Roadbed construction compaction index optimization system and method thereof

Abstract

The invention relates to the technical field of roadbed construction, and discloses a roadbed construction compaction index optimization system and method. Comprising the following steps: screening demand data from first original data, extracting environment data and construction data from second original data, grouping compaction indexes into an index set, preferably selecting a target set corresponding to preposed data through a compaction index optimization model, analyzing optimal indexes from the target set, and calculating the optimal indexes according to the optimal indexes. The evaluation grade of the optimal index is formulated; compared with the prior art, the method can perform multi-dimensional representation on user requirements, construction requirements and environment conditions in roadbed construction, avoids the problem of limitation of a compaction index optimization result, can perform rapid and accurate optimization operation on the compaction indexes through the artificial intelligence model, and improves the construction efficiency. The inaccuracy caused by the fact that the compaction indexes are optimized according to artificial experience and a data query mode is avoided, and the optimization effect of the roadbed construction compaction indexes is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of roadbed construction, and more specifically, to a roadbed construction compaction index optimization system and method thereof. Background Art

[0002] The compaction index of the roadbed is a key indicator for evaluating its bearing capacity and stability, and is also the basis for evaluating the quality of roadbed construction. Different types of compaction indicators correspond to the compaction quality evaluation effects of different dimensions of roadbed construction. In order to conduct a comprehensive and accurate evaluation of the compaction quality of roadbed construction, it is necessary to use the compaction index optimization system to select appropriate compaction indicators from a large number of compaction indicators according to different compaction quality analysis requirements.

[0003] The patent application with reference publication number CN118607973A discloses a method and system for optimizing compaction indexes for roadbed construction, which extracts features from collected data, evaluates the correlation between the extracted features and construction compaction effect indexes through statistical analysis or machine learning algorithms, and determines the weights of each feature index in combination with the analytic hierarchy process, thereby scientifically and objectively considering the weight distribution of intelligent roadbed construction compaction indexes, establishing an optimization model based on the determined feature index weights and feature indexes, and performing decision prediction and optimization on the established optimization model, thereby developing an intelligent decision support system, and generating decision and optimization plans based on the developed system, which can be applied to different actual environments and meet the needs of complex and changeable actual scenarios; The existing compaction index optimization system collects construction information and external environment information during roadbed construction, and performs a large amount of analysis and calculation on the collected information, and then relies on a model to optimize the compaction index. For example, in the above-mentioned patent application, it collects and performs a large amount of calculation on data related to soil characteristics, meteorological conditions and engineering equipment, and combines the optimization model to optimize the compaction index. This optimization method does not take into account the personalized demand data of actual roadbed construction compaction evaluators, making the reference factors involved in the subsequent compaction index optimization not comprehensive enough, resulting in limitations in the optimization results of the compaction index, and the accuracy of the compaction index optimization results cannot be guaranteed.

[0004] In view of this, the present invention proposes a roadbed construction compaction index optimization system and method to solve the above problems. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solutions: a roadbed construction compaction index optimization system, applied to an optimization server, comprising: A first data acquisition module is used to receive first original data from a user terminal, and filter out demand data from the first original data, where the demand data includes calculation accuracy, effective optimization duration, and index optimization value; A second data acquisition module is used to obtain second original data from the construction end, and extract environmental data and construction data from the second original data; An indicator set generation module, used to group compaction indicators in the indicator library based on grouping criteria to generate an indicator set; The target set optimization module is used to aggregate demand data, environmental data and construction data into pre-data, and select the target set corresponding to the pre-data through the trained compaction index optimization model; The target set parsing module is used to parse out the preferred indicators from the target set and formulate evaluation levels for the preferred indicators. The evaluation levels include simple level, general level and difficult level.

[0006] Furthermore, the screening method for calculating accuracy, effective optimization time and index optimization value is as follows: The time when the user first inputs data is taken as the starting time, the current time is taken as the ending time, and the period between the starting time and the ending time is recorded as the demand period; Querying the database for all data input by the user during the required period, and arranging all the data in the order of input time to obtain first original data; The first semantics of all the first original data are identified one by one by using natural language processing technology, and the first semantics are split into a text part and a numeric part by using word segmentation technology; The first semantics of the text parts, namely, precision, duration, and quantity, are recorded as precise semantics, duration semantics, and quantity semantics, respectively, and the precise semantics, duration semantics, and quantity semantics are sorted according to the order of input time to obtain a first queue, a second queue, and a third queue; The precise semantics of the last digit in the first queue, the duration semantics of the last digit in the second queue, and the quantity semantics of the last digit in the third queue are respectively combined with the corresponding digital parts to generate calculation accuracy, effective preferred duration, and indicator preferred value.

