Photovoltaic case analysis quota inference method based on case reasoning
By constructing a structured database and a case indexing system, utilizing Neo4j graph database technology, and dynamically adjusting parameter weights based on case reasoning methods, the problem of lag in traditional photovoltaic quota setting was solved, enabling rapid and accurate quota prediction and management, and improving the reliability of engineering decisions.
Patent Information
- Application Number
- CN202511453865.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-13
- Publication Date
- 2026-01-13
AI Technical Summary
Traditional photovoltaic industry quota setting methods are based on static rule bases, which cannot dynamically respond to market changes. They lack systematic mining and correlation analysis of historical engineering data, resulting in repeated trial and error and waste of resources. The cost of manual intervention is high, and the subjectivity and consistency are poor, making it difficult to meet the needs of rapid quotation and real-time cost control.
By constructing a structured database and a case index system, utilizing Neo4j graph database technology, and based on case reasoning methods, the system retrieves historical cases similar to new projects, dynamically adjusts parameter weights, outputs the optimal quota index scheme, and constructs a feedback closed-loop mechanism through a diagnostic report module to automatically record and provide feedback on problems and fit data in the actual project progress.
It enables the rapid and accurate provision of quota data for new projects, reduces repetitive work and trial-and-error costs, improves forecast accuracy and decision reliability, and avoids the lag caused by manual adjustments.
Smart Images
Figure CN121329328A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of artificial intelligence and construction engineering management technology, specifically to a photovoltaic case analysis quota inference method based on case reasoning. Background Technology
[0002] Currently, the traditional photovoltaic industry quota setting is based on fixed engineering quantity specifications, using techniques such as time measurement method and experience estimation method to form a static "rule base".
[0003] However, the static rule base is outdated and struggles to dynamically respond to changes such as market material price fluctuations and construction technology innovations. Furthermore, the massive amount of historical engineering data has not been systematically mined, lacking correlation analysis and knowledge accumulation of historical cases. This prevents the optimization of new project quotas through experience transfer from similar projects, leading to repeated trial and error and resource waste. Simultaneously, static rules are disconnected from dynamic needs: the static rule base cannot adapt to dynamic factors such as market fluctuations and technological advancements, and existing quotas, based on fixed bill of quantities specifications, have long update cycles, failing to meet the specific requirements of different roof types. High manual intervention costs: photovoltaic quota compilation heavily relies on expert experience, requiring manual comparison of various special scenarios, resulting in low automation and difficulty in meeting the needs of rapid quotation and real-time cost control. A contradiction exists between subjectivity and consistency: different experts may have different judgments on the quotas for the same project; for example, the efficiency conversion for corrugated steel roofs and cement roofs may differ, leading to unstable cost estimation results. Summary of the Invention
[0004] To address the aforementioned issues, the purpose of this invention is to propose a photovoltaic case analysis quota inference method based on case reasoning. By converting historical photovoltaic project data into a structured database and retrieving historical cases similar to new projects through a case indexing system, the method reuses their quota indicators and outputs the optimal quota indicator scheme that meets the needs of the new project. This invention can fully utilize past experience to quickly provide reasonable quota data for new projects, reducing repetitive work and trial-and-error costs.
[0005] This was achieved through the following technical solutions: A case-based reasoning method for photovoltaic case analysis and quota inference includes the following steps: S1. Collecting historical project information for each photovoltaic construction project through a quota inference management platform. This historical project information includes basic information, construction characteristic dimensions, and environmental characteristic dimensions. Each historical project is then categorized and stored according to its characteristic dimensions to construct a case library. S2. Based on the historical project information from step S1 and the overall difficulty coefficient of each historical project, establishing a correlation rule model between the overall difficulty coefficient, construction characteristic dimensions, environmental characteristic dimensions, and key parameters in the construction process of any historical project and the quota indicators. The key factors affecting the quota indicators of any historical project are then selected using this correlation rule model. S3. Utilize Neo4j graph database technology to construct a case index system. Simultaneously, based on the key factors influencing the quota indicators of any historical project, input multiple relevant parameters of any new project into the case index system. Dynamically adjust the matching weight coefficient of each relevant parameter of the current new project according to its importance. S4. The case index system retrieves cases from the case library based on the input parameters and corresponding weight coefficients, recommends multiple cases similar to the current new project and their corresponding quota indicators, compares the quota indicator differences between the current new project and the retrieved similar cases, adjusts the quota indicators, and outputs the optimal quota indicator scheme for the current new project. This invention, by transforming historical engineering data into a structured case library, can improve the accuracy of construction quota prediction and efficient management.
