A scheduling method and system for construction work in subway track area

Abstract

The present invention discloses a scheduling method and system for construction work in a subway track area, which belongs to the field of data processing technology, and includes: collecting and managing the storage data of the management system, extracting the text information of the storage data to generate comprehensive data; judging whether there are data problems in the comprehensive data, establishing a data cleaning model, and inputting the comprehensive data without data problems into the data cleaning model to generate corresponding standard data; generating multiple standard databases, extracting the operation records of the standard data, obtaining the completion efficiency of all operation records, combining the obtained operation plans, risk warnings and construction teams to generate the best operation plan corresponding to the category information; obtaining the best construction working hours of the best operation plan, formulating a scheduling schedule, and outputting the scheduling information of the construction work in the management system. The present invention can generate a scheduling schedule for construction work in a subway track area, and simultaneously obtain the best operation plan for the construction work, thereby improving the efficiency of the construction work.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and in particular relates to a scheduling method and system for construction work in a subway track area. Background Art

[0002] With the rapid development of urban rail transit, the safety and efficiency of subway track area construction work has become increasingly important. The traditional subway track area construction work scheduling method mainly relies on manual operation, such as command and coordination through the subway dispatching center. With the development of computer technology, semi-automatic scheduling methods have gradually emerged. In order to more comprehensively improve the efficiency and safety of construction work, it is necessary to systematically analyze and apply the scheduling method of subway track area construction work.

[0003] For example, Chinese patent application CN116523197A discloses a rail area dispatching and command system, equipment and medium. The system includes a day shift operation module, a dispatching command module, an image progress module, a plan fulfillment module, a driving monitoring module and a plane information module. The day shift operation module is used to create and update the content of the day shift operation; the dispatching command module is used to obtain all equipment information, select the target equipment to issue equipment dispatching commands, and generate reports to record the dispatching information; the image progress module is used to present the progress of professional operations in the form of a bar chart, and use the subway line map as the base map to present the overall construction progress in different colors; the plan fulfillment module is used to obtain the progress comparison information of bridges, switches, tracks, and main line track laying, as well as the reasons for planned delays and the corresponding avoidance or solutions; the driving monitoring module is used to obtain and display the driving route, and reflect the rail vehicle position information in real time on the route map; the plane information module is used to fill in and view engineering information. Another example is Chinese patent CN109118054B, which discloses a subway operation and construction scheduling management system, including a data storage and reader, a protection area setting rule base module, a protection area generation module, a construction plan publishing module, a protection area preservation and version management module, and a protection area conflict detection module for plan reporting. By flexibly setting protection areas under various construction conditions, construction safety is guaranteed, the work efficiency of setting a large number of planned protection areas and the utilization rate of construction resources are improved, human errors are prevented, the burden on construction management personnel is reduced, and statistical data is facilitated.

[0004] However, the above-mentioned prior art only describes the process of the scheduling system. In actual situations, the analysis and application of historical data of construction operations can more accurately and quickly provide operation methods and construction guarantee basis for the scheduling plan of construction operations. Summary of the invention

[0005] In order to solve the above problems, the present invention provides a scheduling method and system for construction work in a subway track area to solve the problems in the prior art.

[0006] In order to achieve the above-mentioned invention object, the present invention proposes a scheduling method for construction work in a subway track area, comprising:

[0007] Collecting the storage data of the management system based on a preset time period, extracting text information of the storage data based on storage tags to generate comprehensive data;

[0008] Based on the preset invalid characters, it is determined whether the comprehensive data has data problems. If so, the text information corresponding to the preset invalid character type in the comprehensive data is deleted, a data cleaning model is established, the comprehensive data without data problems is input into the data cleaning model, and the comprehensive data is cleaned and corrected to generate corresponding standard data;

[0009] Integrate and classify all standard data to generate multiple standard databases and corresponding category information, extract the operation records of the standard data in the standard database based on the storage tags, obtain the completion efficiency of all operation records respectively, set the operation records with the completion efficiency greater than or equal to a preset first threshold as reference records, set the operation records with the completion efficiency less than the first threshold as warning records, summarize all reference records and warning records respectively and generate corresponding operation plans and risk warnings, calculate the skill values ​​of the construction personnel in the operation records, generate a construction team based on the skill value distribution, and combine the operation plan, the risk warning and the construction team to generate the best operation plan corresponding to the category information;

[0010] The optimal construction time of the optimal work plan is obtained based on the reference record, a scheduling plan is formulated based on the optimal construction time, construction frequency and the category information, and scheduling information of the construction work is output in the management system based on the scheduling plan and the optimal work plan.

[0011] Further, generating the comprehensive data includes the following steps:

[0012] The storage tags include user information tags, equipment tags and job tags. The input process of the management system is located based on the storage tags. The login information and record information of the input process are stored based on the preset time period. The login information is set as the text information corresponding to the user information tag, and the record information is set as the text information corresponding to the job tag. A mechanical equipment set is established. Based on the mechanical equipment set, the job equipment information is extracted from the record information and set as the text information of the equipment tag. All text information is set as the comprehensive data based on the corresponding storage tag.

