Intelligent settlement and settlement method and system for power distribution network construction project

Abstract

The invention provides an intelligent settlement and settlement method and system for a power distribution network construction project, and is suitable for the field of power distribution network project construction management. The method comprises project initialization, equipment component informatization, construction process modularization, engineering quantity list generation, intelligent settlement and settlement cost calculation, the construction process modularization of the power distribution network project is realized, the existing construction process in a new project selection module or the construction process of a user-defined module is selected, and the existing construction process comprises clear engineering quantity. After the new project is completed, the project amount of the construction procedures in the custom module is input into the existing database, when the construction procedures in the module are more and more comprehensive, the manual participation demand is smaller, the accuracy of a final result can be better guaranteed, meanwhile, drawing information is recognized through a deep learning model, and the project amount in a drawing is output; the intelligent settlement and settlement system for the power distribution network construction project can analyze the difference reason between the project cost budget and the actual cost through a large language model.

Description

Technical Field

[0001] The present invention relates to an intelligent settlement and final accounting method and system for a distribution network construction project, and is applicable to the field of distribution network project construction management. Background Art

[0002] With the popularization of intelligent systems and the rapid development of artificial intelligence, more and more management processes need to be participated in by intelligent systems to improve the management efficiency of enterprises. In the field of distribution network project construction management, the project settlement and final settlement process still requires a lot of manual calculation, price inquiry, and preparation of settlement and final settlement documents. It is not only difficult to ensure the accuracy of the settlement and final settlement results, but also difficult to integrate with the management process of other projects, and it is not convenient to compare and analyze and form effective conclusions to assist company managers in making decisions. Summary of the invention

[0003] The purpose of the present invention is to solve the problems of low efficiency and difficulty in ensuring the accuracy of results caused by the manual control of the traditional settlement and final settlement process, and to provide an intelligent settlement and final settlement method and system for distribution network construction projects, which has the advantages of high intelligence, accurate results and strong compatibility, and can be widely used in the field of distribution network project construction management.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] An intelligent settlement method for a distribution network construction project specifically comprises the following steps:

[0006] S101 project initialization;

[0007] The project initialization includes establishing a new project, inputting the project name, location, construction unit, construction unit, scale, planned time, geological survey data and construction plan;

[0008] S102 Equipment component informatization;

[0009] The equipment component informatization includes inputting the information of new equipment, tools and components used in the new project into the project and generating one-to-one corresponding QR codes. All equipment, tools and components used in the new project are registered during the process of entering and leaving the warehouse, and the registration information is stored in the database;

[0010] S103 construction process modularization;

[0011] The modularization of the construction process includes entering the engineering quantities of each construction process in the completed project into a database, and counting them into a civil construction module, an overhead line construction module, a cable laying module, a power distribution equipment installation module, a transportation module and a disassembly module according to their classification;

[0012] S104 generates a bill of quantities;

[0013] The generation of the bill of quantities includes selecting construction procedures, cost budgeting, intelligent calculation, dynamic adjustment and optimization, and bill output;

[0014] S105 intelligent settlement;

[0015] The intelligent settlement includes data integration and analysis and generation of settlement reports.

[0016] Furthermore, in step S103, the civil construction module can be divided into two categories: on-site cast foundation and prefabricated foundation installation; the overhead line construction module includes pole tower assembly, pole segment adjustment, pole head material installation, wire installation, conductor erection and grounding device installation; the power distribution equipment installation module includes transformer installation, pole-mounted equipment installation and meter box installation; the transportation module includes manual transportation and automobile transportation.

[0017] Furthermore, in step S104, the selection of construction procedures includes selecting existing construction procedures in each module involved in the new project; the intelligent calculation includes identifying drawing information and outputting the project quantities in the drawings; the cost budget includes combining market price information and the company's internal cost database to convert the project quantities into comprehensive unit prices including direct costs and indirect costs, and generate cost budget details; the dynamic adjustment and optimization includes adjusting the construction plan and the bill of quantities according to design changes and actual on-site conditions; the list output includes generating a bill of quantities file containing serial number, project number, project name, project quantity, unit of measurement, comprehensive unit price and total price information after the dynamic adjustment and optimization.

