Concrete production mixing proportion dynamic regulation method and device and storage medium

Abstract

This invention discloses a method, equipment, and storage medium for dynamic control of concrete mix proportions, relating to the field of concrete production technology. The method includes: real-time acquisition of four types of data during concrete production: raw materials, mixing process, environmental conditions, and the performance of fresh concrete; multi-source fusion processing of each type of data to generate corresponding state sub-vectors; concatenation of these sub-vectors into a production state vector; generation of an initial mix proportion control amount based on the deviation between the production state vector and the target state vector, combined with a control gain matrix; obtaining the final mix proportion control amount after applying control constraints; and online correction of the mix proportion parameters to achieve dynamic closed-loop control. This invention upgrades concrete production from static setting to multi-source sensing and dynamic optimization, significantly reducing slump fluctuations, segregation and bleeding risks, and strength dispersion, while improving pumpability and long-term stability.

Description

Technical Field

[0001] This invention belongs to the field of concrete production technology, and in particular relates to a method, equipment and storage medium for dynamic control of concrete production mix proportions that integrates multi-source sensor data. Background Technology

[0002] Concrete, as the most widely used structural material in civil engineering, directly determines the construction quality and service life of engineering structures through its workability, pumpability, strength, and durability. Among these factors, mix design is the core aspect of concrete performance control, directly affecting the fluidity, cohesiveness, water retention, and hardened mechanical and durability properties of fresh concrete.

[0003] Currently, commercial concrete mixing plants generally adopt a production model of "static mix proportion + post-production manual adjustment": before production starts, the baseline mix proportion is determined once based on the inspection report of incoming raw materials, relevant technical specifications, and experience parameters; when performance fluctuations occur during production, the tester or operator usually makes remedial adjustments by adding water or admixtures based on the slump test results. However, concrete production is a typical time-varying, nonlinear, and multivariate coupled industrial process, and the actual operating conditions have significant dynamic uncertainties.

[0004] (1) The state of raw materials varies over time: The moisture content of aggregates is affected by factors such as rain, sun exposure, and material extraction from different parts of the stockpile, and can fluctuate by 2% to 5% in a short period of time; the aggregate gradation continues to drift with changes in quarry ore source, wear of screening equipment, and different extraction depths; the content of effective components of admixtures shows a decreasing trend with storage temperature and storage time; the temperature of cementitious materials fluctuates significantly with the seasons, and the temperature of cement entering the machine can reach more than 60℃ in summer.

[0005] (2) Dynamic changes during the mixing process: wear of mixer blades, status of discharge gate, and load differences during mixing cycle lead to changes in torque and power characteristics, which in turn affect the uniformity of mixing and the coating effect of slurry.

[0006] (3) Environmental disturbances: Meteorological conditions such as ambient temperature, relative humidity, and wind speed directly affect the hydration rate and moisture evaporation rate of concrete, especially significantly aggravating the slump loss over time during high temperature and low humidity seasons.

[0007] (4) Differentiated construction requirements: Different pumping heights, pumping distances, and pouring locations place different requirements on the pumpability and workability of concrete, making it difficult to achieve dynamic adaptation of static mix proportions.

[0008] Under the coupled effects of the aforementioned multi-source disturbance factors, statically pre-set mix proportions are difficult to maintain the stability of fresh concrete performance in actual continuous production. Engineering practice shows that the production mode of "static mix proportions + post-production manual adjustments" mainly faces the following five technical bottlenecks:

[0009] (1) Instability of fluidity: The slump at the outlet often fluctuates by more than 30-50 mm, which makes it difficult to vibrate and reduces the density of some batches, while some batches show segregation and bleeding.

[0010] (2) Deterioration of cohesiveness: coarse aggregate separates from mortar, the homogeneity of concrete decreases, the cross-sectional properties are uneven after hardening, and the structural bearing capacity is weakened;

[0011] (3) Poor water retention: Water seepage on the surface forms a water film, which increases the surface water-cement ratio, reduces impermeability and durability, and easily causes plastic shrinkage cracks;

[0012] (4) Insufficient pumpability: Excessive pumping resistance leads to pipe blockage and bursting, forcing the interruption of construction and affecting the progress of the project and the life of the pumping equipment;

[0013] (5) Rapid loss over time: The rate of slump loss during transportation and waiting for pouring exceeds expectations, significantly compressing the construction time window and forcing secondary water addition on site, which seriously damages strength and durability.

[0014] To address the aforementioned issues, existing technologies primarily rely on human experience for delayed adjustments, which has limitations in the following aspects:

[0015] Response lag: There is a significant time lag between the discovery of the problem and the implementation of adjustments, during which multiple batches of substandard concrete have already been produced;

[0016] The adjustment is too crude: the adjustment amount lacks quantitative basis, which can easily lead to over-adjustment or under-adjustment, or even cause the water-cement ratio to get out of control and the strength to plummet due to blindly adding water.

[0017] Untraceable: Adjustments rely on operators' subjective judgment, lack standardized records, and are difficult to support quality backtracking and process optimization;

[0018] Multi-parameter coordination is difficult: multiple parameters such as water consumption, admixture dosage, sand ratio, and water-cement ratio are coupled and have an impact, making it difficult to achieve coordinated optimization manually.

[0019] In recent years, a few studies have attempted to introduce a single sensor (such as a microwave moisture meter) to feedforward compensation for water consumption, or to make a rough estimate of slump based on stirring power. However, these studies remain at the level of single-variable correction and have failed to establish a multi-dimensional perception system covering raw materials, mixing process, environmental conditions, and fresh mix performance. Furthermore, they have not formed a systematic control architecture of "state identification - deviation calculation - constraint control - closed-loop feedback".

[0020] Therefore, the concrete production industry urgently needs a technical solution that can integrate multi-source sensor information, identify production status in real time, and dynamically optimize mix proportion parameters, in order to break through the traditional mode of static setting and manual adjustment, and achieve refined and intelligent closed-loop control of concrete workability, pumpability, and strength stability. Summary of the Invention

[0021] To address the aforementioned deficiencies in the existing technology, the present invention aims to provide a method, equipment, and storage medium for dynamic control of concrete production mix proportions, thereby solving at least one of the following technical problems:

[0022] Static mix proportions cannot adapt to dynamic working conditions in real time—overcoming the serious deviation between the actual production state and the preset mix proportion caused by factors such as time-varying raw materials, environmental disturbances, and process drift;

[0023] Manual post-event adjustments are characterized by lag and crudeness – overcoming the common defects of the “material-based adjustment” model, such as delayed response, lack of quantitative basis for adjustment, and easy to cause water-cement ratio loss;

[0024] Single-variable correction cannot achieve multi-parameter synergistic optimization—it changes the limitation of only compensating for water consumption or admixtures in isolation, and solves the problem of linkage control under the coupling conditions of multiple parameters such as water consumption, admixture dosage, sand ratio, and water-cement ratio;

[0025] Production status is imperceptible and control behavior is untraceable—abandon the extensive management approach that lacks status awareness, relies on experience for decision-making, and makes it difficult to trace the causes of quality anomalies;

[0026] The lack of effective integration of multi-source information and the low level of intelligent decision-making – breaking through the limitations of single sensor information dimensions, solving the bottleneck of the difficulty in uniformly representing multi-source heterogeneous data and failing to form a systematic closed-loop control.

