Proportioning design method and system for manufacturing building material based on marine dredged mud

Abstract

The invention relates to the technical field of data processing, and discloses a proportion design method and system for manufacturing a building material based on marine dredged mud. The method comprises the following steps: collecting a dredged mud sample, and measuring characteristic parameters to form a parameter matrix; calculating the similarity between the parameter matrix and the standard library, and determining an applicable building material type; pretreating the dredged mud, measuring the strength and drawing a curve graph; constructing convolutional neural network prediction intensity; adjusting ratio parameter iterative prediction to form a strength database; and screening an optimal ratio to prepare a sample for testing, and determining final ratio data. According to the method, the appropriate building material type and the optimal matching parameter are rapidly determined according to the characteristics of the dredged mud, the experiment workload is reduced, the accuracy and efficiency of matching design are improved, and meanwhile the resource utilization rate of the dredged mud is maximized.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and particularly to a method and system for proportion design of building materials made from marine dredged mud. Background Art

[0002] Marine dredged mud is a large amount of bottom mud material generated during marine engineering construction. Traditionally, it has been mainly disposed of by dumping into the sea area or piling up on land, which not only occupies a large amount of land resources but may also cause environmental pollution. In recent years, with the increasingly strict environmental protection requirements and the promotion of the concept of resource utilization, the conversion of marine dredged mud into building materials has become a research hotspot. Currently, there are research and application cases at home and abroad on using dredged mud to prepare building materials such as bricks, ceramsite, cement concrete, and subgrade fillers. These methods usually use traditional orthogonal test methods or single-factor test methods for proportion design, and determine the optimal proportion through the preparation of a large number of test samples and performance tests. Some researchers have also tried to use mathematical models such as response surface method to assist in proportion optimization.

[0003] However, the existing proportion design methods for making building materials from marine dredged mud have obvious deficiencies. Firstly, the traditional test methods require the preparation of a large number of test samples, which is time-consuming and laborious and has low efficiency. Secondly, the physical and chemical properties of dredged mud are complex and variable, and the characteristics of dredged mud from different sources and treatment methods vary greatly. It is difficult for existing methods to quickly adapt to this change. Thirdly, most of the existing mathematical models only consider limited influencing factors and fail to comprehensively reflect the complex non-linear relationship between the characteristics of dredged mud, proportion parameters, and final performance. Fourthly, existing methods often focus on the optimization of a single performance index and lack comprehensive consideration and balance of multiple performance indexes. Finally, the proportion design in the existing technology is mostly a preliminary setting based on experience and specifications, and the proportion screening process lacks systematicness and pertinence, resulting in low resource utilization efficiency. Summary of the Invention

[0004] This application provides a method and system for proportion design of building materials made from marine dredged mud, which is used to quickly determine suitable building material types and optimal proportion parameters according to the characteristics of dredged mud, reduce the experimental workload, improve the accuracy and efficiency of proportion design, and maximize the resource utilization rate of dredged mud.

[0005] In a first aspect, the present application provides a method for designing the proportion of building materials made from marine dredged mud. The method for designing the proportion of building materials made from marine dredged mud includes: collecting marine dredged mud samples, measuring the particle size distribution, water content, organic matter content, and heavy metal content of the marine dredged mud samples, and generating a dredged mud characteristic parameter matrix; calculating the similarity between the dredged mud characteristic parameter matrix and the building material standard parameter library to determine the type of building materials suitable for the dredged mud; preprocessing the dredged mud, measuring the compressive strength values under different cement dosages, and plotting a strength-cement dosage curve; based on the strength-cement dosage curve, constructing a three-layer convolutional neural network structure, where the input layer includes dredged mud parameters and admixture variables, and the output layer is the predicted strength value; gradually adjusting the proportion parameters of cement, aggregate, and admixture through the convolutional neural network, recording the strength prediction results of each iteration, and forming a proportion-strength database; screening the top three groups of proportions with the highest strength from the proportion-strength database, preparing physical samples and conducting 28-day strength tests to determine the building material proportion data.

[0006] In a second aspect, the present application provides a proportion design system for building materials made from marine dredged mud. The proportion design system for building materials made from marine dredged mud includes:

[0007] A collection module for collecting marine dredged mud samples, measuring the particle size distribution, water content, organic matter content, and heavy metal content of the marine dredged mud samples, and generating a dredged mud characteristic parameter matrix;

[0008] A calculation module for calculating the similarity between the dredged mud characteristic parameter matrix and the building material standard parameter library to determine the type of building materials suitable for the dredged mud;

[0009] A processing module for preprocessing the dredged mud, measuring the compressive strength values under different cement dosages, and plotting a strength-cement dosage curve;

[0010] A construction module for constructing a three-layer convolutional neural network structure based on the strength-cement dosage curve, where the input layer includes dredged mud parameters and admixture variables, and the output layer is the predicted strength value;

[0011] A recording module for gradually adjusting the proportion parameters of cement, aggregate, and admixture through the convolutional neural network, recording the strength prediction results of each iteration, and forming a proportion-strength database;

[0012] A proportion module for screening the top three groups of proportions with the highest strength from the proportion-strength database, preparing physical samples and conducting 28-day strength tests to determine the building material proportion data.

[0013] In a third aspect of the present invention, a computer device is provided, including: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor invokes the instructions in the memory to cause the computer device to execute the above-mentioned proportion design method for manufacturing building materials based on marine dredged mud.

[0014] In a fourth aspect of the present invention, a computer-readable storage medium is provided, in which instructions are stored. When it runs on a computer, it causes the computer to execute the above-mentioned proportion design method for manufacturing building materials based on marine dredged mud.

[0015] In the technical solution provided by this application, through a systematic method for generating a characteristic parameter matrix of dredged mud, key characteristic indicators of dredged mud are comprehensively collected and quantified; secondly, intelligent matching between dredged mud and building material types is achieved by means of similarity calculation technology, avoiding blind trial and error and improving resource allocation efficiency; thirdly, by constructing a strength-cement dosage curve graph, the correlation law between cement dosage and strength development is revealed, providing directional guidance for proportion optimization; fourthly, a three-layer convolutional neural network structure is applied to process multi-dimensional feature data, giving full play to the algorithm advantages of the convolutional neural network in non-linear mapping and feature extraction, enabling the model to accurately capture the complex relationship between dredged mud characteristics, proportion parameters and strength. This proportion prediction model based on deep learning has stronger generalization ability and prediction accuracy compared with traditional statistical methods; fifthly, through the iterative optimization mechanism of the neural network, efficient adjustment and optimization of proportion parameters are realized, greatly reducing the experimental workload and saving time and cost; sixthly, by integrating the preparation of physical samples and the strength test results, a method system combining theoretical prediction and actual verification is established to ensure the reliability and practicality of the proportion design results. Especially in the application of the three-layer convolutional neural network, this method fully considers the matching between algorithm characteristics and actual application requirements. By reasonably designing the network structure and parameters, the model can not only process the high-dimensional characteristic parameters of dredged mud but also accurately predict the strength performance under different proportions, solving the problem of modeling high-dimensional non-linear relationships that are difficult to handle by traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is a schematic diagram of an embodiment of the proportion design method for manufacturing building materials based on marine dredged mud in an embodiment of this application;

[0018] Figure 2 It is a schematic diagram of an embodiment of a ratio design system for making building materials based on marine dredged mud in an embodiment of the present application;

[0019] Figure 3 It is a structural schematic block diagram of a computer device in an embodiment of the present invention. Specific embodiments

[0020] The embodiments of the present application provide a ratio design method and system for making building materials based on marine dredged mud. Terms such as "first", "second", "third", "fourth", etc. (if any) in the specification, claims and above-mentioned drawings of the present application are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order different from that illustrated or described here. In addition, the terms "including" or "having" and any variation thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0021] For ease of understanding, the specific process of the embodiments of the present application will be described below. Please refer to Figure 1 An embodiment of the ratio design method for making building materials based on marine dredged mud in the embodiments of the present application includes:

[0022] Step S101: Collect marine dredged mud samples, measure the particle size distribution, moisture content, organic matter content and heavy metal content of the marine dredged mud samples, and generate a dredged mud characteristic parameter matrix;

[0023] Step S102: Calculate the similarity between the dredged mud characteristic parameter matrix and the building material standard parameter library to determine the applicable building material types of the dredged mud;

[0024] Step S103: Pretreat the dredged mud, measure the compressive strength values under different cement dosages, and draw a strength-cement dosage curve;

[0025] Step S104: Based on the strength-cement dosage curve, construct a three-layer convolutional neural network structure. The input layer includes dredged mud parameters and admixture variables, and the output layer is the predicted strength value;

[0026] Step S105: Gradually adjust the ratio parameters of cement, aggregate and admixture through the convolutional neural network, record the strength prediction results of each iteration, and form a ratio-strength database;

[0027] Step S106: Screen the top three groups of formulations with the highest strength from the formulation-strength database, prepare physical samples, and conduct 28-day strength tests to determine the building material formulation data.

