To improve the working performance of filling slurry, increasing strength and density of backfill body is a research direction of mine filling method. Filling ratio experiments showed that adding appropriate amount of water reducers during the preparation of the filling material could increase the compressive strength of the filling body. In order to obtain economic and efficient water reducers and parameters, four kinds of water reducers, i.e., naphthalene, amino, wood, and calcium binding aliphatic were used for new filling materials. A match experiment with certain mine backfilling materials was carried out, and a GA-SVM prediction model was established to optimize the selection. The four kinds of water reducers were used as the input data and the 28 days compressive strengths of filling body were confirmed to be the synthesized output data. Some training and validating samples were established through indoor experiment; a support vector machine (SVM) regression model was established. Then, the model parameters were optimized through the genetic algorithm (GA). The results show that the best tailing concentrations of the four kinds of water reducers were 0, 0.35%, 0.30%, and 0.60%, and that the compressive strength of filling body could be 4.20 MPa. Compared with the experiment results, the relative error of the prediction result can be controlled within about 1%. This model provides a new method to optimize sedimentation parameters.

filling slurry  /  water-reducers  /  compressive strength  /  support vector machine  /  genetic algorithm