.0931400.00000075.00000.039191-0.0391910.00000036.00000.077803-0.0778030.00000076.00000.032402-0.0324020.00000037.00000.074824-0.0748240.00000077.00000.029093-0.0290930.00000038.00000.072308-0.0723080.00000078.00000.028529-0.0285290.00000039.00000.078359-0.0783590.00000079.00000.028185-0.0281850.00000040.00000.088107-0.0881070.00000080.00000.033504-0.0335040.000000 Computation of Lyapunov exponent values for perturbed dynamical system using time span [0, 80], , , and Λ =0.3.. Table 5Computation of Lyapunov exponent values for perturbed dynamical system using time span [0, 80], , , and .TimeTime1.00000.429957-0.4299570.00000041.00000.008508-0.0085080.0000002.00000.102987-0.1029870.00000042.00000.015528-0.0155280.0000003.00000.105834-0.1058340.00000043.00000.011117-0.0111170.0000004.00000.075147-0.0751470.00000044.00000.016902-0.0169020.0000005.00000.047670-0.0476700.00000045.00000.011421-0.0114210.0000006.00000.057671-0.0576710.00000046.00000.013995-0.0139950.0000007.00000.027924-0.0279240.00000047.00000.013327-0.0133270.0000008.00000.054766-0.0547660.00000048.00000.011889-0.0118890.0000009.00000.002755-0.0027550.00000049.00000.017396-0.0173960.00000010.00000.041722-0.0417220.00000050.00000.008887-0.0088870.00000011.00000.021969-0.0219690.00000051.00000.014753-0.0147530.00000012.00000.026222-0.0262220.00000052.00000.011163-0.0111630.00000013.00000.029764-0.0297640.00000053.00000.015915-0.0159150.00000014.00000.014382-0.0143820.00000054.00000.011800-0.0118000.00000015.00000.026150-0.0261500.00000055.00000.013352-0.0133520.00000016.00000.013993-0.0139930.00000056.00000.013103-0.0131030.00000017.00000.028627-0.0286270.00000057.00000.011603-0.0116030.00000018.00000.006438-0.0064380.00000058.00000.016611-0.0166110.00000019.00000.023054-0.0230540.00000059.00000.009227-0.0092270.00000020.00000.015694-0.0156940.00000060.00000.014185-0.0141850.00000021.00000.016505-0.0165050.00000061.00000.011157-0.0111570.00000022.00000.022535-0.0225350.00000062.00000.015214-0.0152140.00000023.00000.009274-0.0092740.00000063.00000.011888-0.0118880.00000024.00000.019365-0.0193650.00000064.00000.012915-0.0129150.00000025.00000.011572-0.0115720.00000065.00000.012866-0.0128660.00000026.00000.021503-0.0215030.00000066.00000.011396-0.0113960.00000027.00000.009008-0.0090080.00000067.00000.015930-0.0159300.00000028.00000.017431-0.0174310.00000068.00000.009467-0.0094670.00000029.00000.014126-0.0141260.00000069.00000.013712-0.0137120.00000030.00000.013538-0.0135380.00000070.00000.011085-0.0110850.00000031.00000.019831-0.0198310.00000071.00000.014661-0.0146610.00000032.00000.008352-0.0083520.00000072.00000.011796-0.0117960.00000033.00000.016795-0.0167950.00000073.00000.012569-0.0125690.00000034.00000.011123-0.0111230.00000074.00000.012606-0.0126060.00000035.00000.018490-0.0184900.00000075.00000.011215-0.0112150.00000036.00000.010574-0.0105740.00000076.00000.015312-0.0153120.00000037.00000.015120-0.0151200.00000077.00000.009611-0.0096110.00000038.00000.013597-0.0135970.00000078.00000.013288-0.0132880.00000039.00000.012401-0.0124010.00000079.00000.010958-0.0109580.00000040.00000.018383-0.0183830.00000080.00000.014195-0.0141950.000000 Computation of Lyapunov exponent values for perturbed dynamical system using time span [0, 80], , , and . Multistability This section explores the phenomenon of multistability in systems (24). Whenever a perturbation term is introduced in a planar system, the system has the potential to exhibit multistability, which indicates the possibility of many alternative dynamic behaviors. These behaviors are observed under different initial conditions but with the same parameter values. The system exhibits extraordinary multistability, allowing for many dynamic patterns to emerge simultaneously. This spectrum of behaviors includes chaotic, quasi-periodic, and periodic behavior, which can occur under different initial conditions. Figure 17 illustrates the dynamical responses of system (24) under different initial conditions and parameter settings. In Fig. 17a, the system exhibits periodic behavior for the initial conditions (0.2, 0.2), (0.2, 0.1), and (0.1, 0.1), represented by green, magenta, and yellow curves, respectively. This behavior is observed for the parameter values , , , and Λ = 0.3. In Fig. 17b, the system transitions to a quasi-periodic state under the same set of initial conditions, but with parameters , , , and Λ = 0.5. A further change in parameters leads to more complex dynamics. Specifically, Fig. 17c demonstrates the onset of chaotic behavior for , , , and , while maintaining the same initial conditions. Finally, Fig. 17d confirms fully developed chaotic dynamics for , , , and . These results clearly show that the system’s behavior evolves from periodic to quasi-periodic and eventually to