A method and system for designing mix proportions of cement emulsified asphalt mortar for high-speed railways

Abstract

This application relates to a method and system for designing the mix proportion of cement emulsified asphalt mortar for high-speed railways. The method includes the following steps: inputting design factors of the cement emulsified asphalt mortar to be designed into a design model; the design factors include raw material types, quantities, and environmental parameters; predicting and outputting performance indicators corresponding to the cement emulsified asphalt mortar to be designed through the design model; the performance indicators include fluidity, spreadability, working time, compressive strength, elastic modulus, and fatigue life; adjusting the design factors of the cement emulsified asphalt mortar to be designed according to the performance indicators and inputting them into the design model to obtain the performance indicators corresponding to the adjusted design factors; repeating the input and adjustment steps multiple times until the optimal design factors that meet the performance requirements are obtained and output. The system and equipment are used to execute the above method. This application has the advantages of high design efficiency, short cycle, low cost, and high accuracy.

Description

Technical Field

[0001] This application relates to the technical field of civil engineering construction, specifically to a method and system for designing the mix proportion of cement emulsified asphalt mortar for high-speed railways. Background Technology

[0002] Cement emulsified asphalt mortar (CA mortar) is one of the key structural layers of CRTSⅠ and CRTSⅡ type slab track for high-speed railways. It is a viscoelastic organic-inorganic composite material formed by mechanically mixing emulsified asphalt, cement, fine aggregate (sand), water, and appropriate additives, and through the combined effects of cement hydration hardening and asphalt demulsification. It functions as a support, load transfer, vibration damping, and noise reduction material. Due to the complex service environment of my country's high-speed railways, the mechanical properties of cement emulsified asphalt mortar vary, with different mortars used in warm, cold, and frigid regions. Furthermore, different mix designs are required to address requirements for frost resistance, fatigue resistance, and corrosion resistance.

[0003] Traditional methods for designing mix proportions of cement emulsified asphalt mortar mainly rely on experiments and experience, which have problems such as long design and testing cycles, high costs, and insufficient accuracy. Summary of the Invention

[0004] To address the problems existing in the prior art, this application aims to provide a method and system for designing the mix proportions of cement emulsified asphalt mortar for high-speed railways. This application offers advantages such as high design efficiency, short cycle time, low cost, and high accuracy.

[0005] The mix design method for high-speed railway cement emulsified asphalt mortar described in this application includes the following steps:

[0006] Design factor input: Input the design factors of the cement emulsified asphalt mortar to be designed into the design model; the design factors include raw material type, dosage and layout environment parameters;

[0007] Predicted output: The design model predicts and outputs the performance indicators corresponding to the cement emulsified asphalt mortar to be designed; the performance indicators include fluidity, spreadability, working time, compressive strength, elastic modulus and fatigue life;

[0008] Input adjustment: Based on the performance indicators, adjust the design factors of the cement emulsified asphalt mortar to be designed and input them into the design model to obtain the performance indicators corresponding to the adjusted design factors;

[0009] Optimal design output: Repeat the input adjustment steps multiple times until the optimal design factors that meet the performance requirements are obtained and output.

[0010] Preferably, the design model is obtained through the following steps:

[0011] Data preparation: Obtain multiple sets of test data for cement emulsified asphalt mortar, including the design factors and corresponding performance indicators of cement emulsified asphalt mortar;

[0012] Data preprocessing: The experimental data from multiple sets are cleaned and normalized, converted into a two-dimensional matrix, and divided into training set, validation set and test set.

[0013] Model construction: A convolutional neural network was selected as the design model.

[0014] Model training: The constructed design model is trained using the experimental data of the training set until the preset convergence condition or number of iterations is reached;

[0015] Model optimization: The trained design model is validated and tested using the experimental data from the validation and test sets, and the design model is optimized based on the validation and test results until the performance of the design model meets the requirements, thus obtaining the design model.

[0016] Preferably, the step of adjusting the input includes, according to the performance indicators, adjusting the design factors of the cement emulsified asphalt mortar to be designed and inputting them into the design model, including:

[0017] Preset standard values ​​for each performance indicator, and calculate the fitness value Va of the performance indicator according to the following formula. fit :

[0018] Va fit =w1*(Va Fa / St Fa )+w2*(Va Spr / St Spr )+w3*(Va W_t /

[0019] St W_t )+w4*(Va C_s / St C_s )+w5*(Va E_m / St E_m )+w6*(Va F_l / St F_l );

[0020] Among them, Va Fa St represents the fluidity. Fa Va represents the standard value of fluidity. Spr Indicates extensibility, St Spr Va represents the standard value of scalability. W_t Indicates available working hours, StW_t Va represents the standard value for working time. C_s St represents compressive strength. C_s Va represents the standard value of compressive strength. E_m St represents the elastic modulus. E_m Va represents the standard value of the elastic modulus. F_l Indicates fatigue life, St F_l The standard value of fatigue life is represented by w1, w2, w3, w4, w5 and w6, which represent the first weight, second weight, third weight, fourth weight, fifth weight and sixth weight corresponding to the fluidity, spreadability, working time, compressive strength, elastic modulus and fatigue life, respectively.