[0013] Furthermore, establishing a data cleaning model includes the following steps:

[0014] Generate a dictionary label of the data cleaning model based on the storage label, divide the comprehensive data into multiple groups of substrings based on preset special characters, extract information corresponding to the dictionary label in the substring and set it as dictionary data, if the information corresponding to the dictionary label is not extracted from the substring, set the substring as the data to be analyzed, establish a first dictionary, build a tree structure based on the first dictionary, the dictionary data and the data to be analyzed, determine in turn whether the first dictionary has corresponding character information in the dictionary data and the data to be analyzed, if so, set the first dictionary as a tree node of the tree structure, and set the corresponding character information as node information, and generate branches based on the number of the tree nodes and the corresponding node information;

[0015] A weight value is assigned to the first dictionary, an assigned value of the branch is calculated based on the weight value, and the data cleaning model is established based on the branch and the corresponding assigned value.

[0016] Furthermore, generating corresponding standard data includes the following steps:

[0017] A branch corresponding to a maximum allocated value is selected and set as a standard branch, a character type of the standard branch is obtained and set as a standard type, and the completeness of the data type in the comprehensive data is obtained based on the standard type. If the completeness is less than a first preset value, the comprehensive data is deleted; if the completeness is greater than the first preset value and less than a second preset value, character information of the standard type is supplemented in the data to be analyzed in the comprehensive data, and the supplemented comprehensive data is set as the standard data; if the completeness is greater than or equal to the second preset value, character information corresponding to the standard type in the comprehensive data is obtained and set as the standard data.

[0018] Furthermore, generating a standard database includes the following steps:

[0019] Establishing a text classification model, inputting the standard data into the text classification model, wherein the text classification model obtains text information of the job label in the standard data and defines it as construction data, wherein the text classification model sets a first analysis formula to obtain a first character vector of the construction data, sets a second analysis formula to obtain a second character vector of the construction data, obtains a first error between the first character vector and the second character vector, clusters the construction data based on the first error to generate a plurality of clusters, and combines the standard data corresponding to the construction data in the clusters and sets them as the standard database;

[0020] A keyword group of the construction data in the cluster is obtained, and the keyword group is set as the category information corresponding to the standard database.

[0021] Furthermore, obtaining the completion efficiency of the job record includes:

[0022] The completion efficiency P of the operation record in the standard data is calculated based on a first formula, wherein the first formula is: in, is the average operation time of all standard data in the standard database, t is the operation time of the operation record, is the average number of construction workers in all standard data in the standard database, m1 is the number of construction workers in the standard data, is the average number of construction equipment of all standard data in the standard database, and m2 is the number of construction equipment in the standard data.

[0023] Furthermore, generating the optimal operation plan includes the following steps:

[0024] Establishing a text summarization model, wherein the text summarization model trains and learns all reference records, and divides the reference records into a plurality of first clauses, compares the first clauses in two reference records in pairs, obtains sentence vectors of the two first clauses, obtains clause similarity of the two first clauses based on the sentence vectors, and retains any one of the first clauses if the clause similarity is greater than or equal to a third preset value, otherwise retains two first clauses, repeats the above steps until the first clauses in all reference records are extracted, and sets all the retained first clauses as the operation plan;

[0025] The risk warning is generated based on the warning record using the same steps, and the operation plan, the risk warning and the skill value of the construction personnel are combined and set as the optimal operation plan corresponding to the standard database.

[0026] Furthermore, outputting the scheduling information of the construction work includes the following steps:

[0027] The average value of the operation time in the reference record is set as the optimal construction working hours, the category information is set as the category label corresponding to the optimal operation plan, a learning and training model is constructed based on the category label, the learning and training model extracts a specific phrase of the category label, a first word vector is generated based on the specific phrase, the character information of the construction operation is set as input data, a feature vector of the input data is obtained, the feature vector is input into the learning and training model, the correlation between the feature vector and the first word vector is calculated respectively, the specific phrase corresponding to the first word vector whose correlation is greater than or equal to a fourth preset value is set as a candidate phrase, the character similarity between the candidate phrase and the input data is calculated, the optimal operation plan corresponding to the candidate phrase with the maximum character similarity is set as the plan to be constructed, and the plan to be constructed and the scheduling plan table are set as the scheduling information of the construction operation.

[0028] Furthermore, calculating the skill value of the construction personnel in the operation record includes:

[0029] The number of occurrences of the same construction worker in the reference record and the warning record is counted respectively, and the skill value W of the construction worker is calculated based on a second formula, where the second formula is: Wherein, n is the number of the standard database, r1 is the number of times the construction personnel appear in the reference record, and r2 is the number of times the construction personnel appear in the warning record.

[0030] The present invention also provides a scheduling system for construction work in a subway track area, which is used to implement the above-mentioned scheduling method for construction work in a subway track area. The system mainly includes:

[0031] A data collection module collects storage data of the management system based on a preset time period, and extracts text information of the storage data based on storage tags to generate comprehensive data;

[0032] A data cleaning module, which determines whether the comprehensive data has data problems based on preset invalid characters, and if so, deletes text information corresponding to the preset invalid character type in the comprehensive data, establishes a data cleaning model, inputs the comprehensive data without data problems into the data cleaning model, and cleans and corrects all comprehensive data to generate corresponding standard data;

[0033] A data integration module integrates and classifies all standard data to generate multiple standard databases and corresponding category information, extracts the operation records of the standard data in the standard database based on the storage tags, obtains the completion efficiency of all operation records respectively, sets the operation records whose completion efficiency is greater than or equal to a preset first threshold as reference records, sets the operation records whose completion efficiency is less than the first threshold as warning records, summarizes all reference records and warning records respectively and generates corresponding operation plans and risk warnings, calculates the skill values ​​of construction personnel in the operation records, generates a construction team based on the skill value allocation, and combines the operation plan, the risk warning and the construction team to generate an optimal operation plan corresponding to the category information;

[0034] A scheduling generation module obtains the optimal construction time of the optimal work plan based on the reference record, formulates a scheduling plan based on the optimal construction time, construction frequency and the category information, and outputs scheduling information of the construction work in the management system based on the scheduling plan and the optimal work plan.