[0018] Furthermore, in step S105, the data integration includes extracting from the database the specifications, dimensions, quantity and material information of the components used in the new project, the model specifications and quantity of the equipment and tools used, and comparing them with the project design drawings, construction records, material purchase lists and the registration information to verify the consistency and completeness of each data; the generation of the settlement report includes generating a settlement report containing a project overview and detailed unit price and total price information.

[0019] Furthermore, the deep learning model used in the intelligent computing is trained based on the two-dimensional drawing recognition and coordinate extraction technology of SegFormer, and the two-dimensional drawing recognition and coordinate extraction technology based on SegFormer includes the following steps:

[0020] 1) Dataset processing;

[0021] The data set processing includes assigning different semantic meanings to different color labels, converting the data set blueprint into a high-resolution image, using gray bars to avoid distortion, and then using tools to annotate the image and accurately align the label with the edge of the primitive. The data set must contain multiple categories such as independent single targets, single target intersection, dual target intersection, and multi-target connection. Then, the original image of the data set is denoised by a non-local mean filtering algorithm. The processed image meets the following conditions:

[0022]

[0023] Where u is the denoised image, υ is the noisy image, I is the search area, and ω(i,j) is the weight, i.e., the similarity between the i and j region blocks, expressed by the Euclidean distance formula.

[0024]

[0025] Where n(i)-the sum of all weights, h-filter coefficient, - Gaussian weighted Euclidean distance between the i and j region blocks, a - standard deviation of the Gaussian kernel, and then expand the training dataset by randomly transforming the original image;

[0026] 2) Coordinate processing;

[0027] The coordinate processing includes determining the conversion ratio Scale of the X and Y coordinates of the pixel map point coordinates (u, v) and the actual blueprint coordinates (x, y) x Scale y ,

[0028]

[0029] Then, coordinate transformation is performed according to the coordinate transformation formula (4).

[0030]

[0031] In the formula, y max - total image width measured in mm, (u0,v0) and (x0,y0) refer to the different coordinates of the same point in the pixel image and the actual blueprint, respectively;

[0032] 3) Image processing;

[0033] The image processing includes extending the image edge and filling it with a white background, enhancing the semantic information area by a dilation algorithm, and extracting the target primitive contour by a threshold extraction and a hybrid algorithm;

[0034] 4) Hough detection;

[0035] The Hough detection includes converting the image into a grayscale image, detecting edges, and then performing probabilistic Hough line detection and Hough circle detection;

[0036] 5) SQL Server query and output coordinates;

[0037] The SQL Server query and output coordinates include using the SQL Server database service to retrieve contour lines and intersections processed by the Hough transform, and identifying key coordinates of each graphic element;

[0038] 6) Hyperparameter setting and iterative training;

[0039] The hyperparameter setting and iterative training, including setting the learning rate, batch size, optimizer, pre-trained weights, learning rate decay type, weight decay, loss function and cycle count parameters, perform deep learning training on the SegFormer model.

[0040] An intelligent final accounting method for a distribution network construction project, based on steps S101 to S105, further includes the following steps:

[0041] S106 Final cost calculation;

[0042] The calculation of final settlement costs includes the calculation of design fees, supervision fees, construction unit management fees, equipment and tools purchase fees, and the generation of final settlement reports;

[0043] The design fee calculation includes obtaining the billing standard and method of the design fee according to the design contract, calculating the design fee and generating a design fee calculation table, wherein the design fee calculation table includes the billing basis, rate, adjustment coefficient, calculation formula and final amount;

[0044] The supervision fee, the charging method and rate of the supervision fee are determined according to the project supervision contract, the supervision fee is calculated and a supervision fee calculation table is generated, and the supervision fee calculation table includes the charging basis, rate, adjustment coefficient, calculation formula and final amount;