[0007] Furthermore, the environmental data includes construction temperature value, construction humidity value, construction wind speed value and soil type; The extraction method of construction temperature value, construction humidity value, construction wind speed value and soil type is: The construction period of the roadbed construction is queried through the database, and the construction period is divided into A consecutive sub-periods of equal length; The temperature sensor, humidity sensor and wind speed sensor are used to detect the A sub-periods. The temperature, humidity and wind speed values ​​at each detection time are obtained. Real-time temperature values, Real-time humidity values ​​and Real-time wind speed value; Remove the maximum and minimum values ​​of the real-time temperature, real-time humidity and real-time wind speed respectively, and Real-time temperature values, Real-time humidity values ​​and The real-time wind speed values ​​are accumulated one by one and averaged to obtain A sub-temperature values, A sub-humidity values ​​and A sub-wind speed values; The maximum and minimum values ​​of the sub-temperature value, the sub-humidity value and the sub-wind speed value are added and averaged respectively to obtain the construction temperature value, the construction humidity value and the construction wind speed value; Soil data for roadbed construction is queried from the database, keywords of soil data are identified through natural language processing technology, and the keywords are recorded as soil types; Soil types include clay type, sand type, silt type and gravel type.

[0008] Furthermore, the construction data include the light wheel pressure fluctuation value, the number of compaction times, the compaction duration and the compaction temperature; The method for extracting the light wheel pressure fluctuation value is: The database is used to query the time when the roller performs the first and last bare wheel compaction work during the construction period, which are recorded as the compaction start time and compaction end time respectively; From the start time to the end time of compaction, the light wheel pressure of the roller is detected in real time by the pressure sensor, and the time when the light wheel pressure is greater than or equal to the preset pressure threshold for the first time is recorded as the effective start time; From the effective start time to the compaction end time, the time when the light wheel pressure is last less than the preset pressure threshold is recorded as the effective end time, and the period between the effective start time and the effective end time is recorded as the compaction period; During the compaction period, randomly mark non-adjacent query times, and find out The light wheel pressure at the query time is obtained query pressure; After subtracting the query pressure at the next query time from the query pressure at the previous query time, we can get Pressure difference; Will The pressure differences are accumulated and averaged to obtain the light wheel pressure fluctuation value.

[0009] Furthermore, the grouping criteria are: there are no repeated compaction indicators in the indicator set, and the number of compaction indicators in the indicator set is more than two.

[0010] Furthermore, the method for generating the indicator set is: Arrange all compaction indicators in the indicator library in order of storage time to generate an indicator queue; Taking the first compaction indicator in the indicator queue as the starting point and the last compaction indicator in the indicator queue as the end point, randomly select an indefinite amount of compaction indicators from the indicator queue, and combine the selected compaction indicators to obtain E sets to be verified; Count the number of compaction indicators in the E sets to be verified one by one, and remove the sets to be verified with a number of compaction indicators less than 3, to obtain F sets to be identified; Repeatedly identify the compaction indicators in the F to-be-identified sets one by one, eliminate the to-be-identified sets with repeated compaction indicators, and record the remaining to-be-identified sets as indicator sets to obtain H indicator sets.

[0011] Furthermore, the training method of the compaction index optimization model is: Collect multiple sets of pre-data for roadbed construction and the index sets corresponding to the pre-data in advance, and convert the soil types into characters in sequence, converting clay types into NT, sand types into ST, silt types into FT, and gravel types into SS; Mark each set of previous data as a training feature, number the indicator set of each set of training features, and record the number value of the indicator set; The training features are used as the input data of the compaction indicator optimization model, and the number values ​​are used as the output data of the compaction indicator optimization model. The labeled training features are divided into a training set and a test set. The training set is used to train the compaction indicator optimization model, and the test set is used to test the compaction indicator optimization model. An error threshold is preset. When the mean of the prediction errors of all training features in the test set is less than the error threshold, the compaction indicator optimization model is obtained.

[0012] Furthermore, the preferred numbering value is outputted by compacting the index optimization model, and the index set whose numbering value in the index set is consistent with the preferred numbering value is recorded as the target set.

[0013] Furthermore, the method for formulating the simple level, general level and difficult level is as follows: The conventional collection time and the standard number of personnel for the S preferred indicators are queried one by one through the database, and S difficulty levels are calculated based on the S conventional collection time and the S standard number of personnel; The number of preferred indicators whose difficulty is greater than the difficulty threshold is counted and recorded as the difficulty value, and one-fifth of the number of preferred indicators and one-half of the number of preferred indicators are recorded as the first critical value and the second critical value respectively; When the difficulty value is less than or equal to the first critical value, an easy level is formulated; When the first critical value is less than the difficulty value, and the difficulty value is less than or equal to the second critical value, a general level is formulated; When the difficulty value is greater than the second critical value, a difficulty level is determined.

