Method and system for constructing road roller construction process knowledge base based on improved decision tree

Abstract

The present invention provides a method and system for constructing a knowledge base for road roller construction technology based on an improved decision tree. The method constructs a road roller construction technology database based on existing case data of road roller construction projects. The method determines whether the various data on construction objects and construction requirements in the road roller construction technology data are normally distributed. If so, the existing data is kept unchanged; otherwise, the non-normally distributed data is converted to normally distributed data. The Gini systems of the various judgment attributes are calculated and arranged in order to initially form a decision tree. The converted data in the decision tree are inversely transformed. The inversely transformed decision tree is pruned and optimized to form a final improved decision tree model. The improved decision tree model is used to classify and analyze actual data to obtain a road roller construction plan. The present invention can ensure that all road roller construction process data are normally distributed, addressing the limitations of existing decision trees and assisting engineers in making decisions about road roller construction.

Description

Technical Field

[0001] The present invention belongs to the technical field of road roller construction, and in particular relates to a method and system for constructing a road roller construction process knowledge base based on an improved decision tree. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] With the advent of the big data era in recent years, the civil engineering industry has faced new opportunities and challenges. Traditional construction decision-making processes are cumbersome and wasteful of manpower and resources. Researchers are beginning to experiment with new technologies to streamline the decision-making process and accurately and rapidly determine process flows.

[0004] The traditional method for determining the various processes and parameters for roller construction in the civil engineering industry is cumbersome, heavily reliant on the engineer's experience, and highly subjective. Each parameter requires consulting relevant documentation, and the engineer makes the overall decision, which is not conducive to the specific implementation of the project and can easily lead to delays or errors in the construction process.

[0005] Although some researchers have used decision trees to generate decisions using data, decision trees have their own limitations in handling non-normal data and are prone to falling into local optimality. It is difficult to ensure that the construction process data of road rollers are all normally distributed, which has great limitations. Summary of the Invention

[0006] In order to solve the above problems, the present invention proposes a method and system for constructing a road roller construction process knowledge base based on an improved decision tree. The present invention can ensure that the road roller construction process data can be guaranteed to be normally distributed, solve the limitations of the existing decision tree, and assist engineers in making decisions on road roller construction.

[0007] According to some embodiments, the present invention adopts the following technical solutions:

[0008] A method for constructing a road roller construction process knowledge base based on an improved decision tree comprises the following steps:

[0009] Build a roller construction process database based on existing case data of roller construction projects;

[0010] Determine whether the construction object and construction requirement data in the roller construction process data are normally distributed. If so, keep the existing data unchanged. Otherwise, perform Box-Cox transformation on the non-normal distribution data to convert it into normal distribution data.

[0011] Calculate the Gini system of each judgment attribute and arrange them in order to initially form a decision tree;

[0012] Perform relevant inverse transformation on the data that has been transformed by Box-Cox in the decision tree;

[0013] The decision tree after inverse transformation is pruned and optimized to ensure that the minimum number of instances is greater than the set value, thus forming an improved decision tree model;

[0014] Obtain parameters related to the roller's construction object and construction requirements;

[0015] According to the construction object and construction requirement related parameters, the improved decision tree model is used to perform classification analysis to obtain the roller construction plan and form a roller construction process knowledge base.

[0016] As an optional implementation, the construction object of the road roller includes soil type, bottom length, bottom width, depth and slope coefficient of the construction object.

[0017] As an optional implementation method, the parameters related to the roller construction requirements include construction requirement attributes, roller model, and the roller amplitude, frequency, rolling speed and number of rolling passes in the three stages of initial compaction, repeated compaction and final compaction. The construction requirement attributes include compaction degree, flatness and uniformity.

[0018] As an optional implementation method, the two major attributes of construction objects and construction requirements are used as judgment attributes in the decision tree model. Different judgment attributes constitute the root node and internal nodes of the decision tree, and the levels divided by the judgment attributes are the branches of the decision tree.

