Method, system and equipment for dynamically monitoring quality of premixed concrete

Abstract

The invention discloses a method, a system and equipment for dynamically monitoring the quality of premixed concrete, and relates to the field of control. The method comprises the following steps: installing a plurality of sensors at preset positions, wherein the sensors are used for collecting target parameters in real time; data preprocessing is carried out on the first quality parameter sequence, the second quality parameter sequence, the third quality parameter sequence and the fourth quality parameter sequence in the mode that trend analysis is carried out on the average value sampled by the sliding window, and data preprocessing results are used for sequence fusion to obtain a target parameter set; constructing a plurality of three-classification decision random forests based on the target parameter set to obtain a quality evaluation set; and performing voting integration on the quality evaluation set according to a probabilistic algorithm to obtain a quality monitoring index. By means of the scheme, the real-time performance, precision and reliability of quality monitoring of the premixed concrete can be improved.

Description

Technical Field

[0001] The present application relates to the field of control, and in particular to a method, system and equipment for dynamically monitoring the quality of premixed concrete. Background Art

[0002] Premixed concrete is a form of concrete that is mixed evenly in a factory according to a specific formula and then transported to the construction site. Its quality is directly related to the safety and durability of the building structure. Therefore, the monitoring of the quality of premixed concrete is extremely important in construction projects. High-quality premixed concrete can ensure the stability and earthquake resistance of buildings, extend the service life of buildings, and reduce the cost of maintenance and repair. Existing premixed concrete quality monitoring technologies mainly rely on offline detection methods, such as physical and chemical property tests of concrete samples in the laboratory. These tests usually include the measurement of parameters such as compressive strength, slump, density and frost resistance. Although these methods can provide detailed quality data, they have the following problems: poor timeliness, offline detection usually requires a certain amount of time to collect samples, transport, test and analyze, which means that quality problems cannot be discovered and corrected in time during the construction process. Limitations, offline detection can only analyze limited samples and cannot fully reflect the quality of the entire concrete batch. High manual dependence, offline detection requires manual sampling and operation, which is easily affected by human factors, resulting in insufficient reliability and consistency of the test results. These problems make the existing premixed concrete quality monitoring technology have certain limitations in practical applications, and it cannot meet the needs of modern construction projects for efficient, accurate and real-time quality control.

[0003] At present, in order to solve the above problems, some online monitoring methods have been developed in the industry. These methods usually involve installing sensors during the production and transportation of concrete to monitor the physical and chemical parameters of concrete in real time, such as temperature, humidity, stress and strain. Although these methods can improve the real-time and coverage of monitoring to a certain extent, there are still some defects; the amount of data collected by the sensors is large and complex, and advanced data processing and analysis technologies are required to extract effective information from them. The accuracy of some sensors is limited and cannot accurately reflect the actual quality of concrete. Sensors are easily disturbed by the construction site environment, such as vibration, dust and humidity changes, resulting in data instability. Compared with traditional offline detection methods, these online monitoring methods have significantly improved in real-time and coverage, but there are still deficiencies in data processing, accuracy, environmental adaptability and cost.

[0004] Therefore, there is an urgent need for a technical solution that can improve the real-time, accuracy and reliability of pre-mixed concrete quality monitoring. Summary of the invention

[0005] In order to solve the deficiencies of the prior art, the embodiments of the present application provide a method, system and device for dynamically monitoring the quality of premixed concrete. The present application solves the technical problems that the prior art still has deficiencies in data processing, accuracy, environmental adaptability and cost.

[0006] The embodiment of the present application provides a dynamic monitoring method for the quality of premixed concrete, comprising: installing a plurality of sensors at preset positions, the sensors being used to collect target parameters in real time; measuring the durability index of concrete in real time to obtain a first quality parameter sequence; measuring the workability index of concrete in real time to obtain a second quality parameter sequence; measuring the density index of concrete in real time to obtain a third quality parameter sequence; measuring the uniformity index of concrete in real time to obtain a fourth quality parameter sequence; performing data preprocessing on the first, second, third and fourth quality parameter sequences respectively based on trend analysis of the mean of sliding window sampling, and using the results of the data preprocessing for sequence fusion to obtain a target parameter set; constructing multiple three-classification decision random forests based on the target parameter set to obtain a quality assessment set; and performing voting integration on the quality assessment set according to a probability algorithm to obtain a quality monitoring index.

