A concrete reference mix proportion intelligent generation and verification system

Abstract

This invention discloses an intelligent system for generating and verifying benchmark concrete mix proportions, belonging to the field of concrete technology. It includes a parameter input and management module, an AI benchmark mix proportion calculation module, a trial mixing verification and data feedback module, and a mix proportion optimization and production application module. The parameter input and management module includes a design target parameter input module, a raw material parameter input module, and a parameter verification and storage module. This intelligent system for generating and verifying benchmark concrete mix proportions allows technicians to input concrete design targets and raw material parameters, then access a historical mix proportion database. A neural network algorithm is used to match similar projects and calculate the benchmark usage of each raw material. After laboratory trial mixing tests to assess workability and mechanical properties and receiving feedback data, the AI ​​automatically adjusts the parameters to generate the final mix proportion. This enables one-click production by linking the mixing plant, overcoming the pain points of traditional trial mixing methods that rely on experience and have long cycles, significantly improving the accuracy of mix proportion adaptation and reducing the number of trials.

Description

Technical Field

[0001] This invention relates to the field of concrete technology, specifically to an intelligent system for generating and verifying concrete benchmark mix proportions. Background Technology

[0002] The quality of the concrete reference mix design directly determines the final performance of the concrete. A reasonable mix design ensures that the concrete meets design strength, durability, and workability requirements while achieving good economic efficiency. Conversely, an inappropriate mix design may lead to quality problems such as insufficient strength, cracking, poor durability, and poor workability, which not only affect project safety but also increase later maintenance costs. To reduce reliance on manual experience and quickly match similar project cases during the reference mix design process, significantly shortening the generation time and reducing the number of trial mixes, an intelligent generation system is needed, such as application number 201620619302.1, application date 2016- Chinese invention patent application No. 06-20 discloses a concrete mix proportion optimization system. It calculates the output of each layer through an excitation function, calculating layer by layer until the output layer, thereby obtaining the output result. It then further converts the output vector into the amount of various high-performance concrete components required by the user, aiming to optimize all performance and costs, and optimizes the mix proportion of high-performance concrete. Another Chinese invention patent application with application number 202511035514.5 and application date of 2025-07-25 discloses a concrete benchmark mix proportion optimization method and system, which can realize the formulation and dynamic optimization of concrete benchmark mix proportion schemes in complex scenarios.

[0003] Due to the uncertainties at the construction site, the actual mix proportions need to be adjusted according to the actual construction location and also need to be adjusted in real time based on past project experience. If efficient real-time adjustments cannot be made, it will lead to substandard concrete compatibility and performance, increase costs and safety hazards, and lose the core value of intelligent system closed-loop optimization. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent generation and verification system for concrete reference mix proportions, in order to solve the problems mentioned in the background art, such as the inability to perform efficient real-time adjustments, which leads to substandard concrete construction adaptability and performance, increased costs and safety hazards, and loss of the core value of intelligent closed-loop optimization.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A smart system for generating and verifying concrete benchmark mix proportions includes a parameter input and management module, an AI benchmark mix proportion calculation module, a trial mixing verification and data feedback module, and a mix proportion optimization and production application module. The parameter input and management module includes a design target parameter input module, a raw material parameter input module, and a parameter verification and storage module. It accurately collects, verifies, and stores concrete design targets and raw material parameters, providing reliable data support for AI calculations. The AI ​​benchmark mix proportion calculation module relies on algorithm models and historical data to achieve rapid and accurate generation of benchmark mix proportions, fundamentally solving the problem of traditional adaptation relying on experience. The system comprises a database management module, a neural network algorithm matching module, and a baseline dosage calculation module. The trial mixing verification and data feedback module includes a trial mixing scheme generation module, a performance testing module, and a data feedback module. Through laboratory trial mixing and testing, it verifies the feasibility of the baseline mix proportion, providing real data support for subsequent optimization and forming a preliminary closed loop of calculation-verification. The mix proportion optimization and production application module includes an AI automatic optimization module, a mix proportion review and determination module, and a mixing plant control system association module. Based on trial mixing feedback data, it completes iterative optimization of the mix proportion, ultimately outputting a formal mix proportion that can be directly used for production, and achieving seamless integration with the mixing plant.

[0007] Preferably, the design target parameter input module supports technicians in inputting core concrete design requirements, including grade, frost resistance grade, impermeability grade, slump, and spread, and provides both standardized options and custom input modes.

[0008] The raw material parameter input module is responsible for collecting key parameters of raw materials, including sand, stone, cement, water, and additives, covering sand fineness modulus, mud content, crushed stone particle size and gradation, cement type and strength grade, and additive type and manufacturer parameters.

[0009] Preferably, the parameter verification and storage module performs logical verification on the input parameters to avoid invalid data input; at the same time, it stores compliant parameters in the system database to support historical data tracing and retrieval.

[0010] Preferably, the historical mix proportion database management module integrates historical mix proportion data, raw material parameters, construction environment, and performance test results of various projects, establishes a standardized data index, and supports rapid retrieval by project type and parameter range;

[0011] The neural network algorithm matching module calls a preset neural network algorithm model, uses the currently entered design target parameters and raw material parameters as search conditions, matches similar engineering cases in the database, and analyzes the correlation of key parameters.

[0012] Preferably, the benchmark dosage calculation module automatically calculates the benchmark dosage of cement, sand, stone, water, and additives based on the algorithm matching results, generates a preliminary concrete benchmark mix proportion, and clarifies the proportion and dosage range of each component.

