Power distribution network engineering automatic settlement method and system based on modular assembly

Abstract

The application discloses a power distribution network engineering automatic settlement method and system based on modular assemblies, relates to the technical field of power distribution network engineering cost settlement, and when there is a conflict in the settlement rules, maps each conflict rule to a game participant, defines a payment function, the payment function contains rule authority, inter-rule consistency and deviation from actual field, solves Nash equilibrium through optimal response dynamic iteration, and generates settlement values after conflict resolution according to the equilibrium state; according to the assembly set and the engineering quantity, the settlement values are combined to call the benchmark unit price in the standard feature library to calculate the settlement total price. The application realizes the standardized packaging and automatic identification of the power distribution network engineering construction unit by constructing the modular assembly standard feature library and establishing a two-dimensional matching mechanism of the process topology sequence and the material, can accurately adapt to the material matching, unit splitting and merging scene of the standard engineering and non-standard fine-tuning engineering, and greatly improves the automation degree of engineering quantity identification and statistics.

Description

Technical Field

[0001] This invention relates to the field of distribution network engineering cost settlement technology, and in particular to an automatic settlement method and system for distribution network engineering based on modular assemblies. Background Technology

[0002] Distribution network engineering is characterized by dispersed construction sites, diverse construction scenarios, complex material types, and frequent engineering changes. Currently, the settlement work of domestic distribution network engineering projects relies heavily on manual sorting and calculation and offline process verification. The industry generally adopts the method of manually comparing the design drawings, construction visas, and change documents to count the quantities of each item, and manually applying quotas and pricing rules to complete the cost calculation. The whole process has a high degree of manual intervention, the operation process is cumbersome, and the settlement cycle is lengthy.

[0003] The existing settlement technology system lacks a standardized engineering unit encapsulation mechanism and has not formed a fixed logic for matching construction procedures and materials, making it impossible to automatically associate and match scattered construction materials with standardized engineering units. On-site construction involves a large number of non-standard material combinations and minor adjustments to procedures. Traditional settlement models struggle to automatically identify, split, and merge material combinations, relying solely on manual experience for verification and statistics. This easily leads to problems such as missed or incorrect calculations of project quantities and mismatched material pairings, making it difficult to guarantee the accuracy of settlement data.

[0004] Meanwhile, the settlement data for power distribution network projects comes from diverse sources. Design data, visa data, change order data, and actual construction consumption data often exhibit discrepancies in scope and numerical conflicts. Existing technologies lack a unified mechanism for aligning and verifying quantities, and multi-source data cannot achieve accurate mapping and comparison at the process level, resulting in low efficiency in discrepancy identification and a high rate of missed detections. Various settlement and pricing rules have different levels and applicable scenarios, easily leading to value conflicts for the same accounting object. Traditional methods rely on subjective manual adjudication of conflicts, lacking quantitative judgment criteria, resulting in inconsistent adjudication standards, large errors from human intervention, and easily triggering settlement disputes, affecting the fairness and standardization of project settlement.

[0005] In addition, existing intelligent settlement systems can only realize basic data entry and simple billing statistics. They lack the ability to adaptively resolve rule conflicts, autonomously iterate and optimize non-standard engineering units, and achieve full-process traceability settlement. They cannot adapt to the complex and ever-changing working conditions of power distribution network construction sites and cannot meet the current needs of lean, intelligent and automated settlement management for power distribution network projects, thus hindering the digital upgrade process of power distribution network project cost management. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides an automatic calculation method and system for power distribution network engineering based on modular assemblies. The following technical solution is adopted:

[0007] An automatic calculation method for power distribution network projects based on modular assemblies includes the following steps:

[0008] Step 1: Establish a standard feature library for modular assemblies. Store a process topology sequence template and a material association diagram template for each assembly. The process topology sequence template defines the order and parallel relationship of material installation, and the material association diagram template defines the spatial connection and functional dependency relationship between materials.

[0009] Step 2: Collect settlement basis data for the target project. The settlement basis data includes design drawings, visa forms, change orders, contract settlement rules, and actual material consumption list. Extract the time sequence set of actual materials with timestamps from the construction records.

[0010] Step 3: Transform the actual material time sequence set into the actual process topology sequence and material association diagram, and perform sequence comparison and subgraph matching with the process topology sequence template and material association diagram template of the assembly in the standard feature library. Calculate the matching degree with weighted average, and use greedy cover iteration to identify the modular assembly set that constitutes the project and the splitting and merging relationship.

[0011] Step 4: Using the process nodes in the process topology sequence as a benchmark, align and verify the design quantity, visa quantity, change quantity, and actual consumption quantity, and mark the discrepancies.

[0012] Step 5: When there are conflicts in the settlement rules, each conflicting rule is mapped to a game participant, and a payoff function is defined. The payoff function includes the authority of the rule, the consistency between the rules, and the deviation from the actual situation. The Nash equilibrium is solved dynamically and iteratively through the optimal response, and the settlement value after conflict resolution is generated according to the equilibrium state.

[0013] Step 6: Based on the set of assemblies and quantities identified in Step 3, and combined with the settlement values ​​generated in Step 5, calculate the total settlement price by calling the benchmark unit price in the standard feature library;

[0014] Step 7: Generate and archive the settlement report. The settlement report includes the component matching process, verification of discrepancies, and game resolution records.