[0014] A method for optimizing compaction indexes for roadbed construction is implemented based on the above-mentioned system for optimizing compaction indexes for roadbed construction, and includes: S1: receiving first original data from a user terminal, and filtering out demand data from the first original data, the demand data including calculation accuracy, effective optimization duration, and index optimization value; S2: Acquire the second original data of the construction end, and extract the environment data and the construction data from the second original data; S3: based on the grouping criteria, group the compaction indicators in the indicator library to generate an indicator set; S4: Aggregate demand data, environmental data and construction data into pre-data, and select the target set corresponding to the pre-data through the trained compaction index optimization model; S5: Analyze the preferred indicators from the target set and formulate evaluation levels for the preferred indicators. The evaluation levels include simple level, general level and difficult level.

[0015] The technical effects and advantages of the roadbed construction compaction index optimization system and method of the present invention are as follows: The present invention receives the first original data from the user end, filters out the demand data from the first original data, obtains the second original data from the construction end, and extracts the environmental data and construction data from the second original data, so as to make a multi-dimensional comprehensive representation of the user needs, construction requirements and environmental conditions in the roadbed construction, thereby providing comprehensive and diversified data support for the compaction quality evaluation and analysis of the roadbed construction, avoiding the problem of limitations in the subsequent compaction index optimization results, improving the accuracy of the compaction index optimization results, and based on the grouping criteria, grouping the compaction indicators in the indicator library to generate an indicator set, so that an indefinite number of classification and combination operations can be performed on a large number of compaction indicators of different types, so that the indicator set can be used for any different number and type of demand data , environmental data and construction data are matched accordingly, providing a screening basis for the subsequent optimization of compaction indicators, by aggregating demand data, environmental data and construction data into pre-data, and optimizing the target set corresponding to the pre-data through the trained compaction indicator optimization model, parsing the preferred indicators from the target set, and formulating the evaluation level of the preferred indicators. It can not only quickly and accurately optimize the compaction indicators through the artificial intelligence model, avoiding the inaccuracy caused by relying on manual experience and data query methods to optimize the compaction indicators, but also indicate the difficulty of subsequent collection and evaluation of the selected compaction indicators, thereby providing comprehensive, accurate and intuitive opinion guidance for the optimization operation of the compaction indicators of roadbed construction, and improving the optimization effect of the compaction indicators of roadbed construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the architecture of a roadbed construction compaction index optimization system provided in the first embodiment of the present invention; Figure 2 A schematic diagram of a module of a roadbed construction compaction index optimization system provided in the first embodiment of the present invention; Figure 3 A schematic flow chart of a method for optimizing compaction indicators for roadbed construction provided in Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Example 1: Please refer to Figure 1 and Figure 2 As shown, the roadbed construction compaction index optimization system described in this embodiment is applied to the optimization server, including: A first data acquisition module receives first original data from a user terminal and filters out required data from the first original data; The user end refers to the end that needs to use the system to optimize the compaction index of the roadbed construction, and serves as a data interaction and association platform between the system and the user. The first original data refers to all the basic data generated and collected in the user end, so that the first original data can be used as a comprehensive representation of the individual needs and situations of different user ends, and as a basis for subsequent screening of demand data; Demand data refers to the demand for the user end to realize the unconventional personalized index optimization operation of the compaction index optimization operation of the roadbed construction, and serves as an accurate representation of the compaction index optimization of the user end, and is also one of the factors affecting the subsequent compaction index optimization results; The required data include calculation accuracy, effective optimization time and index optimization value; Calculation accuracy refers to the minimum accuracy of the data calculated in the compaction index when the user side optimizes the compaction index for roadbed construction. It can be used as one of the reference factors for subsequent optimization of the compaction index. The greater the calculation accuracy, the higher the user side's requirements for the accuracy of the data calculation in the compaction index.

[0019] The effective optimization time refers to the maximum limit on the time for collecting relevant data in the compaction indicators when the user side optimizes the compaction indicators for roadbed construction. It can be used as one of the reference factors for the subsequent optimization of the compaction indicators. The longer the effective optimization time is, the lower the user side's requirements for the time for collecting relevant data in the compaction indicators.

[0020] The index optimization value refers to the expression of the specific number of indicators in the compaction index when the user end optimizes the compaction index for roadbed construction, so that it can be used as one of the reference factors for subsequent optimization of the compaction index. The larger the index optimization value, the more requirements the user end has for the specific number of compaction indicators.

[0021] The screening method for calculating accuracy, effective optimization time and index optimization value is: The time when the user first inputs data is taken as the starting time, the current time is taken as the ending time, and the period between the starting time and the ending time is recorded as the demand period; All data input by the user during the required period are queried through the database, and all data are arranged in the order of input time to obtain the first original data; by arranging in chronological order, the orderliness of the data can be ensured, and the phenomenon of data chaos and disorder can be avoided, which is convenient for subsequent data search and analysis; The first semantics of all the first original data are identified one by one through natural language processing technology, and the first semantics are split into a text part and a number part through word segmentation technology; the first semantics is used to concisely express the real meaning expressed by the first original data, that is, to achieve a simplification effect of complex data, and the text part and the number part are part of the first semantics, which are used to accurately express the text and numbers of the first original data; The first semantics of the text parts, namely, precision, duration, and quantity, are recorded as precise semantics, duration semantics, and quantity semantics, respectively, and the precise semantics, duration semantics, and quantity semantics are sorted according to the order of input time to obtain a first queue, a second queue, and a third queue; The precise semantics of the last digit in the first queue, the duration semantics of the last digit in the second queue, and the quantity semantics of the last digit in the third queue are respectively combined with the corresponding digital parts to generate calculation accuracy, effective preferred duration, and indicator preferred value.