[0013] Preferably, the test mixing scheme generation module automatically generates a laboratory test mixing operation scheme based on the benchmark mix ratio, specifying the operation requirements for mixing time, material feeding sequence, and test block preparation standards, to ensure the standardization of the test mixing process;

[0014] The performance testing module performs workability and mechanical property tests on the trial-mixed concrete, records test data, test time, and test environment, and supports real-time input of test results.

[0015] Preferably, when the data feedback module is working, it automatically uploads the test data to the system, compares and analyzes it with the design target parameters, identifies items that do not meet the standards, and provides a clear direction for AI optimization.

[0016] Preferably, when the AI ​​automatic optimization module is working, it automatically adjusts the mixing ratio parameters through an algorithm model to generate an optimized formal mixing ratio for the problems found in the trial mixing test.

[0017] Preferably, the mix proportion review and confirmation module supports technicians to manually review the AI-optimized official mix proportion, which can be fine-tuned according to the special needs of the project, and the final mix proportion is locked after the review is passed.

[0018] Preferably, when the associated module of the mixing plant control system is working, it will automatically synchronize the final determined formal mix proportion data to the mixing plant control system, realize the one-click distribution of mix proportion parameters, ensure accurate control of the amount of each raw material used in the production process, and achieve a closed loop of intelligent generation-verification optimization-one-click production.

[0019] Compared with the prior art, the beneficial effects of the present invention are:

[0020] This intelligent generation and verification system for concrete benchmark mix proportions employs a novel structural design. Technicians input concrete design targets and raw material parameters, then access a historical mix proportion database. A neural network algorithm matches similar projects to calculate the benchmark usage of each raw material. After laboratory trial mixing to test workability and mechanical properties and receiving feedback data, AI automatically adjusts parameters to generate the final mix proportion. This is then linked to the mixing plant for one-click production. The system overcomes the pain points of traditional trial mixing, which relies on experience and has long cycles, through a closed-loop model of data matching, trial mixing feedback, and intelligent optimization. This significantly improves the accuracy of mix proportion matching, reduces the number of trials and testing costs, standardizes the production process, ensures concrete performance meets standards, avoids potential quality issues, promotes the intelligent upgrading of concrete preparation, and improves engineering construction efficiency and quality control. Specific details are as follows:

[0021] (1) The intelligent generation and verification system for concrete benchmark mix proportions has a parameter input and management module as the core data entry point of the system. It ensures the comprehensiveness and standardization of data collection by standardizing the input of design objectives and key parameters of raw materials, avoiding omissions or deviations in manual input. Relying on the parameter verification function, it performs logical verification and range verification on the input data, effectively filtering invalid data and providing accurate and reliable data support for subsequent AI calculations, thus ensuring the accuracy of benchmark mix proportion generation from the source. At the same time, the module classifies and stores compliant parameters and supports traceability, which not only facilitates the retrieval of historical data and matching with similar projects, but also accumulates high-quality data resources for system algorithm iteration, reduces trial mix rework caused by data problems, indirectly shortens the project cycle, reduces test costs, and lays a solid data foundation for the closed-loop optimization and intelligent operation of the entire system.

[0022] (2) This intelligent generation and verification system for concrete benchmark mix proportions, the AI ​​benchmark mix proportion calculation module relies on a massive historical mix proportion database and neural network algorithm, can quickly match similar engineering cases without relying on human experience, greatly shorten the benchmark mix proportion generation cycle, and efficiently respond to the personalized parameter requirements of different projects; through the algorithm to deeply explore the intrinsic relationship between design goals, raw material parameters and mix proportions, the generated benchmark dosage is more scientific and adaptable, effectively improving the initial accuracy of mix proportions and reducing the blindness of subsequent trial mixing adjustments; at the same time, the standardized data retrieval and calculation process avoids the subjective bias of traditional manual design, lays a solid foundation for subsequent trial mixing verification and optimization, indirectly reduces trial mixing costs and construction period losses, and highlights the core advantages of the system's intelligence and precision.

[0023] (3) This intelligent generation and verification system for concrete benchmark mix proportions, through the generation of standardized trial mixing schemes by the trial mixing verification and data feedback module, ensures the uniformity of laboratory trial mixing operation specifications and avoids the impact of human operation differences on verification results; accurately detects the workability and mechanical properties of concrete, intuitively verifies the feasibility of benchmark mix proportions, accurately locates non-compliant items, and provides clear and real data basis for AI optimization; relies on the data feedback mechanism to realize the effective connection between calculation and verification, constructs the key link of system closed-loop optimization, and reduces the blindness of subsequent mix proportion adjustments; at the same time, the data retention and traceability function not only facilitates the traceability of engineering quality, but also supplements the historical database with high-quality cases, helps the algorithm model to continuously iterate, and further improves the adaptation accuracy and reliability of the entire system.

[0024] (4) This intelligent generation and verification system for concrete benchmark mix proportions, with the mix proportion optimization and production application module using AI automatic optimization function, accurately adjusts parameters for trial mixing problems to ensure that the formal mix proportions meet actual needs and guarantee that concrete performance meets standards; supports manual review and fine-tuning, taking into account both intelligent efficiency and special engineering needs, and improving mix proportion flexibility; through seamless connection with the mixing plant control system, it realizes one-click distribution of mix proportions and precise production, standardizes the production process and reduces human error; it completes the entire process of calculation-verification-optimization-production in a closed loop, giving full play to the advantages of the intelligent system, greatly shortening the cycle from mix proportion determination to production implementation, reducing rework costs and quality risks, and comprehensively improving the efficiency and quality control level of engineering construction. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the system workflow of the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0027] Example 1: Please refer to Figure 1 The present invention provides the following technical solution: a concrete benchmark mix proportion intelligent generation and verification system, including a parameter input and management module, an AI benchmark mix proportion calculation module, a trial mixing verification and data feedback module, and a mix proportion optimization and production application module.