[0027] This invention solves the above-mentioned technical problems through the following technical solution: a method for dynamic control of concrete production mix proportions, comprising:

[0028] During the concrete production process, real-time data on raw material status, mixing process status, environmental conditions, and fresh concrete performance are collected.

[0029] The raw material state data, mixing process state data, environmental condition data, and fresh concrete performance data are respectively subjected to multi-source sensor fusion processing within the same data type to generate raw material state sub-vectors, mixing process state sub-vectors, environmental condition sub-vectors, and fresh concrete performance sub-vectors.

[0030] The raw material state subvector, the mixing process state subvector, the environmental condition subvector, and the fresh mixing performance subvector are spliced ​​together to form the production state vector.

[0031] Calculate the state deviation vector based on the production state vector and the preset target state vector;

[0032] Based on the state deviation vector and the preset control gain matrix, an initial mix ratio control amount is generated;

[0033] Apply control constraints to the initial mix proportion control amount to generate the final mix proportion control amount;

[0034] Based on the final mix proportion adjustment amount, at least two control parameters in the concrete production mix proportion are corrected online to achieve dynamic closed-loop control of the concrete production mix proportion.

[0035] This invention achieves comprehensive perception of key physical quantities in the concrete production process by real-time acquisition of four types of sensor data: raw material status data, mixing process status data, environmental condition data, and fresh concrete performance data. Based on this, multi-source sensor fusion processing is performed on each type of data to generate corresponding category-specific state sub-vectors. These sub-vectors are then concatenated to form a unified production state vector. This technique overcomes the limitations of existing technologies, such as the dimensionality of information from a single sensor and the difficulty in fusing multi-source heterogeneous data. It transforms scattered, heterogeneous raw data into a structured and standardized representation of the production state, providing clear and complete state input for subsequent precise control, thus achieving a leap from "point-based perception" to "multi-dimensional state cognition."

[0036] The state deviation vector quantitatively characterizes the gap between the current production state and the ideal target, and can keenly capture the impact of dynamic factors such as aggregate moisture content fluctuations, changes in ambient temperature and humidity, and disturbances during the mixing process on concrete performance. Mix proportion control no longer relies on static preset values, but rather dynamically generates adjustment amounts based on real-time perceived deviations, enabling the mix proportion to adaptively adjust with changes in working conditions. This effectively solves the technical bottleneck of traditional static mix proportions being unable to adapt to the time-varying nature of production conditions.

[0037] The initial mix ratio control amount is generated based on the state deviation vector and the preset control gain matrix, transforming the control decision from manual experience-based judgment to quantitative calculation by a mathematical model. A clear mapping relationship is established between the control amount and the state deviation, eliminating the subjectivity and arbitrariness of manual adjustments. At the same time, the entire control process is synchronized with the production cycle, enabling real-time response and completely solving the problems of response lag and difficulty in accurately controlling the adjustment range under the manual "material inspection and adjustment" mode.

[0038] The system performs online corrections to the control parameters in the concrete production mix design based on the final mix design adjustment. Here, "control parameters" can encompass multiple adjustable core parameters in the mix design, allowing the system to coordinately adjust these core parameters according to the specific distribution of state deviations. For example, when segregation risk is detected, the system can simultaneously adjust multiple parameters related to cohesiveness (such as admixture dosage and sand ratio) rather than correcting a single variable in isolation. This improves cohesiveness while ensuring fluidity, overcoming the limitations of existing technologies where single-variable corrections cannot adequately address multiple performance indicators.

[0039] The entire control process of this invention is based on real-time data acquisition, fusion, state construction, and deviation calculation. The triggering conditions, deviation magnitude, basis for generating control quantities, and final execution results of each control can all be recorded and traced by the system. This quantifiable control record provides complete data support for the analysis of the causes of quality anomalies and the optimization of process parameters. It changes the situation where control behavior is untraceable and experience is difficult to pass on under the traditional manual adjustment mode, laying the foundation for continuous improvement of concrete production and digital upgrading of quality control.

[0040] Furthermore, the raw material condition data includes at least one of aggregate moisture content, aggregate gradation, aggregate temperature, and aggregate apparent density;

[0041] The mixing process status data includes at least one of the following: mixing torque, mixing power, mixer speed, and mixing time;

[0042] The environmental operating condition data shall include at least one of the following: ambient temperature, ambient humidity, and wind speed or evaporation factor.

[0043] The performance data of the freshly mixed concrete shall include at least two of the following: fluidity index, cohesiveness index, water retention index, pumpability index, and loss over time index.

[0044] This invention defines the specific content of various types of sensor data, providing a clear data source for the construction of production state vectors, ensuring that the system can comprehensively capture key factors affecting concrete performance, and laying a reliable data foundation for subsequent precise control.

[0045] Furthermore, the liquidity indicator is characterized by the collapse rate;

[0046] The cohesiveness index is a segregation risk index, which is obtained through the following method: a concrete sample is obtained at the discharge port, and the sample surface image is acquired after collapse or expansion; the sample surface image is segmented to extract the coarse aggregate region; the area ratio of the coarse aggregate region to the edge region is calculated, and then the absolute difference between the area ratios of the coarse aggregate region and the edge region is calculated to obtain the segregation risk index; wherein, the edge region refers to the area outside the coarse aggregate region;

[0047] The water retention index is the pressure bleeding coefficient, which is obtained by placing the concrete sample in a pressure bleeding test device and measuring the bleeding volume under a specified pressure and time; calculating the ratio of the bleeding volume to the bleeding test time to obtain the pressure bleeding coefficient.

[0048] The pumpability index is a pumping resistance index, and the specific calculation formula is as follows: ;

[0049] in, Indicates pumping resistance; L represents the equivalent length of the pumping pipeline; N b Indicates the equivalent number of elbows in the pumping pipeline; Indicates the pressure-induced water leakage coefficient; , , The corresponding weighting coefficients are obtained through engineering calibration tests or by regression analysis of historical production data.