[0019] The roller model and the key parameters of the roller in the three stages of initial compaction, re-compaction and final compaction are used as the decision results of the decision tree. Each leaf node reached by the decision tree through each fork represents a roller construction plan, and each construction plan corresponds to a leaf node of the decision tree.

[0020] As an optional implementation method, the specific process of determining whether the various data on construction objects and construction requirements in the roller construction process data are normally distributed includes: using the Anderson-Darling normality test to determine the distribution type, arranging the sample data X in ascending order, and using a discrete expression to calculate the distance parameter between the empirical distribution function and the specified distribution function. If the distance parameter is less than the set test critical value, it is considered that the distribution function of the corresponding variable conforms to the normal distribution; otherwise, it does not conform to the normal distribution.

[0021] As an optional implementation, the specific process of performing Box-Cox transformation on non-normal distribution data includes: using the Box-Cox power function transformation method to transform the non-normal distribution data into normal distribution data. For the non-normal distribution random variable X, there is a transformation formula:

[0022]

[0023] The value of λ is determined by the maximum likelihood estimation.

[0024] As an optional implementation method, the Gini coefficient of each judgment attribute is calculated and arranged in order to initially form a decision tree. The specific process includes: taking each attribute of the construction object and construction requirement related parameters as a judgment attribute, and the Gini coefficient calculation formula is:

[0025]

[0026] Where Gini(X) represents the Gini coefficient of a certain level of attribute X. There are n different cases in this level. p(x k ) represents the frequency of the k-th case in all grade samples. Based on this, the Gini coefficients of all grades in the judgment attribute X are calculated, and the Gini coefficients of all grades are weighted averaged according to their sample numbers to obtain the Gini coefficient of the judgment attribute X.

[0027] Arrange them in ascending order, with the attribute with the highest Gini coefficient as the initial root node, and the hierarchy of that attribute as the fork extending from the root node;

[0028] The Gini coefficients of the remaining attributes are calculated in the sub-knowledge base corresponding to each fork. The attribute with the lowest Gini coefficient is used as the first internal node after the root node forks. The calculation is done step by step. In the process of building the decision tree, the Gini coefficient is extended downward from low to high.

[0029] As a further limitation, if a case lacks some judgment attributes, the case will be discarded as incomplete data.

[0030] As an optional implementation method, the construction plan includes the roller model and key parameters of the roller in the three stages of initial compaction, secondary compaction and final compaction.

[0031] A road roller construction process knowledge base construction system based on an improved decision tree includes:

[0032] A database construction module is configured to construct a road roller construction process database based on existing case data of road roller construction projects;

[0033] The data conversion module is configured to determine whether the data of the construction object and construction requirements in the road roller construction process data are normally distributed. If so, the existing data is kept unchanged; otherwise, the non-normal distribution data is converted into normal distribution data by Box-Cox transformation;

[0034] The decision tree model building module is configured to calculate the Gini system of each judgment attribute and arrange them in order to initially form a decision tree;

[0035] The data inverse transformation module is configured to perform relevant inverse transformation on the data that has undergone Box-Cox transformation in the decision tree;

[0036] The decision tree model optimization module is configured to perform pruning optimization on the decision tree after the inverse transformation to ensure that the minimum number of instances is greater than the set value, thereby forming the final improved decision tree model;

[0037] A parameter acquisition module is configured to obtain parameters related to the construction object and construction requirements of the road roller;

[0038] The classification analysis module is configured to perform classification analysis based on the construction object and construction requirement-related parameters using the improved decision tree model to obtain the roller construction plan and form a roller construction process knowledge base.

[0039] A computer-readable storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor of a terminal device and executing the steps of the above method.

[0040] A terminal device includes a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, and the instructions are suitable for being loaded by the processor and executing the steps of the above method.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] The present invention can analyze the implicit information and rules of the data based on historical case data, and process and utilize the implicit rules, thereby reducing the degree of experience participation of engineers, improving the rationality and intelligence of the decision-making process, and realizing the intelligent integrated construction of the road roller. At the same time, by converting non-normal data, it can ensure that the road roller construction process data can be guaranteed to be normally distributed, solving the limitations of the existing decision tree, assisting engineers in making decisions on road roller construction, forming a final recommended construction plan, and also forming a road roller construction process knowledge base to guide actual construction.

