The intense market dynamics, the expansion of sales channels, and the prominent role of strategic customers who have forward-looking behavior have made dynamic pricing one of the most challenging decision-making problems in multi-channel supply chains. Aiming to face this complexity, this study presents a comprehensive framework in which the uncertainty in the basic parameters of utility and operating costs are modeled as trapezoidal fuzzy numbers and, by using a robust probabilistic programming approach, the decisions obtained from the model are stable and reliable against severe environmental fluctuations. In this model, the demand of strategic and non-strategic customers is considered separately and based on utility functions based on their behavior in different sales channels, so that the effect of strategic customers' forward-looking behavior and their different sensitivity to pricing is reflected in the results. To solve the proposed nonlinear model, three meta-heuristic algorithms including Genetic Algorithm (GA), Gray Wolf Algorithm (GWO) and a novel hybrid method (GWGO) were designed and implemented. Numerical experiments at different scales showed that the proposed model can provide optimal prices that remain profitable and stable even under high uncertainty; also, the GWGO hybrid algorithm provided the best balance between response quality and convergence speed. Sensitivity analysis showed that strategic customers have a significant impact on profitability and the increase in uncertainty associated with this group has a greater impact on profit structure and pricing behavior. This study provides a powerful analytical framework for designing flexible, robust and adaptive pricing policies to customers' behavior in turbulent environments.