[0021] Preset about the fitness value Va fit The fitness standard range is defined as follows: if all performance indicators are greater than the corresponding standard values, and the fitness value Va is within the range specified in the standard range, then... fit If the performance index is within the fitness standard range, the design factor corresponding to the performance index is determined to be the optimal design factor. If it is not satisfied, the performance index that has not reached the standard value is selected, the performance index that has not reached the standard value is analyzed, the design factors that affect the performance index that has not reached the standard value are obtained, and adjustments are made.

[0022] Preferably, the initial weight values ​​of the first weight, second weight, third weight, fourth weight, fifth weight, and sixth weight are configured respectively, and the fitness value Va is calculated. fit At that time, the weight adjustment factor is obtained, and the weights of each performance index are adjusted according to the weight adjustment factor and the preset weight adjustment strategy. The weight adjustment factor includes the terrain of the high-speed railway layout and the maximum design speed of the train.

[0023] The weight adjustment strategy includes:

[0024] In response to the fact that the high-speed railway is located in a plain, the first and third weights are increased, and the fourth and fifth weights are decreased.

[0025] In response to the fact that the high-speed railway is located in a hilly terrain, the fourth and fifth weights are increased, while the first and second weights are decreased.

[0026] In response to the train's maximum design speed being greater than or equal to a preset speed threshold, the fifth and sixth weights are increased, while the first, second, and third weights are decreased.

[0027] In response to the train's maximum design speed being less than the speed threshold, the third and fourth weights are increased, while the fifth and sixth weights are decreased.

[0028] Preferably, in the step of constructing the model, the type and amount of raw materials are defined as adjustable factors, and the environmental parameters are defined as non-adjustable factors. The environmental parameters include the number of days of immersion, the number of days of freeze-thaw cycles, the construction environment temperature, and the application environment temperature coefficient.

[0029] A correlation analysis is performed between the deployment environment parameters and each of the performance indicators. For each deployment environment parameter, each of the performance indicators is traversed to obtain all performance indicators that are related to the deployment environment parameter. A correspondence relationship is established between the deployment environment parameter and the related performance indicators and recorded as a correlation group.

[0030] The design model includes sequentially cascaded input layers, convolutional pooling layers, fully connected layers, and output layers. The input layer includes input nodes in a number equal to the number of items of the design factors. A separate branch network is created for each association group, and the outputs of multiple branch networks are merged in the fully connected layer.

[0031] Preferably, the construction ambient temperature is obtained through the following steps:

[0032] Obtain the construction period for the target road section through the construction plan.

[0033] Historical ambient temperature data of the target road section is obtained. The ambient temperature during the construction period is predicted based on the historical ambient temperature data. The average value of the predicted ambient temperature during the construction period is calculated, which is the construction ambient temperature.

[0034] The application environment temperature coefficient is obtained through the following steps:

[0035] Historical ambient temperature data of the target road segment within a period is obtained, and the highest temperature, lowest temperature and maximum temperature difference per unit time of the target road segment within the period are analyzed. The highest temperature, lowest temperature and maximum temperature difference per unit time are weighted and calculated to obtain the application ambient temperature coefficient.

[0036] Preferably, the input adjustment and optimal design output include the following steps:

[0037] All possible design factors for high-speed railway cement emulsified asphalt mortar are obtained, and a genetic algorithm is used to initialize the population, with each individual in the population corresponding to a set of possible design factors;

[0038] Each individual in the population is input into the design model to obtain the performance index corresponding to each individual, and the fitness value Va corresponding to each individual is further calculated. fit ;

[0039] A correlation analysis is performed on the adjustable factors and various performance indicators to obtain the adjustable factors that have a strong correlation with each performance indicator and to establish a correlation adjustment relationship. The correlation results are obtained. The strong correlation relationship is indicated by the correlation analysis result between the adjustable factors and the performance indicators being no less than a preset correlation degree threshold. The correlation results include the performance indicator, the adjustable factors that have a strong correlation with the performance indicator, and the correlation between the performance indicator and the adjustable factors that have a strong correlation.

[0040] Individuals with fitness values ​​greater than a preset screening threshold in the population are selected as parents. Performance indicators of the selected parents that do not meet the standard value are designated as unqualified indicators. In the crossover operation, adjustable factors that have a strong correlation with the unqualified indicators are selected as mutation objects. The direction of mutation is determined based on the correlation between the unqualified indicators and the adjustable factors, and the crossover operation is performed to generate multiple offspring individuals to form a new population. The new population is used to replace the initial population. This process is repeated iteratively multiple times until the optimal design factors that meet the energy requirements are obtained and output.

[0041] This application discloses a high-speed railway cement emulsified asphalt mortar mix design system, comprising:

[0042] The input module is used to input the design factors of the cement emulsified asphalt mortar to be designed into the design model; the design factors include the type and amount of raw materials and the environmental parameters of the layout.

[0043] The prediction module is used to predict and output the performance indicators corresponding to the cement emulsified asphalt mortar to be designed through the design model; the performance indicators include fluidity, spreadability, working time, compressive strength, elastic modulus and fatigue life.

[0044] The input adjustment module is used to adjust the design factors of the cement emulsified asphalt mortar to be designed according to the performance index and input them into the design model to obtain the performance index corresponding to the adjusted design factors. This process is repeated multiple times until the optimal design factors that meet the performance requirements are obtained.

[0045] The output module is used to output the optimal design factors.

[0046] A computer device according to this application includes a processor and a memory connected by a signal. The memory stores at least one instruction or at least one program. When the at least one instruction or the at least one program is loaded by the processor, it executes the high-speed railway cement emulsified asphalt mortar mix design method as described above.