[0035] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0036] The present invention first collects and manages the comprehensive data of historical construction operations in the management system, and cleans the comprehensive data by establishing a data cleaning model, deletes unavailable missing data or invalid data, and generates standard data, which can improve the usability of the data. Then, a text classification model is established to classify and integrate the standard data into a standard database of multiple category information, which can obtain standard data of the same type of construction operations, and is conducive to the classification and analysis of information of standard data in different standard databases. Finally, a text summary model is established to summarize the text information of each standard database to generate corresponding operation plans and risk warnings, which can quickly and effectively generate the best operation plan as a reference for the process of construction operations.

[0037] The present invention also establishes a learning and training model, which is conducive to quickly selecting the corresponding optimal operation plan according to the scheduling plan table and completing the scheduling information of the construction operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A flowchart of the steps of a scheduling method for construction work in a subway track area according to the present invention;

[0039] Figure 2 A schematic diagram of the result of generating branches from the tree structure in the present invention;

[0040] Figure 3 The present invention is a structural diagram of a dispatching system for construction work in a subway track area. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0042] It is understood that the terms "first", "second", etc. used in this application may be used herein to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, without departing from the scope of this application, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script.

[0043] like Figure 1 As shown, a scheduling method for construction work in a subway track area includes:

[0044] Step S1: collecting and managing the storage data of the management system based on a preset time period, and extracting text information of the storage data based on storage tags to generate comprehensive data.

[0045] Specifically, in this embodiment, the management system refers to a management system for rail work management, which includes information such as personnel login information, operation procedures, construction units and operation records. The storage data refers to the use data trace records of the management system, such as the login information of the on-duty personnel, the construction plan and operation records. The comprehensive data refers to the effective data used to analyze the construction work in the management system obtained through the storage tag. For example, the storage tag is set as a user information tag, an equipment tag and an operation tag, and the corresponding comprehensive data may include personnel information, job title, equipment data used during construction work and record information of construction work. By obtaining comprehensive data through a preset time period, the comprehensive data can be updated and added in real time. For example, if the time period is set to 30 days, the storage data is collected and the comprehensive data is extracted every 30 days.

[0046] Step S2: Based on the preset invalid characters, determine whether the comprehensive data has data problems. If so, delete the text information corresponding to the preset invalid character type in the comprehensive data, establish a data cleaning model, input the comprehensive data without data problems into the data cleaning model, and clean and correct all comprehensive data to generate corresponding standard data.

[0047] Specifically, in this embodiment, invalid characters refer to garbled or incorrectly input character strings. Data type problems in the comprehensive data, such as typos, are identified and deleted by presetting invalid characters. The data cleaning model is used to determine whether there is missing data or noise data in the comprehensive data, where missing data refers to data with incomplete data information, and noise data refers to erroneous data and duplicate data. By establishing a data cleaning model, the data quality of the comprehensive data can be ensured and the accuracy of data analysis can be improved. Standard data is generated by the data cleaning model, where standard data is reliable data without missing data, erroneous data and duplicate data.

[0048] Step S3: Integrate and classify all standard data to generate multiple standard databases and corresponding category information, extract the operation records of standard data in the standard database based on the storage tags, obtain the completion efficiency of all operation records respectively, set the operation records with completion efficiency greater than or equal to a preset first threshold as reference records, set the operation records with completion efficiency less than the first threshold as warning records, summarize all reference records and warning records respectively and generate corresponding operation plans and risk warnings, calculate the skill values ​​of construction personnel in the operation records, generate construction teams based on skill value allocation, and combine the operation plans, risk warnings and construction teams to generate the best operation plan corresponding to the category information.

[0049] Specifically, in this embodiment, the standard data can generate multiple standard databases through integration and classification, wherein the category information refers to the classification information corresponding to the standard database, such as rail flaw detection, insulation inspection and contact network maintenance, etc. The corresponding operation record in the standard data is extracted by selecting the type of storage tag. The operation record refers to the record information generated in the management system corresponding to the storage tag, that is, the information corresponding to the operation tag, which contains valid information such as the operator, operation time, operation steps and operation summary. The completion efficiency refers to the completion rate of the construction process in each operation record, which can be calculated by combining various factors such as construction hours, construction difficulty and construction consumption, wherein the construction hours are calculated by the start time and end time in the operation time. The reference record refers to the standard data with higher completion efficiency, and the construction reference plan of the corresponding standard database can be generated by summarizing the reference record, that is, the operation plan. The warning record refers to the standard data with lower completion efficiency, and the construction avoidance plan of the corresponding standard database can be generated by summarizing the warning record, that is, the risk warning. Skill value is used to evaluate the ability tendency of construction workers in the operation record. Construction team refers to the combination of construction workers. The skill value of construction workers can effectively allocate the construction work type of construction workers. For example, in the same construction work type, the skill values ​​of construction workers are sorted from high to low, and the construction team is composed of construction workers with higher skill values ​​and construction workers with lower skill values. This can help construction workers complete construction work on time and supplement the construction ability of construction workers. The best operation plan means that for each type of construction operation, there is a corresponding operation plan, risk warning and construction team.

