CN-122021080-A - Concrete mixing proportion calculation method based on orthogonal theory

CN122021080ACN 122021080 ACN122021080 ACN 122021080ACN-122021080-A

Abstract

The invention relates to the technical field of concrete and discloses a concrete mixing ratio calculation method based on an orthogonal theory, which comprises the following steps of defining an optimization target, a mixing ratio factor, a target factor mapping knowledge base, a global fitness function and a disturbance amplitude set; the method comprises the steps of determining an initial global optimal solution through physical tests and fitness calculation, identifying a performance short board based on the optimal solution of the previous round, calling a knowledge base to conduct dynamic factor layering, dividing factors into a core, a secondary and an exploration factor set, obtaining layered disturbance vectors based on layering, selecting an orthogonal table, generating a new test set based on the optimal solution of the previous round and the layered disturbance vectors, calculating a score of the new test set to update the global optimal solution until a termination condition is met, and outputting a final optimal matching scheme. According to the invention, by combining dynamic factor layering with orthogonal tests, the short plate with accurate focusing performance is realized, multi-objective collaborative optimization is realized, and the efficiency, the accuracy and the robustness of the mix proportion design are remarkably improved.

Inventors

ZHANG CHUNYANG
SUN JIANGANG
Yin yizhi
ZHOU HENG
ZHAO FENGQIANG
BAO XUDONG
SONG SHICHUN
MA JIANFENG

Assignees

中建三局集团有限公司

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (10)

