铀矿尾渣胶结充填体塌落度影响因素及预测模型

采用生石灰中和后的铀矿尾渣为试验骨料,水泥为胶结剂,试验了不同渣浆质量浓度和灰渣质量比条件下尾渣胶结充填体试件塌落度。利用自适应神经模糊推理系统(adaptive neuro-fuzzy inference system,ANFIS)预测不同渣浆质量浓度和灰渣质量比条件下塌落度。试验结果和模型预测结果表明,在同等条件下,铀矿尾渣胶结充填体的塌落度与渣浆质量浓度和灰渣质量比均成反比;随着渣浆质量浓度变小,铀矿尾渣胶结充填体流动性变大,如果再降低灰渣质量比,铀矿尾渣胶结充填体会出现明显离析现象;铀矿尾渣胶结充填体塌落度ANFIS模型的预测精度较高,结果唯一且可靠;如果不同矿山铀尾渣物理参数相似性较高,则已知的铀矿尾渣胶结充填体塌落度ANFIS模型可以预测该矿山尾渣胶结充填体的塌落度。

Influence Factors and Prediction Model of the Slump of Uranium Tailings Cemented Backfill

The slump tests of cement uranium tailings backfills with different cement-tailing ratio under the conditions of slurry concentration respectively were conducted by using a slump tester. On the basis of the test results,the adaptive neuro-fuzzy inference system(ANFIS) model for predicting slump was established on slurry concentration and cement-tailings ratio. The results show that under the same condition, the slump is inversely proportional to cement-tailings ratio and slurry concentration respectively; The liquidity of cement uranium tailings backfill increase with decrease of slurry concentration, and if the cement-tailings ratio is decreased further, cement uranium tailings backfill disintegrates; Adaptive neuro-fuzzy inference system model for predicting the slump of uranium tailings cemented backfill have greatly accuracy of prediction and sole reliable result; If the physics parameters of uranium tailings of different mine have high similitude, the adaptive neuro-fuzzy inference system(ANFIS) model for predicting slump can be used in another uranium mine, the accuracy and reliability of prediction model can meet mine experiment and engineering practice requirements.