[0028] It can be understood that the execution entity of this application can be a formulation design system for making building materials based on marine dredged mud, or a terminal or a server. Specifically, it is not limited here. In the embodiments of this application, the server is used as the execution entity for illustration.

[0029] Specifically, collect marine dredged mud samples from different depths and different regions, and record the location data of the sampling points through a positioning system. Conduct a comprehensive physicochemical property determination on the collected dredged mud samples, use a laser particle size analyzer to measure the particle size distribution, and obtain the D10, D50, and D90 particle size values, which respectively represent the particle size sizes corresponding to 10%, 50%, and 90% on the cumulative volume distribution curve. Use the drying-weight method to measure the moisture content, use the potassium dichromate oxidation method to measure the organic matter content, and use atomic absorption spectrometry to measure the heavy metal content. Standardize these measurement results to eliminate the dimensional differences between different parameters and construct a dredged mud characteristic parameter matrix. For example, after the determination of a marine dredged mud sample in a certain sea area, the D50 value is 0.075 mm, the moisture content is 42%, the organic matter content is 3.8%, and the lead content is 32 mg / kg. Organize these data in a row-column form to form a characteristic parameter matrix, where the rows represent different samples and the columns represent different characteristic parameters.

[0030] Construct a standard parameter library containing building material types such as cement concrete, blocks, permeable bricks, ceramsite, and impermeable layer materials, and record the requirement indicators of various building materials for raw materials. Extract key eigenvalues from the dredged mud characteristic parameter matrix, calculate the Euclidean distance with the parameters of each building material type in the building material standard parameter library, and obtain a similarity score table. The Euclidean distance calculation measures the similarity between data by comparing the square root of the sum of the squares of the differences in each dimension of two groups of data. The smaller the value, the more similar. Normalize the values in the similarity score table, design an analytic hierarchy decision tree, and screen out building material types with high adaptability. Establish a gradient library for the applicability of dredged mud building materials, including the main applicable types and alternative types. For example, after similarity calculation of a certain dredged mud sample, the adaptability with ceramsite is 0.86, and the adaptability with permeable bricks is 0.72. Therefore, it is determined that this dredged mud is mainly applicable to the preparation of ceramsite. Carry out centrifugal dehydration and drying treatment on the dredged mud, and control the moisture content below 25%. Determine the mineral composition and chemical composition of the pretreated dredged mud by X-ray fluorescence analyzer, and record the content data of calcium, silicon, aluminum, and iron elements. Using the pretreated dredged mud as the base material, add cement at gradients of 5%, 10%, 15%, 20%, 25%, and 30% to prepare 100mm×100mm×100mm standard cube specimens. Cure the standard cube specimens under the conditions of temperature 20±2°C and relative humidity above 95%, and take samples for testing at 3 days, 7 days, and 28 days respectively. Use a pressure testing machine to apply load to the standard cube specimens until they are damaged, and record the compressive strength values of the specimens with different cement dosages at different ages. Pair the compressive strength values with the corresponding cement dosage data, and draw a strength-cement dosage curve with the cement dosage as the abscissa and the compressive strength as the ordinate. This curve intuitively shows the strength change trend after mixing dredged mud with different proportions of cement, providing a basis for subsequent ratio optimization. Extract the data points from the strength-cement dosage curve, and form a training data set with the cement dosage and the corresponding compressive strength values. Combine the key parameters in the dredged mud characteristic parameter matrix to construct a neural network input feature vector, including the particle size distribution characteristics, moisture content, organic matter content, heavy metal content, and cement dosage of the dredged mud. Design a three-layer convolutional neural network structure. The first convolutional layer uses 16 3×3 convolutional kernels to extract features, the second convolutional layer uses 32 3×3 convolutional kernels to deepen the features, and the third convolutional layer uses 64 3×3 convolutional kernels to comprehensively integrate the features. Map the output features of the convolutional layer through a fully connected layer, connect to a hidden layer with 256 neurons, and then connect to a single output neuron, representing the predicted strength value. Use the training data set to train the three-layer convolutional neural network, adopt the mean square error loss function and the Adam optimizer to adjust the network weights. Through the training process, obtain the trained three-layer convolutional neural network structure, with the input layer including dredged mud parameters and admixture variables, and the output layer being the predicted strength value.

[0031] Set the range of cement dosage at 5%-35%, the range of aggregate ratio at 40%-70%, and the range of admixture ratio at 5%-15% to construct a ratio parameter matrix. Divide the ratio parameter matrix into multiple ratio points at a step size of 1% to generate a ratio candidate set, where each ratio point contains specific values of cement, aggregate, and admixture ratios. Sequentially extract ratio points from the ratio candidate set, combine them with dredged mud parameters, and assemble them into input data vectors. Input the input data vectors into a convolutional neural network to obtain corresponding strength prediction output values, and record the corresponding relationship between ratio points and strength values. According to the strength prediction output values, adjust the direction of ratio parameters, gradually explore towards the high-strength region, generate new ratio points and record their strength prediction values. Organize all the tested ratio points and their strength prediction values into a structured data table, establish indexes according to ratio numbers and strength values to form a ratio-strength database. Sort the ratio-strength database in descending order of strength prediction values, extract the top three groups of ratio parameters with the highest strength, and mark them as the preferred ratio groups. Weigh the corresponding weights of pretreated dredged mud, cement, aggregate, and admixture respectively according to the parameter ratios in the preferred ratio groups to prepare three groups of physical test samples. Cure the three groups of physical test samples in a standard environment with a temperature of 20±2°C and a relative humidity of over 95%, and take samples for strength testing at 3 days, 7 days, and 28 days respectively. Record the actual compressive strength values of the three groups of physical test samples at each age, compare them with the strength values predicted by the convolutional neural network, and calculate the prediction errors. Based on the actual compressive strength values and prediction errors, comprehensively score the three groups of ratios, taking into account factors such as 28-day strength, strength growth rate, and prediction accuracy. Select the ratio with the highest comprehensive score from the three groups of ratios to form a standardized ratio table and determine the building material ratio data.