[0045] The construction unit management fee is calculated based on the construction unit's internal management fee calculation method and the number and position information of the dispatched project management personnel, and a management fee calculation table is generated. The management fee calculation table includes the billing basis, adjustment coefficient, calculation formula and final amount;

[0046] The equipment and tools purchase cost calculation includes extracting a list of equipment and tools actually used in the new project from the database, including name, specification model, unit, quantity, and purchase unit price information, collecting relevant purchase contracts, invoices, and payment vouchers, and comparing the information in the database with the purchase information based on the large language model to ensure the authenticity and integrity of the data; calculating the purchase cost of the equipment and tools actually used by the purchase quantity and purchase unit price, and generating a purchase cost detail list, which includes name, specification model, unit, quantity, unit price, and amount fields;

[0047] The final settlement report generated includes information on the project name, location, scale, construction unit, construction unit, design unit, and supervision unit, describes the construction process, major changes, and delays of the project, lists the calculation process, basis, and amount of each type of expense in separate items, calculates the total final settlement amount based on the settlement price information, design fees, supervision fees, construction unit management fees, and equipment and tool purchase fees, compares the cost budget, summarizes the cost situation of the project based on the large language model, and puts forward targeted improvement suggestions.

[0048] An intelligent settlement and final accounting system for distribution network construction projects, comprising a project initialization module, a process planning module, a resource management module, a project quantity calculation module, a settlement and final accounting management module and a system integration and expansion module:

[0049] The project initialization module includes a module for setting parameters of a new project and storing basic data;

[0050] The process planning module includes providing the construction processes entered by the user and the customized construction processes in the wiring diagram design module, civil construction module, overhead line construction module, cable laying module, power distribution equipment installation module, transportation module and disassembly module;

[0051] The resource management module includes one-click generation of a resource library list based on the in-and-out registration information of materials and mechanical resources, and inputting the comprehensive unit price of each construction process of the completed project into the cost database;

[0052] The engineering quantity calculation module includes intelligently identifying the content of the drawings and automatically calculating the engineering quantity according to the project design drawings;

[0053] The settlement and final settlement management module includes generating a settlement and final settlement report based on the settlement and final settlement method according to the engineering quantity calculation results and the data of the cost database, and analyzing the source items of cost differentiation according to the cost budget;

[0054] The system integration and expansion module includes providing an API interface to make the system compatible with other enterprise resource planning systems and integrate third-party tools and applications.

[0055] The beneficial effects of the present invention are: 1. By modularizing the construction process of the distribution network project, the new project can select the existing construction process in the module or the construction process of the custom module to complete the project construction. The existing construction process already contains clear engineering quantities, and the construction process of the custom module is calculated according to the custom rules. After the new project is completed, the engineering quantities of the construction process in the custom module are entered into the existing database. When the number and comprehensiveness of the construction processes in the module are, the demand for manual participation is smaller, and the accuracy of the final result can be guaranteed; 2. The drawing information is identified through the deep learning model, and the engineering quantities in the drawings are output to provide a comparison basis for the actual engineering quantities; 3. The settlement and final settlement system designed based on the intelligent settlement and final settlement method of the distribution network construction project analyzes the reasons for the difference between the project cost budget and the actual cost through a large language model and puts forward corresponding suggestions; 4. By providing an API interface, it is convenient for compatibility with other enterprise resource planning systems and integration of third-party tools and applications, thereby ensuring the scalability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a flow chart of the intelligent final settlement method for the distribution network construction project of the present invention; DETAILED DESCRIPTION

[0057] The specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments given here are only used to illustrate and explain the present invention and cannot be used to limit the present invention. It should be noted that many specific details are set forth in the following description to facilitate a full understanding of the present invention, but the present invention may also have other embodiments and variations thereof, and therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0058] This embodiment is based on a 10kV transmission line project of a certain power grid, a 1.2 km long overhead line, and the conductor model is JKLYJ-10-120mm 2 ,The specific steps of the intelligent settlement method for distribution network construction projects are as follows:

[0059] S101 project initialization;

[0060] Create a new project, input the project name, location, construction unit, construction unit, scale, planned time, geological survey data and construction plan;

[0061] S102 Equipment component informatization;

[0062] The information of new equipment, tools and components used in the new project is input into the project, and a one-to-one corresponding QR code is generated. All equipment, tools and components used in the new project are registered during the process of entering and leaving the warehouse, and the registration information is stored in the database. The equipment and tools used in the project have been entered into the database. After the prefabricated foundation and prefabricated pole leave the factory, they are entered into the database one by one according to the number, specification size, factory time and price data, and a corresponding QR code is generated;

[0063] S103 construction process modularization;

[0064] The engineering quantities of each construction process in the completed project are entered into the database, and are included in the civil construction module, overhead line construction module, cable laying module, power distribution equipment installation module, transportation module and disassembly module according to their classification;

[0065] The preliminary design uses drone aerial photography and image recognition of surface features to automatically generate a preliminary design drawing of the distribution network routing. The formal design is manually adjusted and modified based on the preliminary design. The average pole spacing determined by the formal design drawing of this embodiment is 50m, and a total of 25 ZZ001 medium-voltage single-circuit straight poles of φ190×12m are used, which are connected to the supporting prefabricated foundation with flange anchor bolts. The size of the supporting prefabricated foundation is a square column foundation of 2.8m×1m×1m;

[0066] According to the above-mentioned wiring diagram design content, the civil construction module and the overhead line construction module are already in the database. After positioning and measurement, the excavation size of the ZZ001 medium-voltage single-circuit straight pole φ190×12m is 2.8m×1.2m×1.2m, and a square hole rotary drilling is used to form it in one step. After the prefabricated foundation is transported to the construction site, it is hoisted by a crane. After the hoisting is completed, it is backfilled and compacted, and the remaining soil is transported out. After the prefabricated poles and accessories are transported to the construction site, the poles are supported, erected, aligned, repainted, and temporary wires are installed and disassembled, and then crossarms, clamps, hardware and insulators are installed, followed by laying out, stringing, tightening, adjusting sag, measuring, cutting, connecting, hanging and binding. The stringing process is assisted by drones, and finally the grounding device is installed; the distribution equipment installation module includes transformer installation, pole equipment installation and meter box installation;

[0067] S104 generates a bill of quantities;

[0068] Generate a bill of quantities, including the selection of construction procedures, intelligent calculation, cost budgeting, dynamic adjustment and optimization, and list output; select the existing construction procedures in each module involved in the new project for summary. The summary results are shown below:

[0069] Table 1 Summary of construction procedures for new projects

[0070]

[0071]

[0072] Intelligent calculation, including identification of drawing information and output of engineering quantities in the drawings. In this embodiment, 25 ZZ001 medium-voltage single-circuit linear rods of φ190×12m are used, which are connected to the supporting prefabricated foundation with flange anchor bolts. The size of the supporting prefabricated foundation is a square column foundation of 2.8m×1m×1m, and the length of the conductor is 1223m×1.05×4; cost budgeting, including combining market price information and the internal cost database of the enterprise, converting the engineering quantity into a comprehensive unit price including direct cost and indirect cost, and generating cost budget details, as shown in the following table.

[0073] Table 2 New project cost budget details

[0074]

[0075]

[0076] Adjust the construction plan and bill of quantities according to the design changes and actual site conditions, and then generate a bill of quantities file containing the serial number, project number, project name, project quantity, measurement unit, comprehensive unit price and total price information. The bill of quantities is shown in the following table;

[0077] Table 3 New Project Bill of Quantities

[0078]

[0079]

[0080] S105 intelligent settlement;

[0081] The smart settlement includes data integration and generation of settlement reports;

[0082] The data integration includes extracting from the database the specifications, dimensions, quantity and material information of the components used in the new project, the models, specifications and quantities of the equipment and tools used, and comparing them with the project design drawings, construction records, material purchase lists and the registration information to verify the consistency and completeness of the data;

[0083] The generating of the settlement report includes generating a settlement report including a project overview and detailed unit price and total price information.