[0015] Optionally, in step 3, the process topology sequence template is a directed acyclic graph, where nodes represent material installation processes and edges represent sequential dependencies; the material association graph template is an undirected graph, where nodes represent materials and edges represent spatial co-location or functional association relationships; sequence alignment uses normalized similarity based on graph edit distance, and subgraph matching uses similarity metric based on the maximum common subgraph.

[0016] Optionally, the greedy cover iteration in step 3 includes the following sub-steps:

[0017] Step 31: Calculate the overall matching degree of the current remaining materials with all assemblies in the standard library;

[0018] Step 32: Select the assembly with the highest matching degree that exceeds the threshold as the recognition instance;

[0019] Step 33: Deduct the materials contained in the corresponding assembly from the remaining material set, and update the actual process topology sequence and material association diagram;

[0020] Step 34: Repeat steps 31 to 34 until all materials are covered or cannot be matched. Uncovered materials are aggregated into new assemblies and fed back to the standard feature library.

[0021] Optionally, the alignment verification method in step 4 is as follows: map the quantity data in the design drawings, visa forms, and change orders to the corresponding nodes of the process topology sequence according to the described process parts, forming a multi-source quantity vector for each process node, calculating the difference for each node and generating a difference report containing process identifiers.

[0022] Optionally, the payment function in step 5 is:

[0023] ;in Assign a fixed authority value to the source level of the i-th settlement rule. The reward is the consistency reward between the current strategy of the i-th participant and the strategies of other players in the game. Penalty for deviations from actual on-site consumption. This is the authority weighting coefficient. This is the consistency weighting coefficient. The bias penalty weight coefficient; the optimal response dynamically selects one participant randomly each round, while fixing the strategies of other participants to maximize their own payout.

[0024] Optionally, the method for generating the settlement value after conflict resolution in step 5 is as follows:

[0025] If the deviation of each participant's strategy value is within the preset tolerance when Nash equilibrium is reached, the convergent value is taken as the settlement value; otherwise, the weighted average of each participant's payment value at the equilibrium point is calculated as the settlement value; at the same time, the confidence level of the settlement value is output, and manual review is triggered when the confidence level is lower than the threshold.

[0026] Optionally, in step 6, based on each assembly instance identified in step 3, the corresponding comprehensive unit price and associated quota coefficient are retrieved from the standard feature library; the settlement value generated in step 5 is used as the determined quantity or unit price, multiplied by the comprehensive unit price and combined with the quota coefficient to calculate the individual cost; and the total cost of all assemblies is summed to obtain the total settlement price of the project.

[0027] Optionally, in step 1, the standard feature library supports multi-level nesting of assemblies. The materials of an assembly can contain lower-level assemblies, and the process topology sequence is automatically generated by merging the sequences of the expanded subassemblies.

[0028] Optionally, the settlement report in step 7 includes: the assemblies and matching degree values ​​selected in each greedy coverage iteration, a comparison table of multi-source engineering quantities for each process node, the evolution path of each rule strategy in the game resolution, the final Nash equilibrium state and the generated settlement value.

[0029] The automatic settlement system for distribution network engineering based on modular assemblies is used to realize the automatic settlement method for distribution network engineering based on modular assemblies. The system includes a standard feature library management module, a data acquisition module, an assembly identification module, a two-dimensional matching and assembly identification module, a process node alignment verification module, a game theory conflict resolution module, a settlement cost calculation module, a report generation module, and a central dispatch module.

[0030] The standard feature library management module is used to establish and maintain a standard feature library for assemblies containing process topology sequence templates and material association diagram templates; the data acquisition module is used to acquire multi-source settlement data and extract the actual material time series set; the assembly identification module is used to construct process topology sequences and material association diagrams from the actual material data, perform sequence comparison and subgraph matching, and output the assembly set and splitting and merging scheme through greedy coverage iteration; the process node alignment verification module is used to align multi-source engineering quantities according to process nodes and generate a difference report; the game theory conflict resolution module is used to construct a rule-based payoff function, dynamically solve the Nash equilibrium through the optimal response, and generate the settlement value after conflict resolution based on the equilibrium state; the settlement cost calculation module is used to automatically calculate the total price based on the assembly set, engineering quantity, and settlement value; the report generation module is used to generate and archive traceable reports containing process data; and the central scheduling module is used to control the data interaction and process execution between the modules.

[0031] In summary, the present invention has at least one of the following beneficial technical effects:

[0032] This invention provides an automatic settlement method and system for power distribution network engineering based on modular assemblies. By constructing a standard feature library of modular assemblies and establishing a two-dimensional matching mechanism of process topology sequence and material association, it realizes the standardized encapsulation and automatic identification of construction units in power distribution network engineering. It gets rid of the reliance on traditional manual experience calculation and can accurately adapt to material matching and unit splitting and merging scenarios of standard engineering and non-standard fine-tuning engineering, greatly improving the automation level of engineering quantity identification and statistics, and effectively reducing manual calculation errors.

[0033] By leveraging process nodes to achieve unified alignment and verification of multi-source settlement data, a refined comparison of design, approval, change, and actual consumption data is realized. This automatically identifies discrepancies in quantities, improving the comprehensiveness and accuracy of engineering data verification. A multi-rule conflict resolution model is constructed by introducing game theory. Combining rule authority, rule consistency, and actual on-site deviations, a quantitative payment function is built. The optimal settlement value is iteratively solved using Nash equilibrium, replacing traditional subjective manual adjudication methods. This unifies conflict judgment standards, enabling intelligent and quantitative resolution of settlement rule conflicts, effectively reducing settlement disputes. Attached Figure Description

[0034] Figure 1 This is a flowchart illustrating the automatic settlement method for power distribution network engineering based on modular assemblies according to the present invention.