[0050] Step S4: Obtain the best construction time of the best work plan based on the reference records, formulate a scheduling plan based on the best construction time, construction frequency and category information, and output the scheduling information of the construction work in the management system based on the scheduling plan and the best work plan.

[0051] Specifically, in this embodiment, the optimal construction time refers to the construction time required for the optimal operation plan. The corresponding construction time is obtained from all reference records. The optimal construction time is set by calculating the average value of the construction time. The construction frequency refers to the frequency interval of each construction operation. The rail company usually plans and sets the construction frequency of each construction operation, such as daily inspection, weekly inspection and monthly inspection. The scheduling schedule refers to the scheduling arrangement of all construction operations. For example, the construction frequency of "rail flaw detection" in the construction operation is 1 time / 30 days, and the optimal construction time is 3 hours / section. The construction operation refers to the type of construction operation corresponding to the category information. The optimal operation plan corresponding to the construction operation is quickly extracted. The scheduling information of the construction operation can be effectively completed through the scheduling schedule and the optimal operation plan. Among them, the scheduling information refers to the management system sending the construction operation to the corresponding construction personnel end, construction unit end and equipment management end, which effectively cooperates to complete the construction operation in the subway track area.

[0052] As a preferred technical solution of the present invention, generating comprehensive data includes the following steps:

[0053] The storage tags include user information tags, equipment tags and job tags. The input process of the management system is located based on the storage tags. The login information and record information of the input process are stored based on a preset time period. The login information is set as the text information corresponding to the user information tag, and the record information is set as the text information corresponding to the job tag. A mechanical equipment set is established. Based on the mechanical equipment set, the job equipment information is extracted from the record information and set as the text information of the equipment tag. All text information is set as comprehensive data based on the corresponding storage tags.

[0054] Specifically, in this embodiment, the input process refers to the page port for inputting information in the management system. The login end inputs login information and record information through different input processes, such as user name, user function, department, construction personnel, operation time, construction equipment and construction process, etc., wherein the login information refers to the user information of using the management system, that is, the user name, user function and department in the user information tag, and the record information refers to the information used to fill in the input during the construction process, including but not limited to the construction personnel, operation time, construction equipment and construction process. The mechanical equipment set refers to the set of equipment names used for construction, and the operation equipment information is obtained from the record information through text extraction, wherein text extraction refers to string comparison extraction, for example, the mechanical equipment set contains the noun "jack", if the string "jack" exists in the record information, then "jack" is set as the equipment information corresponding to the equipment tag. The text information corresponding to the user information tag, the equipment tag and the operation tag in the storage tag is combined into comprehensive data, wherein the comprehensive data refers to the text information directly obtained from the page port corresponding to each storage tag in the management system.

[0055] As a preferred technical solution of the present invention, establishing a data cleaning model includes the following steps:

[0056] Generate dictionary labels of the data cleaning model based on the stored labels, divide the comprehensive data into multiple groups of substrings based on preset special characters, extract information corresponding to the dictionary labels in the substrings and set them as dictionary data, if the information corresponding to the dictionary labels is not extracted from the substrings, set the substrings as data to be analyzed, establish a first dictionary, build a tree structure based on the first dictionary, dictionary data and data to be analyzed, determine whether the first dictionary has corresponding character information in the dictionary data and the data to be analyzed, if so, set the first dictionary as a tree node of the tree structure, and set the corresponding character information as node information, and generate branches based on the number of tree nodes and the corresponding node information.

[0057] A weight value is assigned to the first dictionary, and the assigned value of the branch is calculated based on the weight value, and a data cleaning model is established based on the branch and the corresponding assigned value.

[0058] Specifically, in this embodiment, the dictionary label is a classification label used for data correction and data cleaning in the data cleaning model, and the dictionary label includes various storage labels, such as user information labels, equipment labels, and job labels. Special characters refer to characters such as spaces, commas, and periods used to segment text, and the text in the comprehensive data can be segmented into multiple substrings. Dictionary data refers to text data corresponding to character labels in the substrings, and the data to be analyzed is data in the substrings that are to be analyzed for usability. The first dictionary refers to a feature information set, and the dictionary labels in the comprehensive data can be refined and set through the first dictionary, including but not limited to "start time", "end time", "process steps", "construction conclusions", "construction losses" and "personnel division of labor". The tree structure refers to the integrity structure for building data cleaning. By setting tree nodes, the dictionary data and the data to be analyzed can be refined and accurately combined or cleaned to avoid mixing the information corresponding to various names. Simply understand that the tree node is the information corresponding to the node label, and the node information is the information contained in the tree node. For example, the comprehensive data R1 is "the operation start time is 5 am, the construction steps are to first check the number of damaged lighting is 5, and then check the fire hydrant pump is powered on normally", and the dictionary tags are "user information tag", "equipment tag" and "operation tag", among which the dictionary data corresponding to the "user information tag" and "equipment tag" are not detected, and "the operation start time is 5 am, the construction steps are to first check the number of damaged lighting is 5, and then check the fire hydrant pump is powered on normally" all belong to the dictionary data R11 corresponding to the "operation tag", so it is necessary to construct a tree structure through the dictionary data and the first dictionary, such as Figure 2As shown, the dictionary data R11 contains character information corresponding to the "start time", "process steps" and "construction conclusion" of the first dictionary, so the first dictionary it contains is set as a tree node, and the corresponding character information "5 am", "lighting, fire hydrant pump" and "the number of lighting damages is 5, the water pump is powered normally" are set as the node information corresponding to the tree node, so that the branch L1 of the comprehensive data R1 can be generated.