[0006] Preferably, in step S1, the basic information includes at least: project name, sub-item names, total engineering workload, daily engineering workload, and number of construction days; the construction characteristic dimensions include at least: roof type, photovoltaic panel model, and installation plan; and the environmental characteristic dimensions include: rainfall, temperature range, and wind speed. By subdividing historical photovoltaic project data into multiple characteristic dimensions, it is helpful to accurately predict the quota indicators for new projects.
[0007] Preferably, in step S2, the overall difficulty coefficient of each historical project is calculated as follows: using the collected information of each historical project, the daily basic difficulty value D is calculated based on the actual and planned completion amounts of the daily tasks for any historical project. Simultaneously, the daily variable adjustment coefficient K is calculated based on the daily construction environment variable level and the weight value of each resource allocation variable. The obtained daily basic difficulty value is then multiplied by the daily variable adjustment coefficient to obtain the daily task difficulty coefficient D. i Then, by calculating the daily task weight W based on the ratio of the daily planned workload to the current total planned workload, we can determine the daily task weight W. i And based on the calculated daily task difficulty coefficient D i Calculate the overall difficulty coefficient D of the current project. tBy calculating the overall difficulty coefficient of each historical project, a valuable benchmark can be provided for similar projects in the future.
[0008] Preferably, the daily basic difficulty value D is calculated as follows: D = Daily planned task completion amount / Daily actual task completion amount, where D > 1 indicates that the actual difficulty of the task that day is higher than the expected task difficulty, and D < 1 indicates that the actual difficulty of the task that day is lower than the expected task difficulty. By calculating the daily basic difficulty value D, the overall difficulty coefficient of each historical project can be accurately calculated subsequently.
[0009] Preferably, resource allocation variables include: daily construction environment, daily personnel attendance rate, daily tool wear and tear, and daily material supply timeliness rate. By calculating the impact of each resource allocation variable on historical projects, the accuracy of predicting quota indicators for new projects can be improved.
[0010] Preferably, the daily variable adjustment factor K is calculated as follows: By calculating the daily variable adjustment factor K, the overall difficulty coefficient of each historical project can be accurately calculated subsequently.
[0011] Preferably, the difficulty impact coefficient of the daily construction environment is quantified by a comprehensive score of the daily construction environment. The difficulty impact coefficients of daily personnel attendance rate, daily tool wear and tear, and daily material supply timeliness rate are calculated as follows: Difficulty Impact Coefficient = 1 / Efficiency. Quantifying the impact of each resource allocation variable on historical projects helps to clearly understand the magnitude of each resource allocation variable's influence on historical projects.
[0012] Preferably, the overall difficulty coefficient D of any project t The calculation method is as follows: The overall difficulty level of the project is D. t It can quickly provide reasonable quota indicators for new projects.
[0013] Preferably, in step S2, a daily work volume quota suggestion for each historical project is generated based on the correlation between the information of each historical project and the overall difficulty coefficient of each historical project using an association rule model. Generating daily work volume quota suggestions for corresponding historical projects through an association rule model allows for the formation of flexible quota indicators that can be adjusted to adapt to different project needs.
[0014] Preferably, in step S4, after obtaining the optimal quota index scheme, the diagnostic report module in the quota inference management platform will continuously track and record the performance of the quota quantities in the optimal quota index scheme in the actual project progress, automatically identify and record relevant problems and adaptability data, and feed the adaptability data back to the quota inference management platform in real time. By recording the performance in the actual project progress through the diagnostic report module, the prediction accuracy and decision reliability of subsequent new project application scenarios can be significantly improved, avoiding the lag caused by manual adjustments and follow-ups.
[0015] The beneficial effects of this invention compared to the prior art are: The technical solution of this invention transforms historical photovoltaic project data into a structured database and retrieves historical cases similar to new projects through a case index system. It reuses these historical cases' quota indicators to output an optimal quota indicator scheme that adapts to the needs of the new project. This allows the invention to fully utilize historical construction project experience, quickly providing reasonable quota data for new projects and reducing repetitive work and trial-and-error costs. Simultaneously, the quota inference management platform constructs a comprehensive feedback loop mechanism through a diagnostic report module. This mechanism continuously tracks and records the performance of the optimal quota indicator scheme's quota quantities in actual project progress, automatically identifies and records relevant problems and adaptation data, and feeds the adaptation data back to the quota inference management platform in real time. This data is then entered into the case library as typical cases, significantly improving the prediction accuracy and decision-making reliability of subsequent new project application scenarios and avoiding the lag caused by manual adjustments. Attached Figure Description
[0016] Figure 1 This is a flowchart of a photovoltaic case analysis quota inference method based on case reasoning. Detailed Implementation
[0017] The following will refer to the appendices in the embodiments of the present invention. Figure 1 The technical solutions in the embodiments of the present invention will be described in detail below.