[0022] It should be noted that since the user may make input errors or have different requirements at different times when entering relevant demand information, in order to maximize the accuracy of the user's preferred data in the compaction indicator, it is necessary to take the data with the latest input time as the object.

[0023] A second data acquisition module acquires second original data from the construction end, and extracts environmental data and construction data from the second original data; The construction end refers to the end that can provide the system with comprehensive roadbed construction data required for compaction index optimization operations, so that the construction end can provide the system with sufficient roadbed construction data. The second original data refers to the data generated in the construction end that can comprehensively represent the specific situation of the roadbed construction and serve as the basis for extracting subsequent environmental data and construction data. Environmental data refers to the data used by the construction end to evaluate the environment in which the roadbed is located when optimizing the compaction index of the roadbed construction. It is also one of the factors that affect the subsequent compaction index optimization results. Environmental data include construction temperature, construction humidity, construction wind speed and soil type; The construction temperature value refers to the ambient temperature of the location during the overall construction process of the roadbed. Different ambient temperatures will have different effects on the compaction operation of the roadbed construction, thereby affecting the optimization results of the subsequent compaction indicators. The construction humidity value refers to the environmental humidity at the location during the overall construction process of the roadbed. Different environmental humidity will have different effects on the compaction operation of the roadbed construction, thereby affecting the optimization results of the subsequent compaction indicators. The construction wind speed value refers to the environmental wind intensity at the location during the overall construction process of the roadbed. Different wind intensities will have different effects on the evaporation of water in the roadbed construction, thus affecting the optimization results of the subsequent compaction indicators. Soil type refers to the type of soil at the location where the roadbed construction is located. Different soil types will have different effects on the compaction degree of roadbed construction; The extraction method of construction temperature value, construction humidity value, construction wind speed value and soil type is: The construction period of the roadbed construction is queried through the database, and the construction period is divided into A continuous and equal sub-periods; the construction period refers to the period from the beginning of the roadbed construction to the end of the roadbed construction. By dividing the construction period into sub-periods, the consistency of the length of each sub-period can be ensured, and the time period with a large span can be finely divided, reducing the amount of data in each sub-period, which is convenient for subsequent data query and calculation; The temperature sensor, humidity sensor and wind speed sensor are used to detect the A sub-periods. The temperature, humidity and wind speed values ​​at each detection time are obtained. Real-time temperature values, Real-time humidity values ​​and Real-time wind speed value; Remove the maximum and minimum values ​​of the real-time temperature, real-time humidity and real-time wind speed respectively, and Real-time temperature values, Real-time humidity values ​​and The real-time wind speed values ​​are accumulated one by one and averaged to obtain A sub-temperature values, A sub-humidity values ​​and A sub-wind speed values; The expression of the sub-temperature value is: ; In the formula, For the The sub-temperature value of each sub-period, =1,2...A, For the The sub-period Real-time temperature values; The expression of the sub-humidity value is: ; In the formula, For the The humidity value of each sub-period, For the The sub-period Real-time humidity value; The expression of the sub-wind speed value is: ; In the formula, For the The sub-wind speed value of each sub-period, For the The sub-period Real-time wind speed value; The maximum and minimum values ​​of the sub-temperature value, the sub-humidity value and the sub-wind speed value are added and averaged respectively to obtain the construction temperature value, the construction humidity value and the construction wind speed value; The expression of construction temperature value is: ; In the formula, is the construction temperature value, is the maximum value of the sub-temperature value, is the minimum value of the sub-temperature value; The expression of construction humidity value is: ; In the formula, is the construction humidity value, is the maximum value of the sub-humidity value, is the minimum value of the sub-humidity value; The expression of construction wind speed value is: ; In the formula, is the construction wind speed value, is the maximum value of the sub-wind speed value, is the minimum value of the sub-wind speed value; The soil data of roadbed construction is queried through the database, the keywords of the soil data are identified through natural language processing technology, and the keywords are recorded as soil types.

[0024] Specifically, soil types include clay type, sand type, silt type, and gravel type. Each type of soil will have different effects on the compaction performance of subsequent roadbed construction, thereby affecting the results of subsequent compaction index optimization.