[0050] The time-loss index is characterized by the slump time-loss rate, and the specific calculation formula is as follows:

[0051] ;

[0052] in, This indicates the slump loss rate over time; Indicates the slump at the exit station; Indicates the measured slump during the transportation or pouring stage; Indicates the duration of time.

[0053] This invention defines specific methods for obtaining various performance indicators of fresh concrete. By clarifying that slump represents fluidity, segregation risk indicators are obtained through image analysis, pressure bleeding coefficient is calculated through bleeding tests, pumping resistance is calculated using a weighted formula, and time loss rate is represented by slump change rate, quantifiable and reproducible measurement methods are provided for various engineering performance indicators. This provides objective quantitative standards for properties such as cohesiveness and water retention, which previously relied on manual experience, enhancing the accuracy and consistency of condition assessment. It also provides clear input for the subsequent normalization of risk indicators.

[0054] Furthermore, both the initial mix proportion adjustment amount and the final mix proportion adjustment amount include corrections for unit water consumption, admixture dosage, sand ratio, and water-cement ratio; the initial mix proportion adjustment amount is generated through a linear gain relationship.

[0055] ;

[0056] in, This indicates the initial mix proportion adjustment amount; This represents the preset control gain matrix, obtained through engineering calibration tests or by regression analysis of historical production data; This represents the state deviation vector.

[0057] By clarifying that both the initial and final mix proportion control quantities include four correction parameters—unit water consumption, admixture dosage, sand ratio, and water-cement ratio—the specific objects of multi-parameter synergistic control are established. Simultaneously, by generating the initial control quantity through a linear gain relationship, the mapping relationship between state deviation and control quantity is mathematically and modeled, providing a calculable quantitative basis for multi-parameter linkage optimization and overcoming the limitations of relying solely on manual experience to achieve multi-parameter synergistic adjustment.

[0058] Furthermore, the control constraints include at least: a single control amplitude constraint and a water-cement ratio safety range constraint; wherein, the single control amplitude constraint is:

[0059] ;

[0060] in, This indicates the final mix proportion adjustment amount; Indicates the maximum permissible adjustment range;

[0061] The safe range constraint for the water-to-binder ratio is:

[0062] ;

[0063] in, Indicates the water-to-glue ratio; Indicates the water-to-binder ratio correction amount; , These represent the minimum and maximum permissible water-cement ratio, respectively.

[0064] By setting constraints on the amplitude of single adjustments, system oscillations or production out of control caused by excessive adjustments in a single instance are prevented. Furthermore, by setting safety range constraints on the water-cement ratio, the ratio is strictly guaranteed to remain within the strength design requirements, preventing strength drops due to blind adjustments. These constraints, while pursuing optimized workability, provide a safety guarantee for the consistency of concrete strength and long-term durability.

[0065] Furthermore, the regulation method also includes:

[0066] A comprehensive risk index is constructed based on the state deviation vector, and its specific expression is as follows:

[0067] ;

[0068] in, This represents a comprehensive risk indicator; This represents the collapse satisfaction index obtained by transforming the fluidity deviation in the state deviation vector. , , , These represent the risk indicators obtained by normalizing the cohesiveness deviation, water retention deviation, pumpability deviation, and time loss deviation in the state deviation vector, respectively. , , , , This represents the corresponding weighting coefficient;

[0069] The adjustment priority and magnitude of the adjustment parameters in the initial mix proportion adjustment amount are adjusted according to the comprehensive risk index:

[0070] When the slump meets the index When the water consumption is below the corresponding preset threshold, the correction amount for unit water consumption or water-cement ratio will be adjusted first.

[0071] When the cohesion risk indicator When the corresponding preset threshold is exceeded, the adjustment of the admixture dosage correction amount or sand ratio correction amount shall be prioritized;

[0072] When water retention risk index When the corresponding preset threshold is exceeded, the amount of admixture correction is adjusted first.

[0073] When pumpability risk indicators When the corresponding preset threshold is exceeded, the adjustment of the admixture dosage correction amount and / or sand ratio correction amount shall be prioritized;

[0074] Risk indicators of loss over time When the corresponding preset threshold is exceeded, the adjustment of the admixture dosage correction amount or admixture type will be prioritized;

[0075] When the comprehensive risk index exceeds the corresponding preset threshold, the control amplitude in the control constraint condition is reduced.

[0076] By transforming the state deviation vector into a comprehensive risk index, a unified quantitative assessment of multi-dimensional risks is achieved. Furthermore, by clarifying the priority control parameters corresponding to different risk indicators exceeding limits (e.g., prioritizing adjustments to admixture dosage or sand ratio for segregation risk, and prioritizing adjustments to admixture dosage and sand ratio for pumping risk), control decisions can be tailored to the current primary risk type, improving the accuracy and efficiency of control. Simultaneously, reducing the control amplitude when the comprehensive risk exceeds the limit reflects an adaptive correlation between risk level and control intensity.

[0077] Furthermore, the control method also includes a self-learning optimization step:

[0078] For N batches of concrete produced continuously or using a sliding window method, the slump fluctuation coefficient and strength dispersion coefficient are statistically analyzed; wherein, the slump fluctuation coefficient is the ratio of the standard deviation to the mean of the slump during the statistical period, and the strength dispersion coefficient is the ratio of the standard deviation to the mean of the compressive strength of concrete during the statistical period.

[0079] Based on the evaluation results of the collapse fluctuation coefficient and / or intensity dispersion coefficient, update the weight coefficients in the comprehensive risk index, and / or modify the regulation gain matrix based on historical production data to achieve long-term stability optimization.

[0080] By statistically analyzing the slump fluctuation coefficient and intensity dispersion coefficient of continuous production batches, and updating the weight coefficients of the comprehensive risk index and / or correcting the control gain matrix based on the evaluation results, the system acquires long-term adaptive optimization capabilities. This limitation enables the control model of this invention to continuously improve itself with the accumulation of production data, adapting to long-term drift factors such as changes in raw material characteristics, equipment wear, and seasonal environmental fluctuations. This achieves an upgrade from a "static preset model" to a "dynamic self-optimizing system," ensuring the continuous stability of the control effect.

[0081] Furthermore, the control method also includes determining the rate of change of mixing torque and the rate of change of mixing power based on the mixing torque and mixing power in the mixing process state data, respectively;

[0082] The mixing torque change rate and mixing power change rate are used to help characterize the mixing uniformity and cohesiveness, and the generation of the initial mix proportion adjustment amount is corrected accordingly.

