Concrete mix proportion intelligent correction method, device and storage medium

Abstract

This invention discloses an intelligent correction method, device, and storage medium for concrete mix proportions, relating to the intersection of building material preparation and artificial intelligence technology. The method includes: calculating the corrected equivalent density of the cementitious material based on the density and mass percentage of each component, introducing an interaction coefficient; calculating a dynamic safety margin based on the corrected equivalent density and combining the measurement standard deviation of the unit water consumption and the equivalent density of the cementitious material, thereby constructing an adaptive volumetric physical feasible region; when the initially predicted mix proportion parameters do not satisfy the feasible region, constructing an energy function that assigns higher weight to the water-cement ratio deviation, and solving for preliminary correction parameters under the constraints of the feasible region; calculating derived parameters based on the preliminary correction parameters, and performing secondary optimization when the derived parameters are not within the engineering specification range. This invention elevates engineering experience to a mathematically optimal solution under the principle of energy minimization, achieving theoretical optimality and adaptive robustness of the correction path.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of building material preparation and artificial intelligence technology. Specifically, it relates to a method, device and storage medium for intelligent correction of concrete mix proportions based on energy minimization and dynamic constraints. It is particularly suitable for physical compliance correction and engineering parameter optimization of mix proportion prediction results generated by data-driven models. Background Technology

[0002] With the rapid development of artificial intelligence technology, machine learning-based concrete mix design methods have gradually become a research hotspot. These data-driven models can learn the complex mapping relationship between raw material parameters and concrete performance from a large amount of historical experimental data, thereby quickly generating mix design schemes that meet specific performance indicators and significantly improving design efficiency.

[0003] However, current mainstream data-driven models are essentially black-box models. Their decision-making process relies solely on statistical data patterns and lacks an intrinsic understanding of the physical conservation laws of concrete materials (such as the volumetric principle). Therefore, while the mix proportions output by these models may exhibit statistically superior performance, they often fail to meet physical volume requirements in practical applications. For example, the sum of the volumes of all components may not equal the unit volume of concrete, rendering the design unsuitable for direct use in engineering production. This problem severely restricts the engineering application of artificial intelligence technology in the field of concrete material preparation.

[0004] To address the aforementioned issues, existing technologies typically employ manual, empirical adjustments to the predicted results. When volume non-compliance is detected, the conventional approach is to follow the engineering rule of "maintaining the water-cement ratio and adjusting the water usage." While this adjustment path aligns with the fundamental understanding that "the water-cement ratio determines strength," it is essentially a trial-and-error method, lacking theoretical support in a mathematical sense. With multiple feasible adjustment paths coexisting, it cannot be proven that the current empirical choice is the optimal option among all feasible solutions, making it difficult to fundamentally guarantee the reliability of the adjustment results.

[0005] Furthermore, existing technologies generally employ fixed safety factors or rigid boundary conditions for setting volume constraints. This approach fails to adequately consider uncertainties in real-world engineering environments, such as random errors in raw material weighing and batch-to-batch fluctuations in cementitious material density. This "one-size-fits-all" constraint method can easily lead to two extremes: when the safety factor is set too high, the design becomes overly conservative, resulting in unnecessary material waste; when the safety factor is set insufficiently, there is a risk of non-compliant volume design.

[0006] Regarding fundamental physical models, existing techniques typically employ the standard harmonic mean formula to calculate the equivalent density of cementitious materials. This formula assumes that the volume contributions of each component during mixing are linearly additive, neglecting the nonlinear interaction effects arising from particle size distribution optimization and micro-aggregate filling of different powder materials (such as cement, fly ash, and mineral powder) during the mixing process. This oversimplified physical model renders all subsequent volume calculations based on insufficient precision, further exacerbating the discrepancy between predicted results and actual conditions.

[0007] In summary, existing data-driven mix design methods have significant shortcomings in terms of volume compliance, optimality of correction paths, adaptability of constraint boundaries, and accuracy of basic physical models. There is an urgent need for a technical solution that can integrate physical laws, quantify uncertainties, and achieve theoretically optimal corrections to promote the reliable application of artificial intelligence technology in the field of concrete materials. Summary of the Invention

[0008] To address the aforementioned deficiencies in existing technologies, the present invention aims to provide a method, device, and storage medium for intelligent correction of concrete mix proportions, thereby resolving at least one of the following technical problems: non-compliant volume of data-driven mix proportion prediction results, path-dependent correction lacks mathematical optimality support, volume constraint boundaries cannot adapt to engineering uncertainties, equivalent density models of cementitious materials ignore nonlinear interaction effects between materials, and correction schemes struggle to consider the global rationality of derived parameters.

[0009] This invention solves the above-mentioned technical problems through the following technical solution: a method for intelligent correction of concrete mix proportions, comprising:

[0010] Obtain engineering input parameters and uncertainty indicators; the engineering input parameters include the density and mass percentage of each component of the cementitious material, and the uncertainty indicators include the measurement standard deviation of unit water consumption and the measurement standard deviation of the equivalent density of the cementitious material;

[0011] Based on the density and mass percentage of each component of the cementitious material, an interaction coefficient characterizing the nonlinear volume effect between components is introduced to calculate the corrected equivalent density of the cementitious material.

[0012] The aggregate volume characterization quantity is defined based on the corrected cementitious material equivalent density, and the dynamic safety margin is calculated based on the aggregate volume characterization quantity and the uncertainty index.

[0013] An adaptive volume physical feasible region is constructed with the constraint that the aggregate volume characterization quantity is not less than the dynamic safety margin.

[0014] Determine whether the initial predicted mix proportion parameters satisfy the adaptive volume physical feasible region. If not, construct the mix proportion perturbation energy function and solve it under the constraint of satisfying the adaptive volume physical feasible region to obtain the preliminary corrected mix proportion parameters.

[0015] Calculate the derived parameters based on the initially revised mix proportion parameters, and determine whether the derived parameters are within the engineering specification range. If so, output the initially revised mix proportion parameters as the final revised scheme. Otherwise, add the engineering specification range constraint to the optimization problem, and resolve the mix proportion perturbation energy function under the condition of simultaneously satisfying the adaptive volume physical feasible region and the engineering specification range constraint to obtain the final revised scheme.

[0016] To address the shortcomings of existing technologies that commonly employ standard harmonic average formulas in calculating the equivalent density of cementitious materials, neglecting the nonlinear volume effects inherent in the mixing of different powder materials, this invention introduces a correction term characterizing the interactions between components in the equivalent density calculation. Through this technique, the invention can more accurately depict the true volumetric behavior of blends of cement, fly ash, mineral powder, and other materials, fully reflecting the synergistic effects of particle size distribution optimization and micro-aggregate filling on the system's bulk density. Compared to the traditional linear superposition assumption, the equivalent density model constructed in this invention provides more accurate and reliable foundational data for subsequent aggregate volume characterization calculations, fundamentally improving the accuracy of the entire volume balance calculation.