[0025] Construction data refers to the operating data of the equipment used for roadbed construction when the construction end is optimizing the compaction index of roadbed construction. It is also one of the factors that affect the subsequent compaction index optimization results. Construction data include light wheel pressure fluctuation value, compaction times, compaction duration and compaction temperature; The light wheel pressure fluctuation value refers to the variation of the light wheel pressure on the roadbed when the light wheel of the roller is compacting the roadbed, which can be used to indicate the compaction stability of the roller. The larger the light wheel pressure fluctuation value, the worse the compaction stability of the roller. The method for extracting the light wheel pressure fluctuation value is: The database is used to query the time when the roller performs the first and last bare wheel compaction work during the construction period, which are recorded as the compaction start time and compaction end time respectively; From the start time to the end time of compaction, the light wheel pressure of the roller is detected in real time through the pressure sensor, and the moment when the light wheel pressure is greater than or equal to the preset pressure threshold for the first time is recorded as the effective start time; the preset pressure threshold refers to the minimum value of the light wheel pressure that can keep the roller in a normal and stable rolling state, which can provide a basis for the collection range of the light wheel pressure to ensure that the collected light wheel pressure can correspond to the roller being in a normal and stable rolling state; From the effective start time to the compaction end time, the time when the light wheel pressure is last less than the preset pressure threshold is recorded as the effective end time, and the period between the effective start time and the effective end time is recorded as the compaction period; During the compaction period, randomly mark non-adjacent query times, and find out The light wheel pressure at the query time is obtained query pressure; After subtracting the query pressure at the next query time from the query pressure at the previous query time, we can get Pressure difference; The expression of pressure difference is: ; In the formula, For the The pressure difference, =1,2... , For the The query pressure at each query time, For the The query pressure at each query time; Will The pressure differences are accumulated and averaged to obtain the light wheel pressure fluctuation value; The expression of the light wheel pressure fluctuation value is: ; In the formula, is the light wheel pressure fluctuation value, For the A pressure difference.

[0026] The number of compaction times, compaction duration and compaction temperature respectively refer to the amount of compaction of the roadbed by the roller, the duration of compaction and the compaction temperature when the roller is compacting the roadbed, thereby indicating the specific compaction parameters of the roller; the number of compaction times, compaction duration and compaction temperature are all obtained through database query.

[0027] The indicator set generation module groups the compaction indicators in the indicator library based on the grouping criteria to generate an indicator set; The index library refers to a collection of various compaction indicators that can evaluate the quality of roadbed construction compaction, so that the compaction indicators stored in the index library can meet the multi-dimensional indicator selection needs of roadbed construction compaction evaluation under various types, different needs, and different combinations; Since the compaction indicators for evaluating the compaction quality of roadbed construction corresponding to different demand data, environmental data and construction data are different, and the number of compaction indicators used to evaluate the compaction quality of roadbed construction usually needs to be greater than 2, therefore, in order to ensure that the compaction indicators in the indicator library can meet the corresponding compaction optimization requirements of a large amount of different demand data, environmental data and construction data, it is necessary to classify and combine the compaction indicators in the indicator library according to different demand data, environmental data and construction data, and combine the compaction indicators corresponding to each group of different demand data, environmental data and construction data into an indicator set.

[0028] Compaction indicators include but are not limited to compaction degree, compaction meter value, elastic modulus, California bearing ratio, base bed coefficient, water content, porosity, relative density, shear strength, Beckmann deflection, etc., so that a large number of compaction indicators can be grouped in different quantities and types; When grouping compaction indicators in the indicator library, it is necessary to group them according to the restrictions of the grouping criteria to ensure that when the compaction indicators are grouped, the compaction indicators in each indicator set are kept in a unique state, and the number of compaction indicators in each indicator set is at least three; The grouping criteria are: there are no repeated compaction indicators in the indicator set, and the number of compaction indicators in the indicator set is more than two; that is, it is ensured that the compaction indicators in the indicator set can maintain uniqueness, and the number of compaction indicators is at least three; The method for generating the indicator set is: Arrange all compaction indicators in the indicator library in order of storage time to generate an indicator queue; Taking the first compaction indicator in the indicator queue as the starting point and the last compaction indicator in the indicator queue as the end point, randomly select an indefinite amount of compaction indicators from the indicator queue, and combine the selected compaction indicators to obtain E sets to be verified; Count the number of compaction indicators in the E sets to be verified one by one, and remove the sets to be verified with a number of compaction indicators less than 3, to obtain F sets to be identified; Repeatedly identify the compaction indicators in the F to-be-identified sets one by one, eliminate the to-be-identified sets with repeated compaction indicators, and record the remaining to-be-identified sets as indicator sets to obtain H indicator sets.