[0084] S106 Final cost calculation;

[0085] The calculation of final settlement costs includes the calculation of design fees, supervision fees, construction unit management fees, equipment and tools purchase fees, and the generation of final settlement reports;

[0086] The design fee calculation includes obtaining the billing standard and method of the design fee according to the design contract, calculating the design fee and generating a design fee calculation table, wherein the design fee calculation table includes the billing basis, rate, adjustment coefficient, calculation formula and final amount;

[0087] The supervision fee, the charging method and rate of the supervision fee are determined according to the project supervision contract, the supervision fee is calculated and a supervision fee calculation table is generated, and the supervision fee calculation table includes the charging basis, rate, adjustment coefficient, calculation formula and final amount;

[0088] The construction unit management fee is calculated based on the construction unit's internal management fee calculation method and the number and position information of the dispatched project management personnel, and a management fee calculation table is generated. The management fee calculation table includes the billing basis, adjustment coefficient, calculation formula and final amount;

[0089] The equipment and tools purchase cost calculation includes extracting a list of equipment and tools actually used in the new project from the database, including name, specification model, unit, quantity, and purchase unit price information, collecting relevant purchase contracts, invoices, and payment vouchers, and comparing the information in the database with the purchase information based on the large language model to ensure the authenticity and integrity of the data; calculating the purchase cost of the equipment and tools actually used by the purchase quantity and purchase unit price, and generating a purchase cost detail list, which includes name, specification model, unit, quantity, unit price, and amount fields;

[0090] The final settlement report generated includes information on the project name, location, scale, construction unit, construction unit, design unit, and supervision unit, describes the construction process, major changes, and delays of the project, lists the calculation process, basis, and amount of each type of expense in separate items, calculates the total final settlement amount based on the settlement price information, design fees, supervision fees, construction unit management fees, and equipment and tool purchase fees, compares the cost budget, summarizes the cost situation of the project based on the large language model, and puts forward targeted improvement suggestions.

[0091] An intelligent settlement and final accounting system for distribution network construction projects, comprising a project initialization module, a process planning module, a resource management module, a project quantity calculation module, a settlement and final accounting management module and a system integration and expansion module:

[0092] The project initialization module includes a module for setting parameters of a new project and storing basic data;

[0093] The process planning module includes providing the construction processes entered by the user and the customized construction processes in the wiring diagram design module, civil construction module, overhead line construction module, cable laying module, power distribution equipment installation module, transportation module and disassembly module;

[0094] The resource management module includes one-click generation of a resource library list based on the in-and-out registration information of materials and mechanical resources, and inputting the comprehensive unit price of each construction process of the completed project into the cost database;

[0095] The engineering quantity calculation module includes intelligently identifying the content of the drawings and automatically calculating the engineering quantity according to the project design drawings;

[0096] The settlement and final settlement management module includes generating a settlement and final settlement report based on the settlement and final settlement method according to the engineering quantity calculation results and the data of the cost database, and analyzing the source items of cost differentiation according to the cost budget;

[0097] The system integration and expansion module includes providing an API interface to make the system compatible with other enterprise resource planning systems and integrate third-party tools and applications.

[0098] This second embodiment relies on a 10kV transmission line project of a certain power grid, trains a deep learning model based on SegFormer's two-dimensional drawing recognition and coordinate extraction technology, and extracts drawing information, including the following steps:

[0099] 1) Dataset processing;

[0100] Different color labels are given different semantic meanings. The dataset blueprint is converted into a high-resolution image. Gray bars are used to avoid distortion. Then, the image is annotated using tools to accurately align the label with the edge of the primitive. The dataset must contain multiple categories such as independent single targets, single target intersection, dual target intersection, and multi-target connection. Then, the original image of the dataset is denoised using the non-local mean filtering algorithm. The processed image meets the following conditions:

[0101]

[0102] Where u is the denoised image, υ is the noisy image, I is the search area, and ω(i,j) is the weight, i.e., the similarity between the i and j region blocks, expressed by the Euclidean distance formula.