[0018] like Figure 1The diagram shows a flowchart of a photovoltaic case analysis quota inference method based on case reasoning. First, the quota inference management platform categorizes and stores each historical project according to its characteristic dimensions, constructing a case library. Then, it establishes a rule model linking the overall difficulty coefficient, construction characteristic dimension, environmental characteristic dimension, and key parameters in the construction process of any historical project with quota indicators. Next, it constructs a case index system and inputs multiple relevant parameters of any new project into the system, dynamically adjusting the matching weight coefficient of each parameter based on its importance. Finally, the case index system retrieves cases from the case library based on the input parameters and corresponding weight coefficients, outputting the optimal quota indicator scheme for the current new project. This invention, by transforming historical engineering data into a structured case library, can improve the accuracy of construction quota prediction and efficient management.
[0019] The method specifically includes the following steps: S1. Collect historical project information for each photovoltaic construction project through the quota inference management platform. The historical project information includes: basic information, construction characteristic dimensions and environmental characteristic dimensions. Based on the characteristic dimensions of each historical project information, classify and store each historical project to build a case library.
[0020] Specifically, in step S1, the basic information includes: project name, sub-item name, total engineering workload, daily engineering workload, and number of construction days; the construction characteristic dimensions include: roof type, photovoltaic panel model, laying scheme, support scheme, installation method, operation method, roof area, quality requirements, cable burial and waterproofing scheme, etc.; the environmental characteristic dimensions include: rainfall, temperature range and wind force. This method, by subdividing historical photovoltaic project data into multiple characteristic dimensions, helps to accurately predict the quota indicators of new projects.
[0021] S2. Based on the information of each historical project in step S1 and the overall difficulty coefficient of each historical project, establish a correlation rule model between the overall difficulty coefficient, construction characteristic dimension, environmental characteristic dimension, key parameters in the construction process, and quota indicators of any historical project. Use the correlation rule model to filter out the key factors affecting the quota indicators of any historical project. Among them, the quota indicators include: the number of project workers, the project duration, the project machinery shifts, and the project material requirements.
[0022] In this embodiment, the method for calculating the overall difficulty coefficient of each historical project in step S2 is as follows: using the collected information of each historical project, calculate the daily basic difficulty value D based on the actual and planned completion amounts of the daily tasks for any historical project. Simultaneously, calculate the daily variable adjustment coefficient K based on the daily construction environment variable level and the weight value of each resource allocation variable. Multiply the obtained daily basic difficulty value by the daily variable adjustment coefficient to obtain the daily task difficulty coefficient D. i Then, by calculating the daily task weight W based on the ratio of the daily planned workload to the current total planned workload, we can determine the daily task weight W. i And based on the calculated daily task difficulty coefficient D i Calculate the overall difficulty coefficient D of the current project. t Resource allocation variables include: daily construction environment, daily personnel attendance rate, daily tool wear and tear, and daily material supply timeliness. This invention calculates the overall difficulty coefficient of each historical project to provide a valuable benchmark for similar projects in the future.
[0023] Specifically, the daily base difficulty value D is calculated as follows: D = Daily planned task completion amount / Daily actual task completion amount. Where D > 1 indicates that the actual difficulty of the task that day is higher than the expected difficulty, and D < 1 indicates that the actual difficulty of the task that day is lower than the expected difficulty. The smaller the value of D, the lower the difficulty of the task that day. The daily variable adjustment coefficient K is calculated as follows: The difficulty impact coefficient of the daily construction environment is quantified by the comprehensive score of the construction environment on that day, with the score ranging from 1 to 5 points. The higher the score, the more severe the construction environment and the greater the construction difficulty. The difficulty impact coefficients of the daily personnel attendance rate, daily tool wear level, and daily material supply timeliness rate are calculated as follows: Difficulty impact coefficient = 1 / efficiency. Efficiency is calculated as the daily personnel attendance rate, daily tool wear level, or daily material supply timeliness rate. The daily personnel attendance rate is calculated as: actual number of personnel on duty / planned number of personnel on duty. The daily tool wear level is calculated as: number of tools that can be used normally / planned number of tools to be used. The daily material supply timeliness rate is calculated as: amount of materials supplied on time / planned amount of materials required.