[0007] In a possible implementation, data preprocessing is performed on the first, second, third and fourth quality parameter sequences respectively based on a trend analysis of the mean of sliding window sampling, and the results of the data preprocessing are used for sequence fusion to obtain a target parameter set, including: using a window of a preset size to perform sliding operations on the first, second, third and fourth quality parameter sequences based on a preset step size, and calculating the mean in the window at each sliding position to obtain the corresponding first, second, third and fourth window mean sequences; using a linear regression model to fit the first, second, third and fourth window mean sequences respectively to obtain a first regression coefficient, a second regression coefficient and an error coefficient of the linear regression model; using the linear regression model after obtaining the coefficients to calculate the trend value corresponding to each data item in the first, second, third and fourth window mean sequences respectively to obtain the corresponding first, second, third and fourth trend sequences; splicing the first, second, third and fourth quality parameter sequences with the first, second, third and fourth window mean sequences and the first, second, third and fourth trend sequences to construct a target parameter set.

[0008] In one possible implementation, the mean value in the window at each sliding position is calculated using the formula: ,in, Indicates time The window mean of Indicates location The sequence value at Indicates the window size.

[0009] In a possible implementation, multiple three-classification decision random forests are constructed based on the target parameter set to obtain a quality assessment set, including: constructing three-classification random forests for the 12 sequences in the target parameter set respectively, and integrating the classification results of the decision trees in the random forest corresponding to each sequence to obtain a quality assessment set including the three-classification probability values ​​and confidences corresponding to the 12 sequences.

[0010] In one possible implementation, the classification results of the decision trees in the random forest corresponding to each sequence are integrated, and the formula is: ,in, represents the integration result, Indicates A decision tree for input samples The prediction result is The probability of represents the number of decision trees, Represents the prediction result of the mth decision tree for the input sample x.

[0011] In a possible implementation, the quality assessment set is voted and integrated according to a probabilistic algorithm to obtain a quality monitoring indicator, including: taking a weighted average for each category, and the result of the weighted average is used to adjust the weight of the corresponding sequence category according to the corresponding confidence level; and using a Bayesian method to update the probability value of the adjusted category to obtain a quality monitoring indicator.

[0012] In one possible implementation, the adjusted probability value is updated using the Bayesian method, and the formula is: ,in, represents the posterior probability of category k, represents the prior probability of category k, This indicates that this is the classification probability of the i-th sequence for category k. represents the classification probability of the i-th sequence for category j, Represents the classification probability values ​​of 12 sequences Perform the product operation, Represents the normalization calculation to control the sum of the posterior probabilities of all categories to be 1.

[0013] A possible implementation method also includes: the quality monitoring indicators include qualified, warning and abnormal; when the quality monitoring indicators are warning or abnormal, the corresponding target parameter set and the corresponding first, second, third and fourth quality parameter sequences are filled into the preset prompt word template to obtain the target prompt word; the target prompt word is input into the large language model to obtain a quality monitoring report including the cause of the abnormality and the improvement plan.

[0014] The embodiment of the present application also provides a dynamic monitoring system for the quality of premixed concrete, including: a data acquisition unit and a monitoring unit; wherein the data acquisition unit is used to measure the durability index of concrete in real time to obtain a first quality parameter sequence; measure the workability index of concrete in real time to obtain a second quality parameter sequence; measure the density index of concrete in real time to obtain a third quality parameter sequence; measure the uniformity index of concrete in real time to obtain a fourth quality parameter sequence; the monitoring unit is used to perform data preprocessing on the first, second, third and fourth quality parameter sequences respectively based on trend analysis of the mean of sliding window sampling, and the results of the data preprocessing are used to perform sequence fusion to obtain a target parameter set; construct multiple three-classification decision random forests based on the target parameter set to obtain a quality assessment set; and vote and integrate the quality assessment set according to a probability algorithm to obtain a quality monitoring indicator.