[0047] This application discloses a computer-readable storage medium storing at least one instruction or at least one program, which, when loaded by a processor, executes the high-speed railway cement emulsified asphalt mortar mix design method as described above.

[0048] The present application describes a method and system for designing mix proportions of high-speed railway cement emulsified asphalt mortar. Its advantages lie in the fact that the present application combines convolutional neural network technology, which can quickly predict the performance indicators of each mix proportion under different application environments. Based on the predicted performance indicators, the optimal mix proportion that meets the performance requirements is obtained iteratively through a genetic algorithm. This enables the rapid and accurate design of CA mortar mix proportions, which helps to improve design efficiency, shorten the design cycle, and reduce costs. Attached Figure Description

[0049] Figure 1 This is a flowchart illustrating the steps of a high-speed railway cement emulsified asphalt mortar mix design method as described in this embodiment;

[0050] Figure 2 This is a flowchart illustrating the steps involved in constructing the design model as described in this embodiment;

[0051] Figure 3 This is a schematic diagram of the structure of the computer device described in this embodiment.

[0052] Figure labeling: 101 - Processor, 102 - Memory. Detailed Implementation

[0053] like Figure 1 As shown in the figure, the mix design method for high-speed railway cement emulsified asphalt mortar according to this application includes the following steps:

[0054] Design factor input: Input the design factors of the cement emulsified asphalt mortar to be designed into the design model; the design factors include raw material type, dosage and layout environment parameters;

[0055] Predicted output: The design model predicts and outputs the performance indicators corresponding to the cement emulsified asphalt mortar to be designed; the performance indicators include fluidity, spreadability, working time, compressive strength, elastic modulus and fatigue life;

[0056] Input adjustment: Based on the performance indicators, adjust the design factors of the cement emulsified asphalt mortar to be designed and input them into the design model to obtain the performance indicators corresponding to the adjusted design factors;

[0057] Optimal design output: Repeat the input adjustment steps multiple times until the optimal design factors that meet the performance requirements are obtained and output.

[0058] like Figure 2 As shown, the design model is obtained through the following steps:

[0059] Data Preparation: Multiple sets of experimental data on cement emulsified asphalt mortar were acquired. This data included design factors and corresponding performance indicators. For example, Python web scraping technology was used to obtain relevant experimental data on cement emulsified asphalt mortar from data websites such as CNKI, Wanfang, Baidu Wenku, Science Direct, SAGE Journals, Wiley Online Library, and Springer Link. On the other hand, cement emulsified asphalt mortar specimens with different mix proportions were prepared in the laboratory, and tests were conducted on their workability, mechanical properties, durability, and fatigue performance. Data collection was carried out, and the data should cover various properties of cement emulsified asphalt mortar under different raw materials, mix proportions, and installation environments, including the aforementioned fluidity, spreadability, workable time, compressive strength, elastic modulus, and fatigue life. The types and amounts of raw materials and installation environment parameters were used as characteristic data.

[0060] Data preprocessing: Data cleaning and normalization are performed on the multiple sets of experimental data to convert the experimental data into a two-dimensional matrix and divide the multiple sets of experimental data into training set, validation set and test set; Specifically, the experimental data collected in the previous steps are cleaned to remove invalid data such as outliers and duplicates, and the cleaned data is normalized using the Min-Max normalization method to transform each feature data to the range of [0,1].

[0061] Assuming there are N sample data, the data includes the types and contents of raw materials, such as cement content, emulsified asphalt content, sand content, water content, water-reducing agent content, and retarder content. In addition, it includes environmental parameters, such as the number of days of immersion, the number of freeze-thaw cycles, the construction environment temperature, and the application environment temperature coefficient. Specifically, the number of days of immersion and the number of freeze-thaw cycles can be statistically analyzed on a one-year basis, counting the number of days of immersion and the number of days of freeze-thaw cycles at the deployment site within a year.

[0062] The ambient temperature during construction was obtained through the following steps:

[0063] The construction schedule of the target road segment can be obtained by reading the construction plan. The construction plan will record the construction schedule of each road segment in detail. The construction schedule of the target road segment can be obtained directly by reading the construction plan.

[0064] To predict the ambient temperature during construction, historical ambient temperature data for the target road segment is obtained, such as data from the past week or month. For example, a decision tree model can be used. Specifically, a blank decision tree model is built using Python's scikit-learn library. Historical ambient temperature data for the target road segment from the past week is obtained from meteorological data. This historical data undergoes standard data cleaning and preprocessing to convert it into data recognizable by the decision tree model. The converted data is then input into the decision tree model for training. After training, the model is optimized using cross-validation, resulting in a decision tree model capable of predicting the ambient temperature of the target road segment. Before actual construction, ambient temperature data for a certain period prior to the construction period is input into this decision tree model to predict the temperature trend during the construction period.

[0065] Since the predicted ambient temperature is time-series data, and the construction process for a single road segment is a short-term continuous operation, it is considered to use the average value to reflect the temperature conditions of the target road segment during the construction period. In this embodiment, the average value of the predicted ambient temperature during the construction period is calculated as the construction ambient temperature, which can better characterize the ambient temperature conditions during the construction period and is more in line with the actual operation.