[0043] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0045] Figure 1 The figure is a flow chart of at least one embodiment of the present invention. DETAILED DESCRIPTION

[0046] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0047] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0048] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0049] A method for constructing a knowledge base for roller construction techniques based on an improved decision tree is proposed. Targeting the field of big data analysis for roller construction, the Anderson-Darling normality test is used to determine whether each data in the database is normally distributed. If not, the non-normal distribution of the case data in the database is converted to a normal distribution using Box-Cox. The Gini coefficient is calculated and sorted to form a CART decision tree. A Box-Cox inverse transformation and pruning process are then performed to form a knowledge base for roller construction techniques, which guides actual construction and provides recommended construction plans.

[0050] like Figure 1 As shown, the specific steps of the method are introduced with a typical embodiment:

[0051] A database of roller construction cases was constructed, containing a vast amount of roller construction process data accumulated over several years. The cases in the database primarily include construction objects, construction requirements, roller models, and key roller parameters for the three stages of initial compaction, secondary compaction, and final compaction. Construction object attributes include soil type, bottom length, bottom width, depth, and slope coefficient. Soil types are categorized into Class I, Class II, Class III, and Class IV soils according to the soil classification table, and the bottom length, bottom width, depth, and slope coefficient of the construction object are also divided into three levels. Construction requirement attributes include compaction, flatness, and uniformity, each of which is divided into three levels. Key roller parameters for the three stages of initial compaction, secondary compaction, and final compaction include amplitude, frequency, rolling speed, and number of rolling passes, each of which is divided into three levels.

[0052] The construction object and construction requirements are the two main attributes in the decision tree. Different attributes form the root and internal nodes of the decision tree, and the levels of the attributes are divided into the branches of the decision tree. The roller model and the key parameters of the roller during the initial, secondary, and final rolling stages serve as the decision results of the decision tree. Each leaf node reached through each branch of the decision tree represents a roller construction plan, and each construction plan corresponds to a leaf node of the decision tree.

[0053] Determine whether the data of construction objects and construction requirements are normally distributed. If they are not normally distributed, the generated decision tree is likely to fall into local optimality. Use Anderson-Darling normality test to determine the distribution type, sort the sample data X in ascending order, and use A 2 Represents the distance parameter between the empirical distribution function and the specified distribution function. This embodiment uses a discrete expression to calculate A 2 , the formula is as follows:

[0054]

[0055] Where N represents the number of samples, Y i is sorted data, F(Y i ) is the cumulative probability distribution function for a specified distribution (here, the standard normal distribution).

[0056] Existing hypothesis: Hypothesis H0: The distribution function of the X variable conforms to the specified distribution;

[0057] Hypothesis H1: The distribution function of the X variable does not conform to the specified distribution.

[0058] At the specified significance level, if parameter A 2 If it is less than the test critical value, the hypothesis H0 is accepted and it is considered that the distribution function of the X variable conforms to the specified distribution, that is, it conforms to the normal distribution. Otherwise, it does not conform to the normal distribution.

[0059] According to the sample size n, A 2 Corrected, we get the test statistic A * , the formula is as follows:

[0060]

[0061] If A * If the value exceeds 0.77, the normality hypothesis is rejected and the distribution is considered non-normal. Otherwise, the distribution is considered normal. Due to the large amount of data, a more lenient judgment standard is used to judge the normal distribution. The test critical value of 0.77 is a relatively large critical value.