[0015] An embodiment of the present application also provides a dynamic monitoring device for the quality of premixed concrete, comprising: a processor, a memory, and a system bus; wherein the processor and the memory are connected via the system bus; the memory is used to store one or more programs, wherein the one or more programs include instructions, and when the instructions are executed by the processor, the processor executes the method described in the above embodiment.

[0016] In the method, system and device for dynamic monitoring of the quality of premixed concrete as provided above, the embodiments of the present application realize real-time monitoring of concrete quality through multi-sensor fusion and intelligent data analysis, and timely discover and correct quality problems. Furthermore, in some embodiments, the random forest model and probability algorithm are used to accurately evaluate the quality status of concrete and provide reliable quality monitoring indicators. Furthermore, in some embodiments, the system automatically generates detailed target prompt words to guide engineering personnel to take effective measures to ensure that the construction quality meets the requirements. Furthermore, in some embodiments, by optimizing the data processing and analysis process, the operating efficiency of the system is improved, and manual intervention and maintenance costs are reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 A schematic flow chart of a method for dynamically monitoring the quality of premixed concrete provided in an embodiment of the present application; Figure 2 A schematic flow chart of another method for dynamically monitoring the quality of premixed concrete provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present application.

[0020] Those skilled in the art will appreciate that the terms "first", "second" and the like in the embodiments of the present application are only used to distinguish different steps, devices or modules, etc., and do not represent any specific technical meaning, nor do they represent the necessary logical order between them. It should also be understood that in the embodiments of the present application, "multiple" may refer to two or more, and "at least one" may refer to one, two or more. It should also be understood that for any component, data or structure mentioned in the embodiments of the present application, in the absence of explicit limitation or contrary revelation given in the context, it can generally be understood as one or more. In addition, the term "and / or" in the present application is only a kind of association relationship describing the associated objects, indicating that there can be three kinds of relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present application generally indicates that the associated objects before and after are an "or" relationship. It should also be understood that the description of each embodiment in the present application emphasizes the differences between the embodiments, and the same or similar parts can refer to each other. For the sake of brevity, they will not be repeated one by one.

[0021] At the same time, it should be understood that, for ease of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present application and its application or use. The techniques, methods and devices known to ordinary technicians in the relevant fields may not be discussed in detail, but where appropriate, the techniques, methods and devices should be considered as part of the specification. It should be noted that similar numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0022] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0023] Figure 1 A schematic flow chart of a method for dynamically monitoring the quality of premixed concrete provided in an embodiment of the present application.

[0024] like Figure 1 As shown, at step S101, a variety of sensors are installed at preset positions, and the sensors are used to collect target parameters in real time; measure the durability index of concrete in real time to obtain a first quality parameter sequence; measure the workability index of concrete in real time to obtain a second quality parameter sequence; measure the density index of concrete in real time to obtain a third quality parameter sequence; and measure the uniformity index of concrete in real time to obtain a fourth quality parameter sequence.

[0025] Specifically, in order to ensure the stability of concrete quality, this system focuses on monitoring the following indicators: water-cement ratio (w / c) (first quality parameter sequence), which is a key parameter affecting the strength and durability of concrete. Slump (second quality parameter sequence), which reflects the workability of concrete. Aggregate gradation (third quality parameter sequence), which affects the density and strength of concrete. Mixing time (fourth quality parameter sequence), which directly affects the uniformity of concrete.

[0026] This application uses a variety of sensors to collect various parameters in the concrete production process in real time, including water volume sensors, ash volume sensors, slump sensors, and mixing time sensors. These sensors are installed at key locations in the mixing station to ensure the accuracy and real-time nature of data collection. Furthermore, the water-cement ratio (w / c) is calculated as: ,in, and They are the mass of water and cement in concrete respectively. Real-time monitoring through water sensor and ash sensor and , the real-time water-cement ratio can be calculated.