[0066] The application environment temperature coefficient is obtained through the following steps:

[0067] Historical ambient temperature data of the target road segment over a period of time, such as one year, is obtained. The highest temperature, lowest temperature, and maximum temperature difference per unit time within this period are analyzed. A weighted calculation is performed on these data to obtain the application ambient temperature coefficient. Specifically, analysis shows that the design of cement emulsified asphalt mortar typically meets the requirements of a normal temperature range, such as 0°C to 40°C, and this normal temperature range does not significantly damage the performance of the cement emulsified asphalt mortar. Analysis of actual data shows that extreme temperatures, such as extremely low and high temperatures, or excessively large maximum temperature differences per unit time, such as within a day, significantly damage the performance of cement emulsified asphalt mortar. Extreme temperatures can cause significant damage to the performance of cement emulsified asphalt mortar. For example, extremely low temperatures can lead to reduced strength, decreased toughness, and volume shrinkage causing cracks; extremely high temperatures can cause softening and deformation; and drastic temperature changes per unit time can lead to thermal fatigue and internal... The formation of cyclic stress and decreased bonding performance can lead to problems. Analysis of actual data reveals that the environmental temperature data that significantly impacts cement emulsified asphalt mortar includes the aforementioned maximum temperature, minimum temperature, and maximum temperature difference per unit time. Using these three temperature data points individually as input to the design model, given the existing multiple inputs, would greatly increase model complexity, computational load, and prediction accuracy. Therefore, in this embodiment, the maximum temperature, minimum temperature, and maximum temperature difference per unit time are weighted and summed to obtain an application environment temperature coefficient that comprehensively represents the three temperature data points. By comprehensively considering the three temperature data points affecting the construction performance of cement emulsified asphalt mortar, the increase in model complexity can be minimized, making it more beneficial for practical applications.

[0068] After obtaining the design factors mentioned above, assuming that there are a total of M design factors, the M design factors from the N sample data are converted into an N*M two-dimensional matrix. For example, the two-dimensional matrix is ​​shown in Table 1 below:

[0069] Table 1. Two-dimensional matrix

[0070]

[0071] The obtained sample data is divided into a training set, a validation set, and a test set. For example, the training set accounts for 80% of the total data, while the validation set and test set each account for 10%. The training set is used to train the model, the validation set is used to fine-tune the model and prevent overfitting, and the test set is used to evaluate the final performance of the model.

[0072] Model Construction: A convolutional neural network is selected as the design model. For example, a convolutional neural network model is constructed using the deep learning framework TensorFlow or PyTorch, including an input layer, multiple convolutional layers, pooling layers, fully connected layers, and an output layer. The input layer of the design model is constructed according to the number of rows and columns of the two-dimensional matrix. At least two convolutional layers are designed to extract features from the data. Each convolutional layer contains multiple convolutional kernels. Each convolutional kernel slides on the input data and calculates the convolution result to extract local features. The size and number of convolutional kernels are designed according to the complexity of the input features. For example, sampling 3*3 drives the convolutional kernels.

[0073] Sampling the ReLU activation function maps the output of the convolution kernel to a nonlinear space, thereby enhancing the expressive power of the model.

[0074] A pooling layer is established after each convolutional layer to reduce the dimensionality of the convolutional layer output and extract the main features. Max pooling is selected as the pooling function to reduce the dimensionality of the data.

[0075] After the output of the pooling layer is flattened, it is connected to multiple fully connected layers. The fully connected layers perform nonlinear transformations through multiple layers of neurons to learn the complex relationship between design factors and performance indicators. The ReLU activation function is used for each fully connected layer.

[0076] Finally, based on the required performance indicators, corresponding output nodes are designed. For example, six output nodes are configured for each of the following: flowability, scalability, working time, compressive strength, elastic modulus, and fatigue life, to obtain the output layer.

[0077] Model training: The constructed design model is trained using the experimental data in the training set until the preset convergence condition or number of iterations is reached; the model is trained using the training set data until the predetermined convergence condition or number of iterations is reached. During training, the output value of the model is calculated through forward propagation, and the gradient is calculated through the backpropagation algorithm to update the model's weights and bias parameters.

[0078] During model training, the mean squared error loss function is used to evaluate the gap between the model's predicted output performance metrics and the actual collected data, and the Adam optimization algorithm is used to update the model's weights and bias parameters, thereby conducting model training.

[0079] Model optimization: The trained design model is validated and tested using the experimental data from the validation and test sets, and the design model is optimized based on the validation and test results until the performance of the design model meets the requirements, thus obtaining the design model.

[0080] The model is optimized based on the data and performance metrics on the validation dataset. This can be achieved by adjusting hyperparameters using random search or Bayesian optimization methods, adding regularization terms, and using early stopping techniques to prevent overfitting. At the same time, cross-validation is performed on the training, validation, and test sets to evaluate the model's performance.

[0081] The hyperparameters include learning rate, number of iterations, batch size, etc. The initial learning rate is set to 0.0001, the number of iterations is no less than 100, and the batch size is a power of 2, which is 128. The regularization term is added to limit the complexity of the model's parameters. Early stopping is used mainly to save training time.

[0082] The above steps yield a design model that can predict the performance index corresponding to the input design factor.