[0035] Figure 2 This is a schematic diagram of a traceable settlement report according to a specific embodiment of the present invention. Detailed Implementation

[0036] The present invention will be further described in detail below with reference to the accompanying drawings.

[0037] This invention discloses an automatic settlement method and system for power distribution network engineering based on modular assemblies.

[0038] Reference Figure 1 and Figure 2 Example 1, an automatic settlement method for power distribution network projects based on modular assemblies, includes the following steps:

[0039] Step 1: Establish a standard feature library for modular assemblies. Store a process topology sequence template and a material association diagram template for each assembly. The process topology sequence template defines the order and parallel relationship of material installation, and the material association diagram template defines the spatial connection and functional dependency relationship between materials.

[0040] Step 2: Collect settlement basis data for the target project. The settlement basis data includes design drawings, visa forms, change orders, contract settlement rules, and actual material consumption list. Extract the time sequence set of actual materials with timestamps from the construction records.

[0041] Step 3: Transform the actual material time sequence set into the actual process topology sequence and material association diagram, and perform sequence comparison and subgraph matching with the process topology sequence template and material association diagram template of the assembly in the standard feature library. Calculate the matching degree with weighted average, and use greedy cover iteration to identify the modular assembly set that constitutes the project and the splitting and merging relationship.

[0042] Step 4: Using the process nodes in the process topology sequence as a benchmark, align and verify the design quantity, visa quantity, change quantity, and actual consumption quantity, and mark the discrepancies.

[0043] Step 5: When there are conflicts in the settlement rules, each conflicting rule is mapped to a game participant, and a payoff function is defined. The payoff function includes the authority of the rule, the consistency between the rules, and the deviation from the actual situation. The Nash equilibrium is solved dynamically and iteratively through the optimal response, and the settlement value after conflict resolution is generated according to the equilibrium state.

[0044] Step 6: Based on the set of assemblies and quantities identified in Step 3, and combined with the settlement values ​​generated in Step 5, calculate the total settlement price by calling the benchmark unit price in the standard feature library;

[0045] Step 7: Generate and archive the settlement report. The settlement report includes the component matching process, verification of discrepancies, and game resolution records.

[0046] By adopting the above technical solutions, through standardized engineering unit modeling, multi-source data structured acquisition, two-dimensional matching of graph and text topology, process-level data alignment, game theory rule conflict quantification and resolution, modular billing summary and full-process traceability archiving, the entire process of distribution network engineering settlement is automated, quantified and standardized.

[0047] Step 1 establishes a standard feature library for modular assemblies. Based on typical construction processes and material assembly relationships in power distribution networks, various standardized construction units are structurally encapsulated, with assemblies as the smallest standard unit to solidify construction logic and material matching logic. A process topology sequence template is used to constrain the overall construction sequence logic, avoiding statistical deviations in quantities caused by on-site process errors. A material association diagram template is used to solidify the assembly dependencies between materials, ensuring that scattered materials can be categorized and integrated according to the engineering unit dimension.

[0048] Step 2 completes the unified collection and structured processing of multi-source settlement data, integrating the entire chain of data including engineering design, construction process, contractual agreements, and material consumption. At the same time, it extracts construction time sequence information, so that the settlement data has both spatial material attributes and temporal construction attributes, providing a complete data foundation for subsequent topology matching, time sequence alignment, and conflict tracing.

[0049] Step 3 enables intelligent matching of non-standard construction data and standard engineering units on site, transforming disordered on-site material time-series data into a structured topology model. Through two-dimensional similarity weighted matching and greedy iterative coverage mechanism, it automatically completes the identification, splitting and merging of standard components, adapting to complex working conditions such as fine-tuning, addition and subtraction of on-site materials, and non-standard combinations.

[0050] Step 4 relies on unified process nodes to achieve the same source alignment of multi-source engineering quantity data, breaking the problems of inconsistent standards and mismatched locations in traditional settlement data, realizing accurate benchmarking of design, visa, change and actual consumption data, and automatically discovering engineering quantity discrepancies.

[0051] Step 5 addresses the industry pain point of inconsistent values ​​among multiple settlement rules by employing game theory quantitative modeling to transform the rule conflict problem into a multi-participant strategy equilibrium problem. Iterative optimization of strategy values ​​is achieved based on multi-dimensional evaluation indicators, resulting in a unique and reliable settlement value that replaces the manual subjective adjudication mode.

[0052] Step 6 uses the intelligently identified assembly units as the basis for billing, combined with the accurate settlement value after conflict resolution, and links with the standard pricing system to complete the automatic calculation of itemized costs and total costs, ensuring that the billing basis is unified and the calculation logic is standardized.

[0053] Step 7: Collect all process data to form a structured settlement report, enabling full traceability of the matching process, verification results, and conflict resolution process, and meeting the requirements for project settlement audit and archiving management.

[0054] In Example 2, the process topology sequence template in step 3 is a directed acyclic graph, where nodes represent material installation processes and edges represent sequential dependencies; the material association graph template is an undirected graph, where nodes represent materials and edges represent spatial co-location or functional association relationships; sequence alignment uses normalized similarity based on graph edit distance, and subgraph matching uses similarity metric based on the maximum common subgraph.

