Coastal erosion is, more than ever, a global problem. By adopting a high-efficient, cost effective and reliable numerical model, it would help predict and manage erosion, as well as alleviate many coastal problems. This thesis reports the results of a though out investigation on the popular one dimension long-term shoreline change model--- GENESIS, analyze its suitability, sensitivity and technical difficulties likely to encounter while using the model, with the aim to predict the effect of coastal structure on shoreline changes. Prior to perform a modeling task, this report provides constructive recommendation on the setting of the length of shoreline to be covered in the modeling, boundary conditions, grid space, transport parameters K1 and K2 and revision of wave angle, followed by verification using results of several physical scale models, in order to enhance the reliability of the modeling and the parameters employed. Finally, reasonable ranges of K values are proposed. For modeling shoreline changes induced by a detached breakwater with normal incident waves, an empirical equation is proposed to determine the K ratio(K2/K1), which offer a useful guide in achieving the results with in a tolerance limits of 12%~-7%. When consider oblique wave incident to single detached breakwater, K1=0.6 is used and the ratio of K2/K1 ≈ 0.25~0.5. For modeling the effect of a single groin, the present study suggests K1=0.6 and K2/K1 ≈ 1~2. On the basis of these principles for setting the K values, the results are then applied to model the shoreline changes due to the installation of detached breakwater and groin. From the results of this study, for normal wave incident to single detached breakwater, it shows that for a small ratio of the offshore distance to the length of the breakwater S/B or a larger wave height, the salient dimension will increase and wave period has almost no effect on the results produced; for small S/B ratio, the maximum downcoast retreat increase, and its quantity is almost not affected by the wave conditions imposed. For oblique wave incident to single detached breakwater, it shows that for a larger wave angle, a small S/B or a larger wave height, the salient dimension will increase and wave period has almost no effect on the results produced; for larger wave angle or small S/B ratio, the maximum downcoast retreat increase, and its quantity is almost not affected by the wave height and wave period. For modeling the effect of a single groin, it shows that for larger wave angle or length of groin, the maximum downcoast retreat increase, and its quantity is almost not affected by the wave height and wave period.