[0024] Among them, the overall difficulty coefficient D of any project t The calculation method is as follows: Let i represent the i-th day of construction for any project, with a value ranging from [1, n], and n be the total number of days of construction for any project. When the overall difficulty coefficient of the project is D... t When the value is higher than 1, it indicates that the overall difficulty of the current project is relatively high, and resource allocation needs to be appropriately adjusted in subsequent quotas. This is achieved through the overall project difficulty coefficient D. t It can quickly provide reasonable quota indicators for new projects.
[0025] In this embodiment, in step S2, the association rule model generates daily engineering quantity quota suggestions for each historical project based on the correlation between each historical project information and the overall difficulty coefficient of each historical project. The quota suggestions include project personnel allocation and machinery scheduling, etc. At the same time, based on the overall difficulty coefficient of the current project, the total engineering quantity quota of the current project is optimized, including the construction period estimation and total resource requirements, etc., which can form flexible quota indicators that can be adjusted flexibly to adapt to different project needs.
[0026] S3. Construct a case index system using Neo4j graph database technology. At the same time, based on the key factors affecting the quota indicators of any historical project, input multiple relevant parameters of any new project into the case index system, such as project type, construction technology, resource constraints and environmental factors. And dynamically adjust the matching weight coefficient of each relevant parameter of the current new project according to the importance of each relevant parameter of the current new project.
[0027] In step S3, Neo4j graph database technology is a graph database management system that uses nodes, relationships, and attributes to intuitively represent and store data. In this invention, a case index system is constructed using Neo4j graph database technology, and a multidimensional semantic association network is established to support rapid retrieval of parameters such as project type, construction process, and resource requirements.
[0028] S4. The case indexing system retrieves cases from the case library based on multiple input parameters and corresponding weight coefficients, recommends several cases similar to the current new project and their corresponding quota indicators, and compares the quota indicator differences between the current new project and the multiple similar cases obtained from the retrieval, adjusts the quota indicators, and outputs the optimal quota indicator scheme for the current new project. Among them, when multiple relevant parameters are input into the case indexing system, the system will automatically extract relevant condition parameters and, by comparing the accuracy with the case library and the bill of quantities specifications, can quickly generate quota indicators that adapt to the needs of the new project scenario, providing strong support for engineering decision-making.
[0029] In this embodiment, when multiple similar cases are retrieved in the case indexing system, the adaptive optimization module in the quota inference management platform supports local adjustments to the retrieved similar cases to ensure the practicality and accuracy of the case library. For example, a retrieved similar case is: flat roof + 25℃ environment, while the current new project is: flat roof + 10℃ environment. The adaptive optimization module needs to adjust the quota indicators and overall difficulty coefficient of the similar case to adapt to the 10℃ environmental difference, thereby improving the reference value of the similar case and ensuring that the quota indicators of the retrieved similar cases can be reused.
[0030] In this embodiment, after obtaining the optimal quota index scheme, the diagnostic report module in the quota inference management platform will continuously track and record the performance of the quota quantity in the optimal quota index scheme in the actual project progress, automatically identify and record relevant problems and adaptability data, and feed the adaptability data back to the quota inference management platform in real time. This can significantly improve the prediction accuracy and decision reliability of subsequent new project application scenarios, and avoid the lag of manual adjustment and follow-up.
[0031] When the diagnostic report module tracks and records data, it first determines the core performance dimensions that need to be tracked and recorded based on the optimal quota index scheme output in step S4, including: time quota, material loss quota, and machinery shift quota, and ensures that the tracked content is directly related to the quota target. Secondly, it collects actual project data in real time from multiple channels through system integration and manual assistance. For example, it automatically collects the actual usage time of machinery through the GPS positioning of the crane and the time recorder; and it automatically obtains the material warehousing, outbound, and loss data through the material management system. Finally, the diagnostic report module compares the collected actual project data with the quota quantity in the optimal quota index scheme in real time, calculates the deviation rate, and records the time, link, and related factors of the deviation, forming a tracking file of "quota quantity - actual performance - deviation analysis".
[0032] In summary, this invention transforms historical photovoltaic project data into a structured database and retrieves similar historical cases for new projects through a case indexing system. It reuses these historical cases' quota indicators to output the optimal quota indicator scheme adapted to the needs of new projects. This allows the invention to fully utilize historical construction project experience, quickly providing reasonable quota data for new projects and reducing repetitive work and trial-and-error costs. Simultaneously, the quota inference management platform constructs a comprehensive feedback loop mechanism through a diagnostic report module. This mechanism continuously tracks and records the performance of the optimal quota indicator scheme's quota quantities in actual project progress, automatically identifies and records relevant problems and adaptation data, and feeds the adaptation data back to the quota inference management platform in real time. This data is then entered into the case library as typical cases, significantly improving the prediction accuracy and decision-making reliability of subsequent new project application scenarios, avoiding the lag caused by manual adjustments, and demonstrating significant progress.