[0028] The parameter input and management module includes a design target parameter input module, a raw material parameter input module, and a parameter verification and storage module. It accurately collects, verifies, and stores concrete design targets and raw material parameters, providing reliable data support for AI calculations.

[0029] The design target parameter input module supports technicians in inputting core concrete design requirements, including grade, frost resistance grade, impermeability grade, slump, and spread, and provides both standardized options and custom input modes.

[0030] When the aforementioned design target parameter input module is in operation, a standardized parameter input framework is first established. Based on the core design requirements of concrete, an input system covering key performance indicators is developed. This system explicitly includes core indicators such as concrete grade (common grades like C15-C100), frost resistance grade (F50-F300), impermeability grade (P4-P30), slump (10mm-240mm), and spread. It also provides both standardized options and custom input modes. For routine engineering needs, technicians can quickly select from preset options, improving input efficiency. For special projects or personalized design targets under special environments (such as those under extreme temperature and humidity conditions),... (Specific performance requirements), supporting manual input of specific parameter values, balancing versatility and flexibility; during parameter input, the module has a built-in logic verification rule engine to perform compliance and rationality checks on the input data in real time; on the one hand, it verifies the range of parameter values ​​(such as the slump value must not exceed the commonly used reasonable range in engineering) to avoid invalid data due to misinput; on the other hand, it verifies the matching logic between parameters (such as the reasonable range of slump corresponding to high-grade concrete, the compatibility relationship between frost resistance and impermeability grade and grade), if there are contradictory or abnormal data (such as low-grade concrete paired with ultra-high impermeability grade), the system will immediately pop up a prompt to guide technicians to check and correct, avoiding data deviation from the source;

[0031] Once the design target parameters are entered and verified, they will be categorized and stored in the system database according to a preset data structure. A unique index will be established to link them to subsequent raw material parameters, mix proportion calculation results, and other data. In addition, the module supports the retrieval and tracing of historical design target parameters. Technicians can quickly query past data by project name, parameter type, and other conditions, providing a reference for the entry of parameters for similar projects, further improving work efficiency, and laying a solid foundation for the "data-driven" core logic of the entire system. As the data input core of the intelligent generation and verification system for concrete benchmark mix proportions, the design target parameter entry module ensures that comprehensive and reliable basic data is provided for subsequent AI calculations.

[0032] The raw material parameter input module is responsible for collecting key parameters of raw materials, including sand, stone, cement, water, and additives, covering sand fineness modulus, mud content, crushed stone particle size and gradation, cement type and strength grade, and additive type and manufacturer parameters.

[0033] The aforementioned raw material parameter input module, serving as the core data support for the intelligent generation and verification system for concrete benchmark mix proportions, operates on the core logic of comprehensive data collection, standardized verification, and categorized storage. It accurately captures key information on raw materials such as sand, stone, cement, water, and additives, providing high-quality data input for subsequent AI algorithm matching and benchmark mix proportion calculation. Its working principle is as follows: The raw material parameter input module first constructs a standardized input system covering all raw materials. For the core raw materials required for concrete preparation, it clarifies the key parameter collection items for each category: Sand requires input of fineness modulus, mud content, moisture content, particle size distribution, etc.; Crushed stone requires recording of particle size range, gradation type, crushing value, and needle-like / flaky content, etc.; Cement requires labeling of model (e.g., P.O42.5, PSA32.5), strength grade, and stability, etc.; Water requires description of water quality type, pH value, and impurity content, etc.; Additives require specifying type (water-reducing agent, retarder, etc.), manufacturer parameters, and effective ingredient content, etc. It also offers two modes: "option selection + manual entry". Regular parameters can be selected from industry standard options, while special raw material parameters can be customized, balancing entry efficiency and scenario adaptability.

[0034] During the data entry process, the module incorporates multi-dimensional validation rules to ensure data validity in real time. On one hand, it validates the value range, setting reasonable intervals for each parameter according to industry standards (e.g., the standard range for sand fineness modulus is 1.6-3.7). If the value exceeds this range, a pop-up window will display a warning and the standard value. On the other hand, it performs logical correlation validation, such as the compatibility between crushed stone particle size and sand fineness modulus, and the compatibility between additive types and cement models, to avoid contradictory data entry. It also supports uploading raw material testing reports as attachments; the system automatically extracts key data and compares it with manually entered values, reducing human input errors.

[0035] Verified raw material parameters are categorized and stored in the system database according to the hierarchical structure of raw material type, parameter category, and entry time. A unique data identifier is generated and associated with the design target parameters. The raw material parameter entry module supports the query, export, and traceability of historical raw material parameters. It can be quickly retrieved by conditions such as project name, raw material category, and entry date. This provides complete data support for AI algorithms to match similar projects and retains key evidence for subsequent project quality traceability, thus laying a solid data foundation for the intelligent operation of the entire system.

[0036] The parameter validation and storage module performs logical validation on the input parameters to avoid invalid data input; at the same time, it categorizes and stores compliant parameters in the system database, supporting historical data tracing and retrieval.