chaotic as the control parameters vary, highlighting its strong sensitivity to both initial conditions and parameter changes. Analysis of chaotic phenomena in the system 24 through multistability plots. Analysis of multistabilty plots Figure 17 illustrates the calculation of multistability of the perturbed system (24) at the parameters value of , and distinct values of the parameter,, and Λ. The various sub-figures are linked to various groups of initial conditions as indicated by the color-coded trajectories. IC 1 (green) S = 0.2, V =0.2 IC 1 (green) S = 0.2, V =0.2 IC 2 (magenta) S = 0.2, V =0.1 IC 2 (magenta) S = 0.2, V =0.1 IC 3 (yellow) S = 0.1, V =0.1 IC 3 (yellow) S = 0.1, V =0.1 Figure 17a indicates that the system exhibits periodic oscillations as smooth and closed curves repeating after every time. This behavior implies that the system has settled upon a single periodic attractor that is stable and the system behaves in a predictable way and is confined to a simple limit cycle. In case the initial conditions are different, the system begins to exhibit more complex patterns as it is illustrated in Fig. 17b,c. Here, the trajectories of different initial conditions are in the same parameter regime yet they are in different regions of the phase space. The intersecting, interwoven loops suggest that there are a few stable attractors, and each of them defines the evolution of the system, depending on the initial point of its location. This is a sign of quasi-periodic and weakly chaotic processes, where the oscillations remain confined, but they are not periodic. Multistability is the property of being able to coexist between different motion patterns with the same parameter. Figure 17d indicates that the trajectories are highly irregular and fill the phase space densely signifying the occurrence of chaotic behavior. The orbits are no longer smooth and repeating but rather a complex web structure which is suggestive of sensitivity to initial conditions. It means that even the tiniest changes in the initial condition may lead to the system being totally different in its long-term behavior. Multistability Multiple attractors coexist. In general, it can be seen in the multistability plots of Fig. 17 that the system can have multiple attractors that coexist at the same set of parameters. According to the initial conditions, the system can change its state to a periodic and a quasi-periodic or chaotic movement, which underlines the richness and sensitivity of nonlinear dynamical systems. Bifurcation diagram A dynamical system changes its orbit structure, and such a change is a result of a variation of one or several parameters that are characteristic of a differential equation system and is called a bifurcation of the differential equation system. A bifurcation diagram shows how the qualitative behavior of a dynamic system varies when a parameter is altered. It is a very useful tool for finding changes between distinct dynamic regimes, such as when motion goes from steady-state to oscillatory or chaotic. In the context of AKNS model, bifurcation occurs when variations in system parameters alter the number and stability of solutions to the underlying differential equations. Specifically, bifurcation behavior in AKNS is analyzed by plotting the state variable S against the variable V. As parameters like , , , and are applied, the system displays distinct dynamical patterns. As shown in Fig. 18, the region of S-axis between 0.2 and 0.4 corresponds to periodic motion, 0.4 to 0.8 shows quasi-periodic behavior, while the interval from 0.8 to 1.5 indicates the presence of chaotic dynamics. Analysis of chaotic phenomena in the dynamical system (24) through bifurcation diagram. Sensitivity analysis Sensitivity analysis is an effective method of assessing the effect of changes in input parameters on the output of the system and the strength of the results in various initial conditions. This technique provides valuable data about the effect of changes in initial conditions on the system dynamics45,46. The sensitivity analysis of the present model indicates that the smallest variations in the initial conditions lead to the radical variations in the system behavior. To carry out this analysis, the Runge-Kutta method is employed to numerically investigate the sensitivity properties of the Eq. 4. The corresponding perturbed dynamical system is expressed as follows: To examine the sensitivity of the proposed system, three sets of initial conditions were considered. The magenta trajectory corresponds to (S, V) = (0.1, 0), the dashed yellow trajectory to (S, V) = (0.2, 0), and the green trajectory to (S, V) = (0.3, 0). In Fig. 19a, the magenta and yellow curves reflect the instances (S, V) = (0.1, 0) and (S, V) = (0.2, 