[0083] By incorporating the dynamic torque and power characteristics during the stirring process into the control decision, a real-time and continuous auxiliary characterization method is provided for the mixing uniformity and cohesiveness. Compared with segregation risk indicators based on sample analysis, the torque / power change rate can be obtained in real time within each stirring cycle, filling the state perception gap during the sampling interval. This allows the system to detect cohesiveness degradation trends earlier and correct the generation of initial control values ​​in a timely manner, enhancing the timeliness and foresight of the control.

[0084] 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 dynamic control method for concrete production mix proportion as described above.

[0085] 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 dynamic control method for concrete production mix proportions as described above.

[0086] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0087] By collecting four types of multi-source data in real time—raw materials, mixing process, environmental conditions, and fresh mix performance—and classifying, fusing, and constructing state vectors, concrete production is upgraded from the traditional "static mix proportion + post-production manual adjustment" to a closed-loop process of "multi-source perception + state judgment + dynamic control," significantly improving the automation and intelligence level of production control.

[0088] Based on the deviation between the real-time constructed production state vector and the target state, the mix proportion adjustment amount is dynamically generated, which can effectively cope with time-varying factors such as aggregate moisture content fluctuation, environmental temperature and humidity changes, and admixture decay, significantly reduce the fluctuation range of the discharge slump, reduce the probability of quality events such as segregation and bleeding, and improve the stability of concrete workability.

[0089] The control parameters cover multiple core proportion parameters such as unit water consumption, admixture dosage, sand ratio, and water-cement ratio. By controlling the gain matrix, the linkage mapping from state deviation to multiple control parameters is realized, overcoming the limitation of existing single-variable correction that is difficult to take into account multiple performance indicators. It can simultaneously optimize cohesiveness, water retention and pumpability while ensuring fluidity. Attached Figure Description

[0090] 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.

[0091] Figure 1 This is a flowchart of the dynamic control method for concrete production mix proportion in an embodiment of the present invention;

[0092] Figure 2 This is an example diagram comparing the slump distribution and fluctuation coefficient before and after dynamic control in an embodiment of the present invention;

[0093] Figure 3 This is an example diagram comparing the distribution and dispersion coefficient of compressive strength before and after dynamic adjustment 28 days in an embodiment of the present invention. Detailed Implementation

[0094] 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.

[0095] 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.

[0096] Example 1

[0097] This embodiment provides a specific implementation of a method for dynamically controlling the mix proportion of concrete production, applied to the daily production process of a commercial concrete mixing plant. For example... Figure 1 As shown, the method for dynamically controlling the mix proportions in concrete production includes the following steps:

[0098] Step S1: During the concrete production process, collect real-time data on the status of raw materials, the status of the mixing process, environmental conditions, and the performance of freshly mixed concrete.

[0099] In this embodiment, the raw material status data includes aggregate moisture content, aggregate gradation, aggregate temperature, and aggregate apparent density; the mixing process status data includes mixing torque, mixing power, mixer speed, and mixing time; the environmental conditions data includes ambient temperature, ambient humidity, and wind speed (or evaporation factor); and the fresh concrete performance data includes fluidity index, cohesiveness index, water retention index, pumpability index, and time loss index.

[0100] Step S1 is implemented by the system's multi-source sensing module, including a moisture content sensor, a temperature sensor, an online gradation monitoring device, a torque sensor, a power acquisition unit, a rotary encoder, a temperature and humidity sensor, an anemometer, an automatic slump measurement device, a machine vision unit, and a pressure seepage testing device.

[0101] Specifically, a microwave moisture content sensor, an infrared temperature sensor, and an online gradation monitoring device are installed in the aggregate bins to acquire aggregate moisture content, aggregate temperature, and aggregate gradation in real time. The apparent density of the aggregate is pre-entered into the system based on the incoming raw material inspection report or recorded after offline measurement before production. A torque sensor is installed on the mixer spindle to acquire the mixing torque in real time; a power acquisition unit is installed on the mixer motor power line to acquire the mixing power in real time; the mixer speed is obtained through feedback from a rotary encoder or motor frequency converter; the mixing time is obtained by the system control module based on the production cycle time. Temperature and humidity sensors and anemometers are installed in the production control room or at the mixing station to acquire ambient temperature, ambient humidity, and wind speed (or evaporation factor) in real time.

[0102] In a specific embodiment of the present invention, the liquidity indicator is determined by the collapse rate. Characterization. An automatic slump measuring device is installed at the discharge port. The concrete mixture is loaded into a standard conical slump cone. After filling and leveling, the slump cone is lifted vertically upwards. The difference between the highest point of the concrete mixture after slumping and the height of the cone is measured to obtain the slump value.

[0103] Cohesiveness index is a segregation risk indicator It can be obtained through the following methods:

[0104] Concrete samples are obtained at the discharge port using a bypass sampler. After the concrete samples are slumped or expanded, machine vision units acquire images of the sample surface. Image segmentation is performed on the sample surface images to extract the coarse aggregate region. The area ratio of the coarse aggregate region to the edge region is calculated, and then the absolute difference between the area ratios of the coarse aggregate region and the edge region is calculated to obtain the segregation risk index. The edge region refers to the area outside the coarse aggregate region. A higher segregation risk index indicates a higher segregation risk and poorer cohesion.

[0105] Water retention index Pressure bleeding coefficient It can be obtained through the following methods:

[0106] Concrete samples obtained by the bypass sampler are placed in a pressure bleeding test device, and bleeding tests are conducted at a specified pressure and for a specified time. The volume of bleeding is measured, and the bleeding test time is recorded. The ratio of the bleeding volume to the bleeding test time is calculated to obtain the pressure bleeding coefficient. The larger the pressure bleeding coefficient, the higher the risk of bleeding and the worse the water retention.

[0107] In another specific embodiment of the present invention, the water retention index can also be expressed as: .

[0108] Pumpability index is the pumping resistance index. The specific calculation formula is as follows:

[0109] (1)

[0110] Where L represents the equivalent length of the pumping pipeline, which is pre-input into the system according to the pumping plan at the construction site; N b This indicates the equivalent number of elbows in the pumping pipeline, which is calculated based on the actual number and angle of elbows in the pumping pipeline at the construction site and then entered into the system. , , This represents the corresponding weighting coefficient, which is obtained through engineering calibration tests or by regression analysis of historical production data.

[0111] The time-lapse index is composed of slump and time-lapse rate. The characterization is obtained through the following methods:

[0112] Record the initial slump when the concrete leaves the site. During the transportation to the construction site or the waiting stage for pouring, the measured slump is obtained again using an automatic slump measuring device. Record the time elapsed between two measurements. Calculate the slump loss rate over time using the following formula:

[0113] (2)

[0114] Slump loss rate over time The larger the value, the faster the time loss and the shorter the construction window.