[0037] The aforementioned parameter verification and storage module, serving as the core of the intelligent generation and verification system for concrete benchmark mix proportions, focuses on accurate verification and noise reduction, along with standardized storage and traceability. This ensures the authenticity and validity of the input design target parameters and raw material parameters, laying a solid data foundation for subsequent AI calculations, trial mixing verification, and production applications. Its working principle is as follows: The parameter verification and storage module handles all data from the parameter input stage, first initiating a multi-dimensional verification mechanism. On one hand, it performs value range verification, setting reasonable ranges for each parameter based on concrete industry standards and engineering practice standards. For example, the sand fineness modulus is limited to 1.6-3.7, and the slump range is set to 10mm-240mm. If the input data exceeds the threshold, the system immediately pops up an error message and marks the standard range, guiding technical personnel to verify and correct it. On the other hand, it conducts logical correlation verification, analyzing the compatibility between design targets and raw material parameters, such as the matching relationship between high-grade concrete and frost resistance and impermeability grade, and the rationality of the combination of crushed stone particle size and sand fineness modulus, avoiding situations such as "low-grade concrete paired with ultra-high impermeability grade" or "incompatible concrete." The system addresses contradictory data such as "adapting to combinations of additives and cement models"; it also supports linking raw material testing reports, design documents, and other attachments, automatically extracting key data for comparison with entered values ​​to further reduce human input errors; compliant data that passes verification is standardized and categorized by parameter type, project affiliation, and entry time in the parameter verification and storage module, with each design target parameter and raw material parameter linked to generate a unique data identifier. Metadata such as the person who entered the data, verification results, and attachment number are recorded simultaneously to ensure data traceability. The storage architecture is compatible with both structured data (such as numerical parameters) and unstructured data (such as test report attachments), and uses secure encryption to store data in the system database to prevent data loss or tampering. Furthermore, the parameter verification and storage module supports rapid multi-condition retrieval; technicians can query historical data by project name, parameter type, entry date, etc., providing accurate data support for AI algorithms to call historical mix proportion databases to match similar projects, and providing complete evidence for project quality traceability and parameter review, ensuring the continuity and reliability of the entire system's data flow.

[0038] The AI ​​baseline mix ratio calculation module relies on algorithm models and historical data to achieve rapid and accurate generation of baseline mix ratios. It fundamentally solves the problem of traditional trial mixing relying on experience. It consists of a historical mix ratio database management module, a neural network algorithm matching module, and a baseline dosage calculation module.

[0039] The historical mix proportion database management module integrates historical mix proportion data, raw material parameters, construction environment, and performance test results of various projects, establishes a standardized data index, and supports quick retrieval by project type and parameter range;

[0040] The aforementioned historical mix proportion database management module first constructs a comprehensive data source collection system, integrating historical mix proportion data from various projects such as building construction, municipal engineering, and bridges. This system covers core information dimensions: concrete design target parameters (grade, frost resistance and impermeability level, slump, etc.), raw material parameters (sand fineness modulus, crushed stone particle size, cement type, etc.), complete mix proportion data (specific dosage and proportion of cement, sand, stone, water, and additives), construction environmental conditions (temperature, humidity, altitude, etc.), and performance test results of trial mixing and actual projects (workability, 7-day, 28-day strength, etc.), ensuring that the data dimensions are complete and closely aligned with the actual project conditions.

[0041] The historical mix proportion database management module performs standardized preprocessing on the collected raw data, including data cleaning (removing outliers and duplicate data), format standardization (regulating parameter units and expression methods according to industry standards), and tag classification (adding keyword tags according to project type, region, construction season, etc.), forming structured data entries. These are then stored in the database according to a hierarchical architecture of project type-parameter dimension-data time, and multi-dimensional indexes are established (such as design target parameter index, raw material parameter combination index, and project scenario index) to ensure efficient data retrieval. During the data retrieval phase, the historical mix proportion database management module responds to AI-based... The retrieval request from the quasi-mix ratio calculation module, based on the currently entered design goals and raw material parameters, quickly filters out similar engineering data entries with high parameter matching degree through a preset indexing mechanism. After being sorted by matching degree, the data is fed back to the neural network algorithm matching module, providing a reference for the baseline usage calculation. At the same time, the historical mix ratio database management module supports dynamic iterative updates, incorporating parameters, mix ratios, performance test results, and other data from new projects into the database in real time, continuously enriching the data sample size, optimizing data distribution, and helping the neural network algorithm to continuously improve the matching accuracy and the scientific nature of the baseline mix ratio generation, providing long-term data support for system closed-loop optimization.

[0042] The historical mix proportion database management module serves as the core data support unit of the intelligent generation and verification system for concrete benchmark mix proportions, providing high-quality historical data resources for AI algorithm matching.