1. The concrete mixing proportion calculating method based on the orthogonal theory is characterized by comprising the following steps of: S1, system initialization and model definition, namely defining an optimization target vector containing a plurality of performance indexes, a matching factor vector containing a plurality of variable parameters, a target factor mapping knowledge base, a global fitness function for calculating global fitness scores of candidate matching schemes, and a disturbance amplitude set containing numerical values of a plurality of levels; S2, initial round evaluation, wherein a processing unit executing the method generates an initial candidate matching scheme set, and determines an initial global optimal solution by executing a physical test and fitness calculation on the initial candidate matching scheme set; s3, performing iterative optimization rounds, namely circularly executing iterative optimization rounds by taking the initial global optimal solution as a starting point, wherein each round of iteration comprises the following steps: Based on each bias fitness score of the global optimal solution determined in the previous round, identifying a performance index with the minimum bias fitness score as a performance short board, carrying out dynamic factor layering based on the performance short board and a target factor mapping knowledge base, mutually exclusive dividing a matching factor vector into a core factor set, a secondary factor set and a exploring factor set, obtaining a layered disturbance vector based on a disturbance amplitude set and factor layering, selecting an orthogonal table, and generating a new test set based on the global optimal solution and the layered disturbance vector determined in the previous round; Calculating global fitness scores of new candidate matching ratio schemes in a new test set, comparing the global fitness scores with global optimal solutions determined in the previous round, and updating to obtain new global optimal solutions; And S4, outputting a final optimal duty ratio, namely outputting the final updated new global optimal solution as a final optimal matching ratio scheme when a preset termination condition is met.
2. The method of claim 1, wherein the step S1 further comprises defining a bias fitness function for calculating bias fitness scores of candidate mix schemes with respect to performance indexes, defining a global fitness function as a calculation rule for weighted summation of all bias fitness scores, setting a disturbance amplitude set as a set including a first disturbance amplitude value, a second disturbance amplitude value and a third disturbance amplitude value, and satisfying that the first disturbance amplitude value is larger than the second disturbance amplitude value, and the second disturbance amplitude value is larger than the third disturbance amplitude value.
3. The method for calculating the concrete mix based on the orthogonal theory according to claim 1, wherein the step S2 specifically comprises the steps of generating an initial candidate mix scheme set, wherein the initial candidate mix scheme set comprises a plurality of groups of candidate mix schemes; Performing a physical test on the candidate matching schemes in the initial candidate matching scheme set to obtain corresponding measured performance values, and calling a global fitness function to calculate global fitness scores corresponding to the candidate matching schemes in the initial candidate matching scheme set; And comparing global fitness scores of all candidate coordination schemes in the initial candidate coordination scheme set, and selecting the candidate coordination scheme with the largest global fitness score as an initial global optimal solution.
4. The method for calculating the concrete mix proportion based on the orthogonal theory according to claim 1, wherein the step S3 specifically comprises the steps of taking the identified performance short board as a query index, calling a target factor mapping knowledge base, and designating a factor subset returned by the target factor mapping knowledge base as a core factor set; Removing the performance short plates from the optimized target vectors to obtain a non-short plate target set, collecting factor subsets mapped by all performance indexes in the non-short plate target set in a target factor mapping knowledge base, merging to obtain a union, removing factors existing in a core factor set from the union, and determining the remaining factors as secondary factor sets; The factors classified in the core factor set and the factors classified in the secondary factor set are removed from the mix factor vector, and all the remaining factors are determined as the exploration factor set.
5. The method of calculating concrete mix according to claim 4, wherein in the step S3, the obtaining of the layered disturbance vector includes traversing factors in the mix factor vector; If the factor belongs to the core factor set, the disturbance amplitude of the factor is distributed to be a first disturbance amplitude value in the disturbance amplitude set, if the factor belongs to the secondary factor set, the disturbance amplitude of the factor is distributed to be a second disturbance amplitude value in the disturbance amplitude set, and if the factor belongs to the exploration factor set, the disturbance amplitude of the factor is distributed to be a third disturbance amplitude value in the disturbance amplitude set.
6. The method for calculating a concrete mix ratio based on the orthogonal theory according to claim 4, wherein in the step S3, the generating a new test set further includes: core factor filling, namely adding factors in a core factor set into an operation factor set until the number of factors in the operation factor set reaches the number of columns of an orthogonal table or the core factor set is empty; If the factor number of the operation factor set does not reach the column number of the orthogonal table, adding the factors in the secondary factor set into the operation factor set until the factor number of the operation factor set reaches the column number of the orthogonal table or the secondary factor set is empty; And (3) searching for factor filling, namely adding the factors in the searching factor set into the operation factor set until the number of the factors in the operation factor set reaches the number of columns of the orthogonal table if the number of the factors in the operation factor set still does not reach the number of columns of the orthogonal table.
7. The method for calculating the concrete mix proportion based on the orthogonal theory according to claim 6, wherein in the step S3, the generation of the new test set comprises the steps of obtaining a current optimal value corresponding to the factor from a global optimal solution determined in the previous round; And calculating a level two value, wherein the level two value is equal to the current optimal value plus the product of the current optimal value and the disturbance amplitude.
8. The method of claim 7, wherein in the step S3, the generating a new test set further comprises traversing each row of the orthogonal table to generate a new candidate mix scheme, wherein if the value of the current column of the current row of the orthogonal table indicates the first state, the value of the operation factor corresponding to the current column in the new candidate mix scheme is set to be a horizontal value; If the value of the current row and the current column of the orthogonal table indicates the second state, setting the value of the operation factor corresponding to the current column in the new candidate matching ratio scheme as a horizontal two-value; And setting the values of all other factors which do not belong to the operation factor group in the new candidate matching ratio scheme to be consistent with the values of the corresponding factors in the globally optimal solution determined in the previous round.
9. The method for calculating the concrete mix ratio based on the orthogonal theory according to claim 1, wherein in the step S3, the updating to obtain the new global optimal solution includes: constructing a temporary set containing the global optimal solution determined in the previous round and all new candidate matching ratio schemes in the new test set; global fitness scores of all candidate fitness schemes in the temporary set are compared, and the candidate fitness scheme with the largest global fitness score is selected and designated as a new global optimal solution.
10. The method for calculating a concrete mix ratio based on the orthogonal theory according to claim 1, wherein in the step S3, the determining whether the current iteration round counter or the stagnation counter meets a preset termination condition comprises: Judging whether the current iteration round counter is larger than or equal to a preset maximum iteration round value or not; Judging whether the stagnation counter is larger than or equal to the preset upper limit of stagnation iteration times; And when the global fitness score lifting amount of the new global optimal solution relative to the global optimal solution determined in the previous round is smaller than a preset stagnation tolerance threshold, the stagnation counter is increased, and when the global fitness score lifting amount is larger than or equal to the stagnation tolerance threshold, the stagnation counter is reset.