[0032] In the embodiments of the present application, through a systematic dredged mud characteristic parameter matrix generation method, the key characteristic indexes of dredged mud are comprehensively collected and quantified; secondly, the intelligent matching of dredged mud and building material types is realized by means of similarity calculation technology, avoiding blind trial and error and improving the resource allocation efficiency; thirdly, by constructing a strength-cement dosage curve graph, the correlation law between cement dosage and strength development is revealed, providing a directional guidance for proportion optimization; fourthly, a three-layer convolutional neural network structure is applied to process multi-dimensional feature data, giving full play to the algorithm advantages of the convolutional neural network in non-linear mapping and feature extraction, enabling the model to accurately capture the complex relationship between the characteristics of dredged mud, proportion parameters and strength. This proportion prediction model based on deep learning has stronger generalization ability and prediction accuracy compared with traditional statistical methods; fifthly, through the iterative optimization mechanism of the neural network, the efficient adjustment and optimization of proportion parameters are realized, greatly reducing the experimental workload and saving time and cost; sixthly, by integrating the preparation of physical samples and the strength test results, a method system combining theoretical prediction and actual verification is established to ensure the reliability and practicability of the proportion design results. Especially in the application of the three-layer convolutional neural network, this method fully considers the matching of algorithm characteristics and actual application requirements. By reasonably designing the network structure and parameters, the model can not only process the high-dimensional characteristic parameters of dredged mud, but also accurately predict the strength performance under different proportions, solving the problem of high-dimensional non-linear relationship modeling that is difficult to handle by traditional methods.

[0033] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0034] (1) Collect marine dredged mud samples at different depths and in different regions, record the position data of the sampling points through a positioning system, and form a sampling record form;

[0035] (2) Use a laser particle size analyzer to measure the particle size of the marine dredged mud samples, obtain the particle size values of D10, D50, D90 and the distribution curve, and record the particle size distribution of the marine dredged mud;

[0036] (3) Use the drying-weight method to measure the moisture content of the marine dredged mud samples, use the potassium dichromate oxidation method to measure the organic matter content, and use the atomic absorption spectrometry method to measure the heavy metal content;

[0037] (4) Standardize the particle size distribution, moisture content, organic matter content and heavy metal content of the marine dredged mud, remove the unit differences, and obtain standardized data;

[0038] (5) Arrange and organize the standardized data according to the sample number and parameter type, and construct a data table containing all samples and parameters;

[0039] (6) Identify the main characteristic parameters in the data table through correlation analysis, and integrate them to form a dredged mud characteristic parameter matrix.

[0040] Specifically, collect marine dredged mud samples at different depths and in different regions. When sampling, use a special mud sampler to take samples at the surface layer (0 - 30 cm), middle layer (30 - 60 cm), and bottom layer (60 - 100 cm) of the sea area dredging project respectively. At least 3 sample points should be collected in each depth area to ensure the representativeness of the samples. At the same time, record the precise coordinates of each sampling point through the GPS positioning system to form a sampling record table, which contains information such as sample number, sampling depth, longitude and latitude coordinates of the sampling point, and sampling date.

[0041] After the collection is completed, send the samples to the laboratory for physicochemical property determination. Use a laser particle size analyzer to measure the particle size of the marine dredged mud samples. Before the measurement, pre-treat the samples, including removing organic matter and dispersing treatment, and then place the treated samples in the laser particle size analyzer for measurement. This instrument uses the principle of laser diffraction and can accurately measure the particle size distribution. The measurement results obtain the D10, D50, and D90 particle size values, which respectively represent the particle sizes corresponding to the 10%, 50%, and 90% positions on the cumulative distribution curve. At the same time, record the complete particle size distribution curve, which shows the proportion distribution of different particle size particles in the dredged mud. Use the drying - weighing method to measure the moisture content of the marine dredged mud samples. The specific operation is to first weigh the wet mud, then dry it to a constant weight at 105 ± 5 °C, weigh it again, and calculate the moisture content according to the mass difference before and after drying. The organic matter content is determined by the potassium dichromate oxidation method. Mix the sample with potassium dichromate and concentrated sulfuric acid for reaction, and calculate the organic matter content based on the amount of potassium dichromate consumed in the oxidation reaction. The heavy metal content is determined by atomic absorption spectrometry. This method first digests the sample, and then uses an atomic absorption spectrometer to measure the contents of heavy metal elements such as lead, cadmium, chromium, and copper.

[0042] After obtaining the above data, standardize the particle size distribution, water content, organic matter content, and heavy metal content of marine dredged mud. The purpose of standardization is to eliminate the differences in dimensions and numerical ranges between different indicators, making each parameter comparable. The Z-score method is used for standardization, that is, subtracting the mean from the original data and then dividing by the standard deviation to convert it into a standard distribution with a mean of 0 and a standard deviation of 1. For example, for the water content data, calculate the mean and standard deviation of the water content of all samples, and then subtract the mean from the water content of each sample and divide by the standard deviation to obtain the standardized water content value. The same standardization process is also carried out for other parameters such as particle size characteristic values, organic matter content, and heavy metal content data to obtain standardized data. Organize the standardized data according to the sample number and parameter type to construct a data table containing all samples and parameters. The rows of the table represent different samples, and the columns represent different parameter indicators, including the standardized D10, D50, D90 particle size values, water content, organic matter content, various heavy metal contents, etc. This structured data table facilitates subsequent data analysis and processing.

[0043] Identify the main characteristic parameters in the data table through correlation analysis. Calculate the correlation coefficients between each parameter, find out the parameter groups with higher correlations, and the key indicators that have a significant impact on the performance of building materials. For example, the analysis may find that the D50 particle size value has a relatively high correlation with the material strength, while the content of certain heavy metals is closely related to the material durability. Integrate these important parameters to form a characteristic parameter matrix of dredged mud, which is an important basis for subsequent determination of building material types and mix design. Taking a dredging project in a certain sea area as an example, 15 dredged mud samples from different locations and depths were collected. After measurement and analysis, it was found that: the D50 value of sample 1 is 0.063 mm (fine silt grade), the water content reaches 48%, the organic matter content is 4.2%, and the lead content is 28 mg / kg; the D50 value of sample 2 is 0.125 mm (fine sand grade), the water content is 35%, the organic matter content is 2.8%, and the lead content is 18 mg / kg. After standardization, all sample data are integrated into a characteristic parameter matrix of 15 rows × 12 columns. Correlation analysis shows that the particle size distribution and organic matter content are the most critical factors affecting the performance of subsequent building materials, and these parameters are given higher weights to form a characteristic parameter matrix of dredged mud.

[0044] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0045] (1) Extract the key characteristic values from the characteristic parameter matrix of dredged mud, including the average particle size, organic matter content range, and heavy metal indicators, to form a characteristic index set of dredged mud;

[0046] (2) Construct a standard parameter library for building materials including cement concrete, blocks, permeable bricks, ceramsite, and impermeable layer materials, and record the physical and chemical index requirements for various building materials;

[0047] (3) Calculate the Euclidean distance between the dredged mud characteristic index set and the parameters of each building material type in the building materials standard parameter library to obtain a similarity score table;

[0048] (4) Normalize the values in the similarity score table, assign applicability weights to each building material type, and generate a weight matrix;

[0049] (5) Based on the weight matrix, design an analytic hierarchy process decision tree, and combine the characteristics of dredged mud, treatment difficulty, and material performance requirements to screen out building material types with high adaptability;

[0050] (6) Sort the selected building material types according to the adaptability, establish a gradient library for the applicability of dredged mud building materials, including the main applicable types and alternative types, determine the building material types suitable for dredged mud, and retain the data of alternative solutions.

[0051] Specifically, extract the key characteristic values from the dredged mud characteristic parameter matrix. This is achieved through data dimensionality reduction and importance analysis, mainly extracting parameters such as the average particle size, particle size distribution uniformity, organic matter content range, and heavy metal indicators that have the greatest impact on the performance of building materials. The average particle size directly uses the D50 value, the particle size distribution uniformity is calculated through the coefficient of non-uniformity, the organic matter content range includes the minimum value, maximum value, and average value, and the heavy metal indicators focus on the contents of common harmful elements such as lead, cadmium, and mercury. These extracted key characteristic values constitute a dredged mud characteristic index set with a lower dimension but a high information density.