[0043] The neural network algorithm matching module calls the preset neural network algorithm model, uses the currently entered design target parameters and raw material parameters as search conditions, matches similar engineering cases in the database, and analyzes the correlation of key parameters;

[0044] The aforementioned neural network algorithm matching module, as the core unit of intelligent generation, relies on historical data and algorithm models to accurately locate suitable cases, providing a scientific basis for benchmark mix proportion calculation. When the neural network algorithm matching module is working, it first receives the current project's core data transmitted from the parameter input and management module, including concrete design targets (grade, frost resistance and impermeability level, slump, etc.) and raw material parameters (sand fineness modulus, crushed stone particle size, cement type, etc.). This data undergoes standardized preprocessing, uniformly converting parameters of different dimensions and units into algorithm-recognizable feature vectors, eliminating the impact of data heterogeneity. Subsequently, the neural network algorithm matching module calls a well-trained neural network algorithm... The network algorithm model (such as a BP neural network) has been trained using massive amounts of engineering data in a historical mix proportion database. It has learned the nonlinear mapping relationship between design objectives, raw material parameters, and mix proportions through deep learning. The model first extracts features from the feature vector of the current project, identifying key influencing factors (such as the correlation weight between grade and cement dosage, and the matching pattern between slump and water-reducing agent dosage). Then, using these features as the core of the retrieval, it traverses the structured data in the historical mix proportion database. During the matching process, the algorithm calculates the similarity between the feature vector of the current project and the feature vector of historical projects (such as Euclidean distance and cosine similarity), filtering out similar project cases with high parameter matching. Simultaneously, it combines auxiliary labels such as project type and construction environment to optimize the matching results, excluding cases with significant scene differences to ensure the targeted nature of the matching.

[0045] Finally, the neural network algorithm matching module sorts the selected similar engineering cases by matching degree, extracts the core data of the mix proportion (the proportion of each raw material, adjustment logic, etc.), and feeds it back to the benchmark usage calculation module. This provides a direct reference for the generation of the benchmark mix proportion of the current project. The whole process does not require manual intervention. The algorithm autonomously completes feature learning and similarity matching, which not only ensures matching efficiency but also improves the initial adaptation accuracy of the benchmark mix proportion, effectively breaking the limitations of traditional trial mixing that relies on experience.

[0046] The benchmark dosage calculation module automatically calculates the benchmark dosage of cement, sand, stone, water, and additives based on the algorithm matching results, generates a preliminary concrete benchmark mix proportion, and clarifies the proportion and dosage range of each component.

[0047] The aforementioned benchmark usage calculation module, as the core computing unit, uses data support, algorithm empowerment, and quantitative output as its core logic to accurately calculate the benchmark usage of each raw material. Its working principle is as follows: The benchmark usage calculation module first receives two core data inputs: one is the current engineering design target parameters (such as grade, frost resistance and impermeability level, slump, etc.) and raw material parameters (such as sand fineness modulus, crushed stone particle size, cement type, etc.) transmitted from the parameter input and management module; the other is similar engineering case data fed back by the neural network algorithm matching module, including the usage of each raw material, performance test results, and suitable scenario information in historical mix proportions after matching degree ranking.

[0048] Subsequently, the benchmark usage calculation module, based on the preset core logic of concrete mix proportions and combined with the parameter correlation patterns mined by the neural network algorithm, constructs a multi-dimensional calculation model. First, it takes the current engineering design goals as the core constraints, such as determining the minimum usage benchmark of cementitious materials (cement) according to the concrete grade, and initially locking the water-cement ratio range according to the slump requirements. Then, it adjusts the calculation logic in combination with the characteristics of raw material parameters, such as optimizing the sand ratio when the sand fineness modulus is too high, and adjusting the total aggregate usage when the crushed stone particle size is large, to ensure that the calculation fits the actual characteristics of the raw materials.

[0049] Meanwhile, the baseline usage calculation module references mix proportion data from similar engineering cases, using weighted calculations to balance historical experience with current project requirements—cases with higher matching degrees receive greater weight, with a focus on key information such as raw material usage ratios and adjustment logic to avoid deviating from actual project needs. For different components such as cement, sand, stone, water, and additives, corresponding calculation models are used: the amount of cementitious materials is determined by combining strength grade and water-cement ratio; the amount of aggregate is calculated using volumetric or mass methods; and the amount of additives is calculated based on slump requirements and material compatibility.

[0050] Finally, the benchmark dosage calculation module integrates all calculation results and outputs clear benchmark dosages and proportional relationships for each raw material, forming a complete benchmark mix proportion for concrete. This mix proportion not only meets the current engineering design goals and material characteristics but also incorporates successful experiences from similar historical projects, providing a scientific and feasible basic template for subsequent laboratory trial mixing and verification, ensuring the accuracy and suitability of the benchmark dosages.

[0051] The trial mixing verification and data feedback module includes a trial mixing scheme generation module, a performance testing module, and a data feedback module. Through laboratory trial mixing and testing, the feasibility of the baseline mix ratio is verified, providing real data support for subsequent optimization and forming a preliminary closed loop of calculation and verification.

[0052] The trial mixing scheme generation module automatically generates a laboratory trial mixing operation plan based on the benchmark mix ratio, which clarifies the operation requirements for mixing time, material feeding sequence, and test block preparation standards, ensuring the standardization of the trial mixing process;

[0053] When the above-mentioned trial mixing scheme generation module is working, it first receives the complete concrete benchmark mix proportion output by the benchmark dosage calculation module, including the specific dosage, proportion relationship and related parameter description of each raw material such as cement, sand, stone, water and additives. At the same time, it simultaneously retrieves the design target parameters of the current project (such as slump, strength grade, etc.) to clarify the core testing direction of the trial mixing verification.

