<p>One widely concerned issue in the field of seawall settlement prediction is the difficulty in establishing prediction models in the context of insufficient data volume. Fractional-order grey model (FGM), as an extension of the traditional grey model incorporating fractional calculus, has been used to solve the insufficient data volume problem in temporal prediction. Due to the non-integer nature of the fractional order of FGM, traditional parameter estimation methods often lead to increased instability and uncertainty and are no longer applicable, necessitating the adoption of more complex algorithms for estimating the fractional order. To solve this issue, a novel GA-FGM-RTA combined model was proposed for predicting seawall settlement with insufficient data volume, where a genetic algorithm (GA) with enhanced search capabilities was employed to optimal the fractional order and a real-time tracing algorithm (RTA) was applied to provide a dynamic prediction. A case study of Haiyan seawall in Zhejiang province, China was selected to validate the proposed model. We also compared the proposed GA-FGM-RTA model with the fractional-order GM(1,1), integer-order GM(1,1), and fractal theory model. Results exhibit that the proposed GA-FGM-RTA combined model outperforms other relevant models under the same algorithm frame.</p>