[0029] The target set optimization module aggregates demand data, environmental data and construction data into pre-data, and selects the target set corresponding to the pre-data through the trained compaction index optimization model; Pre-data is used to summarize the demand data, environment data and construction data as a whole, that is, to summarize the scattered demand data, environment data and construction data in an orderly manner and form the pre-data of the overall structure; The compaction index optimization model refers to a machine learning model that can select a corresponding target set from an index set based on the pre-data consisting of demand data, environmental data and construction data. At this time, the target set is one of the index sets, and multiple compaction indicators in the target set can match the demand data, environmental data and construction data corresponding to the pre-data; The training method of the compaction index optimization model is: Collect multiple sets of pre-data of roadbed construction and index sets corresponding to the pre-data in advance, and convert soil types into characters in sequence. For example, convert clay type into NT, convert sand type into ST, convert silt type into FT, and convert gravel type into SS; Mark each group of pre-data as a training feature, number the indicator set of each group of training features, and record the number value of the indicator set. For example, when there are W groups of pre-data, the W indicator sets corresponding to the W groups of pre-data are numbered in ascending order one by one, so that the number values ​​of the W indicator sets are 1, 2...W; The training features are used as the input data of the compaction indicator optimization model, and the number values ​​are used as the output data of the compaction indicator optimization model. The labeled training features are divided into a training set and a test set. 70% of the training features are used as the training set, and 30% of the training features are used as the test set. The training set is used to train the compaction indicator optimization model, and the test set is used to test the compaction indicator optimization model. An error threshold is preset. When the mean of the prediction errors of all training features in the test set is less than the error threshold, the compaction indicator optimization model is obtained.

[0030] The compaction index optimization model adopts either a support vector machine model or a random forest model.

[0031] After the compaction index optimization model is trained, the collected prior data can be input into the compaction index optimization model, and the number value corresponding to the prior data can be optimized. At this time, based on the preferred number value, an indicator set with a number value consistent with the preferred number value is screened out from the indicator set, that is, as the target set, and the compaction indicators in the target set are used as the compaction indicators corresponding to the prior data of roadbed construction.

[0032] The target set parsing module parses the preferred indicators from the target set and formulates the evaluation level of the preferred indicators; The preferred index refers to all the compaction indexes in the target set selected by the compaction index optimization model, so that the preferred index can be used as the corresponding index data for the user end to evaluate and detect the compaction quality of roadbed construction, and provide guidance for subsequent evaluation and detection; After obtaining the preferred indicators, it is necessary to give specific analysis opinions on the preferred results of the preferred indicators according to the specific number and type of the preferred indicators, so as to provide reasonable and accurate difficulty prompts for the subsequent evaluation and collection of the preferred indicators; The evaluation levels include simple level, general level and difficult level; the simple level means that the subsequent collection and evaluation of the preferred indicators is of low difficulty, the general level means that the subsequent collection and evaluation of the preferred indicators is of medium difficulty, and the difficult level means that the subsequent collection and evaluation of the preferred indicators is of high difficulty; The method for formulating the simple level, general level and difficult level is as follows: The conventional collection time and standard number of personnel for the S preferred indicators are queried one by one through the database, and S difficulty levels are calculated based on the S conventional collection time and the S standard number of personnel; the conventional collection time refers to the time required for personnel to collect the preferred indicators, which is used to express the difficulty level of collecting the preferred indicators on one hand, and the standard number of personnel refers to the number of personnel required to collect the preferred indicators, which is used to express the difficulty level of collecting the preferred indicators on the other hand; The expression of difficulty is: ; In the formula, For the The difficulty of the selected indicator, =1,2...S, For the The regular collection time of the preferred indicators is For the The standard number of personnel for each preferred indicator, , is a weight factor greater than 0, and ; The number of preferred indicators whose difficulty is greater than the difficulty threshold is counted and recorded as the difficulty value, and one-fifth of the number of preferred indicators and one-half of the number of preferred indicators are recorded as the first critical value and the second critical value respectively; the difficulty threshold refers to the minimum value of the difficulty corresponding to the preferred indicators that are evaluated as difficult to collect, so that the specific difficulty levels of different preferred indicators can be distinguished; When the difficulty value is less than or equal to the first critical value, it means that the number of indicators with greater difficulty in collection and evaluation operations among the preferred indicators is small, and a simple level is formulated; When the first critical value is less than the difficulty value, and the difficulty value is less than or equal to the second critical value, it means that the number of indicators with high difficulty in collection and evaluation operations among the preferred indicators is medium, and a general level is formulated; When the difficulty value is greater than the second critical value, it means that there are a large number of indicators with high difficulty in collection and evaluation operations among the preferred indicators, and a difficulty level is formulated.