[0055] By adopting the above technical solution, the process topology sequence template uses a directed acyclic graph structure, with independent material installation processes as basic nodes and directed edges representing the sequential dependencies and parallel constraints between different processes, thus fully restoring the standardized construction process of standard assemblies. This structure can effectively avoid interference from cyclic process logic and is suitable for the process characteristics of linear construction and segmented construction in power distribution networks.

[0056] The material association diagram template adopts an undirected graph structure, with a single material as the basic node and undirected edges representing the dependencies of material space co-location, bracket binding, loop association, and functional matching. It does not involve temporal constraints and only solidifies the static assembly topology.

[0057] The sequence alignment process adopts the normalized similarity measurement principle based on graph edit distance. By statistically analyzing the editing costs of node additions and subtractions, edge relationship shifts, and process sequence misalignments between the on-site process topology and the standard process topology, the temporal matching similarity is normalized to accurately quantify the degree of deviation between the on-site process and the standard process.

[0058] The subgraph matching process adopts the similarity measurement principle based on the maximum common subgraph. By solving the maximum common structure between the on-site material association graph and the standard material association graph, the proportion of common nodes and common edges is statistically analyzed to quantify the matching degree of the material assembly structure and achieve accurate identification of non-standard material combinations.

[0059] Example 3, step 3 of the greedy cover iteration includes the following sub-steps:

[0060] Step 31: Calculate the overall matching degree of the current remaining materials with all assemblies in the standard library;

[0061] Step 32: Select the assembly with the highest matching degree that exceeds the threshold as the recognition instance;

[0062] Step 33: Deduct the materials contained in the corresponding assembly from the remaining material set, and update the actual process topology sequence and material association diagram;

[0063] Step 34: Repeat steps 31 to 34 until all materials are covered or cannot be matched. Uncovered materials are aggregated into new assemblies and fed back to the standard feature library.

[0064] By adopting the above technical solutions, and through a closed-loop logic of iterative screening, material deduction, model updating, and autonomous learning, we can achieve full coverage matching of all materials and autonomous iterative optimization of the standard feature library.

[0065] Step 31: Traverse the current set of remaining unmatched materials and calculate the comprehensive matching degree between the remaining materials and all assemblies in the standard feature library. The comprehensive matching degree is obtained by weighted fusion of process sequence similarity and material structure similarity, which can simultaneously take into account process compliance and material assembly consistency.

[0066] Step 32 sets a fixed matching threshold as an effective identification and judgment condition, filters out the standard assembly with the highest current matching degree and meets the threshold requirement, and judges it as the current optimal engineering instance, ensuring that the matching results of each round have validity and credibility.

[0067] Step 33 completes material coverage deduction and model dynamic update. Remove all materials contained in the matched assembly from the remaining material set, and simultaneously trim the corresponding nodes and edge relationships of the actual process topology sequence and material association graph to ensure that the remaining model and the remaining materials are completely consistent, avoiding repeated matching and redundant calculations.

[0068] Step 34 iterates through the above matching and updating process until all materials on site are matched and covered, or the remaining materials cannot be matched with any standard assembly. For the remaining scattered materials that cannot be matched, the system automatically aggregates them to form new non-standard assemblies, which are then added to the standard feature library. This enables the standard model to autonomously iterate and upgrade with the construction scenario, continuously improving the matching accuracy of subsequent projects.

[0069] Example 4, step 4, the alignment verification method is as follows: the engineering quantity data in the design drawings, visa forms, and change orders are mapped to the corresponding nodes of the process topology sequence according to the described process parts, forming a multi-source engineering quantity vector for each process node, calculating the difference for each node and generating a difference report containing process identifiers.

[0070] By adopting the above technical solution, a unified data mapping benchmark is established, and all unstructured and heterogeneous settlement quantity data are mapped to standardized process nodes in the process topology sequence, enabling comparison of multi-source data at the same node, location, and dimension.

[0071] During implementation, the system automatically parses the quantities in the design drawings, work orders, and change orders to identify the corresponding construction locations and procedures. It then precisely binds various quantity data to the corresponding procedure nodes, constructing a multi-source quantity vector for each procedure node. The quantity vector uniformly aggregates four core data categories: design quantities, work orders, change orders, and actual consumed quantities.

[0072] The system calculates the numerical differences of multi-source engineering quantity data node by node, distinguishing between positive deviation, negative deviation, and no deviation, and simultaneously generates a structured difference report by binding a unique process identifier. This method can achieve difference screening at the level of detailed processes, solving the problem of missing local deviations in traditional overall engineering quantity comparison, and comprehensively improving the accuracy of engineering quantity verification.

[0073] Example 5, the payment function in step 5 is:

[0074] ;in Assign a fixed authority value to the source level of the i-th settlement rule. The reward is the consistency reward between the current strategy of the i-th participant and the strategies of other players in the game. Penalty for deviations from actual on-site consumption. This is the authority weighting coefficient. This is the consistency weighting coefficient. The bias penalty weight coefficient; the optimal response dynamically selects one participant randomly each round, while fixing the strategies of other participants to maximize their own payout.

[0075] By adopting the above technical solutions and constructing a fair and credible rule-based game model through multi-dimensional quantitative indicators, the conflict of settlement rules can be accurately resolved.

