CN-121999908-A - Multi-source data-based coal-based solid waste cemented filling material performance prediction method

CN121999908ACN 121999908 ACN121999908 ACN 121999908ACN-121999908-A

Abstract

The invention discloses a multi-source data-based coal-based solid waste cemented filling material performance prediction method, which relates to the technical field of data processing and comprises the steps of obtaining target coal-based solid waste, obtaining target components and additional components, obtaining a first strength sequence and a second strength sequence, obtaining the real-time content of the target components of the target coal-based solid waste, obtaining a first performance prediction index according to the first strength sequence, obtaining the real-time content of the i additional components of the target coal-based solid waste, obtaining a second performance prediction index according to the second strength sequence, obtaining the comprehensive performance prediction index of each target coal-based solid waste according to the first performance prediction index of the target components of each target coal-based solid waste and the second performance prediction index of each additional component, and sequentially arranging a plurality of target coal-based solid waste forming prediction sequences according to the comprehensive performance prediction index from large to small. The invention has the advantages of accurate prediction, high prediction efficiency and reliable correction.

Inventors

SHI YANNAN
WANG YIYING
WANG HANQIU
WANG JINGCHONG
SHI CHANGLIANG
WANG MANLI
HUANG HUI
ZHAO JIAWEI
WANG JIAO
XING SHIKUN
LI DEKUI
SUN JIAN

Assignees

河南理工大学

Dates

Publication Date: 20260508
Application Date: 20260123

Claims (10)