[0033] In this embodiment, by receiving the first original data from the user end, filtering out the demand data from the first original data, obtaining the second original data from the construction end, and extracting the environmental data and construction data from the second original data, it is possible to make a multi-dimensional comprehensive representation of user needs, construction requirements and environmental conditions in roadbed construction, thereby providing comprehensive and diversified data support for the compaction quality evaluation and analysis of roadbed construction, avoiding the problem of limitations in the subsequent compaction index optimization results, improving the accuracy of the compaction index optimization results, and based on the grouping criteria, grouping the compaction indicators in the indicator library to generate an indicator set, which can perform an indefinite number of classification and combination operations on massive and different types of compaction indicators, so that the indicator set can be used for any different number and type of needs. The corresponding matching of data, environmental data and construction data is carried out to provide a screening basis for the subsequent optimization of compaction indicators. By aggregating the demand data, environmental data and construction data into pre-data, and optimizing the target set corresponding to the pre-data through the trained compaction indicator optimization model, the preferred indicators are parsed from the target set, and the evaluation level of the preferred indicators is formulated. It can not only quickly and accurately optimize the compaction indicators through the artificial intelligence model, avoiding the inaccuracy caused by relying on manual experience and data query methods to optimize the compaction indicators, but also indicate the difficulty of subsequent collection and evaluation of the selected compaction indicators, thereby providing comprehensive, accurate and intuitive opinion guidance for the optimization operation of the compaction indicators of roadbed construction, and improving the optimization effect of the compaction indicators of roadbed construction.

[0034] Example 2: Please refer to Figure 3 As shown, the part not described in detail in this embodiment is described in the first embodiment, and a method for optimizing compaction indexes for roadbed construction is provided, which is applied to an optimization server and is implemented based on a system for optimizing compaction indexes for roadbed construction, including: S1: receiving first original data from a user terminal, and filtering out demand data from the first original data, the demand data including calculation accuracy, effective optimization duration, and index optimization value; S2: Acquire the second original data of the construction end, and extract the environment data and the construction data from the second original data; S3: based on the grouping criteria, group the compaction indicators in the indicator library to generate an indicator set; S4: Aggregate demand data, environmental data and construction data into pre-data, and select the target set corresponding to the pre-data through the trained compaction index optimization model; S5: Analyze the preferred indicators from the target set and formulate evaluation levels for the preferred indicators. The evaluation levels include simple level, general level and difficult level.

[0035] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A roadbed construction compaction index optimization system, applied to an optimization server, characterized in that: include: A first data acquisition module is used to receive first original data from a user terminal, and filter out demand data from the first original data, where the demand data includes calculation accuracy, effective optimization duration, and index optimization value; A second data acquisition module is used to obtain second original data from the construction end, and extract environmental data and construction data from the second original data; An indicator set generation module, used to group compaction indicators in the indicator library based on grouping criteria to generate an indicator set; The target set optimization module is used to aggregate demand data, environmental data and construction data into pre-data, and select the target set corresponding to the pre-data through the trained compaction index optimization model; The target set parsing module is used to parse out the preferred indicators from the target set and formulate evaluation levels for the preferred indicators. The evaluation levels include simple level, general level and difficult level.

2. A roadbed construction compaction index optimization system according to claim 1, characterized in that: The screening method for calculating accuracy, effective optimization time and index optimization value is: The time when the user first inputs data is taken as the starting time, the current time is taken as the ending time, and the period between the starting time and the ending time is recorded as the demand period; Querying the database for all data input by the user during the required period, and arranging all the data in the order of input time to obtain first original data; The first semantics of all the first original data are identified one by one by using natural language processing technology, and the first semantics are split into a text part and a numeric part by using word segmentation technology; The first semantics of the text parts, namely, precision, duration, and quantity, are recorded as precise semantics, duration semantics, and quantity semantics, respectively, and the precise semantics, duration semantics, and quantity semantics are sorted according to the order of input time to obtain a first queue, a second queue, and a third queue; The precise semantics of the last digit in the first queue, the duration semantics of the last digit in the second queue, and the quantity semantics of the last digit in the third queue are respectively combined with the corresponding digital parts to generate calculation accuracy, effective preferred duration, and indicator preferred value.

3. A roadbed construction compaction index optimization system according to claim 2, characterized in that: Environmental data include construction temperature, construction humidity, construction wind speed and soil type; The extraction method of construction temperature value, construction humidity value, construction wind speed value and soil type is: The construction period of the roadbed construction is queried through the database, and the construction period is divided into A consecutive sub-periods of equal length; The temperature sensor, humidity sensor and wind speed sensor are used to detect the A sub-periods. The temperature, humidity and wind speed values ​​at each detection time are obtained. Real-time temperature values, Real-time humidity values ​​and Real-time wind speed value; Remove the maximum and minimum values ​​of the real-time temperature, real-time humidity and real-time wind speed respectively, and Real-time temperature values, Real-time humidity values ​​and The real-time wind speed values ​​are accumulated one by one and averaged to obtain A sub-temperature values, A sub-humidity values ​​and A sub-wind speed values; The maximum and minimum values ​​of the sub-temperature value, the sub-humidity value and the sub-wind speed value are added and averaged respectively to obtain the construction temperature value, the construction humidity value and the construction wind speed value; Soil data for roadbed construction is queried from the database, keywords of soil data are identified through natural language processing technology, and the keywords are recorded as soil types; Soil types include clay type, sand type, silt type and gravel type.