[0083] Furthermore, in this embodiment, in the aforementioned step of adjusting the design factors of the cement emulsified asphalt mortar to be designed according to performance indicators and inputting them into the design model, it is necessary to represent the adaptability of each performance indicator in a quantitative way, as follows:

[0084] Preset standard values ​​for each performance indicator, which are usually determined according to engineering standards or national standards. Calculate the fitness value Va of the performance indicator using the following formula. fit :

[0085] Va fit =w1*(Va Fa / St Fa )+w2*(Va Spr / St Spr )+w3*(Va W_t /

[0086] St W_t )+w4*(Va C_s / St C_s )+w5*(Va E_m / St E_m )+w6*(Va F_l / St F_l );

[0087] Among them, Va Fa St represents the fluidity. Fa Va represents the standard value of fluidity. Spr Indicates extensibility, St Spr Va represents the standard value of scalability. W_t Indicates available working hours, St W_t Va represents the standard value for working time. C_s St represents compressive strength. C_sVa represents the standard value of compressive strength. E_m St represents the elastic modulus. E_m Va represents the standard value of the elastic modulus. F_l Indicates fatigue life, St F_l The standard value of fatigue life is represented by w1, w2, w3, w4, w5 and w6, which represent the first weight, second weight, third weight, fourth weight, fifth weight and sixth weight corresponding to the fluidity, spreadability, working time, compressive strength, elastic modulus and fatigue life, respectively.

[0088] Preset about the fitness value Va fit The fitness standard range is defined as follows: if all performance indicators are greater than the corresponding standard values, and the fitness value Va is within the range specified in the standard range, then... fit If the performance index is within the fitness standard range, the design factor corresponding to the performance index is determined to be the optimal design factor. If it is not satisfied, the performance index that has not reached the standard value is selected, the performance index that has not reached the standard value is analyzed, the design factors that affect the performance index that has not reached the standard value are obtained, and adjustments are made.

[0089] For high-speed railway construction, the first requirement is that all performance indicators of the cement emulsified asphalt mortar meet the standard values, indicating that the mortar meets basic requirements. Then, the fitness value of each performance indicator is quantitatively calculated. Specifically, this involves calculating the ratio of each performance indicator to its corresponding standard value. For each performance indicator, if it is less than the standard value and the difference is significant, the calculated ratio will be small, resulting in a smaller fitness value. If the performance indicator equals the standard value, the ratio will be close to 1, and the fitness value will be close to 6. In other words, the more a performance indicator exceeds the standard value, the larger the fitness value will be. Therefore, a larger fitness value indicates better performance of the cement emulsified asphalt mortar. However, in practical engineering, marginal benefits need to be considered. Specifically, this refers to the increase in construction and material costs resulting from enhancing the performance of the cement emulsified asphalt mortar while meeting performance requirements. Therefore, in this embodiment, the fitness value selection method is interval selection, i.e., by setting a fitness value Va. fit The fitness standard range, for example [1,2], is defined if all performance indicators are greater than the corresponding standard values, and the fitness value Va is... fit If the fitness index is within the fitness standard range, then the design factor corresponding to the performance index is determined to be the optimal design factor, so that the design and selection of the fitness value can meet both the standard and the engineering practice requirements.

[0090] Furthermore, the weights of each performance indicator have a significant impact on the fitness value calculation results. Analysis of the design factors reveals that some are adjustable, such as the type and content of materials used; these can be adjusted in subsequent targeted adjustments. Others are non-adjustable, such as terrain and the train's maximum design speed; these cannot be manually adjusted but still significantly affect the final performance indicators and require consideration. Therefore, in this embodiment, non-adjustable factors are considered first and incorporated into the fitness value calculation process, ensuring that the fitness value reflects these factors.

[0091] The specific method is as follows: Initial weight values ​​for the first, second, third, fourth, fifth, and sixth weights are configured respectively. These initial weight values ​​can be pre-configured based on standard terrain and design speed, etc., before calculating the fitness value Va. fit At that time, the weight adjustment factor is obtained, and the weights of each performance index are adjusted according to the weight adjustment factor and the preset weight adjustment strategy. The weight adjustment factor includes the terrain of the high-speed railway layout and the maximum design speed of the train.

[0092] The weight adjustment strategy includes:

[0093] In response to the fact that the high-speed railway is located in a plain, the first and third weights are increased, and the fourth and fifth weights are decreased. The plain is flat and the mortar layer is not prone to additional dynamic loads due to terrain changes. Therefore, the corresponding weights for compressive strength and modulus of elasticity can be appropriately reduced. In addition, in order to meet the overall construction period requirements of the high-speed railway, the weights for fluidity and working time can be appropriately increased to adapt to large-scale and rapid construction.

[0094] In response to the fact that the high-speed railway is located in a hilly terrain, the fourth and fifth weights are increased, while the first and second weights are decreased. The hilly terrain may have slight undulations, and the mortar layer is prone to additional dynamic loads due to terrain changes. In this terrain, compressive strength and elastic modulus need to be given priority to cope with the possible additional dynamic loads. Therefore, the weights of these two performance indicators are increased, and correspondingly, the less important first and second weights are decreased.

[0095] In response to the train's maximum design speed being greater than or equal to a preset speed threshold, the fifth and sixth weights are increased, while the first, second, and third weights are decreased. In a specific embodiment, the speed threshold can be set to 300 km / h. For high-speed trains with a maximum design speed greater than or equal to the speed threshold, there are higher requirements for the dynamic performance and fatigue life of the track. Therefore, the fifth and sixth weights are increased, and the first, second, and third weights related to construction are correspondingly reduced.