[0115] Step S2: Perform multi-source sensor fusion processing on the raw material state data, mixing process state data, environmental condition data, and fresh concrete performance data within the same data type to generate raw material state sub-vectors, mixing process state sub-vectors, environmental condition sub-vectors, and fresh concrete performance sub-vectors.

[0116] Step S2 is implemented by the system's data fusion module. The data fusion module is used to perform time synchronization, anomaly removal, normalization, and confidence-weighted fusion of the sensor measurements in each type of sensor data (i.e., raw material status data, mixing process status data, environmental conditions data, and fresh concrete performance data) to obtain the corresponding sub-vector.

[0117] Since different sensors may have different sampling frequencies and data transmission delays, the first step is to time-align the sensor measurements within each type of sensor data. Using the system's unified timestamp as a benchmark, interpolation or nearest neighbor matching methods are used to merge the sensor measurements within the same time moment or a very small time window into multi-source data records for that moment, ensuring the consistency of the subsequently fused data in the time dimension. Taking raw material status data as an example, using the system's unified timestamp as a benchmark, interpolation or nearest neighbor matching methods are used to time-synchronize aggregate moisture content, aggregate gradation, aggregate temperature, and aggregate apparent density.

[0118] For each type of sensor data, outlier detection and removal are performed on the synchronized sensor measurements. Statistical methods (such as the 3σ principle) or methods based on physically reasonable ranges (such as aggregate moisture content between 0% and 20%, and temperature between -10℃ and 80℃) are used to identify and remove outlier data points that significantly deviate from the normal range. For removed outliers, forward imputation or interpolation methods can be used to complete the data sequence, ensuring its continuity.

[0119] For each type of sensor data, the measurements from each sensor are converted into dimensionless standardized values ​​to eliminate the influence of differences in physical dimensions and magnitudes on subsequent fusion. For the i-th sensor measurement in a certain type of sensor data, normalization is performed according to the following formula:

[0120] (3)

[0121] in, This represents the normalized value of the measurement of the i-th sensor at time t; This represents the original value of the measurement measured by the i-th sensor at time t; This represents the mean of the historical data of the i-th sensor; This represents the standard deviation of the historical data of the i-th sensor; This represents a tiny positive number to prevent division by zero.

[0122] Based on factors such as the measurement accuracy, stability, and reliability of each sensor, a confidence weight is assigned to each sensor. For each type of sensor data, the sum of the confidence weights of all sensors is 1.

[0123] For the feature splicing and fusion method, the normalized value of each sensor is multiplied by its corresponding confidence weight to form the fused feature vector of this type of sensor data: For weighted summation fusion, the normalized values ​​of each sensor are summed according to their confidence levels to obtain a comprehensive scalar value as the fusion feature of this type of sensor data. The specific fusion method used can be determined based on the correlation between the sensors within this type of sensor data and engineering requirements.

[0124] Through the above fusion process, state sub-vectors corresponding to the four types of sensor data are generated respectively:

[0125] The sensor measurements in the raw material status data are synchronized in time, anomaly removal is performed, normalization is applied, and confidence-weighted fusion is performed to obtain the raw material status sub-vector M. t The sensor measurements in the mixing process status data are synchronized in time, anomaly removal is performed, normalization is applied, and confidence-weighted fusion is performed to obtain the mixing process status sub-vector B. t The environmental condition data is processed by time synchronization, anomaly removal, normalization, and confidence-weighted fusion of sensor measurements to obtain the environmental condition sub-vector E. t The measured values ​​in the fresh concrete performance data are synchronized over time, anomaly removal is performed, normalization is applied, and confidence-weighted fusion is performed to obtain the fresh concrete performance sub-vector Q. t , .

[0126] Step S3: Concatenate the raw material state sub-vector, the mixing process state sub-vector, the environmental condition sub-vector, and the fresh mixing performance sub-vector to form the production state vector.

[0127] Step S3 is implemented by the system's state determination module, and the production state vector S t It can be represented as S t =[M t B t Et Q t This vector comprehensively reflects the overall state of raw materials, mixing process, environmental conditions, and fresh mix performance during concrete production.

[0128] Step S4: Calculate the state deviation vector based on the production state vector and the preset target state vector.

[0129] Step S4 is implemented by the system's state determination module. Based on the engineering target performance requirements and current operating constraints, the target state vector is determined. For example, for pumped concrete, the target slump can be set to 180±20mm, and the target pumping resistance index can be set to not exceed a preset threshold R. p,th .

[0130] State deviation vector △S t =S t -S ref This deviation vector quantitatively characterizes the gap between the current production state and the target state.

[0131] In one specific embodiment of the present invention, the mixing torque and mixing power in the mixing process status data collected in step S1 are respectively calculated as the rate of change of the mixing torque and the rate of change of the mixing power.

[0132] The rates of change in mixing torque and mixing power are used to help characterize the mixing uniformity and cohesiveness. Specifically, statistical analysis is performed on the rates of change over a mixing period to extract fluctuation characteristics (such as standard deviation, range, and zero-crossing frequency) and construct auxiliary indicators for cohesiveness. The higher the value of this index (∈[0,1]), the greater the risk of cohesion.

[0133] When the cohesive auxiliary index C aux Exceeding the corresponding preset threshold T aux This indicates a potential deterioration in the cohesiveness of the mixture, and the system accordingly adjusts the generation of subsequent initial mix proportion adjustments. The adjustment methods include at least one of the following:

[0134] In the subsequent step S5, when generating the initial mix proportion adjustment amount, a reinforcing coefficient is applied to the control parameters related to cohesiveness (admixture dosage correction amount, sand ratio correction amount). Taking the admixture dosage as an example: ,in The preset enhancement coefficient, This is the correction amount for the admixture dosage in the initial mix proportion control. This is the correction amount for the amount of additives.

[0135] Alternatively, the row elements in the gain matrix corresponding to the cohesive deviation can be dynamically adjusted to make the system more sensitive to cohesive risks.

[0136] Or in the dynamic adjustment based on risk indicators in step S6, C aux As an auxiliary input for cohesive risk, the corresponding risk indicator weights or priorities are dynamically adjusted.

[0137] This auxiliary characterization method, along with the segregation risk index based on sample analysis in step S5, is... They complement each other, with the former providing real-time process monitoring and the latter providing precise offline measurement, together enhancing the system's ability to perceive cohesive states and improving the foresight and accuracy of regulation.

[0138] Step S5: Generate the initial mix ratio control amount based on the state deviation vector and the preset control gain matrix.

[0139] Step S5 is implemented by the system's control decision module. Based on the state deviation vector calculated in step S4, and combined with the preset control gain matrix, the initial mix ratio control amount is generated.