[0103]

[0104] Where n(i)-the sum of all weights, h-filter coefficient, - Gaussian weighted Euclidean distance between i and j region blocks, a- standard deviation of Gaussian kernel, and then the training data set is expanded by randomly transforming the original image. The random transformation method is shown in the following table;

[0105] Table 4 Random transformation method

[0106] Way Specific operations Rotation Clockwise 90°, 180°, 270° Flip Up and down, left and right Contrast Enhancement Enhancement factor 1.3 Brightness Enhancement Enhancement factor 1.2 Sharpness Enhancement Enhancement factor 2.2 Chroma Enhancement Enhancement factor 1.2

[0107] 2) Coordinate processing;

[0108] Pixel counts represent lengths, which are different from the coordinate distances in the actual coordinate system. There is a conversion ratio between the X and Y coordinates of the pixel map point coordinates (u, v) and the actual blueprint coordinates (x, y). x Scale y ,

[0109]

[0110] The coordinates (u1, v1) and (x1, y1) correspond to the initial points of the pixel map and the actual blueprint respectively, while the coordinates (u2, v2) and (x2, y2) correspond to the end points of the pixel map and the actual blueprint respectively. A ratio of at least 1px:1mm is required to ensure the accuracy of the coordinate conversion. The larger the ratio, the more accurate the conversion. Then, the coordinate conversion is performed according to the coordinate conversion formula (4).

[0111]

[0112] In the formula, y max -The total width of the image measured in mm, (u0, v0) and (x0, y0) refer to the different coordinates of the same point in the pixel image and the actual blueprint respectively; this embodiment uses a ratio of 1px:1mm;

[0113] 3) Image processing;

[0114] The image processing includes extending the image edge and filling it with a white background, enhancing the semantic information area by a dilation algorithm, and extracting the target primitive contour by a threshold extraction and a hybrid algorithm;

[0115] 4) Hough detection;

[0116] The Hough detection includes converting the image into a grayscale image, detecting edges, and then performing probabilistic Hough line detection and Hough circle detection;

[0117] 5) SQL Server query and output coordinates;

[0118] The SQL Server query and output coordinates include using the SQL Server database service to retrieve contour lines and intersections processed by the Hough transform, and identifying key coordinates of each graphic element;

[0119] 6) Hyperparameter setting and iterative training;

[0120] The hyperparameter setting and iterative training, including setting the initial learning rate, batch size, optimizer, learning rate decay type, weight decay, loss function and cycle count parameters, perform deep learning training on the SegFormer model. The hyperparameter settings of this embodiment are shown in the following table.

[0121] Table 5 Hyperparameter settings

[0122] parameter set up Initial learning rate <![CDATA[1×10 -4 (Minimum 1×10 -6 )]]> Batch size 24 (frozen), 12 (thawed) Optimizer Adam Learning rate decay type cos Weight decay 0.01 Loss Function Cross-entropy loss Cycle Count 100 (freeze), 100 (thaw) Backbone network SegFormer-b2

[0123] In summary, the intelligent settlement and final accounting method and system for distribution network construction projects of the present invention have the advantages of high intelligence, accurate results and strong compatibility in the field of distribution network project construction management.

[0124] It should be understood that the above-mentioned embodiments are one or more embodiments of the present invention, and there are many other embodiments and variations thereof based on the present invention; the variations and modifications made by ordinary technicians in this industry through the present invention without making pioneering innovations all fall within the scope of protection of the present invention.