0), respectively. Figure 19b shows solutions for (S, V) = (0.1, 0) and (S, V) = (0.3, 0). Figure 19c shows solutions for (S, V) = (0.2, 0) and (S, V) = (0.3, 0). Finally, Fig. 19d combines all three cases: (0.1, 0), (0.2, 0), and (0.3, 0). These results indicate that even small variations in the initial conditions can lead to noticeably different system responses, confirming that the proposed model exhibits sensitivity to initial conditions. However, under the given set of parameters and initial conditions, the trajectories remain regular rather than chaotic, demonstrating bounded and predictable oscillatory behavior. Sensitivity examination on initial conditions for the system (25). Comparative analysis with previous studies In this section, we demonstrate a comparative analysis of the solutions that we obtained in the present work compared to the results of previous studies on the investigation of the (2+1)-dimensional AKNS equation. Even though some past works have been able to derive several families of analytical solutions, the methods used as well as the level of dynamical analysis are quite restricted. To vividly demonstrate these differences, a systematic comparison will be given in Table 6, and then the new contributions of the given work will be discussed in detail. Table 6Comparative evaluation of analytical and dynamical tools reported in the literature.ReferencesPPsBDsCBTSSAMSPMsLEs27××××××××28××××××××29××××××××30××××××××31××××××××32××××××××33××××××××Our study\checkmark\checkmark\checkmark\checkmark\checkmark\checkmark\checkmark\checkmark Comparative evaluation of analytical and dynamical tools reported in the literature. Table 6 clearly demonstrates the significant advancement of the present study over existing literature. Although the derivation of analytical solutions (periodic, bell-shaped, singular, and kink-type wave structures) is the primary concern of earlier works, they lack the in-depth dynamical systems perspective. Conversely, the present work goes beyond the construction of solutions by incorporating an in-depth dynamical study of the AKNS model. In particular, the governing equation is reduced to a planar dynamical system, which allows classifying the equilibrium points with the help of PPs and BDs. Moreover, the external forcing also demonstrates the complex CB that is analyzed systematically by the means of TS, PMs, and LEs. Altogether, the article creates a more comprehensive and strict treatment of the analysis of the (2+1)-dimensional AKNS equation, which fills the gap between the study of the exact solutions and the current theory of dynamical systems. Conclusion In this paper, we performed a comprehensive analytical and dynamical analysis of the fourth-order nonlinear AKNS water wave equation. The MKM was successfully applied to obtain a wide range of exact solutions in rational, hyperbolic, and trigonometric forms. These solutions demonstrated the rich nonlinear behavior of the model and verified the effectiveness of the analytical method employed in capturing a broad range of wave dynamics in multidimensional systems. To gain physical insight into the solutions, we plotted some of the wave profiles in Figs. (1–8) via MATLAB. These representations depicted the dynamics of solitary, kink, periodic, and other types of nonlinear waves, providing insight into their characteristics and dynamics. The analytical results confirmed that the AKNS model exhibited a wide range of waveforms, including solitary and periodic waves. It was found that the interplay between nonlinearity and dispersion was essential to the development and maintenance of these coherent wave structures, which are applicable to various phenomena in coastal, shallow-water, and internal wave systems. Besides the analytical findings, a dynamical study was conducted to investigate the qualitative aspects of the system. By applying a Galilean transformation, the partial differential equation was transformed into a system of nonlinear ordinary differential equations, enabling the study of the dynamics in phase space. The application of methods from dynamical systems theory, including phase portraits, bifurcation diagrams, Lyapunov exponents, and multistability analysis, revealed a plethora of dynamical features, including periodic, quasi-periodic, and chaotic motion. Additionally, the system exhibited sensitivity to initial conditions and other pertinent parameters. This had a profound impact on the transition between different dynamical regimes and was crucial in realistic water-wave systems, where minor variations in the system may result in significant changes in the wave structure, amplitude, and stability. In addition, the current work provided a unified approach to the problem, which involved