[0017] To address the shortcomings of existing technologies that use fixed safety factors for volume constraints and cannot adapt to engineering uncertainties, this invention proposes a dynamic safety margin determination method based on parameter measurement uncertainties. This technique directly links the boundary of the volume constraint to uncertainties such as weighing accuracy and material density fluctuations during actual construction, allowing the safety margin to adaptively adjust with changes in on-site control levels. When construction accuracy is high and material stability is good, the system automatically adopts a smaller safety margin to avoid material waste caused by over-design and improve economy. When construction accuracy is low or material fluctuations are large, the system automatically increases the safety margin, reserving sufficient buffer space for volume compliance and reducing quality risks. This mechanism completely changes the traditional rigid constraint mode of "one-size-fits-all," achieving an intelligent dynamic balance between safety and economy, making mix design more scientific and reasonable.

[0018] To address the shortcomings of existing technologies that rely on empirical rules and lack theoretical support for optimality, this invention transforms the mix proportion correction problem into an energy minimization problem constrained by volume. By assigning a significantly higher penalty weight to deviations in the water-cement ratio, this invention mathematically proves that the engineering empirical choice of "prioritizing maintaining a constant water-cement ratio" is a necessary result under the principle of energy minimization, rather than an arbitrary choice based on subjective judgment. This technique ensures that, while meeting volume compliance requirements, the correction scheme minimizes changes to the initial prediction results, maximizing the preservation of the statistical optimization characteristics of the data-driven model. Furthermore, since the water-cement ratio is a core parameter determining concrete strength, prioritizing its stability fundamentally guarantees the strength stability of the corrected mix proportion, making the correction process both rationally intuitive in engineering and rigorously mathematically sound.

[0019] To address the drawback of single-volume constraint corrections potentially causing other engineering parameters, such as sand ratio, to deviate from reasonable ranges, this invention designs a layered verification and optimization mechanism. After initial volume correction, the system automatically performs secondary verification on derived parameters such as sand ratio to ensure they remain within preset engineering specification ranges. When any derived parameter is detected to be out of bounds, the system automatically upgrades the correction problem to multi-constraint optimization, incorporating other engineering constraints such as sand ratio while satisfying volume constraints, and resolving the globally optimal correction solution. This mechanism effectively prevents the problem of "paying attention to one aspect while neglecting others" in local optimization, ensuring that the final output mix design not only complies with volume requirements but also meets engineering requirements in terms of workability, durability, and other comprehensive performance aspects. By forming a closed-loop process of "initial correction—secondary verification—multi-constraint optimization," this invention significantly improves the global feasibility and engineering implementation reliability of the correction solution.

[0020] Furthermore, the measurement standard deviation of the unit water consumption and the measurement standard deviation of the equivalent density of the cementitious material are calculated using statistical process control methods based on online weighing data and / or laboratory test data within a sliding time window, and data points exceeding the range of ±3 times the current period's standard deviation are removed during the calculation process.

[0021] By employing a sliding time window mechanism, the statistical analysis of measurement standard deviation can reflect recent production fluctuations and changes in measurement accuracy in real time, avoiding interference from outdated fluctuation information in long-term historical data on current uncertainty assessments. Outlier removal logic effectively eliminates distortions in standard deviation calculations caused by abnormal data points due to sensor malfunctions, recording errors, or extreme random factors, improving the accuracy and robustness of uncertainty indicators. Combining online weighing data with laboratory testing data ensures both the real-time nature of data acquisition and the accuracy of laboratory data, providing a reliable data foundation for subsequent dynamic safety margin calculations.

[0022] Furthermore, the corrected equivalent density of the cementitious material is calculated according to the following formula:

[0023] ;

[0024] in, This represents the corrected equivalent density of the cementitious material; n represents the number of components in the cementitious material. This represents the mass percentage of the i-th component; This represents the density of the i-th component; The interaction coefficient between component i and component j is obtained by fitting multiple sets of apparent density test data of cementitious material slurry with different blending ratios, or by looking up a table in a pre-set material parameter library.

[0025] By providing explicit mathematical expressions, an operable implementation scheme for equivalent density calculation is offered. The introduction of interaction coefficients effectively characterizes the nonlinear volume effect during the mixing of multi-component powder materials, making the equivalent density calculation results closer to the real physical state. It is stipulated that the interaction coefficients can be obtained by fitting multiple sets of apparent density test data of cementitious material slurries with different blending ratios, ensuring the statistical reliability of coefficient calibration. It allows the coefficients to be obtained from a pre-set material parameter library, improving the convenience of engineering applications. When the interaction coefficients are missing, the model can degenerate into a standard harmonic average formula, ensuring the applicability of the technical solution under various data conditions.

[0026] Furthermore, the engineering input parameters also include air content, and the aggregate volume characterization quantity is defined according to the following formula:

[0027] ;

[0028] in, W represents the aggregate volumetric quality; r represents the water consumption per unit volume; and r represents the water-cement ratio. Indicates the density of water; Indicates air content; This indicates the corrected equivalent density of the cementitious material.

[0029] The concrete volumetric method principle is transformed into a calculable form through a complete mathematical expression, laying a clear mathematical foundation for the subsequent construction of volume constraints. By incorporating air content as an independent parameter into the calculation, the model can accurately reflect the actual pore volume in concrete, avoiding volume calculation errors caused by ignoring air content. The expression clearly demonstrates the influence of various engineering parameters on aggregate volume, providing a clear basis for solving partial derivatives in the subsequent calculation of sensitivity coefficients.

[0030] Furthermore, the dynamic safety margin is calculated using the following formula:

[0031] ;

[0032] in, Indicates dynamic safety margin; Indicates the basic safety margin; Indicates the confidence level coefficient; This represents the standard deviation of a unit of water consumption. The standard deviation of the measurement representing the equivalent density of cementitious materials; This represents a quantity that characterizes aggregate volume. express The absolute value of the partial derivative with respect to unit water consumption W; express The corrected equivalent density of cementitious materials The absolute value of the partial derivative.

[0033] Through explicit mathematical expressions, the theory of uncertainty quantification is transformed into a computationally feasible method for engineering implementation; the setting of the basic safety margin ensures that minimum design redundancy is retained even under the ideal condition of zero measurement error; the introduction of the confidence level coefficient allows the safety margin to be adjusted according to engineering reliability requirements, meeting the needs of projects with different safety levels; the use of the sensitivity coefficient (absolute value of partial derivative) quantifies the impact of different parameter errors on volume compliance and incorporates it into the margin calculation, making the allocation of the safety margin more scientific and reasonable; the overall formula realizes the quantitative transfer of measurement uncertainty to the volume constraint boundary.

[0034] Furthermore, the energy function E for the mix ratio perturbation is constructed as follows:

[0035] ;

[0036] in, , , The initial predicted unit water consumption, water-cement ratio, and cementitious material consumption are represented by , r, and B, which represent the corrected unit water consumption, water-cement ratio, and cementitious material consumption to be determined, and B = W / r. , Indicates the weighting coefficient. ;

[0037] Alternatively, in a simplified form:

[0038] ;

[0039] in, , Indicates the weighting coefficient. .