[0076] The overall logic of the payment function is to simultaneously consider the legal validity of the rules, the synergy of multiple rules, and the authenticity of on-site construction, so as to avoid the bias in the ruling caused by a single evaluation dimension. The authority of the rule source is assigned a fixed value based on the legal validity level of the settlement documents. The higher the level, the greater the authority value, thus ensuring the priority of formal regulations and contract terms. It serves as a strategy consistency reward metric, used to characterize the degree of alignment between the current rule value and the overall values ​​of other rules, thus suppressing settlement disturbances caused by abnormal biases in a single rule. As a field deviation penalty indicator, it quantifies the degree of deviation between the rule value and the actual material consumption on site, and constrains the settlement value to fit the actual construction conditions.

[0077] α, β, and γ are fixed and adjustable weight coefficients, which correspond to the weight ratios of the authority dimension, consistency dimension, and deviation dimension, respectively. They can be adjusted to suit different settlement scenarios and the settlement control focus of different projects.

[0078] The core principle of optimal response dynamic iteration is to optimize the strategy in a single round. In each iteration, a single player is randomly selected, while the existing strategy values ​​of all other players are fixed. By traversing the feasible value range, the strategy that maximizes its own payoff function is selected as the updated strategy for this round. Through multiple iterations, the strategies of each player are continuously optimized until the overall strategy set tends to stabilize, achieving Nash equilibrium convergence.

[0079] Example 6, the method for generating the settlement value after conflict resolution in step 5 is as follows:

[0080] If the deviation of each participant's strategy value is within the preset tolerance when Nash equilibrium is reached, the convergent value is taken as the settlement value; otherwise, the weighted average of each participant's payment value at the equilibrium point is calculated as the settlement value; at the same time, the confidence level of the settlement value is output, and manual review is triggered when the confidence level is lower than the threshold.

[0081] By adopting the above technical solution, after the Nash equilibrium iteration is completed, the system first determines the dispersion of the strategy values ​​of all game participants. If the deviation between strategy values ​​is less than the preset tolerance, it proves that the multi-source rule values ​​are highly consistent and there is no substantial settlement conflict. The convergent stable value is directly taken as the final settlement value, ensuring that the settlement is simple and efficient.

[0082] If the deviation of the strategy value exceeds the preset tolerance, it proves that there is a substantial difference in the value of multiple rules and there is no unified convergence value. The system uses the payment value of each participant in the equilibrium state as the weight. The higher the payment value, the higher the credibility of the corresponding rule strategy and the larger the corresponding weight ratio. The unique settlement value is obtained by weighted average calculation, realizing the integration of the advantages of multiple rules.

[0083] The system synchronously calculates the overall confidence level of the settlement value, which is derived from the strategy convergence degree, average payout level, and overall deviation level. When the confidence level is lower than the system's preset threshold, it is determined that there is uncertainty in the automatic resolution decision, and a manual review process is automatically triggered to manually verify the game process and the value result, thus avoiding settlement errors in extreme scenarios.

[0084] In Example 7, in step 6, based on each assembly instance identified in step 3, the corresponding comprehensive unit price and associated quota coefficient are retrieved from the standard feature library; the settlement value generated in step 5 is used as the determined quantity or unit price, multiplied by the comprehensive unit price and combined with the quota coefficient to calculate the individual cost; the total cost of all assemblies is summed to obtain the total settlement price of the project.

[0085] By adopting the above technical solution, the system accurately retrieves the corresponding standard comprehensive unit price and scenario-based quota adjustment coefficient from the standard feature library for each valid assembly instance finally identified in step 3. The comprehensive unit price includes the labor cost, basic material cost, and mechanical construction cost corresponding to the assembly, and the quota coefficient is adapted to the scenario-based adjustment needs such as construction environment, construction difficulty, and regional rates.

[0086] The unique settlement value output from the game resolution in step 5 is used as the basis for the final confirmed quantity or unit price of the project, replacing the traditional multi-source dispute data. The settlement value is multiplied by the corresponding comprehensive unit price of the assembly, and combined with the quota coefficient to complete the accurate calculation of the cost of the individual assembly.

[0087] The system aggregates the individual costs of all components level by level, eliminates duplicate billing units and invalid matching units, and finally obtains the total settlement cost of the complete distribution network project. The overall billing logic uses standardized components as units and unified values ​​after conflict resolution as the basis, so as to achieve unified billing standards, transparent calculation process, and accurate and controllable results.

[0088] Example 8: In step 1, the standard feature library supports multi-level nesting of assemblies. The material of an assembly can contain lower-level assemblies, and the process topology sequence is automatically generated by merging the sequences of the expanded subassemblies.

[0089] By adopting the above technical solution, the standard feature library supports a multi-level nested assembly storage structure. The upper-level main assembly is no longer bound to a single basic material, but can contain multiple lower-level sub-assemblies. The sub-assemblies can be further nested into finer-grained construction units, which fully conforms to the engineering characteristics of layered construction and itemized construction of power distribution network projects.

[0090] For nested assemblies, the system has the ability to automatically expand and merge. When constructing the process topology sequence, it automatically disassembles the independent process sequences of all lower-level sub-assemblies, automatically merges and sorts them according to construction dependencies, and generates a complete standardized process topology sequence of the upper-level assembly, ensuring the integrity and accuracy of multi-level assembly matching, verification, and billing.