4. A roadbed construction compaction index optimization system according to claim 3, characterized in that: Construction data include light wheel pressure fluctuation value, compaction times, compaction duration and compaction temperature; The method for extracting the light wheel pressure fluctuation value is: The database is used to query the time when the roller performs the first and last bare wheel compaction work during the construction period, which are recorded as the compaction start time and compaction end time respectively; From the start time to the end time of compaction, the light wheel pressure of the roller is detected in real time by the pressure sensor, and the time when the light wheel pressure is greater than or equal to the preset pressure threshold for the first time is recorded as the effective start time; From the effective start time to the compaction end time, the time when the light wheel pressure is last less than the preset pressure threshold is recorded as the effective end time, and the period between the effective start time and the effective end time is recorded as the compaction period; During the compaction period, randomly mark non-adjacent query times, and find out The light wheel pressure at the query time is obtained query pressure; After subtracting the query pressure at the next query time from the query pressure at the previous query time, we can get Pressure difference; Will The pressure differences are accumulated and averaged to obtain the light wheel pressure fluctuation value.

5. A roadbed construction compaction index optimization system according to claim 4, characterized in that: The grouping criteria are: there are no duplicate compaction indicators in the indicator set, and the number of compaction indicators in the indicator set is more than two.

6. A roadbed construction compaction index optimization system according to claim 5, characterized in that: The method for generating the indicator set is: Arrange all compaction indicators in the indicator library in order of storage time to generate an indicator queue; Taking the first compaction indicator in the indicator queue as the starting point and the last compaction indicator in the indicator queue as the end point, randomly select an indefinite amount of compaction indicators from the indicator queue, and combine the selected compaction indicators to obtain E sets to be verified; Count the number of compaction indicators in the E sets to be verified one by one, and remove the sets to be verified with a number of compaction indicators less than 3, to obtain F sets to be identified; Repeatedly identify the compaction indicators in the F to-be-identified sets one by one, eliminate the to-be-identified sets with repeated compaction indicators, and record the remaining to-be-identified sets as indicator sets to obtain H indicator sets.

7. A roadbed construction compaction index optimization system according to claim 6, characterized in that: The training method of the compaction index optimization model is: Collect multiple sets of pre-data for roadbed construction and the index sets corresponding to the pre-data in advance, and convert the soil types into characters in sequence, converting clay types into NT, sand types into ST, silt types into FT, and gravel types into SS; Mark each set of previous data as a training feature, number the indicator set of each set of training features, and record the number value of the indicator set; The training features are used as the input data of the compaction indicator optimization model, and the number values ​​are used as the output data of the compaction indicator optimization model. The labeled training features are divided into a training set and a test set. The training set is used to train the compaction indicator optimization model, and the test set is used to test the compaction indicator optimization model. An error threshold is preset. When the mean of the prediction errors of all training features in the test set is less than the error threshold, the compaction indicator optimization model is obtained.

8. A roadbed construction compaction index optimization system according to claim 7, characterized in that: The preferred numbering value is output by the compaction indicator optimization model, and the indicator set whose numbering value in the indicator set is consistent with the preferred numbering value is recorded as the target set.

9. A roadbed construction compaction index optimization system according to claim 8, characterized in that: The method for formulating the simple level, general level and difficult level is as follows: The conventional collection time and the standard number of personnel for the S preferred indicators are queried one by one through the database, and S difficulty levels are calculated based on the S conventional collection time and the S standard number of personnel; The number of preferred indicators whose difficulty is greater than the difficulty threshold is counted and recorded as the difficulty value, and one-fifth of the number of preferred indicators and one-half of the number of preferred indicators are recorded as the first critical value and the second critical value respectively; When the difficulty value is less than or equal to the first critical value, an easy level is formulated; When the first critical value is less than the difficulty value, and the difficulty value is less than or equal to the second critical value, a general level is formulated; When the difficulty value is greater than the second critical value, a difficulty level is determined.

10. A method for optimizing compaction indexes for roadbed construction, implemented based on a system for optimizing compaction indexes for roadbed construction according to any one of claims 1 to 9, characterized in that: include: S1: receiving first original data from a user terminal, and filtering out demand data from the first original data, the demand data including calculation accuracy, effective optimization duration, and index optimization value; S2: Acquire the second original data of the construction end, and extract the environment data and the construction data from the second original data; S3: based on the grouping criteria, group the compaction indicators in the indicator library to generate an indicator set; S4: Aggregate demand data, environmental data and construction data into pre-data, and select the target set corresponding to the pre-data through the trained compaction index optimization model; S5: Analyze the preferred indicators from the target set and formulate evaluation levels for the preferred indicators. The evaluation levels include simple level, general level and difficult level.

Images