[0096] In response to the train's maximum design speed being less than the speed threshold, the third and fourth weights are increased, while the fifth and sixth weights are decreased. For medium- and low-speed trains, the requirements for dynamic performance are lower, and more attention is paid to the working time to meet the requirements of large-scale construction. In addition, medium- and low-speed trains usually have a large load or tare weight, so more attention is paid to compressive strength. Therefore, the third and fourth weights are increased, while the fifth and sixth weights are decreased.

[0097] In this step, the fitness value calculation is configured based on the non-adjustable factors in the design factors, so that the calculated fitness value is closer to the actual construction situation.

[0098] Furthermore, in the step of constructing the model, the type and amount of raw materials are defined as adjustable factors, and the environmental parameters are defined as non-adjustable factors. The environmental parameters include the number of days of immersion, the number of days of freeze-thaw cycles, the construction environment temperature, and the application environment temperature coefficient.

[0099] A correlation analysis is performed between the deployment environment parameters and the performance indicators. For example, Pearson correlation coefficient analysis is performed on the deployment environment parameters and corresponding performance indicators in the sample data, and the correlation between the two parameters is determined based on the calculation results of the Pearson correlation coefficient.

[0100] For each of the deployment environment parameters, iterate through each of the performance indicators, obtain all performance indicators that are related to the deployment environment parameter, establish a correspondence between the deployment environment parameter and the related performance indicators, and record them as an association group;

[0101] The design model comprises a sequentially cascaded input layer, convolutional pooling layer, fully connected layer, and output layer. The input layer includes input nodes equal in number to the number of design factors. A separate branch network is created for each association group, and the outputs of multiple branch networks are merged in the fully connected layer. For example, if there is a correlation between construction ambient temperature and flowability, then construction ambient temperature and flowability are considered as an association group. A separate branch network is created for this association group in the design model. A branch network is created for each performance index, and finally, multiple branch networks are merged in the fully connected layer to form a complete design model.

[0102] In this embodiment, by analyzing the correlation between unadjustable factors and performance indicators and incorporating it into the design model, the prediction results of performance indicators can fully consider various environmental factors, which helps to improve the prediction accuracy of performance indicators.

[0103] Furthermore, the input adjustment and optimal design output include the following steps:

[0104] All possible design factors for high-speed railway cement emulsified asphalt mortar are obtained. These possible design factors can be designed based on experience or obtained from existing design data. A genetic algorithm is used to initialize the population, and each individual in the population corresponds to a set of possible design factors.

[0105] Each individual in the population is input into the design model to obtain the performance index corresponding to each individual, and the fitness value Va corresponding to each individual is further calculated. fit ;

[0106] The adjustable factors and various performance indicators are correlated using the Pearson correlation coefficient method or the Spearman rank correlation coefficient method. Adjustable factors with strong correlations to various performance indicators are identified, and correlation adjustment relationships are established to obtain correlation results. The strong correlation relationship is expressed as the correlation analysis result between the adjustable factors and the performance indicators is not less than a preset correlation degree threshold, for example, the absolute value of the Pearson correlation coefficient is not less than 0.8. The correlation results include the performance indicator, the adjustable factors with strong correlations to the performance indicator, and the correlation between the performance indicator and the adjustable factors with strong correlations; for example, water content and fluidity have a strong correlation, and the two are positively correlated.

[0107] Individuals with fitness values ​​greater than a preset screening threshold in the population are selected as parents. Performance indicators among the selected parents that do not meet the standard value are designated as unqualified indicators. In the crossover operation, adjustable factors with a strong correlation to the unqualified indicators are selected as mutation targets. The direction of mutation is determined based on the correlation between the unqualified indicators and the adjustable factors, and the crossover operation is performed accordingly. For example, if the mobility of a certain design factor is unqualified, and correlation analysis determines that water content and mobility have a strong and positive correlation, then water content is selected as the mutation target. The crossover operation is performed with increasing water content as the mutation direction, generating multiple offspring individuals to form a new population. This new population is used to replace the initial population, and the iteration is repeated multiple times until the optimal design factor that meets the energy requirements is obtained and output.

[0108] In this embodiment, by performing correlation analysis on adjustable factors and performance indicators, and by using a genetic algorithm to screen for optimal design factors, the correlation analysis results are introduced into the genetic algorithm. That is, the correlated objects are used as mutation objects, and the mutation direction is determined according to the correlation. This makes the mutation of the genetic algorithm more targeted, helps to reduce the number of iterations, and improves accuracy.

[0109] This embodiment also provides a high-speed railway cement emulsified asphalt mortar mix design system, including:

[0110] The input module is used to input the design factors of the cement emulsified asphalt mortar to be designed into the design model; the design factors include the type and amount of raw materials and the environmental parameters of the layout.

[0111] The prediction module is used to predict and output the performance indicators corresponding to the cement emulsified asphalt mortar to be designed through the design model; the performance indicators include fluidity, spreadability, working time, compressive strength, elastic modulus and fatigue life.

[0112] The input adjustment module is used to adjust the design factors of the cement emulsified asphalt mortar to be designed according to the performance index and input them into the design model to obtain the performance index corresponding to the adjusted design factors. This process is repeated multiple times until the optimal design factors that meet the performance requirements are obtained.

[0113] The output module is used to output the optimal design factors.