[0140] The control gain matrix is ​​an m×n dimensional matrix, where m is the number of control parameters (in this invention, m=4, corresponding to unit water consumption, admixture dosage, sand ratio, and water-cement ratio), and n is the dimension of the state vector. The element k in the control gain matrix... ij This represents the influence coefficient of the j-th state deviation component on the i-th control parameter. The control gain matrix can be obtained through engineering calibration experiments (e.g., by slightly perturbing the unit water consumption, admixture dosage, sand ratio, and water-cement ratio during production and measuring the changes in each state index, thereby establishing a mapping relationship between state deviation and control quantity), or it can be obtained based on historical production data through multiple linear regression or machine learning methods.

[0141] Initial mix proportion adjustment amount Represented as a four-dimensional vector:

[0142] (4)

[0143] in, , , , These correspond to corrections for unit water consumption, admixture dosage, sand ratio, and water-cement ratio, respectively. The calculation formulas are as follows:

[0144] (5)

[0145] in, This represents the preset control gain matrix; This represents the state deviation vector.

[0146] Step S6: Apply control constraints to the initial mix proportion control amount to generate the final mix proportion control amount.

[0147] To ensure the safety and stability of the control process and prevent production fluctuations or loss of intensity due to excessive adjustments in a single instance, control constraints need to be applied to the initial mix proportion control amount to generate the final mix proportion control amount. Step S6 is implemented by the system's control decision module. The control constraints in this embodiment include single control amplitude constraints and water-cement ratio safety range constraints.

[0148] The single-time control amplitude constraint stipulates that the absolute value of each control component must not exceed the preset maximum allowable control amplitude. :

[0149] (6)

[0150] Right now , , , .

[0151] If a certain control component exceeds the preset maximum allowable control range, it will be limited to the corresponding maximum allowable control range.

[0152] The water-cement ratio safety range constraint is that the adjusted water-cement ratio must be within the preset safety range according to the strength grade requirements:

[0153] (7)

[0154] in, Indicates the water-to-glue ratio; , These represent the minimum and maximum allowable water-cement ratios, respectively, which are predetermined based on concrete strength design requirements and durability specifications. If the water-cement ratio correction causes the adjusted water-cement ratio to exceed the safe range in formula (7), the water-cement ratio correction is adjusted to limit it to the maximum allowable adjustment that brings the water-cement ratio to the boundary value of the safe range.

[0155] In a preferred embodiment, to enhance the pertinence and adaptability of the control decision, based on the control constraints imposed on the initial mix ratio control amount in step S6, the control priority and control amplitude of the control parameters are dynamically adjusted by combining the comprehensive risk index and its individual risk indicators.

[0156] Comprehensive risk indicators It is constructed based on the state deviation vector, and its specific expression is:

[0157] (8)

[0158] in, This represents the slump satisfaction index derived from the flowability deviation in the state deviation vector, when the measured slump falls within the target slump range [S]. L ,S U Within the specified time, the slump meets the slump index. =1; when the measured slump is lower than S L or higher than S U At that time, the slump meets the index The deviation is rated between 0 and 1; the greater the deviation, the lower the slump satisfaction index. The smaller. , , , These represent risk indicators obtained by normalizing the cohesiveness deviation, water retention deviation, pumpability deviation, and time loss deviation in the state deviation vector, respectively. The values ​​range from [0,1], and the larger the value, the higher the risk of that dimension. , , , , This represents the corresponding weighting coefficient. ~ The sum equals 1, which can be determined based on engineering experience or historical data, or it can be dynamically updated through self-learning optimization steps.

[0159] When a single risk indicator exceeds its corresponding preset threshold, the system identifies this risk as the primary contradiction in current production. In control decisions, priority is given to adjusting the control parameters related to this risk, while adjustments to other parameters can be appropriately reduced or delayed. The specific correspondence is as follows:

[0160] Slump Satisfaction Index Below the preset threshold This indicates insufficient or excessive fluidity; priority should be given to adjusting the unit water consumption correction or water-cement ratio correction. For example, if the slump is too low ( L If the slump is too high, then the unit water consumption should be increased first; if the slump is too high ( >S U If the unit water consumption is low, then priority should be given to reducing the unit water consumption.

[0161] Separation risk indicators Exceeding the preset threshold This indicates poor cohesiveness of the concrete and a risk of segregation. Therefore, adjustments should be made to the admixture dosage or sand ratio. Generally, increasing the admixture dosage enhances the paste viscosity, while adjusting the sand ratio improves aggregate gradation; both can improve cohesiveness.

[0162] Risk indicators of water retention​ Exceeding the preset threshold This indicates poor water retention and a high risk of bleeding in the concrete, necessitating adjustments to the admixture dosage. For example, increasing the dosage of water-reducing agents or thickeners can reduce free water exudation and improve water retention.

[0163] Pumping risk indicators Exceeding the preset threshold This indicates insufficient pumpability of the concrete and a risk of pipe blockage. Therefore, it is crucial to coordinate the adjustment of admixture dosage and sand ratio. Generally, increasing the admixture dosage can reduce pumping resistance, while appropriately increasing the sand ratio can improve lubrication layer formation.

[0164] Risk indicators of loss over time Exceeding the preset threshold This indicates that the concrete slump is decreasing too rapidly over time, shortening the construction window. Adjustments should be made to the admixture dosage or type. For example, increasing the dosage of retarder admixtures or replacing them with admixtures that offer better slump retention can slow down the hydration process and maintain fluidity.

[0165] In practice, when multiple risk indicators exceed the limit at the same time, the system can determine the risk type to be dealt with first based on the degree of deviation of each risk indicator (i.e. the size of the risk value) and the weight coefficient, or generate the control amount by weighted synthesis.

[0166] When the comprehensive risk index exceeds the preset threshold When this occurs, it indicates that the current production status presents a high degree of comprehensive risk. The system as a whole reduces the control amplitude in the control constraints, that is, it reduces the single control amplitude limit set in step S6. Perform dynamic tightening:

[0167] ,0<η<1 (9)

[0168] in, The contraction coefficient can be determined based on the comprehensive risk index exceeding a threshold. The degree is determined dynamically, for example Alternatively, a piecewise function can be used. The control amplitude limit after contraction. It will replace the original control range limit. Amplitude constraints used in this regulatory cycle.

[0169] At the same time, the safety range constraint for the water-to-binder ratio can also be tightened accordingly, for example, by narrowing the range in which the water-to-binder ratio can fluctuate:

[0170] , (10)

[0171] in, , The positive offset amount, determined based on the level of risk, makes the water-cement ratio control more conservative.