Claims

1. A method for intelligent settlement of distribution network construction projects, characterized in that: The following steps are involved: S101 project initialization; The project initialization includes establishing a new project, inputting the project name, location, construction unit, construction unit, scale, planned time, geological survey data and construction plan; S102 Equipment component informatization; The equipment component informatization includes inputting the information of new equipment, tools and components used in the new project into the project and generating one-to-one corresponding QR codes. All equipment, tools and components used in the new project are registered during the process of entering and leaving the warehouse, and the registration information is stored in the database; S103 construction process modularization; The modularization of the construction process includes entering the engineering quantities of each construction process in the completed project into a database, and counting them into a civil construction module, an overhead line construction module, a cable laying module, a power distribution equipment installation module, a transportation module and a disassembly module according to their classification; S104 generates a bill of quantities; The generation of the bill of quantities includes selecting construction procedures, cost budgeting, intelligent calculation, dynamic adjustment and optimization, and bill output; S105 intelligent settlement; The intelligent settlement includes data integration and analysis and generation of settlement reports.

2. The intelligent settlement method for distribution network construction projects according to claim 1 is characterized in that: In step S103, the civil construction module can be divided into two categories: on-site cast foundation and prefabricated foundation installation; the overhead line construction module includes pole tower assembly, pole segment adjustment, pole head material installation, wire installation, conductor erection and grounding device installation; the power distribution equipment installation module includes transformer installation, pole-mounted equipment installation and meter box installation; the transportation module includes manual transportation and automobile transportation.

3. The intelligent settlement method for distribution network construction projects according to claim 1 is characterized in that: In step S104, the selection of construction procedures includes selecting existing construction procedures in each module involved in the new project; the intelligent calculation includes identifying drawing information and outputting the project quantities in the drawings; the cost budget includes combining market price information and the company's internal cost database to convert the project quantities into comprehensive unit prices including direct costs and indirect costs, and generate cost budget details; the dynamic adjustment and optimization includes adjusting the construction plan and the bill of quantities according to design changes and actual on-site conditions; the list output includes generating a bill of quantities file containing serial number, project number, project name, project quantity, unit of measurement, comprehensive unit price and total price information after the dynamic adjustment and optimization.

4. The intelligent settlement method for distribution network construction projects according to claim 1 is characterized in that: In step S105, the data integration includes extracting from the database the specifications, dimensions, quantity and material information of the components used in the new project, the model specifications and quantity of the equipment and tools used, and comparing them with the project design drawings, construction records, material purchase lists and the registration information to verify the consistency and completeness of each data; the generation of the settlement report includes generating a settlement report containing a project overview and detailed unit price and total price information.

5. The intelligent settlement method for distribution network construction projects according to any one of claims 1 and 3, characterized in that: The deep learning model used in the intelligent computing is trained based on the two-dimensional drawing recognition and coordinate extraction technology of SegFormer, and the two-dimensional drawing recognition and coordinate extraction technology based on SegFormer includes the following steps: 1) Dataset processing; The data set processing includes assigning different semantic meanings to different color labels, converting the data set blueprint into a high-resolution image, using gray bars to avoid distortion, and then using tools to annotate the image and accurately align the label with the edge of the primitive. The data set must contain multiple categories such as independent single targets, single target intersection, dual target intersection, and multi-target connection. Then, the original image of the data set is denoised by a non-local mean filtering algorithm. The processed image meets the following conditions: Where u is the denoised image, υ is the noisy image, I is the search area, and ω(i,j) is the weight, i.e., the similarity between the i and j region blocks, expressed by the Euclidean distance formula. Where n(i)-the sum of all weights, h-filter coefficient, The Gaussian weighted Euclidean distance between j regional patches, a-standard deviation of the Gaussian kernel, and then the training dataset is expanded by randomly transforming the original image; 2) Coordinate processing; The coordinate processing includes determining the conversion ratio Scale of the X and Y coordinates of the pixel map point coordinates (u, v) and the actual blueprint coordinates (x, y) x Scale y , Then, coordinate transformation is performed according to the coordinate transformation formula (4). In the formula, y max - total image width measured in mm, (u0,v0) and (x0,y0) refer to the different coordinates of the same point in the pixel image and the actual blueprint, respectively; 3) Image processing; The image processing includes extending the image edge and filling it with a white background, enhancing the semantic information area by a dilation algorithm, and extracting the target primitive contour by a threshold extraction and a hybrid algorithm; 4) Hough detection; The Hough detection includes converting the image into a grayscale image, detecting edges, and then performing probabilistic Hough line detection and Hough circle detection; 5) SQL Server query and output coordinates; The SQL Server query and output coordinates include using the SQL Server database service to retrieve contour lines and intersections processed by the Hough transform, and identifying key coordinates of each graphic element; 6) Hyperparameter setting and iterative training; The hyperparameter setting and iterative training, including setting the learning rate, batch size, optimizer, pre-trained weights, learning rate decay type, weight decay, loss function and cycle count parameters, perform deep learning training on the SegFormer model.