[0062] If it is judged to be normally distributed data, it remains unchanged; if it is judged to be non-normally distributed data, in order to avoid the decision tree falling into local optimality, the Box-Cox power function conversion method is used to convert the non-normally distributed data into normally distributed data. For non-normally distributed random variables X, there is a transformation formula:

[0063]

[0064] The value of λ is determined by the maximum likelihood estimation, and by determining the appropriate value of λ, Y (λ) Obey the normal distribution, that is, Y (λ) ~N(μ,σ 2 ). Then the likelihood function of the random variable X is:

[0065]

[0066] To facilitate calculation, the likelihood function is taken as logarithm, and the log-likelihood function is:

[0067]

[0068] After transformation by the log-likelihood function, determine the estimate of the mean:

[0069]

[0070] After transformation by the log-likelihood function, determine the estimated value of the standard deviation:

[0071]

[0072] Substituting the maximum likelihood estimates of μ and σ into the likelihood function, we can calculate λ. Transform the calculated series of λs to the normally distributed data, plot a probability graph for the transformed data, and select the λ with the smallest P value in the probability graph to complete the Box-Cox transformation.

[0073] The decision tree construction process for the roller construction process is as follows: Calculate the Gini coefficients of eight judgment attributes, including soil type, bottom length of the construction object, bottom width of the construction object, depth of the construction object, slope coefficient of the construction object, compaction requirement, flatness requirement, and uniformity requirement. The lower the Gini coefficient, the higher the purity and the better the data classification effect. At this point, the non-normal distribution data in the eight judgment attributes have been converted to normal distribution data through Box-Cox. That is, all judgment attributes are now normally distributed data. The Gini coefficient calculation formula is:

[0074]

[0075] Where Gini(X) represents the Gini coefficient of a certain level of attribute X. There are n different cases in this level. p(xk ) represents the frequency of the kth case in all grade samples. Based on this, the Gini coefficients of all grades in the judgment attribute X are calculated, and the Gini coefficients of all grades are weighted averaged according to the number of samples to obtain the Gini coefficient of the judgment attribute X.

[0076] Before calculating the Gini coefficient, incomplete data should be processed with default values. If a case is missing data for eight attributes or the model of the road roller, the case should be discarded. After the default value processing is completed, the Gini coefficient of each attribute is calculated and sorted in ascending order. The attribute with the highest Gini coefficient is used as the initial root node, and the classification of this attribute is used as the fork extending from the root node. The root node is judged, and the knowledge base has been classified once. Based on the results of this classification, the Gini coefficients of the remaining seven attributes are calculated for each sub-knowledge base corresponding to the fork. The attribute with the lowest Gini coefficient is used as the first internal node after the root node forks. And so on, the calculation is carried out step by step, and the decision tree is constructed by extending downward from low to high according to the Gini coefficient.

[0077] The data in the initially generated decision tree judgment was subjected to Box-Cox transformation. To facilitate the use of the decision tree, the data that has been Box-Cox transformed was subjected to inverse Box-Cox transformation to convert the normally distributed data back into non-normal data. The formula is as follows:

[0078]

[0079] Where X is non-normal distribution data, Y (λ) For normally distributed data, λ is determined by maximum likelihood estimation.

[0080] After the inverse Box-Cox transformation, the strategy tree undergoes post-pruning. The minimum number of instances requires that each leaf have at least two instances, and leaves with fewer than two instances are discarded. This optimizes the decision tree's results and increases its confidence.

[0081] After the decision tree is generated, the construction object and various parameters of the roller are input. After classification and analysis through the decision tree, the corresponding roller construction plan is obtained, the roller model and the key parameters of the roller in the three stages of initial compaction, re-compaction and final compaction are determined to guide the actual construction.

[0082] Of course, each construction plan can form a knowledge base of roller construction technology, which can be used later.

[0083] This section also provides the following product examples:

[0084] A road roller construction system based on an improved decision tree, comprising:

[0085] A database construction module is configured to construct a road roller construction process database based on existing case data of road roller construction projects;

[0086] The data conversion module is configured to determine whether the data of the construction object and construction requirements in the road roller construction process data are normally distributed. If so, the existing data is kept unchanged; otherwise, the non-normal distribution data is converted into normal distribution data by Box-Cox transformation;

[0087] The decision tree model building module is configured to calculate the Gini system of each judgment attribute and arrange them in order to initially form a decision tree;

[0088] The data inverse transformation module is configured to perform relevant inverse transformation on the data that has undergone Box-Cox transformation in the decision tree;

[0089] The decision tree model optimization module is configured to perform pruning optimization on the decision tree after the inverse transformation to ensure that the minimum number of instances is greater than the set value, thereby forming the final improved decision tree model;

[0090] A parameter acquisition module is configured to obtain parameters related to the construction object and construction requirements of the road roller;

[0091] The classification analysis module is configured to perform classification analysis using the improved decision tree model based on the construction object and construction requirement-related parameters to obtain a road roller construction plan.