[0052] Construct a standard parameter library for various building materials. This library covers common building material types such as cement concrete, blocks, permeable bricks, ceramsite, and impermeable layer materials. Each building material has corresponding raw material requirement indicators, including particle size distribution requirements, upper limit of organic matter content, heavy metal content limits, etc. These indicators are mainly formulated with reference to the national or industry standards of relevant building materials in the "Building Materials Standard Compilation" to ensure that the finally produced building materials meet the quality requirements. Taking permeable bricks as an example, its particle size requirement for raw materials is 0.15 - 4.75 mm, the organic matter content should be less than 3%, and the lead content should be less than 50 mg / kg. Digitalize and structure these indicator data to form a standard parameter library that can be processed by a computer.

[0053] After completing the construction of the above two data sets, start the similarity calculation. Calculate the Euclidean distance between the dredged mud characteristic index set and the parameters of each building material type in the building materials standard parameter library to obtain a similarity score table. The Euclidean distance calculation uses the following formula:

[0054]

[0055] Among them, D(P, Q) represents the distance value between the dredged mud feature set P and a certain building material standard parameter Q, where p a and q a represent the values of the a-th index in the feature set and the standard parameter respectively, k is the total number of feature indicators, and ω a is the weight coefficient of the a-th feature indicator. The weight coefficient is set according to the influence degree of each indicator on the building material performance. For example, the particle size distribution has a greater impact on permeable bricks, so when calculating the similarity of permeable bricks, the weight of the particle size index will be increased accordingly. The smaller the distance value, the more matching the characteristics of the dredged mud are with the requirements of the building material. In the calculated similarity score table, each column represents a building material type, each row represents a dredged mud sample, and the values in the table are the corresponding Euclidean distance values. Since the dimensions and ranges of different indicators are different, these distance values need to be normalized. The normalization process uses the following formula:

[0056]

[0057] Among them, N(d ij ) represents the normalized applicability score, d ij represents the Euclidean distance value between the i-th dredged mud sample and the j-th building material type, d max and d min represent the maximum distance value and the minimum distance value in the entire similarity score table respectively. The normalized score ranges from 0 to 100%, and the higher the score, the better the applicability. These normalized scores form a weight matrix, and each element in the matrix represents the applicability weight of a specific dredged mud sample to a specific building material type.

[0058] Based on the weight matrix, a hierarchical analysis decision tree is designed for the final screening of building material types. The factors at each level of the decision tree include: applicability weight (the first layer), processing difficulty (the second layer), and material performance requirements (the third layer). The processing difficulty considers the complexity and cost of the pretreatment process that the dredged mud needs to go through, and the material performance requirements focus on indicators such as the strength and durability of the final building material. The decision tree scores each building material type by comprehensively evaluating the factors at these three levels, and screens out the building material type with the highest fitness.

[0059] Finally, sort the screening results according to the adaptability level from high to low, and establish a gradient library for the applicability of dredged mud building materials. This library not only includes the main applicable types, but also retains the sub-optimal alternative types and their scoring data, providing a flexible selection space for actual production. For example, after the above analysis process for a certain dredged mud sample, it is finally determined that its main applicable type is ceramsite, with an adaptability score of 88 points; the alternative type is permeable brick, with an adaptability score of 76 points. This design of multi-option results enables reasonable selection between the main and alternative options according to factors such as market demand and equipment conditions in actual production, enhancing the practicality and flexibility of the method.

[0060] Taking a certain marine engineering dredged mud as an example, by extracting its characteristic index set (D50 = 0.12mm, organic matter content = 3.5%, lead content = 30mg / kg), the Euclidean distance is calculated with the standard parameters of five building materials. When calculating, a weight of 0.5 is assigned to the particle size index, a weight of 0.3 is assigned to the organic matter content, and a weight of 0.2 is assigned to the heavy metal content. The calculation results show that the Euclidean distance value between this dredged mud and ceramsite is the smallest (0.28), followed by permeable brick (0.42), and the distance from other types of building materials is relatively far. After normalization, the applicability weight of ceramsite is 85%, that of permeable brick is 65%, and others are all below 40%. Considering the relatively low treatment difficulty of ceramsite production (only simple dehydration and screening) and its moderate performance requirements, it is determined that this dredged mud is mainly suitable for ceramsite production, and the permeable brick is used as an alternative option.

[0061] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0062] (1) Perform centrifugal dehydration and drying treatment on the dredged mud, control the moisture content below 25%, and obtain pretreated dredged mud;

[0063] (2) Determine the mineral composition and chemical composition of the pretreated dredged mud through an X-ray fluorescence analyzer, and record the content data of calcium, silicon, aluminum, and iron elements;

[0064] (3) Using the pretreated dredged mud as the base material, add cement at gradients of 5%, 10%, 15%, 20%, 25%, and 30% to prepare 100mm×100mm×100mm standard cubic specimens;

[0065] (4) Cure the standard cubic specimens under the conditions of a temperature of 20±2°C and a relative humidity above 95%, and take samples for testing at 3 days, 7 days, and 28 days respectively;

[0066] (5) Apply load to the standard cubic specimens with a pressure testing machine until they are damaged, and record the compressive strength values of the specimens with different cement dosages at different ages;

[0067] (6) Pair the compressive strength values with the corresponding cement dosage data. Using the cement dosage as the abscissa and the compressive strength as the ordinate, plot the strength-cement dosage curve.

[0068] Specifically, pre-treat the dredged mud to reduce its moisture content. Marine dredged mud usually has a moisture content as high as 60 - 80%, and direct use will cause problems such as insufficient strength and large dry shrinkage of building materials. Therefore, dehydration treatment must be carried out. The pre-treatment adopts a two-stage dehydration process: first, mechanically dehydrate the dredged mud through a centrifugal dehydration device, with the rotation speed controlled at 1500 - 2000 rpm and the centrifugation time being 10 - 15 minutes. In this stage, the moisture content can be reduced to about 40%. Then, put the centrifuged dredged mud into a drying device for thermal dehydration, with the temperature controlled at 60 - 80°C to avoid the decomposition of organic matter in the dredged mud due to excessive temperature. The drying time is adjusted according to the batch and moisture content until the moisture content is reduced to less than 25%. The moisture content is detected by the sampling and drying method to ensure meeting the requirements of subsequent processing. After completing the moisture content control, determine the mineral composition and chemical composition of the pre-treated dredged mud by an X-ray fluorescence analyzer. Before the determination, grind the dredged mud sample to less than 200 mesh, press it into a disc with a diameter of 32 mm, and then put it into the analyzer for component analysis. X-ray fluorescence analysis is based on the principle that each element generates characteristic fluorescence after being excited by X-rays, and determines the content of each element by measuring the fluorescence intensity. Focus on the content of elements such as calcium, silicon, aluminum, and iron, which have important effects on the cement hydration reaction and strength development. The content data of these elements are recorded and archived as an important basis for subsequent mix design.

[0069] After obtaining the basic characteristic data of the dredged mud, start preparing test samples. Using the pre-treated dredged mud as the base material, add cement at six gradients of 5%, 10%, 15%, 20%, 25%, and 30% to form test mixes with different cement dosages. The cement used is ordinary Portland cement, and the mix design is calculated using the mass ratio. For example, for the mix with a cement dosage of 15%, add 15 kg of cement to every 100 kg of pre-treated dredged mud. According to the requirements of the standard for test methods of cement concrete, add water to stir the mixed materials (the water-cement ratio is controlled at 0.4 - 0.5), pour them into a standard cube mold of 100 mm × 100 mm × 100 mm, compact them using a vibrating table, and level the surface to prepare standard cube test blocks. Prepare 9 test blocks for each cement dosage gradient, which are respectively used for the compressive strength tests at three ages of 3 days, 7 days, and 28 days, with 3 parallel samples for each age.