[0040] The energy function adopts a normalized function form, eliminating the influence of differences in the dimensions and orders of magnitude of different parameters on the optimization results, enabling the energy function to fairly measure the relative deviation of each parameter; the weighting coefficients satisfy the order of magnitude constraint, mathematically ensuring that the deviation of the water-glue ratio dominates in the energy function, so that the correction process will necessarily prioritize keeping the water-glue ratio basically unchanged; the complete energy function form provides a unified optimization framework for scenarios that require simultaneous adjustment of three parameters; the simplified form reduces computational complexity and is suitable for conventional scenarios where only adjusting the water consumption is needed to restore volume compliance.

[0041] Furthermore, the solution under the constraint of satisfying the adaptive volume physical feasible region employs the Lagrange multiplier method to construct the Lagrange function:

[0042] ;

[0043] Where L represents the Lagrange function; E represents the mix proportion perturbation energy function; Represents the Lagrange multipliers; This represents a quantity that characterizes aggregate volume. Indicates dynamic safety margin;

[0044] The preliminary modified mix proportion parameters were obtained by solving the first-order optimality conditions.

[0045] Transforming the constrained optimization problem into an unconstrained one reduces the mathematical complexity of the solution; the introduction of Lagrange multipliers quantifies the influence of constraints on the optimal solution; solving by first-order optimality conditions ensures that the obtained solution satisfies the necessary extreme conditions for constrained optimization problems; this mathematical framework provides a theoretical basis for proving that "keeping the water-glue ratio constant" is a necessary result under the principle of energy minimization; and the explicit solution method provides a clear algorithmic path for computer program implementation.

[0046] Furthermore, when it is necessary to simultaneously satisfy the adaptive volume physical feasible region and the engineering specification interval constraints, the modified problem is constructed as the following multi-constraint optimization problem:

[0047] ;

[0048] ;

[0049] Where E represents the energy function of the mix proportion disturbance; W and r represent the corrected unit water consumption and water-cement ratio to be determined. This represents a quantity that characterizes aggregate volume. Indicates dynamic safety margin; Indicates derived parameters; and This indicates the lower and upper limits allowed for the project corresponding to the derived parameters;

[0050] The sequential quadratic programming algorithm is used to solve the multi-constraint optimization problem, and the final modified scheme is obtained.

[0051] Incorporating the reasonable range of derived parameters such as sand ratio as explicit constraints into the optimization problem fundamentally ensures the compliance of the modified scheme with respect to derived parameters; explicit constraint expressions provide a clear mathematical interface for computer program implementation; allow the addition of other engineering constraints such as strength, workability, and durability according to actual needs, giving the technical solution good scalability; the sequential quadratic programming algorithm is a mature algorithm for solving nonlinear constrained optimization problems, which can efficiently and stably converge to the optimal solution that satisfies all constraints, ensuring the engineering solvability of multi-constraint optimization problems.

[0052] Based on the same concept, the present invention also provides an electronic device, including a memory, a processor, and a computer program or instructions stored in the memory, wherein the processor executes the computer program or instructions to implement the intelligent correction method for concrete mix proportions as described above.

[0053] Based on the same concept, the present invention also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implements the intelligent correction method for concrete mix proportions as described above.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] By introducing the interaction coefficient between materials, an equivalent density model that can characterize the nonlinear volume effect when multi-component powders are mixed is constructed. This changes the calculation bias caused by the traditional linear superposition assumption and provides a more solid physical basis for the entire volume balance analysis.

[0056] A dynamic safety margin method is proposed, which links the volume constraint boundary with the measurement uncertainty of engineering parameters, so that the safety margin can be adaptively adjusted with construction accuracy and material fluctuations, achieving a dynamic balance between safety and economy, and avoiding the risks of excessive conservatism or compliance caused by a fixed safety factor.

[0057] By transforming the mix proportion correction problem into an energy minimization problem, and by assigning a significantly higher penalty weight to deviations in the water-cement ratio, it is mathematically proven that the engineering experience of "prioritizing the maintenance of a constant water-cement ratio" is an inevitable result under the principle of energy minimization. This makes the correction process both rational based on engineering intuition and rigorous based on mathematical theory.

[0058] By adding a secondary verification and multi-constraint optimization mechanism, the rationality of derived parameters such as sand ratio is further ensured on the basis of satisfying volume constraints, forming a closed-loop process of "preliminary correction - secondary verification - multi-constraint optimization", which effectively prevents global parameter mismatch caused by local optimization and significantly improves the engineering feasibility of the correction scheme. Attached Figure Description

[0059] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only one embodiment of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0060] Figure 1 This is a flowchart of the intelligent correction method for concrete mix proportions in an embodiment of the present invention;

[0061] Figure 2 This is a schematic diagram illustrating the principle of the traditional harmonic mean model;

[0062] Figure 3 This is a schematic diagram illustrating the principle of the equivalent density model of cementitious materials in this embodiment of the invention;

[0063] Figure 4 This is a schematic diagram of the energy minimization correction path in an embodiment of the present invention;

[0064] Figure 5 This is a performance comparison chart of different correction methods in terms of perturbation energy in the embodiments of the present invention;

[0065] Figure 6 This is a performance comparison chart of different correction methods in the final aggregate volume in the embodiments of the present invention. Detailed Implementation

[0066] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0068] Example 1

[0069] This invention provides an intelligent correction method for concrete mix proportions, aiming to address the problems of existing data-driven mix proportion predictions, which often result in non-compliant volume, path-dependent corrections, and a lack of theoretical support for optimality. This method constructs a corrected equivalent density of the cementitious material by introducing inter-material interaction coefficients, calculates a dynamic safety margin based on parameter measurement uncertainties, and constructs an adaptive volumetric physical feasible region. When the initial prediction violates the feasible region, it is corrected using the principle of energy minimization. Finally, secondary verification and multi-constraint optimization ensure the overall engineering rationality. (Refer to...) Figure 1 The intelligent correction method for concrete mix proportions includes the following steps:

[0070] Step S1: Obtain engineering input parameters and uncertainty indicators.

[0071] Step S1 provides the foundational data for all subsequent calculations and quantifies the inherent uncertainty of the engineering parameters. Data acquisition can be performed in parallel or selectively through two approaches:

[0072] Approach 1: Real-time data acquisition via online sensors: Deploying high-precision online measurement equipment on the concrete production line, including but not limited to:

[0073] Electronic scale: Used for real-time weighing of unit water consumption and the mass of each component of cementitious material; the sampling frequency can be set according to the production rhythm.

[0074] Densitometer: Used for online detection of the apparent density of cementitious material slurry. Nuclear densitometer or vibration densitometer can be used to monitor material fluctuations in real time.

[0075] Data collected by online sensors is transmitted in real time to the data processing center via the Industrial Internet, forming a real-time data stream.

[0076] Method 2: Historical Database Retrieval: For raw material parameters or historical production batch data that have been tested offline, retrieve them from the Enterprise Resource Planning (ERP) system or Laboratory Information Management (LIM) system. The historical database includes: raw material incoming inspection reports (such as cement density, fly ash density, etc.), mix proportion records of historical production batches and corresponding measured performance data, and material density test data from periodic laboratory sampling.