[0091] Example 9, the settlement report in step 7 includes: the components and matching degree values ​​selected in each greedy coverage iteration, the multi-source engineering quantity comparison table for each process node, the evolution path of each rule strategy in the game resolution, the final Nash equilibrium state and the generated settlement value.

[0092] By adopting the above technical solution, the settlement report fully records the data of the entire greedy coverage iteration process, including the assembly type selected in each iteration, the corresponding comprehensive matching degree value, material deduction records and model update records, and fully restores the assembly identification process.

[0093] The report compiles a multi-source engineering quantity comparison table for all process nodes, and fully preserves the original multi-source data and deviation statistics for each process node, providing data support for the rectification of engineering quantity differences.

[0094] The report records the complete evolution of the game conflict resolution process, including the strategy iteration paths of each participant in the settlement rules, the trend of payout changes, and the final Nash equilibrium parameters, fully demonstrating the quantitative basis for conflict resolution.

[0095] The report ultimately retains the final settlement value and corresponding confidence level results output by the system, realizing full-chain data traceability from data collection, unit identification, engineering quantity verification, conflict resolution to cost accounting.

[0096] Example 10: Automatic settlement system for distribution network engineering based on modular assemblies. This system is used to implement an automatic settlement method for distribution network engineering based on modular assemblies. The system includes a standard feature library management module, a data acquisition module, an assembly identification module, a two-dimensional matching and assembly identification module, a process node alignment verification module, a game theory conflict resolution module, a settlement cost calculation module, a report generation module, and a central dispatch module.

[0097] The standard feature library management module is used to establish and maintain a standard feature library for assemblies containing process topology sequence templates and material association diagram templates; the data acquisition module is used to acquire multi-source settlement data and extract the actual material time series set; the assembly identification module is used to construct process topology sequences and material association diagrams from the actual material data, perform sequence comparison and subgraph matching, and output the assembly set and splitting and merging scheme through greedy coverage iteration; the process node alignment verification module is used to align multi-source engineering quantities according to process nodes and generate a difference report; the game theory conflict resolution module is used to construct a rule-based payoff function, dynamically solve the Nash equilibrium through the optimal response, and generate the settlement value after conflict resolution based on the equilibrium state; the settlement cost calculation module is used to automatically calculate the total price based on the assembly set, engineering quantity, and settlement value; the report generation module is used to generate and archive traceable reports containing process data; and the central scheduling module is used to control the data interaction and process execution between the modules.

[0098] By adopting the above technical solution, the standard feature library management module is responsible for the construction, storage, updating and maintenance of the full set of assembly models, solidifying the process topology templates and material association templates of multi-level nested assemblies, and undertaking the iterative updates of non-standard assemblies on site, continuously optimizing the standard model system.

[0099] The data acquisition module is responsible for collecting, cleaning, and structuring settlement data across the entire domain. It aggregates settlement data from multiple sources and extracts material sets with time-series attributes, providing standardized input data for subsequent model matching and data verification.

[0100] The two-dimensional matching and assembly identification module is responsible for constructing the on-site engineering topology model, performing process sequence comparison and material sub-graph matching, and completing the identification, splitting and merging of assemblies through a greedy iterative coverage algorithm, outputting a set of modular units adapted to the on-site engineering.

[0101] The process node alignment and verification module uses standardized process nodes as a unified benchmark to complete the mapping and binding of multi-source engineering quantity data and the calculation of differences, automatically generating engineering quantity difference reports and realizing intelligent verification of engineering data.

[0102] The game theory conflict resolution module constructs a multi-participant game model, completes the assignment of payoff functions, the iteration of optimal responses, and the solution of Nash equilibrium. It adaptively generates a unique settlement value based on the equilibrium state, thus completing the quantitative resolution of multi-rule conflicts.

[0103] The settlement cost calculation module relies on the identified assembly units and the resolved settlement values, linking standard unit prices and quota coefficients, to automatically calculate individual costs and total costs at different levels.

[0104] The report generation module collects process and result data from the entire process, generates a traceable structured settlement report, and completes electronic archiving to meet auditing and management needs.

[0105] The central scheduling module serves as the core hub of the system, uniformly managing the startup sequence, data interaction, task flow, and anomaly linkage of each module, ensuring the orderly, coherent, and automated execution of the overall settlement process.

[0106] The following describes the implementation principle of the present invention using specific embodiments:

[0107] The 10kV urban branch line renovation power distribution network project was selected as the application scenario. The project includes conventional construction contents such as overhead line replacement, hardware installation, and grounding laying. There are typical settlement scenarios with inconsistencies between the designed engineering quantity, the on-site visa engineering quantity, and the actual material consumption data, as well as conflicts between the quota pricing rules and the contract pricing rules. The method of this invention is used to complete the automatic settlement simulation of the whole process.

[0108] Step 1: Standard Feature Library Retrieval and Engineering Data Acquisition. The system retrieves standard components for overhead lines. This assembly has a preset process topology sequence including four ordered processes: conductor laying, clamp installation, insulator fixing, and grounding construction. The material association diagram includes five core material categories: conductors, tension clamps, suspension clamps, insulators, and grounding electrodes. Simultaneously, raw project settlement data is collected. The designed conductor quantity is 1200m, the on-site certified conductor quantity is 1150m, and the actual material consumption of conductor during construction is 1180m. The contract pricing rules stipulate settlement based on actual quantities, while industry quota rules stipulate settlement based on certified quantities.