[0070] The prepared standard cubic specimens shall be cured according to the specifications. Place the specimens in a standard curing room with the temperature controlled at 20±2°C and the relative humidity maintained above 95% to ensure that the cement hydration reaction proceeds fully. The curing time is divided into three periods: 3 days, 7 days, and 28 days, which represent the early strength, medium-term strength, and design strength respectively. At each specified age, take out the corresponding batch of specimens from the curing room for surface treatment and test preparation. When each age is reached, conduct a compressive strength test on the standard cubic specimens using a compression testing machine. Before the test, check the surface flatness of the specimens and perform grinding if necessary. Place the specimens in the compression testing machine and apply the load uniformly at the loading rate specified in the standard test method for cement concrete until the specimens are damaged. Record the maximum load value at the time of failure and calculate the compressive strength. For the three parallel samples of each cement dosage and age, calculate their average value as the final compressive strength value and record the standard deviation to evaluate the data reliability.

[0071] After completing the compressive strength tests for all ages and cement dosage gradients, pair and organize the compressive strength values with the corresponding cement dosage data. Using the cement dosage as the abscissa and the compressive strength as the ordinate, plot the data for the three ages of 3 days, 7 days, and 28 days respectively to form three strength-cement dosage curves. These curves intuitively show the influence law of different cement dosages on the strength development of dredged mud-based building materials, providing basic data for subsequent neural network modeling.

[0072] Taking a certain marine engineering dredged mud as an example, after pretreatment, the moisture content drops to 23%. The component analysis shows that its main components include silicon dioxide, aluminum oxide, iron oxide, and calcium oxide. Specimens are prepared according to six cement dosage gradients and cured and tested. The results of the compressive strength test at the 28-day age show that the strength is 4.2 MPa at a cement dosage of 5%, 8.5 MPa at 10%, 13.7 MPa at 15%, 19.2 MPa at 20%, 22.8 MPa at 25%, and 24.5 MPa at 30%. After plotting the strength-cement dosage curve, it is found that when the cement dosage exceeds 25%, the strength growth trend significantly slows down, indicating that the benefit of continuously increasing the cement dosage decreases.

[0073] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0074] (1) Extract the data points from the strength-cement dosage curve graph and form a training data set with the cement dosage and the corresponding compressive strength values;

[0075] (2) Combine the key parameters in the dredged mud characteristic parameter matrix to construct a neural network input feature vector, including the particle size distribution characteristics, moisture content, organic matter content, heavy metal content, and cement dosage of the dredged mud;

[0076] (3) Design a three-layer convolutional neural network structure. The first convolutional layer uses 16 3×3 convolutional kernels to extract features, the second convolutional layer uses 32 3×3 convolutional kernels to deepen the features, and the third convolutional layer uses 64 3×3 convolutional kernels to synthesize the features;

[0077] (4) Map the features output by the convolutional layer through a fully connected layer, connect to a hidden layer with 256 neurons, and then connect to a single output neuron to represent the predicted strength value;

[0078] (5) Use the training dataset to train the three-layer convolutional neural network, adopt the mean square error loss function and the Adam optimizer to adjust the network weights;

[0079] (6) Obtain the trained three-layer convolutional neural network structure through the training process. The input layer contains dredged mud parameters and admixture variables, and the output layer is the predicted strength value.

[0080] Specifically, extract the data points in the strength-cement dosage curve graph and convert them into a training dataset. The specific operation is to digitize the strength-cement dosage curve graph obtained in the previous step, and extract the compressive strength values at three ages of 3 days, 7 days, and 28 days corresponding to each cement dosage (5%, 10%, 15%, 20%, 25%, 30%) to form a structured data table. This data table contains the cement dosage as the independent variable, the compressive strength as the dependent variable, and records the age information as the conditional variable. For example, a set of data may be recorded as: cement dosage 20%, age 28 days, compressive strength 19.2 MPa. In this way, 18 data points (6 cement dosages × 3 ages) are extracted from the curve graph, and these data points constitute the basic training dataset. Combine the key parameters in the dredged mud characteristic parameter matrix to construct the input feature vector of the neural network. According to the parameters such as the particle size distribution characteristics (D10, D50, D90 values), moisture content, organic matter content, and heavy metal content of the dredged mud, these parameters are closely related to the cement hydration reaction and the final strength development. In order to construct a comprehensive input feature vector, integrate these parameters with the cement dosage to form a multi-dimensional feature vector. Among them, the input feature vector includes: dredged mud particle size distribution characteristics (3 values), moisture content (1 value), organic matter content (1 value), heavy metal content (usually 3-5 values according to the detected metal types), cement dosage (1 value), and age (1 value), totaling about 10-12 characteristic parameters. These parameters are used as the input layer of the neural network after being standardized.

[0081] Based on the complexity of the dimension and intensity prediction tasks of the input feature vectors, a three-layer convolutional neural network structure is designed. Convolutional neural networks perform excellently in processing data with spatial correlation, and there is a certain correlation among the various characteristic parameters of dredged mud. Therefore, a convolutional neural network is selected for feature extraction and intensity prediction. The first convolutional layer uses 16 3×3 convolutional kernels. The main function of this layer is to extract low-level feature patterns from the original input features. The convolutional kernel size of 3×3 is a commonly used size in computer vision and is suitable for capturing local feature associations. The second convolutional layer uses 32 3×3 convolutional kernels to further deepen the feature representation on the basis of the basic features extracted in the first layer, increasing the degree of feature abstraction. The third convolutional layer uses 64 3×3 convolutional kernels for higher-level feature synthesis and abstraction, thus generating a feature representation highly relevant to the prediction task. After each convolutional operation, an activation function (such as ReLU) is connected to introduce non-linear capabilities, and at the same time batch normalization is used to reduce internal covariate shift and accelerate the training process.

[0082] The features extracted by the convolutional layer need to be mapped through a fully connected layer to be converted into intensity prediction values. After flattening the output of the third convolutional layer, a hidden layer containing 256 neurons is connected. This hidden layer has sufficient capacity to represent complex non-linear mapping relationships. The hidden layer uses the ReLU activation function to ensure that the network has sufficient non-linear expression capabilities. After the hidden layer, it is connected to a single output neuron, which uses a linear activation function (or no activation function) to directly output the predicted compressive strength value. This mapping structure from multi-dimensional features to a single intensity value can effectively capture the comprehensive influence of each input parameter on the intensity.

[0083] After building the network structure, the network is trained using the training dataset prepared previously. The training adopts the supervised learning method, with the input feature vector as the network input and the actually measured compressive strength as the label. The mean squared error (MSE) is selected as the loss function. It calculates the squared difference between the predicted strength value and the actual strength value and is a commonly used loss function in regression problems. The Adam (Adaptive Moment Estimation) optimizer is selected. This is an optimization algorithm with an adaptive learning rate that combines the advantages of momentum and RMSProp, can effectively handle the problem of sparse gradients, and accelerate the convergence process. During the training process, the network weights are gradually adjusted with gradient descent until the loss function reaches a stable and low value, indicating that the network has learned the mapping relationship between the input features and the strength well. After sufficient training, the final three-layer convolutional neural network model is obtained. This model can predict the compressive strength under different conditions based on the characteristic parameters of the dredged mud and the amount of cement used. To verify the performance of the model, usually a part of the data is reserved as the test set, the prediction error of the model on the test set is calculated, and the generalization ability of the model is evaluated. Taking a certain marine dredged mud as an example, the average relative error of the neural network model established using the above method on the test set is controlled within 5%, indicating that the model has good prediction accuracy. This trained neural network becomes the core tool for the mix ratio optimization design in the subsequent steps, can quickly predict the material strength under different mix ratio parameters, avoids a large amount of experimental work, and greatly improves the efficiency and accuracy of the mix ratio design.