[0077] The engineering input parameters obtained in this embodiment include at least the density and mass percentage of each component of the cementitious material, as well as the air content. The cementitious material typically comprises a combination of one or more components such as cement, fly ash, and mineral powder. The density of each component can be obtained from quality certificates provided by the material supplier, actual measurement data from on-site re-inspection, or historical statistical averages.

[0078] The mass percentage represents the relative proportion of each cementitious material component in the current mix design, satisfying the normalization condition. The mass percentage can be extracted directly from the initial predicted mix design to be corrected, or it can be directly input by the user.

[0079] Air content is the volume fraction of air bubbles in a unit volume of concrete. It is usually preset according to the type and strength grade of concrete, and its value range is usually 1% to 3%.

[0080] Uncertainty indicators include at least the measurement standard deviation of water consumption per unit volume and the measurement standard deviation of the equivalent density of the cementitious material. These two standard deviations are calculated using statistical process control methods based on online weighing data and / or laboratory testing data within a sliding time window. Specifically, production data from the most recent 30 production batches, the most recent week, or the most recent month can be selected as the sliding time window. The standard deviation of the measured water consumption per unit volume for all batches within the sliding time window is calculated as its measurement standard deviation, and the standard deviation of the measured equivalent density of the cementitious material for all batches within the sliding time window is also calculated as its measurement standard deviation.

[0081] During the calculation process, data points exceeding the range of ±3 times the current period's standard deviation are removed to avoid interference from outliers in the statistical results. After removing outliers, the standard deviation is recalculated for the remaining data, and this process is repeated iteratively until no new outliers are detected. The final standard deviation is taken as the corresponding measurement standard deviation.

[0082] The measured equivalent density of cementitious materials can be obtained through the following methods:

[0083] For historical batches, the equivalent density of the batch is calculated according to the actual proportion of cementitious materials used in the batch and the measured density of each component, using the standard harmonic average formula (or the modified formula if there is an interaction coefficient); the calculated equivalent density is used as the measured value of the equivalent density of the cementitious materials in that batch.

[0084] Step S2: Based on the density and mass ratio of each component of the cementitious material, an interaction coefficient characterizing the nonlinear volume effect between components is introduced to calculate the corrected equivalent density of the cementitious material.

[0085] In concrete engineering, cementitious materials are typically composed of various powdered materials such as cement, fly ash, and mineral powder mixed in a certain proportion. Traditional equivalent density models use a standard harmonic mean formula, assuming that the volume contributions of each component are independent, essentially a simplified model of linear superposition. However, in actual engineering, due to differences in particle morphology and size distribution among different powdered materials, synergistic effects such as particle size distribution optimization and micro-aggregate filling often occur during the mixing process, causing the actual bulk density of the mixed system to deviate from the theoretical calculation value obtained by the independent superposition assumption.

[0086] To address the aforementioned issues, this invention improves upon the traditional equivalent density model by introducing an interaction coefficient characterizing the nonlinear volume effect between materials, thus constructing a modified equivalent density model for cementitious materials. This model more accurately depicts the true physical behavior of multi-component mixtures. The modified equivalent density of the cementitious material is calculated using the following formula:

[0087] (1)

[0088] in, This represents the corrected equivalent density of the cementitious material; n represents the number of components in the cementitious material. This represents the mass percentage of the i-th component, satisfying... ; This represents the density of the i-th component, expressed in kg / m³. 3 ; This represents the interaction coefficient between the i-th component and the j-th component.

[0089] The physical meaning of the modified model shown in expression (1) is as follows: the first term on the right side of the equation is the reciprocal of the volume contribution of each component independently, i.e., the traditional harmonic average term; the second term is the interaction correction term, which is used to characterize the nonlinear volume change caused by the synergistic effect of particle size distribution optimization and micro-aggregate filling when the two materials are mixed. When the interaction coefficient is positive, it means that the actual volume after mixing is less than the linear superposition value (i.e., the packing is denser); when the interaction coefficient is negative, it means that the actual volume after mixing is greater than the linear superposition value (i.e., the packing is looser).

[0090] Interaction coefficient It can be obtained in the following two ways:

[0091] The first method is the experimental fitting method: At least five groups of cementitious material pastes with typical admixture ratios are selected for apparent density testing. Taking cement and mineral powder as examples, the following mass admixture ratio combinations can be selected:

[0092] Cement:mineral powder = 7:3, cement:mineral powder = 6:4, cement:mineral powder = 5:5, cement:mineral powder = 4:6, cement:mineral powder = 3:7.

[0093] For each mix proportion, the apparent density of the cementitious slurry was measured, and the corresponding actual specific volume (the reciprocal of density) was calculated. Simultaneously, the theoretical specific volume was calculated using the standard blending average formula. The difference between the actual and theoretical specific volumes is the interaction effect contribution value. The interaction effect contribution values ​​for different mix proportions were then compared with the corresponding... Perform linear regression, and the regression coefficients will be the interaction coefficients we are looking for.

[0094] The second method is using a material parameter database lookup: A pre-configured material parameter database, containing interaction coefficients for common material combinations, can be used during system deployment. For example, a combination of silicate cement and S95 grade mineral powder: Combination of silicate cement and Class I fly ash: Combination of mineral powder and fly ash: .

[0095] The material parameter library can be configured according to regional material characteristics, engineering experience, or industry standards, and can be continuously updated and optimized based on measured data during system use.

[0096] To ensure the robustness and backward compatibility of the algorithm in engineering deployment, a degradation mechanism is designed: when the interaction coefficient of the target material combination is not found in the material parameter library and experimental fitting data is lacking, the system automatically degrades all... This causes the modified model to degenerate into a standard harmonic mean formula without interaction terms:

[0097] (2)

[0098] This degradation mechanism ensures that the method of the present invention can operate normally under any data conditions: when there is sufficient material interaction data, a more accurate correction model is used to improve the calculation accuracy; when data is missing, it automatically switches to the traditional model to ensure that the basic functions are not interrupted, which reflects the engineering adaptability and reliability of the algorithm.

[0099] Figure 2 and Figure 3 The comparison demonstrates the working principles of the traditional harmonic mean model and the equivalent density model of this invention, which considers nonlinear interaction effects. The traditional harmonic mean model assumes that the volumes of cement and mineral powder are linearly additive when mixed, and that the total volume after mixing equals the sum of the volumes of each component. The corresponding equivalent density calculation formula is in the standard harmonic mean form. The model of this invention introduces an interaction coefficient to characterize nonlinear effects such as particle size distribution optimization and micro-aggregate filling. The actual total volume V′ after mixing is less than the linear superposition value V1+V2, and the corresponding correction term is... . Figure 2 and Figure 3 This intuitively reveals the physical essence of how the model of this invention can more accurately characterize the true volumetric behavior of multi-component cementitious materials.

[0100] Step S3: Define the aggregate volume characterization quantity based on the corrected cementitious material equivalent density, and calculate the dynamic safety margin based on the aggregate volume characterization quantity and uncertainty index.

[0101] Step S3 aims to construct a slurry volume model and define aggregate volume characterization quantities based on the corrected equivalent density of cementitious materials, and then calculate the dynamic safety margin according to the measurement uncertainty of engineering parameters, laying the foundation for the subsequent construction of the adaptive volume physical feasible region.