[0054] Based on the benchmark mix proportion and design requirements, the trial mixing scheme generation module relies on built-in industry test specifications (such as the concrete mix proportion design code) to construct a standardized trial mixing process framework. First, determine the core operating parameters: calculate the optimal mixing volume for a single trial mix based on the characteristics of the raw materials and the mix proportion dosage (usually 50L or 100L to ensure uniform mixing), and match the corresponding mixer model and specifications; set the mixing time (generally 90-120 seconds, which can be dynamically adjusted according to the aggregate particle size) and the material feeding sequence (following the scientific sequence of "aggregate-cement-additive-water" to avoid cement clumping);

[0055] Subsequently, the trial mixing scheme generation module refined the operational requirements for each stage of the trial mixing process: it clarified the standards for the accuracy of raw material weighing (e.g., the weighing error of cement and additives should not exceed ±1%, and the weighing error of sand and stone should not exceed ±2%), and marked the calibration requirements for weighing tools; it specified the specifications for test block production (e.g., test block size, vibration method, and curing conditions after molding) to ensure that the performance of the test blocks can truly reflect the actual quality of the concrete; it clarified the time nodes for workability testing (slump and spread should be measured within 5 minutes after mixing) and the operation methods, as well as the curing period (7 days, 28 days) and testing standards for mechanical performance test blocks;

[0056] Finally, the trial mixing scheme generation module integrates all parameters and requirements to generate a structured trial mixing scheme document. This document includes the project name, details of the baseline mix proportion, a list of trial mixing equipment, detailed operating procedures, testing items and standards, and data recording tables. It supports direct export and printing for laboratory personnel. The scheme ensures the standardization and uniformity of the trial mixing process, avoiding the impact of human error on verification results, and provides a clear basis for subsequent performance testing and data feedback, ensuring the scientific rigor and effectiveness of the trial mixing verification.

[0057] The performance testing module performs workability and mechanical property tests on the trial-mixed concrete, records test data, test time, and test environment, and supports real-time input of test results.

[0058] When the above-mentioned performance testing module is working, it first receives the testing standards and current engineering design target parameters output by the trial mixing scheme generation module, and clarifies the core testing indicators: workability focuses on slump (standard range of 10mm-240mm) and spread, mechanical properties focus on testing 7d and 28d compressive strength, and at the same time, it associates with special testing items corresponding to design requirements such as freeze-thaw resistance and impermeability. Based on concrete industry test specifications, it determines the selection of testing equipment (such as slump cone, pressure testing machine, impermeability meter, etc.) and equipment calibration requirements to ensure that the accuracy of testing tools meets the standards.

[0059] After the trial mixing is completed, the performance testing module starts the workability test according to the preset procedure: strictly follow the time requirement of "testing within 5 minutes after mixing", operate the slump cone loading, compaction and lifting process according to the standard, and accurately read the slump value; by measuring the maximum diameter and vertical diameter of the concrete after slump expansion, take the average value as the expansion data, and simultaneously record the ambient temperature and humidity during the test to avoid environmental factors from interfering with the objectivity of the data.

[0060] In the mechanical performance testing phase, the performance testing module, according to the test block production standards specified in the trial mixing plan, performs standard curing (temperature 20±2℃, humidity ≥95%) on the molded concrete test blocks. Compressive strength tests are conducted using a pressure testing machine at 7 days and 28 days of curing, recording the maximum pressure value at test block failure. The compressive strength is calculated based on the stress area of ​​the test block. If the design objectives include freeze-thaw resistance and impermeability requirements, specialized testing is initiated simultaneously, completing freeze-thaw cycle and osmotic pressure tests according to the corresponding specifications.

[0061] After testing, the performance testing module automatically compares the measured data with the design target parameters to determine whether the standards are met (e.g., whether the slump is within the design range, whether the strength meets the grade requirements), and generates a test report containing test data, compliance status, and deviation analysis. Simultaneously, the raw data and judgment results are fed back to the AI ​​optimization module in real time via the system interface, providing accurate data support for adjusting mix proportion parameters (e.g., adding water-reducing agent when slump is insufficient, adjusting cement dosage when strength is insufficient), ensuring that the subsequently generated formal mix proportion fully meets the engineering performance requirements.

[0062] When the data feedback module is working, it automatically uploads the test data to the system, compares and analyzes it with the design target parameters, identifies the items that do not meet the standards, and provides a clear direction for AI optimization.

[0063] When the aforementioned data feedback module is working, it first clarifies the scope and source of data collection. The core of the module receives all the test data output by the performance testing module, including concrete workability (measured values ​​of slump and spread, and ambient temperature and humidity), mechanical properties (7-day and 28-day compressive strength data, and records of specimen failure characteristics), as well as specific test results such as frost resistance and impermeability (if any). At the same time, it simultaneously collects key practical data during the trial mixing process, such as actual raw material weighing deviations, mixing time adjustments, and changes in the order of material addition, to ensure the completeness of the feedback data.

[0064] The data feedback module standardizes the collected raw data by unifying the data format and units (e.g., strength unit is standardized to MPa, slump unit is standardized to mm), removing abnormal data (e.g., significant deviations due to operational errors), and organizing the data in a structured manner according to the structure of "test item - data result - compliance status - deviation value". For indicators that do not meet the standards, the module automatically calculates the deviation between the measured value and the design target (e.g., slump is 20mm lower than the design value) and marks the deviation type (e.g., insufficient performance, parameter exceeding the standard), providing a clear problem orientation for AI optimization.

[0065] The processed standardized data is accurately transmitted through the system's built-in interface: on the one hand, the complete test data and deviation analysis results are pushed to the AI ​​optimization module in real time, clearly indicating the core directions that need to be adjusted (such as optimizing the amount of cementitious materials if the strength is insufficient, and adjusting the water-cement ratio if the slump exceeds the standard); on the other hand, it is simultaneously fed back to the historical mix proportion database management module as a supplement to new engineering case data, accumulating experience for matching similar projects in the future.