1. A multi-source data-based coal-based solid waste cemented filling material performance prediction method is characterized by comprising the following steps: Acquiring a plurality of target coal-based solid wastes, acquiring target components with the largest proportion and additional components except the target components in each target coal-based solid waste, acquiring first cemented filling strength corresponding to the target components under different recorded contents based on multi-source data, and acquiring second cemented filling strength corresponding to the target components under the same recorded content and the additional components under different recorded contents; sequentially arranging the corresponding first cemented filling intensities according to the order of the recorded content values of the target components from small to large to obtain a first intensity sequence of the target components, and sequentially arranging the corresponding second cemented filling intensities according to the order of the recorded content values of the additional components from small to large to obtain a second intensity sequence of the additional components; Acquiring the real-time content of target components of each target coal-based solid waste, and acquiring a first performance prediction index according to the real-time content of the target components of each target coal-based solid waste and a first strength sequence of the target components; Acquiring the real-time content of the ith additional component of each target coal-based solid waste, and acquiring a second performance prediction index according to the real-time content of the ith additional component of each target coal-based solid waste and a second intensity sequence of the ith additional component; and acquiring comprehensive performance prediction indexes of the target coal-based solid wastes according to the first performance prediction indexes of the target components of the target coal-based solid wastes and the second performance prediction indexes of the additional components, and sequentially arranging a plurality of target coal-based solid wastes according to the sequence from the large to the small of the comprehensive performance prediction indexes to form a prediction sequence.
2. The method for predicting the performance of the multi-source data based coal-based solid waste cemented filling material according to claim 1, wherein the obtaining the first performance prediction index according to the real-time content of the target component and the first strength sequence of the target component of each target coal-based solid waste comprises: Acquiring first cemented filling intensities corresponding to two record contents adjacent to each other at positions where real-time contents are located in a first intensity sequence of a target component, taking the first cemented filling intensity positioned at the left side as a first left value of the target component, and taking the first cemented filling intensity positioned at the right side as a first right value of the target component; and acquiring a first performance prediction index of the target components of each target coal-based solid waste according to the first left value and the first right value of the target components of each target coal-based solid waste.
3. The method for predicting the performance of the multi-source data-based coal-based solid waste cemented filling material according to claim 2, wherein the obtaining the first performance prediction index of the target component of each target coal-based solid waste according to the first left value and the first right value of the target component of each target coal-based solid waste comprises: subtracting the recorded content of the first left value from the recorded content corresponding to the first right value of the target components of each target coal-based solid waste to obtain a first content reference value, subtracting the recorded content of the first left value from the real-time content of the target components of each target coal-based solid waste to obtain a first content actual value, and dividing the first content actual value by the first content reference value to obtain a first ratio; subtracting the first left value from the first right value of the target components of each target coal-based solid waste to obtain a first intensity reference value, and multiplying the first value by the first intensity reference value to obtain a first variable; And adding the first left value of the target component of each target coal-based solid waste with the first variable to obtain a first performance prediction index of the target component of each target coal-based solid waste.
4. The method for predicting the performance of a coal-based solid waste cemented filling material based on multi-source data according to claim 1, wherein the obtaining a second performance prediction index according to the real-time content of the i-th additional component and the second strength sequence of the i-th additional component of each target coal-based solid waste comprises: Acquiring the second cemented filling strength corresponding to two record contents adjacent to the real-time content of the ith additional component in the second strength sequence of the ith additional component, taking the second cemented filling strength positioned at the left side as a second left value of the ith additional component, and taking the second cemented filling strength positioned at the right side as a second right value of the ith additional component; And obtaining a second performance prediction index of the ith additional component of each target coal-based solid waste according to the second left value and the second right value of the ith additional component of each target coal-based solid waste.
5. The method for predicting the performance of the multi-source data based coal-based solid waste cemented filling material according to claim 4, wherein the obtaining the second performance prediction index of the i-th additional component of each target coal-based solid waste according to the second left value and the second right value of the i-th additional component of each target coal-based solid waste comprises: Subtracting the recorded content of the second left value from the recorded content corresponding to the second right value of the i-th additional component of each target coal-based solid waste to obtain a second content reference value, subtracting the recorded content of the second left value from the real-time content of the i-th additional component of each target coal-based solid waste to obtain a second content actual value, and dividing the second content actual value by the second content reference value to obtain a second ratio; subtracting a second left value from a second right value of the ith additional component of each target coal-based solid waste to obtain a second intensity reference value, and multiplying a second ratio by the second intensity reference value to obtain a second variable; And adding the second left value of the ith additional component of each target coal-based solid waste with the second variable to obtain a second performance prediction index of the ith additional component of each target coal-based solid waste.
6. The method for predicting the performance of the multi-source data based coal-based solid waste cemented filling material according to claim 1, wherein the obtaining the comprehensive performance prediction index of each target coal-based solid waste according to the first performance prediction index of the target component of each target coal-based solid waste and the second performance prediction index of each additional component comprises: Respectively associating first weights with first performance predictors of target components of each target coal-based solid waste, and respectively associating second weights with second performance predictors of each additional component of each target coal-based solid waste; Multiplying the first performance prediction index of the target component of each target coal-based solid waste by a first weight to obtain a main item of each target coal-based solid waste, and multiplying the second performance prediction index of each additional component of each target coal-based solid waste by a second weight to obtain a plurality of initial superposition items of each target coal-based solid waste; If additional components meeting a preset rule exist in the additional components of each target coal-based solid waste, associating contribution degree adjusting parameters with the additional components, and acquiring a correction superposition item according to the contribution degree adjusting parameters and initial superposition items obtained by the second performance prediction indexes of the additional components associated with the contribution degree adjusting parameters; And adding all main items, all initial superposition items and all correction superposition items of the target coal-based solid wastes, and obtaining the comprehensive performance prediction index of each target coal-based solid waste.
7. The multi-source data based coal-based solid waste cemented filling material performance prediction method of claim 6, wherein the first weight comprises: The real-time content of the target component of each target coal-based solid waste is defined as a first weight.
8. The method for predicting the performance of a multi-source data based coal-based solid waste cemented filling material of claim 6, wherein the second weight comprises: The real-time content of each additional component of each target coal-based solid waste is defined as a second weight.
9. The method for predicting the performance of a multi-source data based coal-based solid waste cemented filling material according to claim 6, wherein if there is an additional component satisfying a predetermined rule in each additional component of each target coal-based solid waste, associating a contribution degree adjustment parameter to the additional component comprises: Judging whether interaction exists between the ith additional component and other components in the jth target coal-based solid waste or not based on the multi-source data; if the interaction is the increase of the intensity, multiplying the preset correction proportion by 1, and taking the calculation result as a contribution degree adjusting parameter and correlating with the ith additional component; If the difference exists and the interaction is the intensity reduction, the preset correction proportion is multiplied by-1, and the calculated result is used as a contribution degree adjusting parameter and is related with the ith additional component.
10. The method of claim 6, wherein the obtaining the corrected superposition term from the initial superposition term obtained from the second performance prediction index of the additional component of the contribution adjustment parameter and the associated contribution adjustment parameter comprises: and multiplying the initial superposition item obtained by the second performance prediction index of the additional component related to the contribution adjustment parameter by the contribution adjustment parameter, and obtaining a corrected superposition item.