6. An intelligent final accounting method for a distribution network construction project, characterized in that: Based on the intelligent settlement method for distribution network construction projects according to any one of claims 1 to 4, the method further comprises the following steps: S106 Final cost calculation; The calculation of final settlement costs includes the calculation of design fees, supervision fees, construction unit management fees, equipment and tools purchase fees, and the generation of final settlement reports; The design fee calculation includes obtaining the billing standard and method of the design fee according to the design contract, calculating the design fee and generating a design fee calculation table, wherein the design fee calculation table includes the billing basis, rate, adjustment coefficient, calculation formula and final amount; The supervision fee, the charging method and rate of the supervision fee are determined according to the project supervision contract, the supervision fee is calculated and a supervision fee calculation table is generated, and the supervision fee calculation table includes the charging basis, rate, adjustment coefficient, calculation formula and final amount; The construction unit management fee is calculated based on the construction unit's internal management fee calculation method and the number and position information of the dispatched project management personnel, and a management fee calculation table is generated. The management fee calculation table includes the billing basis, adjustment coefficient, calculation formula and final amount; The equipment and tools purchase cost calculation includes extracting a list of equipment and tools actually used in the new project from the database, including name, specification model, unit, quantity, and purchase unit price information, collecting relevant purchase contracts, invoices, and payment vouchers, and comparing the information in the database with the purchase information based on the large language model to ensure the authenticity and integrity of the data; calculating the purchase cost of the equipment and tools actually used by the purchase quantity and purchase unit price, and generating a purchase cost detail list, which includes name, specification model, unit, quantity, unit price, and amount fields; The final settlement report generated includes information on the project name, location, scale, construction unit, construction unit, design unit, and supervision unit, describes the construction process, major changes, and delays of the project, lists the calculation process, basis, and amount of each type of expense in separate items, calculates the total final settlement amount based on the settlement price information, design fees, supervision fees, construction unit management fees, and equipment and tool purchase fees, compares the cost budget, summarizes the cost situation of the project based on the large language model, and puts forward targeted improvement suggestions.

7. An intelligent settlement and final accounting system for distribution network construction projects, characterized in that: It includes project initialization module, process planning module, resource management module, engineering quantity calculation module, settlement and final settlement management module and system integration and expansion module: The project initialization module includes a module for setting parameters of a new project and storing basic data; The process planning module includes providing the construction processes entered by the user and the customized construction processes in the wiring diagram design module, civil construction module, overhead line construction module, cable laying module, power distribution equipment installation module, transportation module and disassembly module; The resource management module includes one-click generation of a resource library list based on the in-and-out registration information of materials and mechanical resources, and inputting the comprehensive unit price of each construction process of the completed project into the cost database; The engineering quantity calculation module includes intelligently identifying the content of the drawings and automatically calculating the engineering quantity according to the project design drawings; The settlement and final settlement management module includes generating a settlement and final settlement report based on the settlement and final settlement method according to the engineering quantity calculation results and the data of the cost database, and analyzing the source items of cost differentiation according to the cost budget; The system integration and expansion module includes providing an API interface to make the system compatible with other enterprise resource planning systems and integrate third-party tools and applications.

Images