[0102] Based on the corrected equivalent density of the cementitious material calculated in step S2, a slurry volume model including water and cementitious material is first constructed. The volume V occupied by the slurry in a unit volume of concrete is then defined. p Calculate using the following formula:

[0103] (3)

[0104] Where W represents the unit water consumption, in kg / m³ 3 ; r represents the water-to-glue ratio, which is dimensionless; The density of water is usually expressed as 1000 kg / m³. 3 ; Indicates air content; This indicates the corrected equivalent density of the cementitious material, in kg / m³. 3 .

[0105] The physical meaning of expression (3) is that the first term on the right side of the equation is the volume occupied by the unit water consumption, and the second term is the volume occupied by the cementitious material (because the amount of cementitious material B = W / r). The sum of the two is the total volume of the slurry.

[0106] After considering the air volume entrained in concrete, a mass fraction of aggregate volume is defined. Aggregate volumetric properties represent the volume of concrete per unit volume after deducting the volume of paste and air, which is available for aggregate to occupy. It is calculated using the following formula:

[0107] (4)

[0108] in, This indicates the air content, representing the volume fraction of air bubbles entrained in a unit volume of concrete. It can be preset according to the concrete type and strength grade.

[0109] Substituting formula (3) into formula (4), we obtain the complete expression for the aggregate volume characterization:

[0110] (5)

[0111] This invention abandons the traditional approach of using a fixed safety factor and introduces the concept of dynamic safety margin, upgrading the hard constraint of volume balance to an adaptive volumetric physical feasible region defined by inequalities. Dynamic safety margin The calculation formula is based on the dynamic calculation of the measurement uncertainty of engineering parameters and is derived from the sensitivity analysis of volume error.

[0112] (6)

[0113] in, It represents the basic safety margin, which is the minimum design redundancy that must be retained even under ideal conditions where the measurement error is zero. This is a configurable parameter, and its preferred value range is 0.001 to 0.01. A value that is too small may lead to volume compliance risks, while a value that is too large may cause unnecessary material waste. The specific value can be configured based on the importance level of the project and user experience.

[0114] This represents the confidence level coefficient, used to reflect reliability requirements. The preferred value range is 2 to 4, corresponding to 2 in statistics. Up to 4 Confidence level. A larger value indicates a more conservative safety margin and a higher confidence level in volume compliance; a smaller value indicates a more economical safety margin, but allows for slightly higher compliance risk. A larger value can be selected for important structural projects. For general engineering projects, a smaller value can be selected. value.

[0115] This represents the standard deviation of a unit of water consumption. This represents the standard deviation of the measurement of the equivalent density of the cementitious material. See the description in step S1 for how to obtain these two standard deviations.

[0116] express The absolute value of the partial derivative with respect to unit water consumption W, i.e., the sensitivity coefficient; express The corrected equivalent density of cementitious materials The absolute value of the partial derivatives is the sensitivity coefficient. These two sensitivity coefficients reflect the degree of influence of changes in various parameters on aggregate volume. Their physical meaning is that they quantify the amount of aggregate volume change caused by a unit change in water consumption or equivalent density, thereby amplifying the parameter error through the sensitivity coefficient and transforming it into an impact on volume compliance.

[0117] The adaptive adjustment of the constraint boundary can be achieved through the above dynamic safety margin calculation formula (formula (6)). The core mechanism is as follows:

[0118] When low on-site construction precision leads to large fluctuations in unit water consumption ( When the density increases, or when there are large density fluctuations between batches of raw materials, the equivalent density standard deviation may be affected. When the size increases, the system will automatically calculate a larger dynamic safety margin, thus making the mix design more conservative and reserving more buffer space for volume compliance; conversely, when the measurement uncertainty is small ( and When the dynamic safety margin is smaller, the system will adopt a smaller dynamic safety margin to avoid material waste caused by over-design and achieve a more economical mix design.

[0119] This mechanism links volumetric constraint boundaries with actual engineering control levels, so that safety margins are no longer fixed empirical values, but can be adaptively adjusted according to construction accuracy and material stability, thereby achieving an intelligent dynamic balance between safety and economy.

[0120] Step S4: Construct an adaptive volumetric physical feasible region with the constraint that the aggregate volume characterization quantity is not less than the dynamic safety margin.

[0121] Step S4 aims to construct a volumetric physical feasible domain that can adaptively adjust to engineering uncertainties based on the aggregate volume characterization quantity defined in step S3 and the calculated dynamic safety margin, thus providing clear constraint boundaries for subsequent mix proportion correction.

[0122] In concrete mix design, the volumetric method is the fundamental physical law that ensures the sum of the volumes of all components equals the unit volume. Traditional methods typically use fixed boundaries as volume constraints, such as directly requiring the aggregate volume to be equal to a certain constant or greater than a certain fixed threshold. This rigid constraint cannot adapt to the uncertainties brought about by material fluctuations and measurement errors in actual engineering, and can easily lead to overly conservative or risky designs.

[0123] To address the aforementioned problems, this invention incorporates the dynamic safety margin calculated based on measurement uncertainty in step S3 into the volume constraint, constructing an adaptive volumetric physical feasible region. The core of this feasible region lies in the fact that its boundary is no longer a fixed constant, but can be dynamically adjusted according to construction accuracy and material stability, thereby achieving adaptive changes in the constraint conditions.

[0124] An adaptive volumetric physical feasible region is constructed with the constraint that the aggregate volume characterization quantity is not less than the dynamic safety margin. Its mathematical expression is:

[0125] (7)

[0126] The physical meaning of expression (7) is that the set of all mix proportion parameters (W,r) that satisfy the requirement that the aggregate volume characterization quantity is not less than the dynamic safety margin constitutes the volumetric physical feasible region under the current engineering conditions. Only mix proportion schemes located within this feasible region are considered compliant in a volumetric physical sense.

[0127] The constructed volumetric physically feasible region has the following adaptive properties:

[0128] When the on-site construction accuracy is low, resulting in a large standard deviation in the measurement of unit water consumption, the dynamic safety margin automatically increases, the feasible region shrinks accordingly, the volume compliance requirements for the mix design become more stringent, and more safety redundancy is forcibly retained; conversely, when the construction accuracy is high, resulting in a small standard deviation in the measurement of unit water consumption, the dynamic safety margin decreases, the feasible region widens, and more economical mix design is allowed.

[0129] When the density fluctuations between batches of raw materials are large, leading to an increase in the standard deviation of the equivalent density of the cementitious material, the dynamic safety margin also automatically increases, and the feasible region shrinks; when the material stability is good, resulting in a small standard deviation of the equivalent density of the cementitious material, the feasible region is correspondingly widened.

[0130] Users can configure the confidence level coefficient. This allows for adjustment of the slackness of the feasible region. For important structural engineering projects, a larger slackness can be selected. This value makes the feasible region more stringent, ensuring high reliability; for general engineering projects, a smaller value can be selected. This value makes the feasible region relatively flexible, thus improving economic efficiency.