Description

Multi-source data-based coal-based solid waste cemented filling material performance prediction method Technical Field The invention relates to the technical field of data processing, in particular to a coal-based solid waste cemented filling material performance prediction method based on multi-source data. Background In the field of coal mine environmental protection, the recycling of coal-based solid wastes (such as combustion ash, gasification residues, washing gangue and the like) is a key link for reducing environmental pollution and realizing green exploitation. When the solid wastes are utilized to prepare the cementing filling material, the prior experience depends on a linear model of manual trial preparation or a single data source, and the cementing strength performance of the mixed material is difficult to accurately predict because the solid wastes have complex sources and large component fluctuation (comprising various heterogeneous wastes such as combustion, gasification, washing, desulfurization, pyrolysis and the like). Specifically, the content combination of the main component (the solid waste with the largest proportion) and the auxiliary component (the solid waste with other types) in different solid wastes is changed greatly, and the heterogeneous multi-source information such as industry standard, knowledge graph, historical data and the like are not effectively fused, so that a basic strength rule of the main component under different contents and a synergetic/antagonistic action model of the auxiliary component cannot be established. Meanwhile, the existing prediction method ignores two types of nonlinear characteristics, namely non-uniform influence of content change of the same main component on strength (such as strength response rate difference between a low content area and a high content area) and interaction of each component. Finally, the real-time detection of solid waste component data lacks an adaptive calculation mechanism, rapid interpolation between discrete multi-source data points cannot be performed according to actual content, and dynamic contribution weight of the multi-component coupling effect to final strength cannot be quantized, so that blindness of material proportioning design is high, and trial and error cost is high. Disclosure of Invention Aiming at the technical problem of performance prediction misalignment caused by component complexity and action nonlinearity of the coal-based solid waste cemented filling material in the prior art, the invention provides a method for predicting the performance of the coal-based solid waste cemented filling material based on multi-source data. A multi-source data-based coal-based solid waste cemented filling material performance prediction method comprises the steps of obtaining a plurality of target coal-based solid waste, obtaining target components with the largest proportion among the target coal-based solid waste and additional components except the target components, obtaining first cemented filling strength corresponding to the target components under different recorded contents based on multi-source data, obtaining second cemented filling strength corresponding to the combination of the target components with different recorded contents, sequentially arranging the corresponding first cemented filling strength according to the recorded content value of the target components from small to large, obtaining a first strength sequence of the target components, sequentially arranging the corresponding second cemented filling strength according to the recorded content value of each additional component from small to large, obtaining a second strength sequence of each additional component, obtaining real-time content of the target components of each target coal-based solid waste, obtaining a first performance prediction index according to the real-time content of each target coal-based solid waste component and the first strength sequence of the target components, obtaining a real-time performance prediction index of each additional component of each target coal-based solid waste, sequentially arranging the corresponding first additional strength sequences according to the recorded content value of the target components from small to large, and obtaining the real-time performance prediction index of each additional component of each target coal-based solid waste according to the real-time content of each target component, and obtaining the real-time performance prediction index of each additional component according to the real-time performance index of each target coal-time. Optionally, acquiring the first performance prediction index according to the real-time content of the target component and the first strength sequence of the target component of each target coal-based solid waste comprises acquiring the first cemented filling strength corresponding to two record contents adjacent to each other in the position wh