[0114] The system in this embodiment and the method described above belong to the same inventive concept and can be understood with reference to the above description, which will not be repeated here.

[0115] like Figure 3As shown, this embodiment also provides a computer device, including a processor 101 and a memory 102 connected via a bus signal. The memory 102 stores at least one instruction or at least one program. When the at least one instruction or the at least one program is loaded by the processor 101, it executes the high-speed railway cement emulsified asphalt mortar mix design method described above. The memory 102 can be used to store software programs and modules. The processor 101 executes various functional applications by running the software programs and modules stored in the memory 102. The memory 102 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for functions, etc.; the data storage area may store data created according to the use of the device, etc. In addition, the memory 102 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 102 may also include a memory controller to provide the processor 101 with access to the memory 102.

[0116] The methods and embodiments provided in this application can be executed in a computer terminal, server, or similar computing device; that is, the aforementioned computer device may include a computer terminal, server, or similar computing device. The internal structure of the computer device may include, but is not limited to, a processor, a network interface, and memory. The processor, network interface, and memory within the computer device can be connected via a bus or other means.

[0117] The processor 101 (or CPU, Central Processing Unit) is the computing and control core of the computer device. The network interface may optionally include a standard wired interface or a wireless interface (such as Wi-Fi, mobile communication interface, etc.). The memory 102 is the storage device in the computer device used to store programs and data. It is understood that the memory 102 here can be a high-speed RAM storage device, or a non-volatile storage device, such as at least one disk storage device; optionally, it can also be at least one storage device located remotely from the processor 101. The memory 102 provides storage space that stores the operating system of the electronic device, which may include, but is not limited to: Windows (an operating system), Linux (an operating system), Android (a mobile operating system), iOS (a mobile operating system), etc., and this application does not limit this; furthermore, the storage space also stores one or more instructions suitable for being loaded and executed by the processor 101, which may be one or more computer programs (including program code). In the embodiments of this specification, the processor 101 loads and executes one or more instructions stored in the memory 102 to implement the high-speed railway cement emulsified asphalt mortar mix design method described in the above method embodiments.

[0118] This application also provides a computer-readable storage medium storing at least one instruction or at least one program segment. When the at least one instruction or at least one program segment is loaded by the processor 101, it executes the high-speed railway cement emulsified asphalt mortar mix design method as described above. The aforementioned computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0119] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0120] For those skilled in the art, various other corresponding changes and modifications can be made based on the technical solutions and concepts described above, and all such changes and modifications should fall within the protection scope of the claims of this application.

Claims

1. A method for designing the mix proportion of cement emulsified asphalt mortar for high-speed railways, characterized in that, Includes the following steps: Design factor input: Input the design factors of the cement emulsified asphalt mortar to be designed into the design model; the design factors include raw material type, dosage and layout environment parameters; Predicted output: The design model predicts and outputs the performance indicators corresponding to the cement emulsified asphalt mortar to be designed; the performance indicators include fluidity, spreadability, working time, compressive strength, elastic modulus and fatigue life; Input adjustment: Based on the performance indicators, adjust the design factors of the cement emulsified asphalt mortar to be designed and input them into the design model to obtain the performance indicators corresponding to the adjusted design factors; Optimal design output: Repeat the input adjustment steps multiple times until the optimal design factors that meet the performance requirements are obtained and output. The step of adjusting the input, which involves adjusting the design factors of the cement emulsified asphalt mortar to be designed and inputting them into the design model based on the performance indicators, includes: Preset standard values ​​for each performance indicator, and calculate the fitness value of the performance indicator according to the following formula. : ； in, Indicates fluidity. Indicates the standard value of fluidity. Indicates scalability. Indicates the standard value of scalability. Indicates available working hours. This represents the standard value for working time. Indicates compressive strength. This represents the standard value of compressive strength. Indicates the elastic modulus. This represents the standard value of the elastic modulus. Indicates fatigue life. This represents the standard value for fatigue life. and The first weight, second weight, third weight, fourth weight, fifth weight, and sixth weight respectively represent the flowability, spreadability, working time, compressive strength, elastic modulus, and fatigue life. Preset about the fitness value The fitness standard range is defined as follows: if all performance indicators are greater than the corresponding standard values, and the fitness value is... If the performance index is within the fitness standard range, the design factor corresponding to the performance index is determined to be the optimal design factor. If it is not satisfied, the performance index that has not reached the standard value is selected, the performance index that has not reached the standard value is analyzed, the design factors that affect the performance index that has not reached the standard value are obtained, and adjustments are made. The initial weight values ​​for the first weight, second weight, third weight, fourth weight, fifth weight, and sixth weight are configured respectively, and the fitness value is calculated. At that time, the weight adjustment factor is obtained, and the weights of each performance index are adjusted according to the weight adjustment factor and the preset weight adjustment strategy. The weight adjustment factor includes the terrain of the high-speed railway layout and the maximum design speed of the train. The weight adjustment strategy includes: In response to the fact that the high-speed railway is located in a plain, the first and third weights are increased, and the fourth and fifth weights are decreased. In response to the fact that the high-speed railway is located in a hilly terrain, the fourth and fifth weights are increased, while the first and second weights are decreased. In response to the train's maximum design speed being greater than or equal to a preset speed threshold, the fifth and sixth weights are increased, while the first, second, and third weights are decreased. In response to the train's maximum design speed being less than the speed threshold, the third and fourth weights are increased, while the fifth and sixth weights are decreased.