[0172] For example, suppose the pumping risk index of a certain batch of concrete... =0.85, exceeding the preset threshold. =0.7, while other risk indicators are all within the normal range. According to the above priority rules, when generating the final mix proportion control amount, the system will prioritize ensuring the adjustment needs of the admixture dosage correction amount and sand ratio correction amount related to pumping risk. Even if the deviations in other dimensions of the state deviation vector are also large, the adjustment range of their corresponding control parameters will be appropriately reduced, so that control resources can focus on solving the pumping problem.

[0173] If we simultaneously consider the risk indicator R t =0.65 also exceeds the preset threshold R. t,th =0.6, then the system will further limit the single adjustment amplitude. Shrink to 0.8× This imposes stricter restrictions on the adjustment of all control parameters in order to avoid systemic risks.

[0174] Through the above-mentioned dynamic adjustment based on risk indicators, the control decision of this invention not only considers the magnitude of the state deviation, but also incorporates the semantic information of risk, making the control behavior more intelligent and precise. Under the premise of ensuring production safety, it can prioritize the resolution of the main contradictions in current production, and significantly improve the adaptability and robustness of the system.

[0175] Step S7: Based on the final mix proportion adjustment amount, the adjustment parameters in the concrete production mix proportion are corrected online to achieve dynamic closed-loop control of the concrete production mix proportion.

[0176] Step S7 is implemented by the system's execution module. The final mix proportion adjustment amount is sent to the execution module. The execution module includes a water metering execution unit, an admixture metering execution unit, and a sand ratio adjustment unit (or aggregate ratio adjustment unit), realizing multi-parameter linkage control.

[0177] Water metering execution unit: Adjusts the opening of the water supply valve or the frequency of the metering pump according to the unit water consumption correction amount in the final mix proportion control, and corrects the unit water consumption online. If the unit water consumption correction amount is positive, the water consumption is increased; if the unit water consumption correction amount is negative, the water consumption is decreased.

[0178] Admixture metering execution unit: Adjusts the delivery volume or frequency of the admixture metering pump according to the admixture dosage correction amount in the final mix proportion control amount, and corrects the admixture dosage.

[0179] Sand ratio adjustment unit: Adjusts the discharge ratio of aggregate batching bins according to the sand ratio correction amount in the final mix proportion control amount, and corrects the sand ratio (i.e., the mass ratio of fine aggregate to coarse aggregate).

[0180] The water-cement ratio correction is achieved through the coordination of water usage and cementitious material usage. Since the amount of cementitious material typically remains constant across consecutive production batches, the water-cement ratio adjustment is primarily achieved by adjusting the water usage. Therefore, there is a corresponding relationship between the water-cement ratio correction and the unit water usage correction. In actual implementation, the system prioritizes ensuring that the actual water-cement ratio after water usage adjustment meets the safety range constraints and uses the water-cement ratio correction as a monitoring indicator to ensure that the control results meet expectations.

[0181] After the control is executed, the system enters the next sampling cycle, restarts data collection from step S1, updates the production state vector in real time, calculates the state deviation, and generates the control quantity again, forming a dynamic closed-loop control. The triggering conditions, deviation magnitude, basis for control quantity generation, and execution results of each control are all recorded by the system, forming a traceable control log for use in quality analysis and process optimization.

[0182] Step S8: Self-learning optimization steps.

[0183] Step S8 is implemented by the system's control and decision-making module or an added self-learning module to achieve long-term stability optimization.

[0184] For N batches of concrete produced continuously or using a sliding window method, the slump fluctuation coefficient (CV) is statistically analyzed. S With intensity dispersion coefficient CV f Among them, the slump fluctuation coefficient CV S The ratio of the standard deviation to the mean of the slump during the statistical period is the strength dispersion coefficient (CV). f This is the ratio of the standard deviation to the mean of the concrete compressive strength during the statistical period.

[0185] Based on the slump fluctuation coefficient CV S and / or intensity dispersion coefficient CV f The evaluation results will be used to update the weighting coefficients in the comprehensive risk index. ~ And / or adjust the control gain matrix based on historical production data. The update method can be:

[0186] When the slump fluctuation coefficient CV S or intensity dispersion coefficient CV f When the corresponding preset target value is exceeded, the weight of the corresponding risk indicator should be appropriately increased to strengthen regulation; or a large amount of historical data should be collected to compare the state deviation vector ΔS. t , Mixture ratio control amount ΔP tRegression analysis was performed to refit the control gain matrix, making the model more closely match actual production characteristics.

[0187] Through self-learning optimization, the system of this invention can continuously improve itself as production data accumulates, adapting to long-term drift factors such as changes in raw material characteristics, equipment wear and tear, and seasonal environmental fluctuations, thus ensuring the continuous stability of the control effect.

[0188] To verify the technical effectiveness of this invention, N batches (N≥50) of the same strength grade (e.g., C30 pumped concrete) were continuously produced, and comparative experiments were conducted using both traditional manual control methods and the method of this invention. The distribution and fluctuation coefficient (CV) of the slump before and after control were statistically analyzed. S The distribution and coefficient of variation of the 28-day compressive strength. f .

[0189] like Figure 2 As shown, after adopting the dynamic control method of this invention, the distribution of the slump at the outlet is more concentrated, the fluctuation amplitude is significantly reduced, and the slump fluctuation coefficient CV is reduced. S The rate of water separation and bleeding decreased from approximately 12.5% ​​before regulation to approximately 3.2% after regulation, representing a relative decrease of 74.4%; the number of water separation and bleeding events decreased by more than 80%, and the incidence of pump blockage decreased by 75%.

[0190] like Figure 3 As shown, the 28-day compressive strength distribution is also more concentrated after dynamic adjustment, and the strength dispersion coefficient CV f The percentage decreased from approximately 8.5% before regulation to approximately 2.1% after regulation, a relative decrease of 75.3%, indicating that the present invention effectively ensures project quality and durability while improving the workability stability and strength consistency of concrete.

[0191] Through the above steps S5 to S7, the present invention realizes a complete closed loop from state deviation to multi-parameter collaborative control, upgrading concrete production from traditional static setting and manual post-event adjustment to an intelligent control process based on multi-source perception and dynamic optimization, which significantly improves the quality stability and process controllability of concrete production.

[0192] Example 2

[0193] 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 dynamic control method for concrete production mix proportions in this invention.

[0194] 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.

[0195] 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.

[0196] 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 method for dynamic control of concrete production mix proportions in embodiments of the present invention.

[0197] 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.