[0131] In a two-dimensional parameter space defined by unit water consumption W and water-cement ratio r, the aggregate volume characterization quantity It is a continuous function of unit water consumption W and water-cement ratio r. For a given dynamic safety margin, the equation is... A contour line is defined in the parameter space; this contour line forms the boundary of the feasible region. The feasible region is the area enclosed by this contour line that satisfies... The part.

[0132] When the dynamic safety margin increases, the contour lines contract inward, and the feasible region decreases; when the dynamic safety margin decreases, the contour lines expand outward, and the feasible region increases. This geometric relationship intuitively demonstrates the adaptive characteristics of the feasible region.

[0133] Step S5: Determine whether the initial predicted mix proportion parameters satisfy the adaptive volume physical feasible region. If not, construct the mix proportion perturbation energy function and solve it under the constraint of satisfying the adaptive volume physical feasible region to obtain the preliminary corrected mix proportion parameters.

[0134] Step S5 aims to determine whether the initial predicted mix proportion parameters satisfy the adaptive volume physical feasible region constructed in step S4, and if not, to trigger a correction mechanism based on the principle of energy minimization, and obtain the initially corrected mix proportion parameters by solving the constraint optimization problem.

[0135] The obtained initial predicted mix proportion parameters should include at least the initially predicted unit water consumption. Water-to-glue ratio These initial predicted mix proportions can be generated by data-driven models (such as machine learning models) based on historical data, or initially formulated based on experience.

[0136] Substituting the initial predicted mix proportion parameters into formula (5), the corresponding aggregate volume characterization value is obtained. .

[0137] like If the initial predicted mix proportion parameters already meet the requirements of the adaptive volume physical feasible region, no correction is needed, and they can be directly used as the initially corrected mix proportion parameters for subsequent steps.

[0138] like If the initial predicted mix proportion parameters violate the requirements of the adaptive volume physical feasible region, the energy minimization correction mechanism needs to be triggered to optimize and adjust the parameters to bring them into the feasible region.

[0139] To measure the deviation of the modified scheme from the initial predicted scheme, a mix proportion perturbation energy function is defined. This invention provides two forms of the energy function, which can be selected according to engineering implementation requirements:

[0140] Complete form (including the amount of cementitious materials):

[0141] (8)

[0142] Where W, r, and B represent the unit water consumption, water-cement ratio, and cementitious material dosage to be corrected, and ; , This represents the weighting coefficients, reflecting the cost of different engineering parameters deviating from their initial values. The full form also incorporates deviations in cementitious material usage into the energy function, making it suitable for scenarios requiring strict control of cementitious material usage. All terms are normalized (divided by the square of the initial value), eliminating the influence of dimensional and order-of-magnitude differences on energy measurement.

[0143] Simplified form (containing only unit water consumption and water-cement ratio):

[0144] (9)

[0145] in, , This represents the weighting coefficient. The simplified form is suitable for typical scenarios where only adjusting water usage is needed to restore volume compliance.

[0146] Regardless of the form used, the weighting coefficients must satisfy the following order of magnitude relationship: , The physical significance of this asymmetric weighting structure lies in the fact that the water-cement ratio is a core control parameter determining the strength of concrete, and deviations from it incur a cost far exceeding deviations in unit water consumption. By assigning a significantly higher weight to the water-cement ratio deviation term, it mathematically ensures that the correction process prioritizes maintaining the water-cement ratio essentially constant.

[0147] The objective of the modified problem is to minimize the perturbation energy of the initial prediction by the modified scheme, while satisfying the adaptive volume physical feasible region. Based on the definition of aggregate volume characterization, the optimization problem can be formulated as:

[0148] (10)

[0149] This is a nonlinear optimization problem with inequality constraints. The inequality constraints are activated when the optimal solution lies exactly on the boundary of the feasible region; at this point, the problem is equivalent to the problem with equality constraints: .

[0150] This invention can employ various numerical optimization methods to solve the above-mentioned constrained optimization problem, with the Lagrange multiplier method or sequential quadratic programming algorithm being preferred. The following explanation uses the Lagrange multiplier method as an example to illustrate the solution principle:

[0151] For inequality-constrained problems, we can first assume that the constraints are activated, introduce the Lagrange multiplier λ, and construct the Lagrange function:

[0152] (11)

[0153] Based on the first-order optimality condition, taking the partial derivative with respect to r and setting it to zero, the deviation of the water-glue ratio can be expressed as:

[0154] (12)

[0155] in, This represents the optimal solution, i.e., the initially corrected unit water consumption and water-to-binder ratio; Represents the energy function The partial derivative of r does not include The set of terms, whose values ​​are finite. Because Located in the denominator and having a sufficiently large value ( or ), and All are finite values, and the right side of formula (12) approaches zero overall, thus we have: .

[0156] This derivation proves that when the penalty weight for deviation of the water-cement ratio is large enough, the optimal solution to the energy minimization problem must converge to the correction path of "keeping the water-cement ratio basically unchanged". In other words, the correction process will prioritize maintaining the stability of the water-cement ratio, mainly by adjusting the unit water consumption W (and the amount of cementitious material B derived from B = W / r) to meet the volume constraint.

[0157] In practical engineering deployments, mature numerical optimization libraries (such as NLopt, IPOPT, and other optimization solvers) can be used to automatically solve the above optimization problems. Algorithm inputs include: energy function. constraint functions and its gradient, initial point Variable boundaries (e.g., W > 0, r > 0). The algorithm iterates until the convergence condition is met (e.g., the gradient norm is less than a threshold or the iteration step size is less than a threshold), and outputs the optimal solution. .

[0158] If the complete energy function is used, the amount of cementitious material used can also be calculated based on B = W / r after preliminary correction.

[0159] Figure 4 A plane for mixing ratio parameters was constructed with unit water consumption W as the horizontal axis and water-cement ratio r as the vertical axis. Figure 4 curves in The region to the right of the curve represents the boundary of the adaptive volume physically feasible region. The left side represents the physically infeasible region ( Initial predicted mix proportion parameters Located within the infeasible region. Traditional correction methods typically project to the boundary along the vertical direction (keeping r constant), i.e., the conventional projection path; while this invention, based on the principle of energy minimization, applies an order-of-magnitude penalty weight to the deviation of the water-cement ratio ( ), optimal solution While maintaining the water-to-glue ratio as constant as possible while satisfying the boundary constraints, the modified path is approximately horizontally shifted and eventually converges to... The boundary point.

[0160] Step S6: Calculate the derived parameters based on the preliminary revised mix proportion parameters.

[0161] Derived parameters refer to engineering indicators reflecting the comprehensive performance of concrete, further derived from the initially revised mix proportion parameters. In this embodiment, the derived parameters include at least the sand ratio. The sand ratio is one of the key parameters in concrete mix design, defined as the percentage of fine aggregate mass to the total aggregate mass (the sum of the mass of fine aggregate and coarse aggregate). Its value directly affects the workability, strength, and durability of concrete. The sand ratio can be calculated using one of the following methods:

[0162] Method 1 (based on empirical formula): Based on parameters such as concrete strength grade, maximum coarse aggregate size and fineness modulus of fine aggregate, the sand ratio is estimated using empirical formulas in the "Specification for Mix Proportion Design of Ordinary Concrete". This estimated value is used as a reference value for the sand ratio under the current mix proportion.