2. The mix design method for high-speed railway cement emulsified asphalt mortar according to claim 1, characterized in that, The design model was obtained through the following steps: Data preparation: Obtain multiple sets of test data for cement emulsified asphalt mortar, including the design factors and corresponding performance indicators of cement emulsified asphalt mortar; Data preprocessing: The experimental data from multiple sets are cleaned and normalized, converted into a two-dimensional matrix, and divided into training set, validation set and test set. Model construction: A convolutional neural network was selected as the design model. Model training: The constructed design model is trained using the experimental data of the training set until the preset convergence condition or number of iterations is reached; Model optimization: The trained design model is validated and tested using the experimental data from the validation and test sets, and the design model is optimized based on the validation and test results until the performance of the design model meets the requirements, thus obtaining the design model.

3. The mix design method for high-speed railway cement emulsified asphalt mortar according to claim 2, characterized in that, In the model building process, the types and amounts of raw materials are defined as adjustable factors, and the environmental parameters are defined as non-adjustable factors. The environmental parameters include the number of days of immersion, the number of days of freeze-thaw cycles, the construction environment temperature, and the application environment temperature coefficient. A correlation analysis is performed between the deployment environment parameters and each of the performance indicators. For each deployment environment parameter, each of the performance indicators is traversed to obtain all performance indicators that are related to the deployment environment parameter. A correspondence relationship is established between the deployment environment parameter and the related performance indicators and recorded as a correlation group. The design model includes sequentially cascaded input layers, convolutional pooling layers, fully connected layers, and output layers. The input layer includes input nodes in a number equal to the number of items of the design factors. A separate branch network is created for each association group, and the outputs of multiple branch networks are merged in the fully connected layer.

4. The mix design method for high-speed railway cement emulsified asphalt mortar according to claim 3, characterized in that, The ambient temperature during construction was obtained through the following steps: Obtain the construction period for the target road section through the construction plan. Historical ambient temperature data of the target road section is obtained. The ambient temperature during the construction period is predicted based on the historical ambient temperature data. The average value of the predicted ambient temperature during the construction period is calculated, which is the construction ambient temperature. The application environment temperature coefficient is obtained through the following steps: Historical ambient temperature data of the target road segment within a period is obtained, and the highest temperature, lowest temperature and maximum temperature difference per unit time of the target road segment within the period are analyzed. The highest temperature, lowest temperature and maximum temperature difference per unit time are weighted and calculated to obtain the application ambient temperature coefficient.

5. The mix design method for high-speed railway cement emulsified asphalt mortar according to claim 3, characterized in that, The input adjustment and optimal design output include the following steps: All possible design factors for high-speed railway cement emulsified asphalt mortar are obtained, and a genetic algorithm is used to initialize the population, with each individual in the population corresponding to a set of possible design factors; Each individual in the population is input into the design model to obtain the performance index corresponding to each individual, and the fitness value corresponding to each individual is further calculated. ; A correlation analysis is performed on the adjustable factors and various performance indicators to obtain the adjustable factors that have a strong correlation with each performance indicator and to establish a correlation adjustment relationship. The correlation results are obtained. The strong correlation relationship is indicated by the correlation analysis result between the adjustable factors and the performance indicators being no less than a preset correlation degree threshold. The correlation results include the performance indicator, the adjustable factors that have a strong correlation with the performance indicator, and the correlation between the performance indicator and the adjustable factors that have a strong correlation. Individuals with fitness values ​​greater than a preset screening threshold in the population are selected as parents. Performance indicators of the selected parents that do not meet the standard value are designated as unqualified indicators. In the crossover operation, adjustable factors that have a strong correlation with the unqualified indicators are selected as mutation objects. The direction of mutation is determined based on the correlation between the unqualified indicators and the adjustable factors, and the crossover operation is performed to generate multiple offspring individuals to form a new population. The new population is used to replace the initial population. This process is repeated iteratively multiple times until the optimal design factors that meet the energy requirements are obtained and output.

6. A mix design system for high-speed railway cement emulsified asphalt mortar, wherein the mix design method for high-speed railway cement emulsified asphalt mortar as described in any one of claims 1-5 is used, characterized in that, include: The input module is used to input the design factors of the cement emulsified asphalt mortar to be designed into the design model; the design factors include the type and amount of raw materials and the environmental parameters of the layout. The prediction module is used to predict and output the performance indicators corresponding to the cement emulsified asphalt mortar to be designed through the design model; the performance indicators include fluidity, spreadability, working time, compressive strength, elastic modulus and fatigue life. The input adjustment module is used to adjust the design factors of the cement emulsified asphalt mortar to be designed according to the performance index and input them into the design model to obtain the performance index corresponding to the adjusted design factors. This process is repeated multiple times until the optimal design factors that meet the performance requirements are obtained. The output module is used to output the optimal design factors.

7. A computer device comprising a processor and a memory connected by signals, characterized in that, The memory stores at least one instruction or at least one program segment, which is executed when loaded by the processor as described in any one of claims 1-5, regarding the high-speed railway cement emulsified asphalt mortar mix design method.

8. A computer-readable storage medium having stored thereon at least one instruction or at least one program, characterized in that, When the at least one instruction or the at least one program segment is loaded by the processor, the high-speed railway cement emulsified asphalt mortar mix design method as described in any one of claims 1-5 is executed.

Images