[0027] Slump can be measured by a slump sensor. The slump sensor converts the slump value into an electrical signal and transmits it to the central control system. The standard range of slump is the positive and negative error range of a certain preset value. Exceeding this range will trigger an early warning.

[0028] Aggregate grading monitoring can be done by using a particle size analyzer, which uses laser particle size analysis technology to monitor the particle size distribution of aggregates in real time. Assume that the particle size distribution is , then the aggregate gradation meets certain requirements for the cumulative screening mass percentage: ,in, Particle size The quality of aggregate.

[0029] Mixing time directly affects the uniformity and strength of concrete. The mixing time sensor records the mixing time to ensure that the mixing time is within the set range.

[0030] At step S102, data preprocessing is performed on the first, second, third and fourth quality parameter sequences respectively based on trend analysis of the mean of sliding window sampling, and the results of the data preprocessing are used for sequence fusion to obtain a target parameter set.

[0031] In one embodiment, a sliding operation is performed on the first, second, third and fourth quality parameter sequences based on a preset step size using a window of a preset size, and the mean in the window at each sliding position is calculated to obtain the corresponding first, second, third and fourth window mean sequences; a linear regression model is used to fit the first, second, third and fourth window mean sequences respectively to obtain the first regression coefficient, the second regression coefficient and the error coefficient of the linear regression model; the linear regression model after obtaining the coefficients is used to calculate the trend value corresponding to each data in the first, second, third and fourth window mean sequences respectively to obtain the corresponding first, second, third and fourth trend sequences; the first, second, third and fourth quality parameter sequences are spliced ​​with the first, second, third and fourth window mean sequences and the first, second, third and fourth trend sequences to construct a target parameter set.

[0032] Specifically, sliding window sampling is a smoothing technique that reduces fluctuations and noise by calculating the average of several consecutive values ​​in a numerical sequence. Specifically, a window size is set , then the calculation formula of the sliding window sampling value is: ,in, Indicates time The window mean of Indicates location The sequence value at Indicates the window size. By processing each quality parameter sequence through the sliding window sampling method, the data can be effectively smoothed and short-term fluctuations can be eliminated, thereby better reflecting the actual change trend of concrete quality.

[0033] In the implementation process, the water-cement ratio sequence is firstly processed by sliding window sampling to obtain the smoothed first quality parameter sequence. For example, in a concrete production process, the original observation data of the water-cement ratio is as follows: ,In order to smooth the data, the window size is set to 3, and the sliding window adopts the following processing process: By analogy, we can obtain the smoothed first window mean sequence of water-cement ratio: In the same way, the slump, aggregate gradation and mixing time series can be processed by sliding windows to obtain the smoothed second, third and fourth window mean series.

[0034] Based on the smoothed water-cement ratio sequence, a linear regression model can be used for trend analysis. Based on the above embodiment, it is assumed that the smoothed water-cement ratio sequence is: , the following linear regression model can be established: ,in, represents the water-cement ratio at time t, and are the regression coefficients, is the error term. The above sequence is fitted by the least squares method to obtain the regression coefficient and .

[0035] In one embodiment, the result obtained by regression analysis is: , then the trend equation is: Through this trend equation, the trend value at each moment can be calculated. For example:

[0036] By analogy, the first trend value sequence of water-cement ratio after trend analysis is as follows: ,The same method can be used to perform trend analysis on the slump, aggregate gradation and mixing time series to obtain the second, third and fourth trend value series after ,trend adjustment.

[0037] Compared with the original data and the processed data adopted by the sliding window, the sequence after trend analysis is smoother and eliminates short-term fluctuations and noise. Each value in the sequence reflects the long-term trend and can more accurately represent the overall change direction of the data. Through the sequence after trend analysis, future trends can be predicted, so as to make early judgments on the changes in concrete quality.

