Intelligent inventory demand forecasting: using chaos theory and optimization for supply chain resilience

Supply chain inventory management is an essential yet demanding activity with
parameters such as uncertain customer demand, time-varying lead times, and unexpected
production or shipping disruptions. Traditional models like newsvendor and
economic order quantity provide the basis for modeling strategies but are just insufficient
when dealing with the real-world uncertainties. In this research, a cuttingedge
demand forecasting model based on the Red Panda Optimizer with Logistic
Mapping (RPO-LM) is proposed to support better inventory decisions. The methodology
proposes three basic phases: data preparation, model training, and model
evaluation. The prediction accuracy is improved by normalizing historical inventory
data and splitting them into training and test datasets. The RPO-LM model is trained
to reflect short-term fluctuation and long-term trends in demand to remain adaptive
to dynamic inventory levels. To ascertain the accuracy of the predictions, performance
analysis is carried out using metrics such as mean absolute percentage error
and root mean squared error. The proposed model supports anticipatory inventory
management, minimizing stockouts and overstocking, thereby maximizing supply
chain effectiveness. Python simulations validate the effectiveness of the suggested
strategy, demonstrating its superiority over more conventional approaches such as
the least squares polynomial sinusoidal method and auto regressive integrated moving
average. With a mean absolute percentage error of around 4 and a root mean
squared error of about 6, the suggested approach has the highest forecast accuracy.
Outcomes verify that proposed method greatly enhances prediction accuracy and
economic efficiency, qualifying it as an optimal solution for contemporary inventory
control systems.