[0059] For another example, in the comprehensive data R2, "the operation starts at 5:00 a.m., the construction steps are firstly that construction worker A checks that the number of damaged lighting is 5, then construction worker B checks that the fire hydrant pump is powered on normally, and water leakage is prone to occur at section H", the dictionary data R21 and R22 corresponding to the dictionary labels "user information label" and "operation label" are detected respectively, where the dictionary data R21 is "construction worker A" and "construction worker B", and the dictionary data R22 is "5:00 a.m.", "check lighting; check fire hydrant pump" and "the number of damaged lighting is 5; the pump is powered on normally", where "water leakage is prone to occur at section H" is the data to be analyzed, and the node labels of the first dictionary corresponding to the dictionary data R21 and R22 are "start time", "process steps", "construction conclusion" and "personnel division of labor". Compared with the comprehensive data R1 and R2, the branch L2 generated by the tree structure established by the comprehensive data R2 is more refined than the branch L1 generated by the comprehensive data R1, and the record text of the comprehensive data is retained in more detail.

[0060] The weight value refers to the importance of each node. For example, the weight values ​​of the user information label, equipment label and operation label in the dictionary label correspond to 20%, 20% and 60% respectively. There are 3 groups of tree nodes and corresponding node information in the comprehensive data R1 "the construction time is 8 am, and the construction steps are to first check the number of lighting damages is 5, and then check the power supply of the fire hydrant pump is normal". Therefore, the distribution value of the branch generated by the comprehensive data R1 in the tree structure is 0.6*3=1.8; there are 4 tree nodes and corresponding node information in the comprehensive data R2, so the distribution value of the branch generated by the comprehensive data R2 in the tree structure is 0.2*1+0.6*3=2. By setting the tree structure, the dictionary data and the data to be analyzed can be marked out in sequence, and the refined marks can be connected as branches in sequence to generate the distribution value of the comprehensive data, which can accurately locate the missing data, wrong data or duplicate data in the comprehensive data.

[0061] As a preferred technical solution of the present invention, generating corresponding standard data includes the following steps:

[0062] A branch corresponding to the maximum allocated value is selected and set as the standard branch, a character type of the standard branch is obtained and set as the standard type, and the completeness of the data type in the comprehensive data is obtained based on the standard type. If the completeness is less than a first preset value, the comprehensive data is deleted. If the completeness is greater than the first preset value and less than a second preset value, character information of the standard type is supplemented in the data to be analyzed in the comprehensive data, and the supplemented comprehensive data is set as the standard data. If the completeness is greater than or equal to the second preset value, the character information corresponding to the standard type in the comprehensive data is obtained and set as the standard data.

[0063] Specifically, in this embodiment, the maximum allocation value indicates that the branch composed of character tags contained in the comprehensive data is the most complete, and the corresponding comprehensive data is the most complete, so the allocation value is set to the completeness of the data type. If the completeness is less than the first preset value, it means that the comprehensive data is missing a lot, so it is deleted. If the completeness is between the first preset value and the second preset value, it means that the comprehensive data is missing a little, and the corresponding character information can be extracted from the data to be analyzed based on the standard type. For example, if "construction personnel" is missing in the comprehensive data, it can be supplemented by extracting users who fill in "construction records", so that the data type of the standard data has uniqueness, accuracy, completeness, consistency and validity.

[0064] As a preferred technical solution of the present invention, generating a standard database includes the following steps:

[0065] A text classification model is established, and standard data is input into the text classification model. The text classification model obtains text information of the job label in the standard data and defines it as construction data. The text classification model sets a first analysis formula to obtain a first character vector of the construction data, sets a second analysis formula to obtain a second character vector of the construction data, obtains a first error between the first character vector and the second character vector, clusters the construction data based on the first error to generate multiple clusters, and combines the standard data corresponding to the construction data in the clusters and sets them as a standard database.

[0066] Obtain keyword groups of the construction data in the cluster, and set the keyword groups as category information corresponding to the standard database.

[0067] Specifically, in this embodiment, the text classification model refers to a convolutional neural network model. The accuracy of the text classification model can be improved by generating clusters through unsupervised clustering. The text classification model will represent the construction data contained in the input job label by vectors through the first analysis formula and the second analysis formula, that is, the first character vector and the second character vector. The first analysis formula and the second analysis formula are two sets of neural networks with different parameters to generate vectors, wherein the first character vector and the second character vector both refer to the feature vectors of the construction data. The first error refers to the error between the category information of the text information in the text classification model and the correct category information of the category of the text information, that is, the error between the text information and the center point of the cluster, wherein the cluster refers to the classification result output by the text classification model, and the standard data corresponding to the construction data contained in the classification result is set to the same standard database, and the classification of all standard data is realized by multiple standard databases.