analytical solution methods and modern dynamical analysis. It offered valuable insights into the interaction of nonlinearity, dispersion, and external forces and thus improved the understanding of nonlinear wave dynamics in fluid systems. Future work ∙ In future the proposed model can be formulated by adding artificial intelligence (AI) and machine learning (ML) methods to study and forecast the intricate nonlinear dynamics. These smart techniques can be used to streamline parameters, categorize dynamical regimes (periodic, quasi-periodic, and chaotic), and discover concealed patterns in large-scale simulation data. We will also consider stochastic and variable coefficient approximations of the AKNS equation to understand them. Appendix A: Analytical framework and derivation of solutions A.1: Governing model and boundary conditions The (2+1)-dimensional AKNS-type water wave equation studied in this work is given by where Z(x, y, t) denotes the wave profile and B is a real parameter. To ensure mathematical consistency of the reduction procedure, the following assumptions are imposed: Smoothness: Z(x, y, t) is sufficiently differentiable. Smoothness: Z(x, y, t) is sufficiently differentiable. Boundary conditions (localized case):27 Boundary conditions (localized case):27 Alternatively (periodic case): Z and its derivatives are periodic in ξ. Alternatively (periodic case): Z and its derivatives are periodic in ξ. These conditions justify the integration step used later and ensure that the integration constant vanishes. The absence of explicitly stated boundary conditions in the main text is thus addressed here. A.2: Traveling wave transformation and PDE–ODE reduction To construct analytical solutions, the traveling wave transformation is introduced. Using this transformation: Substituting into the governing PDE gives: Integrating once with respect to ξ and using the boundary conditions yields: This is the reduced ordinary differential equation used in the manuscript (Eq. (4)). Thus, the analytical procedure follows: A.3: Construction of exact solutions The reduced equation is solved using the modified Khater method (MKM), as outlined in Section 2 of the manuscript. Step 1: Solution Ansatz Step 2: Auxiliary Equation Step 3: Substitution Substituting into the reduced ODE produces a polynomial in , whose coefficients yield an algebraic system. Step 4: Parameter Determination Solving this system (performed symbolically in Maple as stated in the manuscript) gives the parameter relations, including: Step 5: Families of Solutions Depending on the discriminant different solution families arise: Δ < 0: trigonometric (periodic solutions) Δ < 0: trigonometric (periodic solutions) Δ > 0: hyperbolic (soliton/kink solutions) Δ > 0: hyperbolic (soliton/kink solutions) Special parameter cases: rational and mixed structures Special parameter cases: rational and mixed structures These correspond to the solution sets – presented in the manuscript. A.4: Reconstruction of PDE solutions Each solution K(ξ ) generates a solution of the original PDE via Thus, the analytical solutions presented in the manuscript are not numerically generated, but are obtained explicitly through: PDE → ODE reduction PDE → ODE reduction Analytical solution of ODE (MKM) Analytical solution of ODE (MKM) Back-substitution into original variables Back-substitution into original variables This clarifies the authors’ contribution in deriving the solution families. A.5: Remark on analytical scope The obtained solutions describe traveling-wave structures of the original PDE under the stated transformation and boundary conditions. The derivations are fully analytical, while graphical results in the manuscript are used only for visualization. Author contributions Conceptualization, M.B.R. .; methodology, M.B.R. and S.S.K.; software, M.I. and T.K.; validation, M.B.R. and S.S.K.; formal analysis, M.B.R. , S.S.K., T.K.; investigation, S.S.K.and M.B.R.; data curation, M.B.R.; writing—original draft preparation, M.I.; writing—review and editing, M.B.R., S.S.K. and M.A.; visualization, M.B.R. and S.S.K.; supervision, M.A.; project administration, M.B.R. ; valiadation, M.B.R. ; funding acquisition, T.K. All authors have read and agreed to the published version of the manuscript Funding This article has been produced with the financial support of the European Union under the REFRESH – Research Excellence For Region Sustainability and High-tech Industries project number CZ via the Operational Programme Just Transition. Data availability All data that support the findings of this study are included within the article. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement All data that support the findings of this study are included within the article.