[0092] A computer-readable storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor of a terminal device and executing the steps of the above method.

[0093] A terminal device includes a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, and the instructions are suitable for being loaded by the processor and executing the steps of the above method.

[0094] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0095] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0096] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0097] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0098] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

[0099] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A method for constructing a knowledge base of road roller construction technology based on an improved decision tree, characterized by: The following steps are involved: Build a roller construction process database based on existing case data of roller construction projects; Determine whether the construction object and construction requirement data in the roller construction process data are normally distributed. If so, keep the existing data unchanged. Otherwise, perform Box-Cox transformation on the non-normal distribution data to convert it into normal distribution data. Calculate the Gini system of each judgment attribute and arrange them in order to initially form a decision tree; Perform relevant inverse transformation on the data that has been transformed by Box-Cox in the decision tree; The decision tree after inverse transformation is pruned and optimized to ensure that the minimum number of instances is greater than the set value, thus forming an improved decision tree model; Obtain parameters related to the roller's construction object and construction requirements; According to the construction object and construction requirement related parameters, the improved decision tree model is used to conduct classification analysis to obtain the roller construction plan and form a roller construction process knowledge base; The roller construction object includes soil type, bottom length, bottom width, depth and slope coefficient of the construction object; The parameters related to the roller construction requirements include construction requirement attributes, roller model, and the roller amplitude, frequency, rolling speed, and number of rolling passes in the three stages of initial rolling, secondary rolling, and final rolling. The construction requirement attributes include compaction, flatness, and uniformity. The two major attributes of construction objects and construction requirements are used as judgment attributes in the decision tree model. Different judgment attributes constitute the root node and internal nodes of the decision tree. The levels divided by the judgment attributes are the branches of the decision tree. The roller model and the key parameters of the roller in the three stages of initial, secondary, and final compaction are used as the decision results of the decision tree. Each leaf node reached by the decision tree through each fork represents a roller construction plan, and each construction plan corresponds to a leaf node of the decision tree. The specific process of calculating the Gini coefficients of each judgment attribute and arranging them in order to initially form a decision tree includes: taking each attribute of the construction object and construction requirement related parameters as a judgment attribute, calculating the Gini coefficient, arranging them in ascending order, and using the attribute with the highest Gini coefficient as the initial root node, and the attribute's grade as the branch extending from the root node; Calculate the Gini coefficients of the remaining attributes in the sub-knowledge base corresponding to each bifurcation. The attribute with the lowest Gini coefficient is used as the first internal node after the root node bifurcates. Calculate the coefficients step by step, and extend downward from low to high according to the Gini coefficients in the process of building the decision tree. If a case lacks some judgment attributes, this case will be discarded as incomplete data.

2. The method for constructing a road roller construction process knowledge base based on an improved decision tree according to claim 1, wherein: The specific process of determining whether the various data on construction objects and construction requirements in the road roller construction process data are normally distributed includes: using the Anderson-Darling normality test to determine the distribution type, arranging the sample data X in ascending order, and using a discrete expression to calculate the distance parameter between the empirical distribution function and the specified distribution function. If the distance parameter is less than the set test critical value, it is considered that the distribution function of the corresponding variable conforms to the normal distribution; otherwise, it does not conform to the normal distribution.

3. The method for constructing a road roller construction process knowledge base based on an improved decision tree according to claim 1, wherein: The specific process of Box-Cox transformation of non-normal distribution data includes: using the Box-Cox power function transformation method to transform non-normal distribution data into normal distribution data. For non-normal distribution random variable X, there is a transformation formula: The value of is determined by maximum likelihood estimation.