[0038] In practical applications, the sequence after trend analysis can help engineers and operators better understand the changing trend of concrete quality. By analyzing the long-term trend of water-cement ratio, it can be determined whether the current concrete has too much or too little water, so as to adjust the mix ratio to ensure the strength and durability of the concrete. By analyzing the trend of slump, it can be determined whether the fluidity of the concrete meets the construction requirements to prevent unqualified concrete from appearing during the construction process. By analyzing the trend of aggregate grading, it can be ensured that the aggregate particles are reasonably distributed, thereby ensuring the density and strength of the concrete. By analyzing the trend of mixing time, it can be determined whether the mixing is sufficient, thereby ensuring the uniformity of the concrete.

[0039] After the preprocessing of the sliding window method and the smoothing trend analysis method is completed, the 12 parameter sequences are sequence fused to form a target parameter set. The process of sequence fusion is to combine the feature vectors of each quality parameter sequence to build a more comprehensive and accurate concrete quality assessment model. In one embodiment, a direct splicing method can be used.

[0040] At step S103, multiple three-classification decision random forests are constructed based on the target parameter set to obtain a quality assessment set. In one embodiment, three-classification random forests are constructed for the 12 sequences in the target parameter set, and the classification results of the decision trees in the random forest corresponding to each sequence are integrated to obtain a quality assessment set including the three-classification probability values ​​and confidences corresponding to the 12 sequences.

[0041] Specifically, first, the construction of the target parameter set is based on multiple parameters such as concrete durability, workability, density and uniformity collected by sensors in real time, and is completed through data preprocessing steps such as sliding window mean calculation and trend analysis. After obtaining the target parameter set, a multi-piece three-classification decision random forest model is constructed based on it.

[0042] For each sequence in the target parameter set (including the original sequence, window mean sequence and trend sequence of durability, workability, density and uniformity, a total of 12 sequences), a three-class random forest model is constructed. Each random forest model consists of multiple decision trees. By integrating the classification results of these decision trees, the three-class probability values ​​and confidence levels corresponding to each sequence are finally obtained. The following are the specific steps: Construct a three-class random forest model. For each sequence, the random forest model contains M decision trees. Each decision tree classifies according to the input sample x and outputs a classification result y and the corresponding probability value .in, Represents the mth decision tree.

[0043] The classification results of the integrated decision trees are integrated to calculate the three-class probability value and confidence of the sequence. The specific formula is as follows: ,in, represents the integration result, Indicates A decision tree for input samples The prediction result is The probability of represents the number of decision trees, It represents the prediction result of the mth decision tree for the input sample x. The final classification result is determined by weighted average or majority voting of the decision tree classification results of each sequence.

[0044] Through the above steps, a quality assessment set including three-category probability values ​​and corresponding confidence levels corresponding to 12 sequences can be obtained. At step S104, the quality assessment set is voted and integrated according to the probability algorithm to obtain a quality monitoring indicator. In one embodiment, a weighted average is performed on each category, and the result of the weighted average is used to adjust the weight of the corresponding sequence category according to the corresponding confidence level; the probability value of the adjusted category is updated using the Bayesian method to obtain a quality monitoring indicator.

[0045] Based on the above examples, the three-classification probability values ​​for the 12 sequences and each sequence are , each Contains the probability values ​​of three categories, namely These probability values ​​need to be combined to obtain a total three-category probability value .

[0046] First, the probability value of each category is weighted averaged. The weight of each sequence is set to , the sum of the weights is 1, that is The weighted average calculation formula is: , where k=1, 2, and 3 represent three categories respectively.

[0047] Then, adjust the weight of each sequence based on its confidence. Set the confidence to , then the weight adjustment formula is: , after adjusting the weight Used for weighted average calculations.

[0048] Secondly, the Bayesian method is used to update the probability value. Assume that the initial probability distribution is , combined with the probability value of the sequence for Bayesian update, the formula is: ,in, represents the posterior probability of category k, represents the prior probability of category k, This indicates that this is the classification probability of the i-th sequence for category k. represents the classification probability of the i-th sequence for category j, Represents the classification probability values ​​of 12 sequences Perform the product operation, Represents the normalization calculation to control the sum of the posterior probabilities of all categories to be 1.