[0033] The above embodiments are merely illustrative of the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solutions based on the technical concept proposed in this invention shall fall within the scope of protection of this invention.
Claims
1. A photovoltaic case analysis quota inference method based on case reasoning, characterized in that, The method includes the following steps: S1. Collect historical project information for each photovoltaic construction project through the quota inference management platform. The historical project information includes: basic information, construction characteristic dimensions and environmental characteristic dimensions. Based on the characteristic dimensions of each historical project information, classify and store each historical project to build a case library. S2. Based on the information of each historical project in step S1 and the overall difficulty coefficient of each historical project, establish a correlation rule model between the overall difficulty coefficient, construction feature dimension, environmental feature dimension, key parameters in the construction process, and quota indicators of any historical project. Use the correlation rule model to filter out the key factors that affect the quota indicators of any historical project. S3. Construct a case index system using Neo4j graph database technology. At the same time, based on the key factors affecting the quota indicators of any historical project, input multiple relevant parameters of any new project into the case index system. And dynamically adjust the matching weight coefficient of each relevant parameter of the current new project according to the importance of each relevant parameter of the current new project. S4. The case indexing system retrieves cases from the case library based on multiple input parameters and corresponding weight coefficients, recommends multiple cases similar to the current new project and their corresponding quota indicators, compares the quota indicator differences between the current new project and the multiple similar cases retrieved, adjusts the quota indicators, and outputs the optimal quota indicator scheme for the current new project.
2. The photovoltaic case analysis quota inference method based on case reasoning according to claim 1, characterized in that, In step S1, the basic information should include at least: project name, sub-item name, total engineering workload, daily engineering workload, and number of construction days; the construction characteristic dimensions should include at least: roof type, photovoltaic panel model, and installation plan; the environmental characteristic dimensions should include: rainfall, temperature range, and wind speed.
3. The photovoltaic case analysis quota inference method based on case reasoning according to claim 1, characterized in that, In step S2, the overall difficulty coefficient of each historical project is calculated as follows: using the collected information of each historical project, the daily basic difficulty value D is calculated based on the actual and planned completion amounts of the daily tasks for any historical project. Simultaneously, the daily variable adjustment coefficient K is calculated based on the daily construction environment variable level and the weight value of each resource allocation variable. The obtained daily basic difficulty value is then multiplied by the daily variable adjustment coefficient to obtain the daily task difficulty coefficient D. i ; Then, the daily task weight W is calculated by comparing the daily planned workload with the current total planned workload. i And based on the calculated daily task difficulty coefficient D i Calculate the overall difficulty coefficient D of the current project. t .
4. The photovoltaic case analysis quota inference method based on case reasoning according to claim 3, characterized in that, The daily base difficulty value D is calculated as follows: D = Daily task planned completion amount / Daily task actual completion amount. When D > 1, it means that the actual difficulty of the task on that day is higher than the expected difficulty of the task. When D < 1, it means that the actual difficulty of the task on that day is lower than the expected difficulty of the task.
5. The photovoltaic case analysis quota inference method based on case reasoning according to claim 3, characterized in that, Resource allocation variables include: daily construction environment, daily staff attendance rate, daily tool wear and tear, and daily material supply timeliness.
6. The photovoltaic case analysis quota inference method based on case reasoning according to claim 3, characterized in that, The daily variable adjustment factor K is calculated as follows: .
7. A photovoltaic case analysis quota inference method based on case reasoning according to claim 5 or 6, characterized in that, The difficulty impact coefficient of the daily construction environment is quantified by the comprehensive score of the construction environment on that day. The difficulty impact coefficients of the daily personnel attendance rate, daily tool wear and tear, and daily material supply timeliness rate are calculated as follows: Difficulty impact coefficient = 1 / efficiency.
8. The photovoltaic case analysis quota inference method based on case reasoning according to claim 3, characterized in that, Overall difficulty level of any project: D t The calculation method is as follows: .
9. The photovoltaic case analysis quota inference method based on case reasoning according to claim 1, characterized in that, In step S2, based on the correlation between each historical project information and the overall difficulty coefficient of each historical project, the association rule model will generate the daily engineering quantity quota suggestion for the corresponding historical project.
10. The photovoltaic case analysis quota inference method based on case reasoning according to claim 1, characterized in that, In step S4, after obtaining the optimal quota index scheme, the diagnostic report module in the quota inference management platform will continuously track and record the performance of the quota quantity in the optimal quota index scheme in the actual project progress, automatically identify and record relevant problems and adaptation data, and feed the adaptation data back to the quota inference management platform in real time.