[0163] Method 2 (Reverse Calculation Based on Absolute Volume Method): Assuming the aggregate gradation satisfies the theory of closest packing, the total aggregate volume V is calculated based on parameters such as the porosity and apparent density of fine and coarse aggregates. agg The ratio of fine aggregate to coarse aggregate is used to calculate the sand ratio.

[0164] Step S7: Determine whether the derived parameters are within the engineering specification range; if so, output the initially corrected mix proportion parameters as the final corrected scheme; otherwise, add the engineering specification range constraint to the optimization problem, and re-solve the mix proportion perturbation energy function under the condition of simultaneously satisfying the adaptive volume physical feasible region and the engineering specification range constraint to obtain the final corrected scheme.

[0165] Step S7 aims to perform a global rationality check on the derived parameters. Through a secondary verification mechanism, it is ensured that the modified mix proportion scheme not only meets the core volume constraints, but also complies with the specifications in terms of workability, economy, and other engineering indicators, thereby achieving the global feasibility of the modified scheme.

[0166] The derived parameters (e.g., sand ratio) calculated in step S6 are compared with a preset engineering specification range. The engineering specification range can be configured according to concrete strength grade, project type, construction conditions, etc., for example:

[0167] For ordinary concrete with strength grades of C30 to C50, the reasonable range for the sand ratio can be preset to [0.38, 0.45].

[0168] For high-strength concrete, the sand ratio may be slightly lower, such as [0.35, 0.42];

[0169] For pumped concrete, the sand ratio should be relatively high, such as [0.40, 0.48].

[0170] Users can flexibly configure these intervals in the system according to different engineering needs.

[0171] The specific judgment logic is as follows:

[0172] like If the derived parameters are within the corresponding engineering specification range, the preliminarily corrected mix proportion parameters meet the global rationality requirements, and are directly output as the final corrected scheme. Indicates derived parameters; and This indicates the lower and upper limits allowed for the project corresponding to the derived parameters.

[0173] like or If the derived parameter is out of bounds, the multi-constraint optimization mechanism needs to be triggered to perform secondary optimization on the initially corrected mix proportion parameters.

[0174] When the derived parameters exceed the corresponding engineering specification range, the system automatically upgrades the correction problem to multi-constraint optimization. Based on the original adaptive volume physical feasible region constraint, it adds the engineering specification range constraint corresponding to the derived parameters and re-solves for the minimum value of the energy function. The mathematical expression of the multi-constraint optimization problem is shown in formula (10), and the constraints are as follows: and .

[0175] This optimization problem can be solved using the sequential quadratic programming algorithm. In each iteration, the sequential quadratic programming algorithm constructs a quadratic programming subproblem to approximate the original problem and performs a one-dimensional search along the search direction until the convergence condition is met. The algorithm input includes: energy function. Volume constraint function and its gradient and derived parameter constraint functions and and its gradient, initial point Variable boundaries (e.g., W > 0, r > 0). The algorithm iterates until the convergence condition is met (e.g., the gradient norm is less than a threshold or the iteration step size is less than a threshold), and outputs the optimal solution. This is the final modified scheme that simultaneously satisfies the adaptive volume physical feasible region and the engineering specification interval constraints.

[0176] In addition to the sand ratio, the present invention may also incorporate constraints from other derived parameters as needed, such as:

[0177] Strength constraint: Estimate the concrete strength based on the corrected water-cement ratio to ensure that it is not lower than the design strength grade;

[0178] Workability constraints: Estimate the slump based on the unit water consumption and sand ratio to ensure that construction requirements are met;

[0179] Durability constraints: Durability indicators such as freeze resistance and impermeability are determined based on the amount of cementitious materials used and the water-cement ratio.

[0180] Material usage limits: such as minimum amount of cementitious materials and maximum water usage.

[0181] These constraints can all be expressed as inequalities with respect to W and r, and are all incorporated into the multi-constraint optimization problem for solution, thereby obtaining a final modified scheme that fully meets the engineering requirements.

[0182] After completing step S7, the system outputs the final corrected solution (i.e., the mix proportion parameters that satisfy volume constraints and all derived parameter constraints) to the production control system to guide actual production. Simultaneously, the system writes back the input data, correction process, and final solution to the database for subsequent model optimization and parameter library updates, forming a data closed loop.

[0183] Through the secondary verification and multi-constraint optimization mechanism in step S7, this invention ensures that the modified mix proportion scheme is not only compliant in terms of volume physical meaning, but also meets the specification requirements in terms of key engineering parameters such as sand ratio, thereby significantly improving the overall feasibility and reliability of the scheme in engineering implementation.

[0184] Example 2

[0185] To quantitatively illustrate the technical advantages of this invention over traditional correction methods, this embodiment provides a set of simulation comparison examples to verify the optimality and adaptive robustness of this invention under different construction accuracy conditions.

[0186] Set a set of initial predicted mix proportion parameters: unit water consumption W0 = 230 kg / m³ 3 The water-to-binder ratio r0 = 0.48.

[0187] Two different sets of construction accuracy conditions were set (characterized by measurement standard deviation):

[0188] Working Condition A (High-Precision Construction): , ;

[0189] Condition B (Low-precision construction): , .

[0190] The traditional "hard projection" correction method (comparison method) and the method of this invention are used for correction respectively. The comparison method targets the volume constraint boundary, that is, it forces the requirement. The method of this invention is based on a dynamic safety margin mechanism, with The goal is to achieve the desired results. Table 1 shows a comparison of the correction results and performance indicators of the two methods under different operating conditions.

[0191] As shown in Table 1, the perturbation energy E of the corrected solution obtained by this invention is lower than that of the comparative method under all operating conditions. Under operating condition A, E = 0.0035 for this invention, and E = 0.0021 under operating condition B; while the comparative method has E = 0.0041 under all operating conditions. This indicates that, under the same constraint conditions, the corrected path determined by this invention modifies the initial predicted solution less, thus constituting a better solution in terms of perturbation energy. This result verifies that the corrected method based on the principle of energy minimization in this invention possesses strict mathematical optimality.

[0192] Final aggregate volume of traditional "hard projection" method Under all working conditions, it is forcibly corrected to 0.0000. Its correction target is only to meet the theoretical volume constraint boundary, without considering the measurement uncertainty in the actual construction process. Therefore, it cannot reserve any risk redundancy for the system, and the engineering robustness is seriously insufficient.

[0193] In contrast, the dynamic safety margin mechanism introduced in this invention can adaptively adjust the safety margin according to the construction accuracy. :

[0194] Under high-precision operating condition A, the system automatically calculates... While meeting volume compliance requirements, appropriate design redundancy was retained;

[0195] Under the less precise operating condition B, where uncertainty is greater, the system automatically increases the safety margin to... This allows for more adequate buffer space to be reserved for volume balance.

[0196] This adaptive adjustment mechanism can significantly reduce the risk of volume non-compliance caused by measurement errors and material fluctuations, thereby effectively avoiding the resulting quality fluctuations and rework losses.