[0066] In addition, the data feedback module establishes a full-process data association and traceability mechanism, binding the feedback data with the current project's design target parameters, raw material parameters, benchmark mix ratio, and trial mixing scheme, generating a unique data link identifier, and recording data flow nodes and timestamps; it also supports data visualization queries, allowing technicians to view feedback results, reasons for deviations, and subsequent optimization directions in real time, ensuring the targeting and accuracy of AI optimization, providing complete data support for project quality traceability, and strengthening the core link for the closed-loop operation of the system.

[0067] The mix proportion optimization and production application module includes an AI automatic optimization module, a mix proportion review and confirmation module, and a mixing plant control system association module. Based on the trial mixing feedback data, it completes the iterative optimization of the mix proportion and finally outputs a formal mix proportion that can be directly used for production, and achieves seamless integration with the mixing plant.

[0068] When the AI ​​automatic optimization module is working, it automatically adjusts the mixing ratio parameters through the algorithm model to generate an optimized formal mixing ratio in response to the problems found in the trial mixing test.

[0069] The aforementioned AI automatic optimization module, as the core of the system's closed loop, first receives the trial mixing test data (measured values ​​of workability and mechanical properties) and design target parameters transmitted by the data feedback module. It then compares and analyzes deviations (such as insufficient slump or substandard strength) using algorithms. Relying on the parameter correlation patterns mined by the neural network algorithm and combining optimization experience from similar historical projects, it adjusts the mix proportion parameters (such as water-reducing agent dosage, cement dosage, and water-cement ratio) in a targeted manner. The optimization process strictly follows the core logic of concrete mix proportions and industry standards, ensuring that the adjusted parameters are adapted to the characteristics of raw materials and construction needs. Ultimately, it generates accurate and compliant formal mix proportions, providing a reliable basis for one-click production at the mixing plant and realizing intelligent iterative optimization of the mix proportions.

[0070] The mix proportion review and confirmation module allows technicians to manually review the AI-optimized official mix proportions. Fine-tuning can be done according to the specific needs of the project. Once the review is approved, the final mix proportion is locked.

[0071] When the above-mentioned mix proportion review and confirmation module is working, it first receives the formal mix proportion output by the AI ​​automatic optimization module, and simultaneously retrieves the design target parameters, raw material parameters, trial mixing test data and optimization adjustment records to build a complete data verification chain. Through the built-in algorithm, it automatically verifies whether the proportion of each raw material in the mix proportion meets the industry standard, whether the optimization adjustment logic matches the trial mixing deviation (such as whether the adjustment range of water-reducing agent is reasonable when the slump is insufficient), and verifies the compatibility of the mix proportion with the construction location environment and process requirements.

[0072] Subsequently, the mix proportion review and confirmation module generates a structured review report, clearly presenting the mix proportion details, optimization basis, compliance judgment results, and key data relationships, and pushes it to the technical lead's end. Manual review and correction are supported, allowing technical personnel to fine-tune parameters based on specific engineering needs (such as adaptation to extreme environments). After the review is approved, a confirmation instruction must be entered.

[0073] Finally, the mix proportion review and confirmation module encrypts and stores the confirmed official mix proportion and synchronizes it to the mixing plant control system, generating a unique production identifier. At the same time, it retains complete review records (including reviewer, confirmation time, and adjustment traces), which not only ensures the reliability of the mix proportion but also provides a basis for quality traceability, thus building a solid last line of quality control before production.

[0074] When the associated module of the mixing plant control system is working, it will automatically synchronize the finalized formal mix proportion data to the mixing plant control system, realize the one-click distribution of mix proportion parameters, ensure accurate control of the amount of each raw material used in the production process, and achieve a closed loop of intelligent generation-verification optimization-one-click production.

[0075] When the above-mentioned mixing plant control system associated module is working, it first receives the final formal mix proportion output by the mix proportion review and confirmation module, which includes the precise dosage, proportion relationship and production adaptation requirements of each raw material such as cement, sand, stone, water and additives (such as mixing time, feeding sequence, etc.). At the same time, it synchronizes the core information of the current project (project name, production batch, construction location, etc.) to ensure that the production data corresponds accurately with the project requirements.

[0076] Subsequently, the data standardization conversion process is initiated by the associated module of the batching plant control system. Due to differences in parameter formats and communication protocols among the batching plant control systems, the associated module utilizes built-in multi-protocol adapter interfaces (such as MODBUS and PROFINET) to convert the structured data of the formal mix proportions (such as raw material weight values ​​and proportional parameters) into control commands and data formats recognizable by the batching plant, thus eliminating data compatibility issues between systems. During the conversion process, data integrity and accuracy are rigorously verified to ensure accurate transmission of raw material usage and process parameters.

[0077] After the data conversion is completed, the associated module of the mixing plant control system pushes the control commands and mix proportion data to the mixing plant control system in real time through an encrypted communication link, realizing one-click parameter issuance. At the same time, a two-way data interaction channel is established: on the one hand, clear production instructions are issued to the mixing plant, including the raw material weighing accuracy requirements (such as cement error ±1%, sand and gravel error ±2%), mixing time settings, and feeding sequence specifications; on the other hand, production feedback data from the mixing plant is received in real time, such as the actual weighing value of each raw material, mixing time, equipment operating status, and production completion quantity, dynamically monitoring whether the production process is consistent with the mix proportion requirements.

