IBAS-BP Dynamic Evaluation Of Gravelly Soil Blending Quality Considering Material And Construction Conditions

The evaluation of the blending quality of gravelly soil is an important part of the construction quality control of the core rockfill dam.However,the current quality evaluation of gravelly soil blending lacks consideration of material,and still uses the side station supervision as the main method,uses the P5 content of the sampling point as the main indicator,which will greatly reduce the accuracy and efficiency of blending quality evaluation.This paper puts forward the dynamic evaluation method for the blending quality of gravelly soil under the comprehensive consideration of material and construction conditions,realizes multi-scale and in-depth dynamic evaluation of blending quality from four aspects: material condition,evaluation index system,parameter perception method and dynamic evaluation model.The main research results are as follows:(1)Aiming at the current lack of analytical methods for the collocation of gravelly soil particle and the blending ratio after mixing,this paper uses the MMF model and the standard base entropy to respectively realize the judgment of the particle collocation based on the compaction performance and the quantitative analysis check of the gravelly soil blending ratio.(2)Aiming at the problems that using P5 content as an evaluation index for the blending quality of gravelly soil can neither reflect the blending uniformity,nor can it characterize the blending situation of particle with other size,this paper combines the blending uniformity value(BUV)of gravelly soil,controlled particle size content and coefficient of variation(Cv)to construct a gravelly soil blending quality evaluation index system to achieve an effective characterization of the gravelly soil blending quality.(3)Aiming at the disadvantages that the side station supervision is easy to be interfered by human factors when controlling the construction parameters,and the control efficiency is low,the control accuracy is poor.Based on the geometric relationships of construction machinery,this paper proposes a method to perceive construction parameters of gravelly soil blending with the positioning and posture information,so as to realize the monitoring and collection of blending construction parameters.(4)Aiming at the shortcomings of the random sampling detection method that has low timeliness and high randomness,and it is difficult to reflect the blending quality of the whole work unit,this paper uses the improved beetle antennae search(IBAS)algorithm modified by the backtracking idea and the non-linear step to optimize the initial thresholds and weights of the BP neural network,and then constructs a dynamic evaluation model of blending quality based on IBAS-BP neural network,finally,combined with Kriging interpolation method to realize the dynamic evaluation of the blending quality on the whole work unit.The IBAS-BP dynamic evaluation method for the blending quality of gravelly soil considering material and construction conditions is applied to actual project.The results firstly prove that the MMF model and standard base entropy can effectively analyze the particle collocation of gravelly soil and the blending ratio;secondly,the blending quality evaluation index system constructed can effectively reflect the blending uniformity of gravelly soil and the content of each controlled particle size;thirdly,the construction parameter perception method of gravelly soil blending can effectively monitor and collect the paving thickness and the blending times,its control accuracy can meet the needs of the project;finally,compared with the commonly used regression models,the model constructed in this paper shows consistency,representativeness and superiority,which can realize the accurate interpolation of BUV on the whole work unit,and further guarantee the blending quality of gravelly soil.