[0197] Figure 5 and Figure 6 The traditional "hard projection" method and the method of this invention are compared in bar chart form in terms of perturbation energy E and final aggregate volume. Performance on two indicators. For example... Figure 5 As shown, the perturbation energy of the traditional method is 0.0041, while that of the method of this invention is reduced to 0.0035, indicating that the corrected path of this invention alters the initial prediction less and has better mathematical optimality; as Figure 6 As shown, traditional methods forcibly correct aggregate volume to 0.0000 without considering engineering uncertainties, while the method of this invention, based on dynamic safety margin, adjusts the final aggregate volume to 0.0080, reserving a reasonable safety redundancy for volume compliance.

[0198] In summary, the simulation comparison examples fully demonstrate that, compared with the prior art, the present invention has significant improvements in the optimality, adaptability, and engineering robustness of the solution.

[0199] Example 3

[0200] This invention also provides an electronic device, which includes a memory, a processor, and a computer program or instructions stored in the memory. The processor executes the computer program or instructions to implement the intelligent correction method for concrete mix proportions in this invention.

[0201] Although not shown, the electronic device includes a processor that can perform various appropriate operations and processes based on programs and / or data stored in read-only memory (ROM) or loaded from a storage portion into random access memory (RAM). The processor can be a multi-core processor or may contain multiple processors. In some embodiments, the processor may include a general-purpose main processor and one or more specialized coprocessors, such as a central processing unit, graphics processing unit (GPU), neural network processor (NPU), digital signal processor (DSP), etc. Various programs and data required for device operation are also stored in RAM. The processor, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0202] The processor and memory described above are used together to execute programs / instructions stored in the memory. When the program / instructions are executed by the computer, they can implement the methods, steps, or functions described in the above embodiments.

[0203] Although not shown, embodiments of the present invention also provide a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implements the intelligent correction method for concrete mix proportions in embodiments of the present invention.

[0204] Readable storage media include both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0205] The above description only discloses specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or modifications that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for intelligent correction of concrete mix proportions, characterized in that, The method includes: Obtain engineering input parameters and uncertainty indicators; the engineering input parameters include the density and mass percentage of each component of the cementitious material, and the uncertainty indicators include the measurement standard deviation of unit water consumption and the measurement standard deviation of the equivalent density of the cementitious material; Based on the density and mass ratio of each component of the cementitious material, an interaction coefficient characterizing the nonlinear volume effect between components is introduced to calculate the corrected equivalent density of the cementitious material. The aggregate volume characterization quantity is defined based on the corrected cementitious material equivalent density, and the dynamic safety margin is calculated based on the aggregate volume characterization quantity and the uncertainty index. An adaptive volume physical feasible region is constructed with the constraint that the aggregate volume characterization quantity is not less than the dynamic safety margin. Determine whether the initial predicted mix proportion parameters satisfy the adaptive volume physical feasible region. If not, construct the mix proportion perturbation energy function and solve it under the constraint of satisfying the adaptive volume physical feasible region to obtain the preliminary corrected mix proportion parameters. Calculate the derived parameters based on the initially revised mix proportion parameters, and determine whether the derived parameters are within the engineering specification range. If so, output the initially revised mix proportion parameters as the final revised scheme. Otherwise, add the engineering specification range constraint to the optimization problem, and resolve the mix proportion perturbation energy function under the condition of simultaneously satisfying the adaptive volume physical feasible region and the engineering specification range constraint to obtain the final revised scheme.

2. The intelligent correction method for concrete mix proportions according to claim 1, characterized in that, The standard deviation of the unit water consumption and the standard deviation of the equivalent density of the cementitious material are calculated using statistical process control methods based on online weighing data and / or laboratory test data within a sliding time window, and data points exceeding the range of ±3 times the current period's standard deviation are removed during the calculation process.

3. The intelligent correction method for concrete mix proportions according to claim 1, characterized in that, The corrected equivalent density of the cementitious material is calculated using the following formula: ； in, This represents the corrected equivalent density of the cementitious material; n represents the number of components in the cementitious material. This represents the mass percentage of the i-th component; This represents the density of the i-th component; The interaction coefficient between component i and component j is obtained by fitting multiple sets of apparent density test data of cementitious material slurry with different blending ratios, or by looking up a table in a pre-set material parameter library.

4. The intelligent correction method for concrete mix proportions according to claim 1, characterized in that, The engineering input parameters also include air content, and the aggregate volume characterization quantity is defined according to the following formula: ； in, W represents the aggregate volumetric quality; r represents the water consumption per unit volume; and r represents the water-cement ratio. Indicates the density of water; Indicates air content; This indicates the corrected equivalent density of the cementitious material.

5. The intelligent correction method for concrete mix proportions according to claim 1, characterized in that, The dynamic safety margin is calculated using the following formula: ； in, Indicates dynamic safety margin; Indicates the basic safety margin; Indicates the confidence level coefficient; This represents the standard deviation of a unit of water consumption. The standard deviation of the measurement representing the equivalent density of cementitious materials; This represents a quantity that characterizes aggregate volume. express The absolute value of the partial derivative with respect to unit water consumption W; express The corrected equivalent density of cementitious materials The absolute value of the partial derivative.

6. The intelligent correction method for concrete mix proportions according to claim 1, characterized in that, The energy function E for the mix proportion disturbance is constructed as follows: ； in, , , The initial predicted unit water consumption, water-cement ratio, and cementitious material consumption are represented by , r, and B, which represent the corrected unit water consumption, water-cement ratio, and cementitious material consumption to be determined, and B = W / r. , Indicates the weighting coefficient. ; Alternatively, in a simplified form: ； in, , Indicates the weighting coefficient. .

7. The intelligent correction method for concrete mix proportions according to claim 1, characterized in that, The solution is obtained under the constraint of satisfying the adaptive volume physical feasible region, and the Lagrange multiplier method is used to construct the Lagrange function: ； Where L represents the Lagrange function; E represents the mix proportion perturbation energy function; Represents the Lagrange multipliers; This represents a quantity that characterizes aggregate volume. Indicates dynamic safety margin; The preliminary modified mix proportion parameters were obtained by solving the first-order optimality conditions.

8. The intelligent correction method for concrete mix proportions according to any one of claims 1 to 7, characterized in that, When it is necessary to simultaneously satisfy the adaptive volume physical feasible region and the engineering specification interval constraints, the modified problem is constructed as the following multi-constraint optimization problem: ； ； Where E represents the energy function of the mix proportion disturbance; W and r represent the corrected unit water consumption and water-cement ratio to be determined. This represents a quantity that characterizes aggregate volume. Indicates dynamic safety margin; Indicates derived parameters; and This indicates the lower and upper limits allowed for the project corresponding to the derived parameters; The sequential quadratic programming algorithm is used to solve the multi-constraint optimization problem, and the final modified scheme is obtained.

9. An electronic device comprising a memory, a processor, and a computer program or instructions stored in the memory, characterized in that, The processor executes the computer program or instructions to implement the intelligent correction method for concrete mix proportions as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by the processor, they implement the intelligent correction method for concrete mix proportions as described in any one of claims 1 to 8.

Images