Description

Concrete mixing proportion calculation method based on orthogonal theory Technical Field The invention relates to the technical field of concrete, in particular to a concrete mixing ratio calculating method based on an orthogonal theory. Background The concrete is used as the most widely used building material in civil engineering, the performance of the concrete depends on the design of the mixing proportion, the mixing proportion of the concrete refers to the proportion relation among the components such as cement, water, coarse and fine aggregate, additives and the like, and the scientific and reasonable mixing proportion design has decisive significance for ensuring the mechanical property, working performance, durability and controlling the production cost of the concrete. Currently, concrete mix determination relies on empirical trial and error and specification-based calculations, and engineering technicians typically refer to relevant industry standards (e.g., JGJ 55-2011, general concrete mix design procedure) to gradually determine the final mix by laboratory trial and error, and with the development of computing technology, there are also some aided design methods that attempt to optimize a particular performance by single variable control or simple regression models, such as pursuing only the highest strength or lowest cost. However, the traditional experience trial-matching method is time-consuming and labor-consuming, highly depends on personal experience, and is difficult to ensure that a global optimal solution is obtained, more importantly, concrete is a complex multi-component and multi-factor mutual coupling system, nonlinear constraint relations exist among various performance indexes (such as strength, workability and durability) of the concrete, the traditional optimization method is limited to optimization of a single target or a few targets, the problem of collaborative balance among multiple targets is difficult to treat, when the traditional optimization method faces various novel additives or materials, the number of the blending ratio factors is increased, the experimental design space is huge, the comprehensive influence of a plurality of factors and interaction thereof on the overall performance cannot be efficiently explored by the traditional experimental method, the optimization iteration efficiency is low, and the complex engineering requirements are difficult to quickly respond. Therefore, the invention provides a concrete mixing proportion calculating method based on an orthogonal theory, which solves the defects in the prior art. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a concrete mixing ratio calculation method based on an orthogonal theory, which solves the problems of inaccurate key factor identification, low iteration efficiency and incapability of effectively balancing various performance indexes in multi-objective optimization in mixing ratio calculation. The invention provides a concrete mixing proportion calculating method based on an orthogonal theory, which is characterized by comprising the following steps of: S1, system initialization and model definition, namely defining an optimization target vector containing a plurality of performance indexes, a matching factor vector containing a plurality of variable parameters, a target factor mapping knowledge base, a global fitness function for calculating global fitness scores of candidate matching schemes, and a disturbance amplitude set containing numerical values of a plurality of levels; S2, initial round evaluation, wherein a processing unit executing the method generates an initial candidate matching scheme set, and determines an initial global optimal solution by executing a physical test and fitness calculation on the initial candidate matching scheme set; S3, performing iterative optimization rounds by taking an initial global optimal solution as a starting point loop, wherein each round of iteration comprises the steps of identifying a performance index with a minimum bias fitness score as a performance short board based on each bias fitness score of the global optimal solution determined in the previous round, performing dynamic factor layering based on the performance short board and a target factor mapping knowledge base, mutually exclusive dividing a matching factor vector into a core factor set, a secondary factor set and an exploration factor set, obtaining a layered disturbance vector based on a disturbance amplitude set and factor layering, selecting an orthogonal table, and generating a new test set based on the global optimal solution and the layered disturbance vector determined in the previous round; Calculating global fitness scores of new candidate matching ratio schemes in a new test set, comparing the global fitness scores with global optimal solutions determined in the previous round, and updating to obtain new global optimal solutions; And S4,