[0084] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0085] (1) Set the range of cement dosage to be 5%-35%, the range of aggregate ratio to be 40%-70%, and the range of admixture ratio to be 5%-15%, and construct a mix ratio parameter matrix;

[0086] (2) Divide the mix ratio parameter matrix into multiple mix ratio points at a step size of 1% to generate a mix ratio candidate set, and each mix ratio point contains specific values of cement, aggregate, and admixture ratios;

[0087] (3) Sequentially extract mix ratio points from the mix ratio candidate set, and combine them with the dredged mud parameters to assemble into an input data vector;

[0088] (4) Input the input data vector into the convolutional neural network to obtain the corresponding strength prediction output value, and record the corresponding relationship between the mix ratio point and the strength value;

[0089] (5) According to the strength prediction output value, adjust the direction of the mix ratio parameters, gradually explore towards the high-strength region, generate new mix ratio points and record their strength prediction values;

[0090] (6) Organize all the tested mix ratio points and their strength prediction values into a structured data table, establish an index according to the mix ratio number and strength value, and form a mix ratio - strength database.

[0091] Specifically, according to the building material design specifications, set the reasonable range of mix ratio parameters. Among them, the range of cement dosage is set at 5% - 35%. The lower limit of this range is based on the need to ensure the minimum bonding strength, and the upper limit takes into account cost control and the possible dry shrinkage problems caused by excessive cement dosage; the range of aggregate proportion is set at 40% - 70%, which is based on the basic requirement of aggregate as the skeleton material in building materials; the range of admixture proportion is set at 5% - 15%, mainly considering the influence of admixtures on workability, strength development and durability. Through these three parameter ranges, a three - dimensional mix ratio parameter matrix is constructed, and each dimension of this matrix corresponds to cement dosage, aggregate proportion and admixture proportion respectively.

[0092] To ensure the refinement of mix ratio design, divide the mix ratio parameter matrix by a step size of 1%. That is, for cement dosage, starting from 5%, take a point every 1% until 35%, a total of 31 points; for aggregate proportion, starting from 40%, take a point every 1% until 70%, a total of 31 points; for admixture proportion, starting from 5%, take a point every 1% until 15%, a total of 11 points. This division method will theoretically generate 31×31×11 = 10,571 mix ratio points, and each mix ratio point contains specific values of cement, aggregate and admixture proportions. However, considering that the sum of the proportions of each component in the mix ratio must be 100%, a constraint condition needs to be added: cement dosage + aggregate proportion + admixture proportion + dredged mud proportion = 100%. Filter out the mix ratio points that meet the conditions based on this to form the final mix ratio candidate set. This step is essentially to discretize the continuous mix ratio parameter space for subsequent systematic strength prediction and optimization.

[0093] Extract mix ratio points from the generated mix ratio candidate set in a certain preset order (such as increasing cement dosage). For each extracted mix ratio point, it is necessary to combine it with the characteristic parameters of the dredged mud to assemble a complete input data vector. This vector not only contains mix ratio information (cement dosage, aggregate proportion, admixture proportion), but also contains the dredged mud characteristic parameters obtained in the previous steps (such as particle size distribution, water content, organic matter content, heavy metal content, etc.). Thus, a complete input data vector may contain 10 - 15 parameters, comprehensively describing the material composition and basic characteristics.

[0094] The assembled input data vector is fed into the trained convolutional neural network, and the network directly outputs the corresponding strength prediction value. This process is called forward propagation. Among them, the input data first undergoes feature extraction and transformation through three convolutional layers, and then is mapped to the final strength prediction value through the fully connected layer. For each mixing ratio point, its formulation parameters and the corresponding strength prediction value are recorded to establish the correspondence between the mixing ratio point and the strength value. This method avoids the preparation and testing of a large number of experimental specimens and greatly improves the efficiency of formulation design.

[0095] Based on the preliminary strength prediction results, the formulation parameters are adjusted directionally to explore higher strength regions. This adjustment process follows the gradient ascent strategy, and the calculation formula is:

[0096]

[0097] where, P new represents the new formulation parameter vector, P old represents the current formulation parameter vector, η is the step size coefficient, represents the gradient of the strength function at the current mixing ratio point. The gradient calculation is approximately realized by the finite difference method:

[0098]

[0099] where, δ is a small perturbation value (usually taken as 0.5%-1%), e1, e2, and e3 are the unit vectors in the directions of cement dosage, aggregate ratio, and admixture ratio respectively, and S(P) represents the strength prediction value corresponding to the formulation parameter vector P. By calculating the strength change rates in all directions around a specific mixing ratio point, the direction with the fastest strength growth is determined, and a new mixing ratio point is generated along this direction. The neural network is continued to be used to predict the strength of the newly generated mixing ratio point, and the prediction results are recorded. Such iteration is carried out until the preset termination condition is reached (such as the strength growth rate is lower than the threshold).

[0100] All the tested mixing ratio points and their strength prediction values are organized into a structured data table. This data table contains multiple fields: mixing ratio number, cement dosage, aggregate ratio, admixture ratio, dredged mud ratio, predicted strength value, etc. A main index is established according to the mixing ratio number, and an auxiliary index of the strength value is created at the same time to facilitate querying and analyzing data from different perspectives. The formed mixing ratio-strength database becomes the data basis for subsequent screening of the optimal mixing ratio.

[0101] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0102] (1) Sort the mixing ratio-strength database in descending order according to the strength prediction value, extract the first three groups of formulation parameters with the highest strength, and mark them as the preferred mixing ratio groups;

[0103] (2) Weigh the corresponding weights of the pretreated dredged mud, cement, aggregate, and admixture respectively according to the parameter ratios in the preferred ratio group to prepare three groups of physical test samples;

[0104] (3) Cure the three groups of physical test samples in a standard environment with a temperature of 20 ± 2°C and a relative humidity above 95%, and take samples for strength testing at 3 days, 7 days, and 28 days respectively;

[0105] (4) Record the actual compressive strength values of the three groups of physical test samples at each age, compare them with the predicted strength values of the convolutional neural network, and calculate the prediction error;

[0106] (5) Based on the actual compressive strength values and prediction errors, comprehensively evaluate the three groups of ratios, taking into account factors such as 28-day strength, strength growth rate, and prediction accuracy;

[0107] (6) Select the ratio with the highest comprehensive score from the three groups of ratios to form a standardized ratio table and determine the building material ratio data.

[0108] Specifically, through the database query statement, all the records in the ratio-strength database are sorted in descending order according to the predicted strength value, so that the ratio with the highest strength will be ranked at the top. From the sorted database, the top three groups of ratio parameters with the highest strength prediction values are extracted. These three groups of ratio parameters represent the ratio schemes that are most likely to achieve high strength based on the neural network prediction results. Mark these three groups of ratio parameters as the preferred ratio groups, and accurately record parameters such as the cement dosage, aggregate ratio, admixture ratio, and dredged mud ratio. This method of screening preferred ratios based on strength prediction results avoids a large number of trial-and-error processes in traditional ratio design and improves the design efficiency. According to the parameter ratios in the preferred ratio groups, physical test samples are prepared. For each group of ratios, the corresponding weights of pre-treated dredged mud, cement, aggregate, and admixture are accurately weighed according to their parameter ratios. For example, for a test batch with a total weight of 10 kg, if the ratio is 20% cement, 58% aggregate, 10% admixture, and 12% dredged mud, then 2 kg of cement, 5.8 kg of aggregate, 1 kg of admixture, and 1.2 kg of pre-treated dredged mud need to be weighed. After the material weighing is completed, stirring and molding are carried out according to the standard process for preparing building materials. Among them, first, the dry powder materials (cement, admixture) are mixed evenly, then the pre-treated dredged mud and aggregate are added and mixed, and finally water is added to adjust to the appropriate workability. The entire stirring process usually takes 3 - 5 minutes to ensure uniform mixing of the materials. The mixed materials are poured into a standard cube mold of 100 mm×100 mm×100 mm, vibrated and compacted on a vibrating table, and the surface is leveled to complete the preparation of three groups of physical test samples. At least 9 test blocks are prepared for each group of ratios, which are respectively used for strength tests at three ages of 3 days, 7 days, and 28 days. The three groups of prepared physical test samples are placed in a standard curing room for curing. The curing conditions are strictly controlled in a standard environment with a temperature of 20±2°C and a relative humidity of more than 95%. Such environmental conditions are conducive to the normal progress of the cement hydration reaction. During the curing process, the corresponding test blocks are taken out at 3 days, 7 days, and 28 days for strength tests. Before the strength test, the surface of the test block needs to be treated to ensure that the loading surface is flat, and then a pressure testing machine is used to apply a load to the test block at a specified loading rate until it fails, record the failure load value, and calculate the compressive strength.