[0068] In this embodiment, the keyword group refers to a phrase with a high frequency of occurrence in multiple construction data in the cluster, and the category information refers to the construction name corresponding to the standard database, for example, category information such as cable laying.

[0069] As a preferred technical solution of the present invention, obtaining the completion efficiency of the operation record includes:

[0070] The completion efficiency P of the operation record in the standard data is calculated based on the first formula, which is: in, is the average working time of all standard data in the standard database, t is the working time of the working record, m1 is the average number of construction workers in all standard data in the standard database, m2 is the average number of construction equipment in all standard data in the standard database, and m2 is the number of construction equipment in the standard data.

[0071] Specifically, in this embodiment, the completion efficiency is usually associated with the construction time and the number of construction personnel. In the standard database, the operation records of each standard data correspond to the same construction name, that is, the construction operation type is the same, so the construction difficulty value is relatively consistent. Therefore, the first formula is calculated from the perspective of the ratio to the average value. Usually, the project efficiency = completed workload / consumed resources. The number of construction personnel and the number of construction equipment are extracted from the text information in the standard data. The higher the value of the completion efficiency, the better the feasibility of the corresponding standard data.

[0072] As a preferred technical solution of the present invention, generating the best operation plan includes the following steps:

[0073] A text summarization model is established. The text summarization model trains and learns all reference records, and divides the reference records into multiple first clauses. The first clauses in two reference records are compared in pairs to obtain sentence vectors of the two first clauses. The clause similarity of the two first clauses is obtained based on the sentence vectors. If the clause similarity is greater than or equal to a third preset value, any one of the first clauses is retained, otherwise both first clauses are retained. The above steps are repeated until the first clauses in all reference records are extracted, and all the retained first clauses are set as the operation plan.

[0074] The same steps are used to generate risk warnings based on warning records, and the work plan, risk warnings, and skill values ​​of construction workers are combined and set as the optimal work plan corresponding to the standard database.

[0075] Specifically, in this embodiment, in order to generate the best operation plan corresponding to all standard data in the standard database, it is necessary to establish a text summary model. The text summary model can divide the reference record into multiple first clauses by special characters. In order to more completely compare all reference records, a first clause can be extracted from each of the two reference records for pairwise comparison. The sentence vector can be a text vector generated by a natural language model. The sentence vector refers to the text vector of the two second clauses in the block clause. The similarity of the two text vectors is calculated by the cosine angle formula, which is the clause similarity. If the clause similarity is greater than or equal to the third preset value, it means that the similarity of the two first clauses is high, so only one first clause is retained. If the clause similarity is low, it means that the similarity of the two first clauses is low, and there are different text information, so it is necessary to retain two first clauses at the same time. Repeating the above steps can make the clause similarity of each first clause in all reference records be calculated mutually, and then delete the first clauses with repeated descriptions, and combine all the retained first clauses and set them as the operation plan.

[0076] In this embodiment, based on the same steps, text information corresponding to the risk warning can be generated in all warning records, which can effectively avoid risk problems that may occur during the operation. The skill value of the construction personnel can determine the work ability tendency of the construction personnel. Therefore, the operation plan, risk warning and the skill value of the construction personnel are combined to generate the best operation plan. Each standard database corresponds to an optimal operation plan, so multiple standard databases can generate an optimal operation plan set.

[0077] As a preferred technical solution of the present invention, outputting the scheduling information of the construction work includes the following steps:

[0078] The average value of the operation time in the reference record is set as the optimal construction working hours, the category information is set as the category label corresponding to the optimal operation plan, a learning and training model is constructed based on the category label, the learning and training model extracts a specific phrase of the category label, a first word vector is generated based on the specific phrase, the character information of the construction operation is set as input data, a feature vector of the input data is obtained, the feature vector is input into the learning and training model, the correlation between the feature vector and the first word vector is calculated respectively, the specific phrase corresponding to the first word vector whose correlation is greater than or equal to a fourth preset value is set as a candidate phrase, the character similarity between the candidate phrase and the input data is calculated, the optimal operation plan corresponding to the candidate phrase with the maximum character similarity is set as the plan to be constructed, and the plan to be constructed and the scheduling plan table are set as the scheduling information of the construction operation.

[0079] Specifically, in this embodiment, the average value of the operation time in the reference record refers to the average value of the construction time in all reference records, which can be used to measure the time required for each construction operation. The characters of the construction operation may be difficult to find during the input process, so the overlap of the character information corresponding to the construction operation and the category label can be quickly compared by constructing a learning training model, which is conducive to improving the accuracy and speed of input. The category label refers to the category information of the standard database corresponding to the best operation plan. The category label can specifically refer to the corresponding best operation plan. Therefore, all category labels are input into the learning training model. The learning training model refers to the machine learning model, which can perform morphological analysis on the category label. The specific phrase refers to a phrase with high importance in the category label. The first word vector and the feature vector are both generated by the natural language model, which can be generated by integrating the phrase vector and the character vector in the specific phrase and the input data respectively. The correlation between the first word vector and the feature vector can be calculated by the cosine angle formula. If the correlation is greater than or equal to the fourth preset value, it means that the corresponding category label and the input data have a high correlation, so it is set as a candidate phrase. Select the candidate phrase with the greatest character similarity to the input data from multiple candidate phrases, and obtain the corresponding category label. Obtain the corresponding optimal operation plan from the category label, where the character similarity Q can be calculated by the similarity formula, which is: Where n1 is the number of characters that are the same in the candidate phrase and the input data, and n is the number of characters in the input data. Through this step, the best operation plan corresponding to the construction operation can be quickly extracted from multiple best operation plans.