[0109] Step 2: Two-dimensional matching and greedy iterative assembly component identification. Extract actual construction timeline and material consumption data, construct the on-site process topology sequence and material association graph, and calculate temporal similarity and structural similarity respectively. Temporal normalized similarity. =94.2%, the structural similarity of the maximum common subgraph of materials. =95.6%. Set time-series weights. =0.4, structural weight =0.6, overall matching degree The calculation formula is as follows:

[0110] ;

[0111] Substituting the numerical values: Match = 95.04%. The system presets a matching threshold of 80%. Since the current matching degree is greater than the threshold, it is determined that the current project is fully compatible with the standard components of overhead lines. There is no need to split or merge; the corresponding component instance is directly output. After deducting all matched materials, the remaining material set is empty, and the matching iteration ends.

[0112] Step 3: Verification of Work Node Quantities. Using the conductor laying work node as a unified benchmark, four types of quantity data are bound to form a work node quantity vector: design quantity 1200m, approval quantity 1150m, change quantity 0m, and actual consumed quantity 1180m. The system compares the quantities node by node to obtain the quantity differences. The design quantity deviates from the actual consumption by +20m, and the approval quantity deviates from the actual consumption by -30m, automatically generating a work node-level quantity difference record.

[0113] Step 4: Conflict Resolution and Settlement Value Calculation in Multi-Rule Game. There are two conflict settlement rules, corresponding to two game participants. Weighting coefficients are set as α=0.5, β=0.3, and γ=0.2. Participant 1 follows the contract-based settlement rule, with an authoritative value assigned. =8.5, strategy value 1180m. Participant 2 is a fixed-amount visa settlement rule, authoritatively assigned value. =7.0, strategy value is 1150m.

[0114]

[0115] Obtain consistency scores and deviation penalty scores; Participant 1's consistency score. =7.2, Deviation Penalty Score =0; Participant 2 consistency score =7.6, Deviation Penalty Score =2.1. Substitute into the payment function formula:

[0116] ;

[0117] Calculate Participant 1's payment value: =0.5×8.5+0.3×7.2-0.2×0=6.41.

[0118] Calculate Participant 2's payout: =0.5×7.0+0.3×7.6-0.2×2.1=5.36.

[0119] The system initiates dynamic iteration for the optimal response, and the strategy stabilizes after 12 iterations. A preset strategy tolerance of 2% is used. If the current relative deviation between the two strategies exceeds this tolerance, the final settlement value is calculated using a weighted average of the equilibrium payout values, as shown in the following formula:

[0120] ;

[0121] Substituting the values: Q = (1180 × 6.41 + 1150 × 5.36) ÷ (6.41 + 5.36) ≈ 1166m. The system's comprehensive iterative convergence and deviation level calculation confidence level is 92.7%, which is higher than the preset threshold of 85%. No manual review is required, and the settlement value is confirmed as a valid basis for billing.

[0122] Step 5, Modular assembly cost calculation. The standard comprehensive unit price for overhead line assemblies is 42 yuan / m, with a quota adjustment coefficient of 1.0 for this project. The cost calculation formula is as follows:

[0123] ;

[0124] Substituting the values ​​into the calculation: Total = 1166 × 42 × 1.0 = 48972 yuan. This project only includes a single core component and has no other sub-item costs. The total project cost is 48972 yuan.

[0125] Step 6: Automatic Archiving of Settlement Reports. The system fully retains the comprehensive matching degree data of the components for this project, details of the differences in the quantities of work processes and nodes, game iteration rounds, the calculation process of the payment values ​​for each rule, the Nash equilibrium convergence state, the final settlement value and confidence level, and the detailed accounting of itemized costs, forming a complete and traceable settlement report and completing electronic archiving, such as... Figure 2As shown, the automated settlement process for this project was completed.

[0126] This case fully replicates the implementation logic of each core step of the invention, employing standardized formulas for quantitative calculation throughout, with rigorous and reproducible numerical derivations. It achieves accurate identification of engineering units through topology matching, quantifies engineering quantity deviations through process alignment, and resolves settlement rule conflicts through a game theory model. The entire process is conducted without human intervention, ensuring that settlement values ​​closely align with actual on-site conditions, resulting in accurate and compliant billing results. This effectively solves the problems of significant disputes and low efficiency in traditional settlement methods.

[0127] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. An automatic settlement method for power distribution network engineering based on modular assemblies, characterized in that, Includes the following steps: Step 1: Establish a standard feature library for modular assemblies. Store a process topology sequence template and a material association diagram template for each assembly. The process topology sequence template defines the order and parallel relationship of material installation, and the material association diagram template defines the spatial connection and functional dependency relationship between materials. Step 2: Collect settlement basis data for the target project. The settlement basis data includes design drawings, visa forms, change orders, contract settlement rules, and actual material consumption list. Extract the time sequence set of actual materials with timestamps from the construction records. Step 3: Transform the actual material time sequence set into the actual process topology sequence and material association diagram, and perform sequence comparison and subgraph matching with the process topology sequence template and material association diagram template of the assembly in the standard feature library. Calculate the matching degree with weighted average, and use greedy cover iteration to identify the modular assembly set that constitutes the project and the splitting and merging relationship. Step 4: Using the process nodes in the process topology sequence as a benchmark, align and verify the design quantity, visa quantity, change quantity, and actual consumption quantity, and mark the discrepancies. Step 5: When there are conflicts in the settlement rules, each conflicting rule is mapped to a game participant, and a payoff function is defined. The payoff function includes the authority of the rule, the consistency between the rules, and the deviation from the actual situation. The Nash equilibrium is solved dynamically and iteratively through the optimal response, and the settlement value after conflict resolution is generated according to the equilibrium state. Step 6: Based on the set of assemblies and quantities identified in Step 3, and combined with the settlement values ​​generated in Step 5, calculate the total settlement price by calling the benchmark unit price in the standard feature library; Step 7: Generate and archive the settlement report. The settlement report includes the component matching process, verification discrepancies, and game resolution records.