[0049] For example, in a quality monitoring embodiment, the real-time monitoring data of premixed concrete includes four indicators: durability, workability, density, and uniformity. The data collected by the sensor in real time constitutes the time series of these indicators. Through sliding window mean calculation and trend analysis, 12 target parameter sequences are obtained. Next, a random forest model is constructed to classify each sequence, and its three-classification probability value and confidence are calculated. At a certain moment, the classification probability values ​​of the 12 sequences are as follows: Durability (first) original sequence probability value: [0.2, 0.5, 0.3]; Durability (first) mean sequence probability value: [0.1,0.6, 0.3]; Durability (first) trend sequence probability value: [0.3, 0.4, 0.3]; Workability (second) original sequence probability value: [0.25, 0.5, 0.25]; Workability (second) mean sequence probability value: [0.15, 0.55, 0.3]; Workability (second) trend sequence probability value: [0.2, 0.6, 0.2]; Density (third) original sequence probability value: [0.2, 0.4, 0.4]; Density (third) mean sequence probability value: [0.3, 0.5, 0.2]; density (third) trend sequence probability value: [0.25,0.5, 0.25]; uniformity (fourth) original sequence probability value: [0.3, 0.4, 0.3]; uniformity (fourth) mean sequence probability value: [0.2, 0.5, 0.3]; uniformity (fourth) trend sequence probability value: [0.25, 0.45, 0.3]; for these probability values, set the initial weights equal, that is . Calculate the total three-category probability value after weighted average:

[0050] That is, the total three-class probability value obtained by the weighted average method is [0.225, 0.491, 0.284]. Next, after obtaining the initial weighted average total three-class probability value, confidence adjustment and Bayesian update are required to obtain a more accurate total three-class probability value. Assume that the initial confidence of each sequence is: durability original sequence confidence: 0.8; durability mean sequence confidence: 0.7; durability trend sequence confidence: 0.9; workability original sequence confidence: 0.75; workability mean sequence confidence: 0.85; workability trend sequence confidence: 0.8; density original sequence confidence: 0.7; density mean sequence confidence: 0.9; density trend sequence confidence: 0.75; uniformity original sequence confidence: 0.85; uniformity mean sequence confidence: 0.8; uniformity trend sequence confidence: 0.9.

[0051] Assume that the original weights are equal, that is . Adjust the weight according to the confidence level, and the adjusted weight The calculation formula is: , calculate the weight-adjusted sum: , the adjusted weight for: , for example, the original series of durability with adjusted weights: , similarly, the adjusted weights of other sequences are calculated: durability original sequence: 0.0824; durability mean sequence: 0.0724; durability trend sequence: 0.0934; workability original sequence: 0.0773; workability mean sequence: 0.0876; workability trend sequence: 0.0824; density original sequence: 0.0724; density mean sequence: 0.0934; density trend sequence: 0.0773; uniformity original sequence: 0.0876; uniformity mean sequence: 0.0824; uniformity trend sequence: 0.0934.

[0052] The adjusted weighted average calculation formula is: , calculate the weighted average total probability value of each category: , it can be obtained that the total probability value after confidence adjustment is [0.22863, 0.49185, 0.27974].

[0053] After adjusting the weights according to the confidence level, the Bayesian method is used to update the probability, and finally a more accurate total three-class probability value is obtained. Specifically, the initial prior probability for: Category 1 (High Quality) (Qualified): ; Category 2 (medium quality) (warning): ; Category 3 (Low Quality) (Abnormal): ; Update according to the Bayesian formula: , calculate the joint probability product:

[0054] Calculate the denominator:

[0055] Compute the posterior probability: .

[0056] It can be obtained that the total probability value after Bayesian update is [0.0258, 0.9692, 0.0050]. Based on the maximum a posteriori probability principle, the category with the highest probability is selected as the final quality state. It can be obtained that in this embodiment, the quality monitoring index is the second category, that is, an early warning is given. By adjusting the weighted average and Bayesian update after the weights are adjusted, a more accurate total three-category probability value for concrete quality assessment is finally obtained. This method effectively integrates multiple data sources and model prediction results, and improves the accuracy and reliability of concrete quality monitoring. If abnormal fluctuations are found in certain quality parameters, the system will issue an early warning to guide engineering personnel to take measures to ensure that the concrete quality meets the engineering requirements.