[0080] As a preferred technical solution of the present invention, calculating the skill value of the construction personnel in the operation record includes:

[0081] The number of times the same construction worker appears in the reference record and the warning record is counted respectively, and the construction worker's skill value W is calculated based on the second formula, which is: Among them, n is the number of standard databases, r1 is the number of times the construction personnel appear in the reference records, and r2 is the number of times the construction personnel appear in the warning records.

[0082] Specifically, in this embodiment, the number of occurrences of construction personnel can be detected and calculated based on the personal information of construction personnel such as account name, affiliated unit and contact information. The higher the skill value, the more suitable the construction personnel is for the corresponding construction name and can perform construction work quickly and effectively. At the same time, by setting the allocation method of skill value, a construction team can be generated, which is conducive to construction personnel learning multiple skills.

[0083] It is particularly noteworthy that the present invention can generate a scheduling schedule for construction work in a subway track area, and simultaneously obtain the best work plan for the construction work, and combine and generate scheduling information for the construction work, thereby improving the efficiency of the construction work.

[0084] like Figure 3 As shown, the present invention also provides a scheduling system for construction work in a subway track area, which is used to implement the above-mentioned scheduling method for construction work in a subway track area. The system mainly includes:

[0085] The data collection module collects the storage data of the management system based on a preset time period, and extracts the text information of the storage data based on the storage tags to generate comprehensive data.

[0086] The data cleaning module determines whether there are data problems in the comprehensive data based on preset invalid characters. If so, the text information corresponding to the preset invalid character type in the comprehensive data is deleted, a data cleaning model is established, the comprehensive data without data problems is input into the data cleaning model, and all comprehensive data are cleaned and corrected to generate corresponding standard data.

[0087] The data integration module integrates and classifies all standard data to generate multiple standard databases and corresponding category information, extracts the operation records of standard data in the standard database based on storage tags, obtains the completion efficiency of all operation records respectively, sets the operation records with completion efficiency greater than or equal to a preset first threshold as reference records, sets the operation records with completion efficiency less than the first threshold as warning records, summarizes all reference records and warning records respectively and generates corresponding operation plans and risk warnings, calculates the skill values ​​of construction personnel in the operation records, generates construction teams based on skill value allocation, and combines the operation plans, risk warnings and construction teams to generate the best operation plan corresponding to the category information.

[0088] The scheduling generation module obtains the optimal construction time of the optimal work plan based on reference records, formulates a scheduling plan based on the optimal construction time, construction frequency and category information, and outputs the scheduling information of the construction work in the management system based on the scheduling plan and the optimal work plan.