[0078] If data deviations occur during production (such as raw material weighing exceeding the allowable range) or equipment malfunctions, the associated module of the mixing plant control system will immediately trigger an early warning mechanism, push a prompt message to the technical personnel, and suspend the mixing plant production process. Production will resume after the problem is investigated and corrected. After production is completed, the associated module of the mixing plant control system will associate and store complete production data (mixing ratio parameters, actual material input data, production time, operators, etc.) with the previous design, trial mixing, and review data to generate a full-process production traceability file. This ensures that the concrete production quality is consistent with the mix design requirements and provides complete data support for project quality traceability, realizing a closed-loop control of "design-verification-production".

[0079] The above is the entire working process of the device, and all contents not described in detail in this specification are existing technologies known to those skilled in the art.

Claims

1. A system for intelligent generation and verification of concrete benchmark mix proportions, comprising a parameter input and management module, an AI benchmark mix proportion calculation module, a trial mixing verification and data feedback module, and a mix proportion optimization and production application module, characterized in that: The parameter input and management module includes a design target parameter input module, a raw material parameter input module, and a parameter verification and storage module. It accurately collects, verifies, and stores concrete design targets and raw material parameters, providing reliable data support for AI calculations. The AI ​​benchmark mix ratio calculation module relies on algorithm models and historical data to achieve rapid and accurate generation of benchmark mix ratios, which solves the problem of traditional trial mixing relying on experience. It consists of a historical mix ratio database management module, a neural network algorithm matching module, and a benchmark dosage calculation module. The trial mixing verification and data feedback module includes a trial mixing scheme generation module, a performance testing module, and a data feedback module. Through laboratory trial mixing and testing, the feasibility of the benchmark mix ratio is verified, providing real data support for subsequent optimization and forming a preliminary closed loop of calculation and verification. The mix proportion optimization and production application module includes an AI automatic optimization module, a mix proportion review and determination module, and a mixing plant control system association module. Based on the trial mixing feedback data, it completes the iterative optimization of the mix proportion and finally outputs a formal mix proportion that can be directly used for production, and achieves seamless integration with the mixing plant.

2. The intelligent generation and verification system for concrete reference mix proportions according to claim 1, characterized in that: The design target parameter input module allows technicians to input core design requirements for concrete, including grade, frost resistance grade, impermeability grade, slump, and spread, providing both standardized options and custom input modes. The raw material parameter input module is responsible for collecting key parameters of raw materials, including sand, stone, cement, water, and additives, covering sand fineness modulus, mud content, crushed stone particle size and gradation, cement type and strength grade, and additive type and manufacturer parameters.

3. The intelligent generation and verification system for concrete reference mix proportions according to claim 1, characterized in that: The parameter verification and storage module performs logical verification on the input parameters to avoid invalid data input; at the same time, it stores compliant parameters in the system database, supporting historical data tracing and retrieval.

4. The intelligent generation and verification system for concrete reference mix proportions according to claim 1, characterized in that: The historical mix proportion database management module integrates historical mix proportion data, raw material parameters, construction environment, and performance test results of various projects, establishes a standardized data index, and supports rapid retrieval by project type and parameter range; The neural network algorithm matching module calls a preset neural network algorithm model, uses the currently entered design target parameters and raw material parameters as search conditions, matches similar engineering cases in the database, and analyzes the correlation of key parameters.

5. The intelligent generation and verification system for concrete reference mix proportions according to claim 1, characterized in that: The benchmark dosage calculation module automatically calculates the benchmark dosage of cement, sand, stone, water, and additives based on the algorithm matching results, generates a preliminary concrete benchmark mix proportion, and clarifies the proportion and dosage range of each component.

6. The intelligent generation and verification system for concrete reference mix proportions according to claim 1, characterized in that: The test mixing scheme generation module automatically generates a laboratory test mixing operation plan based on the benchmark mix ratio, which clarifies the operation requirements for mixing time, material feeding sequence, and test block preparation standards, ensuring the standardization of the test mixing process; The performance testing module performs workability and mechanical property tests on the trial-mixed concrete, records test data, test time, and test environment, and supports real-time input of test results.

7. The intelligent generation and verification system for concrete reference mix proportions according to claim 1, characterized in that: When the data feedback module is working, it automatically uploads the test data to the system, compares and analyzes it with the design target parameters, identifies items that do not meet the standards, and provides a clear direction for AI optimization.

8. The intelligent generation and verification system for concrete reference mix proportions according to claim 1, characterized in that: When the AI ​​automatic optimization module is working, it automatically adjusts the mixing ratio parameters through the algorithm model to generate an optimized formal mixing ratio based on the problems found in the trial mixing test.

9. The intelligent generation and verification system for concrete reference mix proportions according to claim 1, characterized in that: The mix proportion review and confirmation module allows technicians to manually review the AI-optimized official mix proportions. Fine-tuning can be made according to specific project requirements, and the final mix proportions are locked after the review is approved.

10. The intelligent generation and verification system for concrete reference mix proportions according to claim 1, characterized in that: When the associated module of the mixing plant control system is working, it will automatically synchronize the final determined formal mix proportion data to the mixing plant control system, realize the one-click distribution of mix proportion parameters, ensure accurate control of the amount of each raw material used in the production process, and achieve a closed loop of intelligent generation-verification optimization-one-click production.

Images