2. The automatic settlement method for power distribution network engineering based on modular assemblies according to claim 1, characterized in that, In step 3, the process topology sequence template is a directed acyclic graph, where nodes represent material installation processes and edges represent sequential dependencies; the material association graph template is an undirected graph, where nodes represent materials and edges represent spatial co-location or functional association relationships; sequence alignment uses normalized similarity based on graph edit distance, and subgraph matching uses similarity metric based on the maximum common subgraph.

3. The automatic settlement method for power distribution network engineering based on modular assemblies according to claim 2, characterized in that, Step 3, the greedy cover iteration, includes the following sub-steps: Step 31: Calculate the overall matching degree of the current remaining materials with all assemblies in the standard library; Step 32: Select the assembly with the highest matching degree that exceeds the threshold as the recognition instance; Step 33: Deduct the materials contained in the corresponding assembly from the remaining material set, and update the actual process topology sequence and material association diagram; Step 34: Repeat steps 31 to 34 until all materials are covered or cannot be matched. Uncovered materials are aggregated into new assemblies and fed back to the standard feature library.

4. The automatic settlement method for power distribution network engineering based on modular assemblies according to claim 3, characterized in that, The alignment verification method in step 4 is as follows: map the quantity data in the design drawings, visa forms, and change orders to the corresponding nodes of the process topology sequence according to the described process parts, forming a multi-source quantity vector for each process node, calculating the difference for each node and generating a difference report containing process identifiers.

5. The automatic settlement method for power distribution network engineering based on modular assemblies according to claim 4, characterized in that, The payment function in step 5 is: ;in Assign a fixed authority value to the source level of the i-th settlement rule. The reward is the consistency reward between the current strategy of the i-th participant and the strategies of other players in the game. Penalty for deviations from actual on-site consumption. This is the authority weighting coefficient. This is the consistency weighting coefficient. The bias penalty weight coefficient; the optimal response dynamically selects one participant randomly each round, while fixing the strategies of other participants to maximize their own payout.

6. The automatic settlement method for power distribution network engineering based on modular assemblies according to claim 5, characterized in that, The method for generating the settlement value after conflict resolution in step 5 is as follows: If the deviation of each participant's strategy value is within the preset tolerance when Nash equilibrium is reached, the convergent value is taken as the settlement value; otherwise, the weighted average of each participant's payment value at the equilibrium point is calculated as the settlement value; at the same time, the confidence level of the settlement value is output, and manual review is triggered when the confidence level is lower than the threshold.

7. The automatic settlement method for power distribution network engineering based on modular assemblies according to claim 6, characterized in that, In step 6, based on each component instance identified in step 3, the corresponding comprehensive unit price and associated quota coefficient are retrieved from the standard feature library; the settlement value generated in step 5 is used as the determined quantity or unit price, multiplied by the comprehensive unit price and combined with the quota coefficient to calculate the individual cost; the total cost of all components is summed to obtain the total settlement price of the project.

8. The automatic settlement method for power distribution network engineering based on modular assemblies according to claim 7, characterized in that, In step 1, the standard feature library supports multi-level nesting of assemblies. The materials of an assembly can contain lower-level assemblies, and the process topology sequence is automatically generated by merging the sequences of the expanded subassemblies.

9. The automatic settlement method for power distribution network engineering based on modular assemblies according to claim 8, characterized in that, The settlement report in step 7 includes: the components and matching degree values ​​selected in each greedy coverage iteration, the multi-source engineering quantity comparison table for each process node, the evolution path of each rule strategy in the game resolution, the final Nash equilibrium state and the generated settlement value.

10. An automatic settlement system for power distribution network projects based on modular assemblies, characterized in that: To implement the automatic settlement method for power distribution network engineering based on modular assemblies as described in claim 9, the system includes a standard feature library management module, a data acquisition module, an assembly identification module, a two-dimensional matching and assembly identification module, a process node alignment verification module, a game theory conflict resolution module, a settlement cost calculation module, a report generation module, and a central dispatch module. The standard feature library management module is used to establish and maintain a standard feature library for assemblies containing process topology sequence templates and material association diagram templates; the data acquisition module is used to acquire multi-source settlement data and extract the actual material time series set; the assembly identification module is used to construct process topology sequences and material association diagrams from the actual material data, perform sequence comparison and subgraph matching, and output the assembly set and splitting and merging scheme through greedy coverage iteration; the process node alignment verification module is used to align multi-source engineering quantities according to process nodes and generate a difference report; the game theory conflict resolution module is used to construct a rule-based payoff function, dynamically solve the Nash equilibrium through the optimal response, and generate the settlement value after conflict resolution based on the equilibrium state; the settlement cost calculation module is used to automatically calculate the total price based on the assembly set, engineering quantity, and settlement value; the report generation module is used to generate and archive traceable reports containing process data; and the central scheduling module is used to control the data interaction and process execution between the modules.

Images