Accurate interval forecasting of rural power load is of great significance to ensure the security, stability and economic operation of the system in the rural area. To solve the problem of the poor adaptability, low accuracy and high calculation cost in the rural power load forecasting, a method considering interval construction is proposed. First, the temporal convolutional network is utilized to mine the temporal features in the deterministic forecasting. Then, the interval forecasting model considering interval construction is transformed into a bilevel programming model. Finally, the whale optimization algorithm is employed to solve the problem, and the optimal deterministic forecasting results are expanded or scaled based on the scaling coefficient method to calculate the rural power load interval forecasting results. The proposed method demonstrates generalizability and accuracy in the scenario of rural area, providing reliable information on load variations for the operation and dispatch of the rural power system.