[0057] Figure 2 A schematic flow chart of another method for dynamically monitoring the quality of premixed concrete provided in an embodiment of the present application. Figure 2 As shown, at step S201, the quality monitoring indicators include warning and abnormality; when the quality monitoring indicators are warning or abnormal, the corresponding target parameter set and the corresponding first, second, third and fourth quality parameter sequences are filled into the preset prompt word template to obtain the target prompt word. The prompt word template can be designed according to specific engineering requirements, including problem description, parameter details and recommended measures.

[0058] At step S202, the target prompt word is input into the large language model to obtain a quality monitoring report including the cause of the abnormality and the improvement plan. For example, when the current concrete quality test result is a warning or abnormality, the model outputs: "The detailed parameters are as follows: Durability: average value 0.45, trend value decreases; Workability: average value 0.35, trend value unchanged; Density: average value 0.50, trend value increases; Uniformity: average value 0.40, trend value decreases.

[0059] Recommended measures: Please check the concrete material ratio and adjust the mixing time and speed to ensure uniform mixing.

[0060] Through this method, the system can generate detailed target prompt words in time when the concrete quality is unqualified, and guide engineering personnel to take effective measures to ensure that the concrete quality meets the engineering requirements.

[0061] Furthermore, an embodiment of the present application also provides a dynamic monitoring system for the quality of premixed concrete, comprising: a data acquisition unit and a monitoring unit; wherein the data acquisition unit is used to measure the durability index of concrete in real time to obtain a first quality parameter sequence; to measure the workability index of concrete in real time to obtain a second quality parameter sequence; to measure the density index of concrete in real time to obtain a third quality parameter sequence; to measure the uniformity index of concrete in real time to obtain a fourth quality parameter sequence; the monitoring unit is used to perform data preprocessing on the first, second, third and fourth quality parameter sequences respectively based on a trend analysis of the mean of sliding window sampling, and the result of the data preprocessing is used to perform sequence fusion to obtain a target parameter set; based on the target parameter set, multiple three-classification decision random forests are constructed to obtain a quality assessment set; the quality assessment set is voted and integrated according to a probability algorithm to obtain a quality monitoring indicator.

[0062] Furthermore, an embodiment of the present application also provides a dynamic monitoring device for the quality of premixed concrete, comprising: a processor, a memory, and a system bus; the processor and the memory are connected via the system bus; the memory is used to store one or more programs, the one or more programs including instructions, and when the instructions are executed by the processor, the processor executes any of the above methods.

[0063] Furthermore, an embodiment of the present application also provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute any of the above methods.

[0064] It can be known from the description of the above implementation mode that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., including several instructions for enabling a computer device (which can be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0065] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0066] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0067] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for dynamic monitoring of the quality of premixed concrete, characterized in that: include: Installing a variety of sensors at preset locations, the sensors are used to collect target parameters in real time; Real-time measurement of the durability index of concrete to obtain a first quality parameter sequence; Real-time measurement of concrete workability indexes to obtain a second quality parameter sequence; Real-time measurement of the density index of concrete to obtain the third quality parameter sequence; Real-time measurement of the uniformity index of concrete to obtain a fourth quality parameter sequence; The first, second, third and fourth quality parameter sequences are respectively preprocessed by way of trend analysis on the mean of sliding window sampling, and the results of the data preprocessing are used for sequence fusion to obtain a target parameter set; multiple three-classification decision random forests are constructed based on the target parameter set to obtain a quality assessment set; the quality assessment set is voted and integrated according to a probabilistic algorithm to obtain a quality monitoring indicator.