[0109] Detailedly record the actual compressive strength values of three groups of physical test samples at each age. Meanwhile, compare these actually measured strength values with the predicted strength values of the convolutional neural network under the same mixing ratio conditions. Calculate the prediction error for each mixing ratio at each age. The calculation method of the prediction error is the difference between the actual strength value and the predicted strength value divided by the actual strength value, and then multiplied by 100% to obtain the relative error percentage. Through the magnitude of the prediction error, the prediction accuracy of the neural network model for different mixing ratios can be evaluated, and this index will also be an important reference factor for subsequent comprehensive scoring. Based on the actually measured compressive strength values and prediction errors, conduct a comprehensive scoring for the three groups of mixing ratios. The comprehensive scoring considers multiple factors: First is the 28-day strength value, which is the most important performance index of building materials; second is the strength growth rate. Calculate the strength development curve through the strength values at three time points of 3 days, 7 days, and 28 days to evaluate the early strength and strength growth potential of the material; third is the prediction accuracy, that is, the degree of coincidence between the actual strength and the predicted strength. This index reflects the reliability of the mixing ratio design method. Assign weights to these three factors (such as the 28-day strength accounting for 60%, the strength growth rate accounting for 25%, and the prediction accuracy accounting for 15%), calculate the weighted total score, and obtain the comprehensive score of each group of mixing ratios.

[0110] Select the mixing ratio with the highest comprehensive score from the three groups of mixing ratios and determine it as the final building material mixing ratio. Organize the parameters of this mixing ratio (cement dosage, aggregate ratio, admixture ratio, dredged mud ratio) in a standard format to form a standardized mixing ratio table, and record the performance parameters corresponding to the mixing ratio (such as expected strength, density, cement dosage, etc.) to complete the determination of the building material mixing ratio data. This method of determining the final mixing ratio based on actual test results not only utilizes the efficient screening ability of the neural network model but also ensures the reliability of the results through physical verification. It is a mixing ratio design method that combines theory and practice.

[0111] The above described the mixing ratio design method for making building materials based on marine dredged mud in the embodiments of the present application. Next, the mixing ratio design system for making building materials based on marine dredged mud in the embodiments of the present application will be described. Please refer to Figure 2 In an embodiment, the mixing ratio design system for making building materials based on marine dredged mud in the embodiments of the present application includes:

[0112] A collection module, used to collect marine dredged mud samples, measure the particle size distribution, moisture content, organic matter content, and heavy metal content of the marine dredged mud samples, and generate a dredged mud characteristic parameter matrix;

[0113] A calculation module, used to calculate the similarity between the dredged mud characteristic parameter matrix and the building material standard parameter library to determine the building material types suitable for the dredged mud;

[0114] A processing module for preprocessing the dredged mud, measuring the compressive strength values under different cement dosages, and plotting a strength-cement dosage curve graph;

[0115] A construction module for constructing a three-layer convolutional neural network structure based on the strength-cement dosage curve graph, with the input layer including dredged mud parameters and admixture variables, and the output layer being the predicted strength value;

[0116] A recording module for gradually adjusting the mixing ratio parameters of cement, aggregate, and admixture through the convolutional neural network, recording the strength prediction results of each iteration, and forming a mixing ratio-strength database;

[0117] A mixing ratio module for screening the top three groups of mixing ratios with the highest strength from the mixing ratio-strength database, preparing physical samples and conducting 28-day strength tests to determine the building material mixing ratio data.

[0118] Through the collaborative cooperation of the above various components, and through a systematic method for generating the characteristic parameter matrix of dredged mud, the key characteristic indicators of dredged mud are comprehensively collected and quantified; secondly, the intelligent matching between dredged mud and building material types is realized by means of similarity calculation technology, avoiding blind trial and error and improving the resource allocation efficiency; thirdly, through the construction of the strength-cement dosage curve graph, the correlation law between cement dosage and strength development is revealed, providing a directional guidance for mixing ratio optimization; fourthly, applying a three-layer convolutional neural network structure to process multi-dimensional feature data gives full play to the algorithm advantages of the convolutional neural network in non-linear mapping and feature extraction, enabling the model to accurately capture the complex relationship between the characteristics of dredged mud, mixing ratio parameters, and strength. This mixing ratio prediction model based on deep learning has stronger generalization ability and prediction accuracy compared with traditional statistical methods; fifthly, through the iterative optimization mechanism of the neural network, the efficient adjustment and optimization of mixing ratio parameters are realized, greatly reducing the experimental workload and saving time and cost; sixthly, by integrating the results of physical sample preparation and strength tests, a method system combining theoretical prediction and actual verification is established to ensure the reliability and practicality of the mixing ratio design results. Especially in the application of the three-layer convolutional neural network, this method fully considers the matching between algorithm characteristics and actual application requirements. By reasonably designing the network structure and parameters, the model can not only process the high-dimensional characteristic parameters of dredged mud but also accurately predict the strength performance under different mixing ratios, solving the problem of high-dimensional non-linear relationship modeling that is difficult to handle by traditional methods.

[0119] Referring to Figure 3 In the embodiment of the present invention, a computer device is further provided. This computer device can be a server, and its internal structure can be as Figure 3As shown in the figure. The computer device includes a processor, a memory, a display screen, an input device, a network interface, and a database connected through a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0120] Those skilled in the art can understand that Figure 3 the structure shown in the figure is only a block diagram of some structures related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0121] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0122] Those of ordinary skill in the art can understand that all or part of the processes in the above embodiment methods can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to memory, storage, database, or other media provided by the present invention and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0123] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0124] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0125] As described above, the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A proportioning design method for making building materials based on marine dredged mud, characterized in that The mix design method for making building materials based on marine dredged mud includes: Collect marine dredged mud samples, measure the particle size distribution, water content, organic matter content and heavy metal content of the marine dredged mud samples, and generate a dredged mud characteristic parameter matrix; Calculate the similarity between the dredged mud characteristic parameter matrix and the building material standard parameter library to determine the applicable building material types for the dredged mud; Pretreat the dredged mud, measure the compressive strength values under different cement dosages, and draw a strength-cement dosage curve; Based on the strength-cement dosage curve, construct a three-layer convolutional neural network structure. The input layer includes dredged mud parameters and admixture variables, and the output layer is the predicted strength value; Gradually adjust the mix ratio parameters of cement, aggregate and admixture through the convolutional neural network, record the strength prediction results of each iteration, and form a mix ratio-strength database; Select the top three groups of mix ratios with the highest strength from the mix ratio-strength database, prepare physical samples and conduct 28-day strength tests to determine the building material mix ratio data.

2. The mixing ratio design method for making building materials based on marine dredged mud according to claim 1, characterized in that, The step of collecting marine dredged mud samples, measuring the particle size distribution, water content, organic matter content and heavy metal content of the marine dredged mud samples, and generating a dredged mud characteristic parameter matrix includes: Collect marine dredged mud samples at different depths and different regions, record the position data of the sampling points through a positioning system, and form a sampling record form; Use a laser particle size analyzer to measure the particle size of the marine dredged mud samples, obtain the D10, D50, D90 particle size values and distribution curves, and record the particle size distribution of the marine dredged mud; Use the drying-weight method to measure the water content of the marine dredged mud samples, use the potassium dichromate oxidation method to measure the organic matter content, and use the atomic absorption spectrometry method to measure the heavy metal content; Standardize the particle size distribution, water content, organic matter content and heavy metal content of the marine dredged mud, remove the unit differences, and obtain standardized data; Arrange and organize the standardized data according to the sample number and parameter type, and construct a data table containing all samples and parameters; Identify the main characteristic parameters in the data table through correlation analysis, and integrate them to form the dredged mud characteristic parameter matrix.

