A fractional-order supply chain model can represent a feature that conventional models may not capture directly: the influence of past conditions on current demand, supply and production. This matters in digital manufacturing, where changes in orders, inventory and production decisions can propagate across several stages of a supply chain. Research into fractional-order system dynamics examines how that memory affects stability—and how control strategies may help manage chaotic behavior in a mathematical model.
Why supply chain memory matters
In an integer-order model, a system’s rate of change is described using local state information. Fractional calculus extends this approach: a fractional derivative incorporates the combined influence of historical states, rather than treating each moment as isolated from what came before. That property makes fractional-order methods relevant to systems in which past demand, production and stock conditions continue to shape current behavior.
Supply chains are also nonlinear systems with uncertainty. A shift in customer demand can travel upstream from retailers to distributors and manufacturers, sometimes changing in magnitude as information is passed along. This amplification is known as the bullwhip effect. When ordering and production responses interact, supply, inventory and delivery patterns can become difficult to predict.
Digital manufacturing adds another process to consider: the transition from digital design to production. A model that connects retailer demand, distributor supply, completed digital designs and manufacturing output can help examine how this process interacts with the wider supply chain.
How a fractional-order digital manufacturing model works
The studied model represents four state variables: retailer demand orders, distributor supply, the quantity of completed digital designs, and production based on those designs. Its relationships describe how the variables influence one another. For example, demand and supply are linked, while production and digital design activity affect the manufacturer’s contribution to the chain.
A shared fractional derivative order, denoted by q, is applied to the model’s equations. When q equals 1, the model reduces to its integer-order form. Values below 1 introduce the fractional memory effect. The researchers use the Caputo definition of the fractional derivative and the Adomian decomposition method to obtain approximate numerical solutions.
To assess system behavior, the analysis uses several complementary tools:
- Bifurcation diagrams show how system behavior changes as a model parameter varies.
- Maximum Lyapunov exponents help identify sensitivity associated with chaotic dynamics.
- Phase portraits and time series show relationships and changes among the modeled variables.
- Spectral entropy and C0 complexity measures assess the complexity of the resulting time series.
Using more than one measure helps distinguish periodic behavior from complex or chaotic dynamics in the modeled system.
Digital production rate and modeled stability
The model treats the rate at which digital designs enter production as a key parameter. In the analyzed case, with the fractional order set to 0.75, changes to this parameter produce different patterns. As the rate decreases, the system passes through period-doubling behavior and enters a chaotic regime below the reported threshold of 2.18. At higher values, the modeled behavior is periodic.
These values describe the particular model and parameter choices studied; they should not be read as universal operating limits for manufacturers. The broader implication is that the relationship between digital design and production can influence the dynamics of the whole modeled chain. In the study’s interpretation, a higher rate of moving designs into production can support quicker replenishment, while a lower rate may leave retailer demand unmet and contribute to unstable ordering, inventory and production patterns.
The complexity measures follow the same general pattern as the bifurcation analysis: higher complexity appears in the modeled chaotic region, while lower values accompany periodic behavior. This agreement suggests that complexity measures can provide another way to characterize changes in system dynamics—not a standalone operational diagnosis.
Fractional order changes the picture
The derivative order also affects the model’s behavior. For the parameter settings examined, the system is mostly chaotic over a range of fractional orders from 0.47 to 0.62, while higher orders in the analyzed range produce periodic behavior. The model also shows differences in production fluctuations as the order changes, including how production behaves when distributor supply is insufficient compared with when supply is adequate.
This is the practical modeling distinction: an integer-order representation may not reflect the same history-dependent behavior as a fractional-order one. Comparing both forms can help technical teams investigate whether memory effects are important to the question being modeled. The findings do not establish that one derivative order or model structure will fit every real supply chain; the appropriate representation depends on the system and its data.
Using feedback to control and synchronize systems
The study also considers feedback controllers for two related tasks: stabilizing a chaotic system and synchronizing two systems that start from different initial conditions. For synchronization, one system acts as the driver and the other as the response. The controller uses the differences between their demand, supply, digital-design and production states to construct corrective inputs.
In the proposed design, those inputs simplify the error dynamics into stable equations. Numerical simulations show the state differences converging toward zero under the designed controllers. The simulations also report faster synchronization for smaller fractional orders in the cases examined. These are model-based results, rather than a demonstration of implementation in an operating supply chain.
Practical implications for automation decision-makers
For industrial engineers, system integrators and operations leaders, the value of this work is as a modeling and analysis framework. It highlights several considerations when building or evaluating digital manufacturing supply chain models:
- Represent historical influence when past demand, production or inventory conditions may continue to affect current decisions.
- Examine interactions between demand, distributor supply, digital design and production rather than treating manufacturing output in isolation.
- Test parameter changes for periodic, bifurcating and chaotic behavior using more than one diagnostic method.
- Evaluate feedback control in the model before drawing conclusions about its suitability for a real process.
A fractional-order supply chain model offers a structured way to study memory, nonlinear dynamics and control in digital manufacturing. The analysis shows that both the derivative order and the rate of digital design entering production can change modeled behavior, while feedback can synchronize simulated systems with different initial states. For decision-makers, the practical lesson is to treat these methods as tools for understanding and testing supply chain dynamics—not as universal thresholds or substitutes for validating a model against the operation it is intended to represent.
Access Official Technical PDF