2. The dynamic monitoring method according to claim 1, characterized in that: in, The first, second, third and fourth quality parameter sequences are respectively preprocessed by performing trend analysis on the mean of the sliding window samples, and the results of the data preprocessing are used for sequence fusion to obtain a target parameter set, including: using a window of a preset size to perform sliding operations on the first, second, third and fourth quality parameter sequences based on a preset step size, and calculating the mean in the window at each sliding position to obtain the corresponding first, second, third and fourth window mean sequences; using a linear regression model to fit the first, second, third and fourth window mean sequences respectively to obtain a first regression coefficient, a second regression coefficient and an error coefficient of the linear regression model; using the linear regression model after obtaining the coefficients to calculate the trend value corresponding to each data in the first, second, third and fourth window mean sequences respectively to obtain the corresponding first, second, third and fourth trend sequences; splicing the first, second, third and fourth quality parameter sequences with the first, second, third and fourth window mean sequences and the first, second, third and fourth trend sequences to construct a target parameter set.

3. The dynamic monitoring method according to claim 2, characterized in that: in, Calculate the mean value in the window at each sliding position. The formula is: ,in, Indicates time The window mean of Indicates location The sequence value at Indicates the window size.

4. The dynamic monitoring method according to claim 1, characterized in that: in, Based on the target parameter set, multiple three-classification decision random forests are constructed to obtain a quality assessment set, including: three-classification random forests are constructed for the 12 sequences in the target parameter set respectively, and the classification results of the decision trees in the random forest corresponding to each sequence are integrated to obtain a quality assessment set including three-classification probability values ​​and confidences corresponding to the 12 sequences.

5. The dynamic monitoring method according to claim 4, characterized in that: in, The classification results of the decision trees in the random forest corresponding to each sequence are integrated, and the formula is: ,in, represents the integration result, Indicates A decision tree for input samples The prediction result is The probability of represents the number of decision trees, Represents the prediction result of the mth decision tree for the input sample x.

6. The dynamic monitoring method according to claim 1, characterized in that: in, The quality assessment set is voted and integrated according to a probability algorithm to obtain a quality monitoring indicator, including: weighted averaging each category, the result of the weighted average is used to adjust the weight of the corresponding sequence category according to the corresponding confidence level; and the probability value of the adjusted category is updated using a Bayesian method to obtain a quality monitoring indicator.

7. The dynamic monitoring method according to claim 6, characterized in that: in, The Bayesian method is used to update the adjusted probability value, and the formula is: ,in, represents the posterior probability of category k, represents the prior probability of category k, represents the classification probability of the i-th sequence for category k, represents the classification probability of the i-th sequence for category j, Represents the classification probability values ​​of 12 sequences Perform the product operation, Represents the normalization calculation to control the sum of the posterior probabilities of all categories to be 1.

8. The dynamic monitoring method according to claim 1, characterized in that: Also includes: The quality monitoring indicators include qualified, early warning and abnormal; When the quality monitoring indicator is a warning or abnormality, the corresponding target parameter set and the corresponding first, second, third and fourth quality parameter sequences are filled into the preset prompt word template to obtain the target prompt word; The target prompt word is input into the large language model to obtain a quality monitoring report including abnormal reasons and improvement plans.

9. A dynamic monitoring system for the quality of premixed concrete, characterized in that: include: A data acquisition unit and a monitoring unit; wherein the data acquisition unit is used to measure the durability index of concrete in real time to obtain a first quality parameter sequence; to measure the workability index of concrete in real time to obtain a second quality parameter sequence; to measure the density index of concrete in real time to obtain a third quality parameter sequence; to measure the uniformity index of concrete in real time to obtain a fourth quality parameter sequence; the monitoring unit is used to perform data preprocessing on the first, second, third and fourth quality parameter sequences respectively based on trend analysis of the mean of sliding window sampling, and the results of the data preprocessing are used for sequence fusion to obtain a target parameter set; multiple three-classification decision random forests are constructed based on the target parameter set to obtain a quality assessment set; the quality assessment set is voted and integrated according to a probability algorithm to obtain a quality monitoring indicator.

10. A dynamic monitoring device for the quality of premixed concrete, characterized in that: include: A processor, a memory, and a system bus; wherein the processor and the memory are connected via the system bus; The memory is used to store one or more programs, wherein the one or more programs include instructions, and when the instructions are executed by the processor, the processor executes the method according to any one of claims 1 to 8.

Images