Some distributors face challenges in selecting suppliers and determining the quantity of orders to meet customer demand adequately. The root cause of this issue lies in the lack of application of accurate forecasting techniques and insufficient consideration of uncertainty factors. Therefore, this study proposes a model that incorporates forecasting techniques and accounts for the uncertainty of input parameters. First, applying and comparing ARIMA, Holt-Winter, and ANN methods to forecast future demand for datasets collected that are both trendy and seasonal. The forecasted demand is one of the input parameters for the proposed Multi-objective Supplier Selection and Order Allocation model. Use the Mixed Integer Linear Programming (MILP) method to resolve uncertainty in supplier purchasing prices and the weighted sum approach to combine the objective functions. The proposed model establishes Supplier Selection and Order Allocation planning for the distributor in the lubricant industry, resulting in a total cost decrease of 12.30% and a damaged product decrease of 5.63% within four weeks compared to the actual. Moreover, the model is capable of solving more complex scenarios, including cases with up to 15 suppliers and 10 product types over 8 weeks across various scenarios. However, the solution time is still considerable, suggesting that exploring metaheuristic optimization algorithms could be beneficial to improve the efficiency of the model in handling such large-scale problems, as well as considering additional constraints related to sustainability and other uncertainty factors to better reflect real-world conditions.