4. The method for constructing a road roller construction process knowledge base based on an improved decision tree according to claim 1, wherein: The specific process of calculating the Gini coefficient of each judgment attribute and arranging them in order to initially form a decision tree includes: taking each attribute of the construction object and construction requirement related parameters as a judgment attribute, and the Gini coefficient calculation formula is: Where Gini(X) represents the Gini coefficient of a certain level of the judgment attribute X. There are n different cases in this level. p(xk) represents the frequency of the kth case in all level samples. The Gini coefficients of all levels in the judgment attribute X are calculated accordingly. The Gini coefficients of all levels are weighted averaged according to the number of samples to obtain the Gini coefficient of the judgment attribute X.

5. The method for constructing a road roller construction process knowledge base based on an improved decision tree according to claim 1, wherein: The construction plan includes the roller model and the key parameters of the roller in the three stages of initial compaction, secondary compaction and final compaction.

6. A road roller construction process knowledge base construction system based on an improved decision tree, characterized by: include: A database construction module is configured to construct a road roller construction process database based on existing case data of road roller construction projects; The data conversion module is configured to determine whether the data of the construction object and construction requirements in the road roller construction process data are normally distributed. If so, the existing data is kept unchanged; otherwise, the non-normal distribution data is converted into normal distribution data by Box-Cox transformation; The decision tree model building module is configured to calculate the Gini system of each judgment attribute and arrange them in order to initially form a decision tree; The data inverse transformation module is configured to perform relevant inverse transformation on the data that has undergone Box-Cox transformation in the decision tree; The decision tree model optimization module is configured to perform pruning optimization on the decision tree after the inverse transformation to ensure that the minimum number of instances is greater than the set value, thereby forming the final improved decision tree model; A parameter acquisition module is configured to obtain parameters related to the construction object and construction requirements of the road roller; The classification analysis module is configured to perform classification analysis based on the construction object and construction requirement-related parameters using the improved decision tree model to obtain the roller construction plan and form a roller construction process knowledge base; The roller construction object includes soil type, bottom length, bottom width, depth and slope coefficient of the construction object; The parameters related to the roller construction requirements include construction requirement attributes, roller model, and the roller amplitude, frequency, rolling speed, and number of rolling passes in the three stages of initial rolling, secondary rolling, and final rolling. The construction requirement attributes include compaction, flatness, and uniformity. The two major attributes of construction objects and construction requirements are used as judgment attributes in the decision tree model. Different judgment attributes constitute the root node and internal nodes of the decision tree. The levels divided by the judgment attributes are the branches of the decision tree. The roller model and the key parameters of the roller in the three stages of initial, secondary, and final compaction are used as the decision results of the decision tree. Each leaf node reached by the decision tree through each fork represents a roller construction plan, and each construction plan corresponds to a leaf node of the decision tree. The specific process of calculating the Gini coefficients of each judgment attribute and arranging them in order to initially form a decision tree includes: taking each attribute of the construction object and construction requirement related parameters as a judgment attribute, calculating the Gini coefficient, arranging them in ascending order, and using the attribute with the highest Gini coefficient as the initial root node, and the attribute's grade as the branch extending from the root node; Calculate the Gini coefficients of the remaining attributes in the sub-knowledge base corresponding to each bifurcation. The attribute with the lowest Gini coefficient is used as the first internal node after the root node bifurcates. Calculate the coefficients step by step, and extend downward from low to high according to the Gini coefficients in the process of building the decision tree. If a case lacks some judgment attributes, this case will be discarded as incomplete data.

7. A computer-readable storage medium, characterized in that: There are a plurality of instructions, which are suitable for being loaded by a processor of a terminal device and executing the steps of the method according to any one of claims 1 to 5.

8. A terminal device, characterized in that: The method comprises a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; and the computer-readable storage medium is used to store a plurality of instructions, wherein the instructions are suitable for being loaded by the processor and executing the steps of the method according to any one of claims 1 to 5.

Images