3. The design method of the proportion for making building materials based on marine dredged mud according to claim 1, characterized in that, The step of calculating the similarity between the dredged mud characteristic parameter matrix and the building material standard parameter library to determine the applicable building material types for the dredged mud includes: Extract key characteristic values from the dredged mud characteristic parameter matrix, including the average particle size, organic matter content range and heavy metal index, to form a dredged mud characteristic index set; Construct a building material standard parameter library containing cement concrete, blocks, permeable bricks, ceramsite and impermeable layer materials, and record the physical and chemical index requirements of various building materials; Calculate the Euclidean distance between the dredged mud characteristic index set and the parameters of each building material type in the building material standard parameter library to obtain a similarity score table; Normalize the values in the similarity score table, assign applicability weights to each building material type, and generate a weight matrix; Based on the weight matrix, design an analytic hierarchy process decision tree, and combine the dredged mud characteristics, treatment difficulty and material performance requirements to screen the building material types with high adaptability; Sort the selected building material types according to the adaptability, establish a gradient library for the applicability of dredged mud building materials, including the main applicable types and alternative types, determine the building material types suitable for dredged mud and retain the data of alternative solutions.

4. The proportioning design method for making building materials based on marine dredged mud according to claim 1, characterized in that, Perform pretreatment on the dredged mud, measure the compressive strength values under different cement dosages, and draw a strength-cement dosage curve, including: Perform centrifugal dehydration and drying on the dredged mud to control the moisture content below 25% to obtain pretreated dredged mud; Determine the mineral composition and chemical composition of the pretreated dredged mud by means of an X-ray fluorescence analyzer, and record the content data of calcium, silicon, aluminum, and iron elements; Using the pretreated dredged mud as the base material, add cement in gradients of 5%, 10%, 15%, 20%, 25%, and 30% to prepare 100mm×100mm×100mm standard cubic specimens; Cure the standard cubic specimens under the conditions of a temperature of 20±2°C and a relative humidity above 95%, and take samples for testing at 3 days, 7 days, and 28 days respectively; Apply a load to the standard cubic specimens using a pressure testing machine until they are damaged, and record the compressive strength values of the specimens with different cement dosages at different ages; Pair the compressive strength values with the corresponding cement dosage data, and draw a strength-cement dosage curve with the cement dosage as the abscissa and the compressive strength as the ordinate.

5. The mixing ratio design method for making building materials based on marine dredged mud according to claim 1, characterized in that Based on the strength-cement dosage curve, construct a three-layer convolutional neural network structure, with the input layer including dredged mud parameters and admixture variables, and the output layer being the predicted strength value, including: Extract the data points from the strength-cement dosage curve, and form a training data set with the cement dosage and the corresponding compressive strength values; Combine the key parameters in the characteristic parameter matrix of the dredged mud to construct a neural network input feature vector, including the particle size distribution characteristics, moisture content, organic matter content, heavy metal content, and cement dosage of the dredged mud; Design a three-layer convolutional neural network structure. The first convolutional layer uses 16 3×3 convolutional kernels to extract features, the second convolutional layer uses 32 3×3 convolutional kernels to deepen the features, and the third convolutional layer uses 64 3×3 convolutional kernels to comprehensively integrate the features; Map the output features of the convolutional layer through a fully connected layer, connect to a hidden layer with 256 neurons, and then connect to a single output neuron to represent the predicted strength value; Use the training data set to train the three-layer convolutional neural network, adopt the mean square error loss function and the Adam optimizer to adjust the network weights; Through the training process, obtain a trained three-layer convolutional neural network structure, with the input layer including dredged mud parameters and admixture variables, and the output layer being the predicted strength value.

6. The proportion design method for manufacturing building materials based on marine dredged mud according to claim 1, characterized in that, Gradually adjust the mixing ratio parameters of cement, aggregate, and admixture through the convolutional neural network, record the strength prediction results of each iteration, and form a mixing ratio-strength database, including: Set the cement dosage range at 5%-35%, the aggregate ratio range at 40%-70%, and the admixture ratio range at 5%-15% to construct a mixing ratio parameter matrix; Divide the mixing ratio parameter matrix into multiple mixing ratio points at a step size of 1% to generate a mixing ratio candidate set, and each mixing ratio point contains specific values of the cement, aggregate, and admixture ratios; Sequentially extract the proportion points from the proportion candidate set, and combine them with the dredged mud parameters to assemble an input data vector; Input the input data vector into the convolutional neural network to obtain the corresponding strength prediction output value, and record the corresponding relationship between the proportion point and the strength value; According to the strength prediction output value, adjust the direction of the proportion parameters, gradually explore towards the high-strength area, generate new proportion points and record their strength prediction values; Organize all the tested proportion points and their strength prediction values into a structured data table, and establish an index according to the proportion number and strength value to form a proportion-strength database.

7. The proportion design method for making building materials based on marine dredged mud according to claim 6, characterized in that, Screen the top three groups of proportions with the highest strength from the proportion-strength database, prepare physical samples and conduct 28-day strength tests to determine the building material proportion data, including: Sort the proportion-strength database in descending order of the strength prediction value, extract the parameters of the top three groups of proportions with the highest strength, and mark them as the preferred proportion groups; Weigh the corresponding weights of the pretreated dredged mud, cement, aggregate and admixture according to the parameter ratios in the preferred proportion groups to prepare three groups of physical test samples; Cure the three groups of physical test samples in a standard environment with a temperature of 20±2°C and a relative humidity above 95%, and take samples for strength tests at 3 days, 7 days and 28 days respectively; Record the actual compressive strength values of the three groups of physical test samples at each age, compare them with the predicted strength values of the convolutional neural network, and calculate the prediction error; Based on the actual compressive strength value and the prediction error, comprehensively score the three groups of proportions, considering factors such as 28-day strength, strength growth rate and prediction accuracy; Select the proportion with the highest comprehensive score from the three groups of proportions to form a standardized proportion table and determine the building material proportion data.

8. A proportioning design system for making building materials based on marine dredged mud, which is used to implement the proportioning design method for making building materials based on marine dredged mud as described in any one of claims 1-7, characterized in that, The proportion design system for making building materials based on marine dredged mud includes: A collection module for collecting marine dredged mud samples, measuring the particle size distribution, moisture content, organic matter content and heavy metal content of the marine dredged mud samples, and generating a dredged mud characteristic parameter matrix; A calculation module for calculating the similarity between the dredged mud characteristic parameter matrix and the building material standard parameter library to determine the building material types suitable for the dredged mud; A processing module for preprocessing the dredged mud, measuring the compressive strength values under different cement dosages, and drawing a strength-cement dosage curve; A construction module for constructing a three-layer convolutional neural network structure based on the strength-cement dosage curve, with the input layer including dredged mud parameters and admixture variables, and the output layer being the predicted strength value; A recording module for gradually adjusting the proportion parameters of cement, aggregate and admixture through the convolutional neural network, recording the strength prediction results of each iteration, and forming a proportion-strength database; A proportion module for screening the top three groups of proportions with the highest strength from the proportion-strength database, preparing physical samples and conducting 28-day strength tests to determine the building material proportion data.

9. A computer device, characterized in that, It includes a memory and a processor. The memory stores a computer program that can run on the processor. It is characterized in that when the processor executes the computer program, it implements the proportion design method for making building materials based on marine dredged mud according to any one of claims 1 to 7.

10. A computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, the processor is caused to execute the proportion design method for making building materials based on marine dredged mud according to any one of claims 1 to 7.

Images