[0089] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0090] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The above-mentioned program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0091] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0092] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for those of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0093] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A scheduling method for construction work in a subway track area, characterized in that: The method comprises the following steps: Collecting the storage data of the management system based on a preset time period, extracting text information of the storage data based on storage tags to generate comprehensive data; Wherein, generating comprehensive data comprises the following steps: the storage tags comprise user information tags, equipment tags and operation tags, locating the input process of the management system based on the storage tags, storing the login information and record information of the input process based on the preset time period, setting the login information as the text information corresponding to the user information tag, setting the record information as the text information corresponding to the operation tag, establishing a mechanical equipment set, extracting the operation equipment information from the record information based on the mechanical equipment set, and setting it as the text information of the equipment tag, and setting all the text information as the comprehensive data based on the corresponding storage tags; Based on the preset invalid characters, it is determined whether the comprehensive data has data problems. If so, the text information corresponding to the preset invalid characters in the comprehensive data is deleted, a data cleaning model is established, the comprehensive data without data problems is input into the data cleaning model, and all comprehensive data are cleaned and corrected to generate corresponding standard data; Wherein, establishing a data cleaning model comprises the following steps: generating a dictionary label of the data cleaning model based on the storage label, dividing the comprehensive data into multiple groups of substrings based on preset special characters, extracting information corresponding to the dictionary label in the substring and setting it as dictionary data, if the information corresponding to the dictionary label is not extracted from the substring, setting the substring as data to be analyzed, establishing a first dictionary, constructing a tree structure based on the first dictionary, the dictionary data and the data to be analyzed, judging in turn whether the first dictionary has corresponding character information in the dictionary data and the data to be analyzed, if so, setting the first dictionary as a tree node of the tree structure, and setting the corresponding character information as node information, and generating branches based on the number of the tree nodes and the corresponding node information; Assigning weight values ​​to the first dictionary, calculating assigned values ​​of the branches based on the weight values, and establishing the data cleaning model based on the branches and the corresponding assigned values; Selecting a branch corresponding to a maximum allocated value and setting it as a standard branch, obtaining a character type of the standard branch and setting it as a standard type, obtaining the completeness of the data type in the comprehensive data based on the standard type, deleting the comprehensive data if the completeness is less than a first preset value, supplementing the data to be analyzed in the comprehensive data with character information of the standard type if the completeness is greater than the first preset value and less than a second preset value, and setting the supplemented comprehensive data as the standard data, and obtaining the character information corresponding to the standard type in the comprehensive data and setting it as the standard data if the completeness is greater than or equal to the second preset value; Integrate and classify all standard data to generate multiple standard databases and corresponding category information, extract the operation records of the standard data in the standard database based on the storage tags, obtain the completion efficiency of all operation records respectively, set the operation records with the completion efficiency greater than or equal to a preset first threshold as reference records, set the operation records with the completion efficiency less than the first threshold as warning records, summarize all reference records and warning records respectively and generate corresponding operation plans and risk warnings, calculate the skill values ​​of the construction personnel in the operation records, generate a construction team based on the skill value allocation, and combine the operation plan, the risk warning and the construction team to generate an optimal operation plan corresponding to the category information; Wherein, a text classification model is established, the text classification model is a convolutional neural network model, clusters are generated by unsupervised clustering, the standard data is input into the text classification model, the text classification model obtains the text information of the job label in the standard data and defines it as construction data, the text classification model sets a first analysis formula to obtain a first character vector of the construction data, sets a second analysis formula to obtain a second character vector of the construction data, obtains a first error between the first character vector and the second character vector, clusters the construction data based on the first error to generate a plurality of clusters, and combines the standard data corresponding to the construction data in the clusters and sets them as the standard database; Acquire a keyword group of the construction data in the cluster, and set the keyword group as the category information corresponding to the standard database; Acquiring the completion efficiency of the operation record includes: The completion efficiency P of the operation record in the standard data is calculated based on a first formula, wherein the first formula is: in, is the average operation time of all standard data in the standard database, t is the operation time of the operation record, is the average number of construction workers in all standard data in the standard database, m1 is the number of construction workers in the standard data, is the average number of construction equipment of all standard data in the standard database, and m2 is the number of construction equipment in the standard data; Calculating the skill value of the construction personnel in the operation record includes: The number of occurrences of the same construction worker in the reference record and the warning record is counted respectively, and the skill value W of the construction worker is calculated based on a second formula, where the second formula is: Wherein, n is the number of the standard database, r1 is the number of times the construction personnel appear in the reference record, and r2 is the number of times the construction personnel appear in the warning record; Generating the optimal operation plan includes the following steps: Establishing a text summarization model, wherein the text summarization model trains and learns all reference records, and divides the reference records into a plurality of first clauses, compares the first clauses in two reference records in pairs, obtains sentence vectors of the two first clauses, obtains clause similarity of the two first clauses based on the sentence vectors, and retains any one of the first clauses if the clause similarity is greater than or equal to a third preset value, otherwise retains two first clauses, repeats the above steps until the first clauses in all reference records are extracted, and sets all the retained first clauses as the operation plan; Using the same steps to generate the risk warning based on the warning record, combining the operation plan, the risk warning and the skill value of the construction personnel and setting them as the optimal operation plan corresponding to the standard database; Obtaining the best construction time of the best operation plan based on the reference record, formulating a scheduling plan based on the best construction time, construction frequency and the category information, and outputting scheduling information of the construction work in the management system based on the scheduling plan and the best operation plan; Outputting the scheduling information of the construction work includes the following steps: The average value of the operation time in the reference record is set as the optimal construction working hours, the category information is set as the category label corresponding to the optimal operation plan, a learning and training model is constructed based on the category label, the learning and training model extracts a specific phrase of the category label, a first word vector is generated based on the specific phrase, the character information of the construction operation is set as input data, a feature vector of the input data is obtained, the feature vector is input into the learning and training model, the correlation between the feature vector and the first word vector is calculated respectively, the specific phrase corresponding to the first word vector whose correlation is greater than or equal to a fourth preset value is set as a candidate phrase, the character similarity between the candidate phrase and the input data is calculated, the optimal operation plan corresponding to the candidate phrase with the maximum character similarity is set as the plan to be constructed, and the plan to be constructed and the scheduling plan table are set as the scheduling information of the construction operation.

2. A dispatching system for construction work in a subway track area, used to implement the method according to claim 1, characterized in that: The system includes the following modules: A data collection module collects storage data of the management system based on a preset time period, and extracts text information of the storage data based on storage tags to generate comprehensive data; A data cleaning module, which determines whether the comprehensive data has data problems based on preset invalid characters, and if so, deletes text information corresponding to the preset invalid characters in the comprehensive data, establishes a data cleaning model, inputs the comprehensive data without data problems into the data cleaning model, and cleans and corrects all comprehensive data to generate corresponding standard data; A data integration module integrates and classifies all standard data to generate multiple standard databases and corresponding category information, extracts the operation records of the standard data in the standard database based on the storage tags, obtains the completion efficiency of all operation records respectively, sets the operation records whose completion efficiency is greater than or equal to a preset first threshold as reference records, sets the operation records whose completion efficiency is less than the first threshold as warning records, summarizes all reference records and warning records respectively and generates corresponding operation plans and risk warnings, calculates the skill values ​​of construction personnel in the operation records, generates a construction team based on the skill value allocation, and combines the operation plan, the risk warning and the construction team to generate an optimal operation plan corresponding to the category information; A scheduling generation module obtains the optimal construction time of the optimal work plan based on the reference record, formulates a scheduling plan based on the optimal construction time, construction frequency and the category information, and outputs scheduling information of the construction work in the management system based on the scheduling plan and the optimal work plan.

Images