[0198] 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 dynamically controlling the mix proportions in concrete production, characterized in that, The control method includes: During the concrete production process, real-time data on raw material status, mixing process status, environmental conditions, and fresh concrete performance are collected. The raw material state data, mixing process state data, environmental condition data, and fresh concrete performance data are respectively subjected to multi-source sensor fusion processing within the same data type to generate raw material state sub-vectors, mixing process state sub-vectors, environmental condition sub-vectors, and fresh concrete performance sub-vectors. The raw material state subvector, the mixing process state subvector, the environmental condition subvector, and the fresh mixing performance subvector are spliced ​​together to form the production state vector. Calculate the state deviation vector based on the production state vector and the preset target state vector; Based on the state deviation vector and the preset control gain matrix, an initial mix ratio control amount is generated; Apply control constraints to the initial mix proportion control amount to generate the final mix proportion control amount; Based on the final mix proportion adjustment amount, at least two control parameters in the concrete production mix proportion are corrected online to achieve dynamic closed-loop control of the concrete production mix proportion.

2. The method of claim 1, wherein the method further comprises: The raw material condition data includes at least one of the following: aggregate moisture content, aggregate gradation, aggregate temperature, and aggregate apparent density. The mixing process status data includes at least one of the following: mixing torque, mixing power, mixer speed, and mixing time; The environmental operating condition data shall include at least one of the following: ambient temperature, ambient humidity, and wind speed or evaporation factor. The performance data of the freshly mixed concrete shall include at least one of the following: fluidity index, cohesiveness index, water retention index, pumpability index, and loss over time index.

3. The method of claim 2, wherein the step of dynamically adjusting the mix ratio of the concrete production is performed by a computer system. The liquidity metric is characterized by collapse. The cohesiveness index is a segregation risk index, which is obtained by: taking concrete samples at the discharge port, and collecting surface images of the samples after collapse or expansion; and extracting the coarse aggregate region by image segmentation of the sample surface images. The area ratio of the coarse aggregate region to the edge region is calculated, and then the absolute difference between the area ratios of the coarse aggregate region and the edge region is calculated to obtain the segregation risk index; wherein, the edge region refers to the region outside the coarse aggregate region. The water retention index is the pressure bleeding coefficient, which is obtained by placing the concrete sample in a pressure bleeding test device and measuring the bleeding volume under a specified pressure and time; calculating the ratio of the bleeding volume to the bleeding test time to obtain the pressure bleeding coefficient. The pumpability index is a pumping resistance index, and the specific calculation formula is as follows: ； wherein, represents the pumping resistance index; L represents the equivalent length of the pumping pipeline; N b represents the equivalent number of bends of the pumping pipeline; represents the pressure bleeding coefficient; , , represents the corresponding weight coefficient, which is obtained through engineering calibration test or is obtained by regression from historical production data; The time-loss index is characterized by the slump time-loss rate, and the specific calculation formula is as follows: ； in, This indicates the slump loss rate over time; Indicates the slump at the exit station; Indicates the measured slump during the transportation or pouring stage; Indicates the duration of time.

4. The method for dynamic control of concrete production mix proportions according to claim 1, characterized in that, Both the initial mix proportion adjustment amount and the final mix proportion adjustment amount include the unit water consumption correction amount, the admixture dosage correction amount, the sand ratio correction amount, and the water-cement ratio correction amount; the initial mix proportion adjustment amount is generated through a linear gain relationship: ； in, This indicates the initial mix proportion adjustment amount; This represents the preset control gain matrix, obtained through engineering calibration tests or by regression analysis of historical production data; This represents the state deviation vector.

5. The method for dynamic control of concrete production mix proportions according to claim 1, characterized in that, The control constraints include at least: single-time control amplitude constraints and water-cement ratio safety range constraints; wherein, the single-time control amplitude constraints are: ； in, This indicates the final mix proportion adjustment amount; Indicates the maximum permissible adjustment range; The safe range constraint for the water-to-binder ratio is: ； in, Indicates the water-to-glue ratio; Indicates the water-to-binder ratio correction amount; , These represent the minimum and maximum permissible water-cement ratio, respectively.

6. The method for dynamic control of concrete production mix proportions according to any one of claims 1 to 5, characterized in that, The control method also includes: A comprehensive risk index is constructed based on the state deviation vector, and its specific expression is as follows: ； in, This represents a comprehensive risk indicator; This represents the collapse satisfaction index obtained by transforming the fluidity deviation in the state deviation vector. , , , These represent the risk indicators obtained by normalizing the cohesiveness deviation, water retention deviation, pumpability deviation, and time loss deviation in the state deviation vector, respectively. , , , , This represents the corresponding weighting coefficient; The adjustment priority and magnitude of the adjustment parameters in the initial mix proportion adjustment amount are adjusted according to the comprehensive risk index: When the slump meets the index When the water consumption is below the corresponding preset threshold, the correction amount for unit water consumption or water-cement ratio will be adjusted first. When the cohesion risk indicator When the corresponding preset threshold is exceeded, the adjustment of the admixture dosage correction amount or sand ratio correction amount shall be prioritized; When water retention risk index When the corresponding preset threshold is exceeded, the amount of admixture correction is adjusted first. When pumpability risk indicators When the corresponding preset threshold is exceeded, the adjustment of the admixture dosage correction amount and / or sand ratio correction amount shall be prioritized; Risk indicators of loss over time When the corresponding preset threshold is exceeded, the adjustment of the admixture dosage correction amount or admixture type will be prioritized; When the comprehensive risk index exceeds the corresponding preset threshold, the control amplitude in the control constraint condition is reduced.

7. The method for dynamic control of concrete production mix proportions according to claim 6, characterized in that, The control method also includes a self-learning optimization step: For N batches of concrete produced continuously or using a sliding window method, the slump fluctuation coefficient and strength dispersion coefficient are statistically analyzed; wherein, the slump fluctuation coefficient is the ratio of the standard deviation to the mean of the slump during the statistical period, and the strength dispersion coefficient is the ratio of the standard deviation to the mean of the compressive strength of concrete during the statistical period. Based on the evaluation results of the collapse fluctuation coefficient and / or intensity dispersion coefficient, update the weight coefficients in the comprehensive risk index, and / or modify the regulation gain matrix based on historical production data to achieve long-term stability optimization.

8. The method for dynamic control of concrete production mix proportions according to claim 1, characterized in that, The control method further includes determining the rate of change of mixing torque and the rate of change of mixing power based on the mixing torque and mixing power in the mixing process status data, respectively. The mixing torque change rate and mixing power change rate are used to help characterize the mixing uniformity and cohesiveness, and the generation of the initial mix proportion adjustment amount is corrected accordingly.

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 method for dynamic control of concrete production 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 method for dynamic control of concrete production mix proportions as described in any one of claims 1 to 8.

Images