Construction cost management method, system and equipment for whole engineering settlement process and storage medium

Abstract

The invention provides a project settlement whole process cost management method, system and device, and a storage medium, and relates to the technical field of engineering. The method comprises the steps of obtaining an engineering project plan, and determining an expected cost of a current process; then, the completed project amount, the completed construction cost and the uncompleted project amount are obtained, and manpower, the construction period and raw material data needed for completing the remaining project are determined according to the uncompleted project amount and the project plan; thirdly, obtaining the price trends of manpower and raw materials, and estimating the predicted cost of the uncompleted project amount according to the price trends; and finally, adding the predicted cost of the unfinished project amount and the cost of the finished project amount to obtain the predicted total cost of the current process, and comparing the predicted total cost with the expected cost in real time. And when the comparison difference value exceeds a preset threshold value, generating a risk early warning signal. According to the method, the execution condition of the cost is dynamically tracked, the follow-up cost is predicted in combination with the market price fluctuation trend, the project cost risk can be found in time, and a basis is provided for project management and control decision.

Description

Technical Field

[0001] The present application relates to the engineering field, and specifically to a method, system, equipment and storage medium for cost management throughout the entire process of engineering settlement. Background Art

[0002] During the review process of project settlement, due to the different positions, positions and purposes of the reviewers, as well as the differences in their personal work levels, the review results often differ to varying degrees. Generally speaking, it is normal for the settlement results to differ to a certain extent. However, if the difference is too large, there may be a situation where the cost is intentionally lowered or overestimated.

[0003] In order to strengthen the management of project costs and avoid the problem of excessive settlement differences, some studies have proposed methods for dynamically managing project costs. For example, Chinese patent application CN201911160731.1 discloses a method for dynamic management of the cost of the entire project settlement process. This method compares the actual investment and planned investment for each process of the project. When deviations occur, the deviation data and reasons are promptly reported to the approval personnel to facilitate adjustment of the planned investment when necessary. The approval personnel will proceed to the next process only after completing the approval.

[0004] This method strengthens the dynamic monitoring of project costs to a certain extent and helps control settlement differences. However, only when the current process is completed and the actual investment is inconsistent with the planned investment, the reasons will be analyzed and the corresponding approval will be carried out. This leads to a lag in the management's control over the cost of the entire project and the inability to provide advance warning. Summary of the invention

[0005] The present application provides a method, system, equipment and storage medium for cost management throughout the entire process of project settlement, which can improve the foresight and timeliness of cost monitoring.

[0006] In a first aspect, the present application provides a method for cost management of the entire process of project settlement, the method comprising: Obtain the project plan of the project and determine the expected construction cost of the current process based on the project plan; Obtain the completed engineering quantity, engineering cost of the completed engineering quantity and unfinished engineering quantity of the current process, and determine the manpower data, engineering cycle and raw material data required for the unfinished engineering quantity based on the unfinished engineering quantity and the engineering project plan of the engineering project; Obtaining labor price trend curves and raw material price trend curves, and determining the predicted cost of the unfinished project volume based on the labor price trend curves, raw material price trend curves, required labor data, project cycle and raw material data; Based on the predicted cost of the unfinished project volume and the project cost of the completed project volume, the predicted cost of the current process is determined, and the predicted cost of the current process is compared with the expected project cost in real time to obtain the comparison difference. When the comparison difference exceeds the preset difference threshold, a risk warning signal is generated.

[0007] By adopting the above technical solution, by obtaining the project plan, the expected project cost of the current process is determined, laying the foundation for subsequent cost prediction and risk control. During the construction process, by obtaining data such as the completed project volume, the project cost of the completed project volume and the unfinished project volume, and combining the project plan, the human data, project cycle and raw material data required to complete the unfinished project volume are accurately estimated, providing a key basis for predicting the cost of the unfinished project volume. Further, the present invention uses the human price trend curve and the raw material price trend curve, combined with the required human data, project cycle and raw material data, to predict the human cost and material cost of the unfinished project volume, and obtain the predicted cost of the unfinished project volume. This prediction method fully considers the dynamic change trend of labor and raw material prices, and greatly improves the accuracy and rationality of cost prediction. After obtaining the predicted cost of the unfinished project volume, the present invention adds it to the actual cost of the completed project volume to obtain the predicted cost of the current process, and compares it with the expected project cost in real time to calculate the comparison difference. By setting a reasonable warning threshold, when the comparison difference exceeds the threshold, a risk warning signal is generated in time. This real-time dynamic comparison and early warning mechanism can detect when the project cost deviates from the expected target at the first time, allowing project managers to take prompt countermeasures to reduce the risk of cost overruns and ensure that the project is completed within the planned budget.

[0008] In a second aspect of the present application, a method system for managing the cost of a project settlement in its entire process is provided, comprising: A data acquisition module is used to acquire the engineering project plan of the engineering project and determine the expected engineering cost of the current process according to the engineering project plan; The module for determining the unfinished workload is used to obtain the completed workload of the current process, the engineering cost of the completed workload and the unfinished workload, and determine the manpower data, engineering cycle and raw material data required for the unfinished workload based on the unfinished workload and the engineering project plan; The cost forecasting module is used to obtain the labor price trend curve and the raw material price trend curve, and determine the forecast cost of the unfinished project volume based on the labor price trend curve, the raw material price trend curve, the required labor data, the project cycle and the raw material data; The early warning module is used to determine the predicted cost of the current process based on the predicted cost of the unfinished project volume and the project cost of the completed project volume, and to compare the predicted cost of the current process with the expected project cost in real time to obtain the comparison difference. When the comparison difference exceeds the preset difference threshold, a risk early warning signal is generated.

[0009] In a third aspect of the present application, a computer storage medium is provided, wherein the computer storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the above method steps.

[0010] In the fourth aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the above method.

[0011] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. This application obtains the project plan to determine the expected construction cost of the current process, laying the foundation for subsequent cost forecasting and risk control. During the construction process, by obtaining data such as the completed project volume, the construction cost of the completed project volume, and the unfinished project volume, and combining it with the project plan, the manpower data, construction cycle, and raw material data required to complete the unfinished project volume are accurately estimated, providing a key basis for predicting the cost of the unfinished project volume.

[0012] 2. The present invention uses the labor price trend curve and the raw material price trend curve, combined with the required manpower data, engineering cycle and raw material data, to predict the labor cost and material cost of the unfinished engineering quantity, and obtain the predicted cost of the unfinished engineering quantity. This prediction method fully considers the dynamic change trend of labor and raw material prices, and greatly improves the accuracy and rationality of cost prediction. After obtaining the predicted cost of the unfinished engineering quantity, the present invention adds it to the actual cost of the completed engineering quantity to obtain the predicted cost of the current process, and compares it with the expected engineering cost in real time to calculate the comparison difference. By setting a reasonable warning threshold, when the comparison difference exceeds the threshold, a risk warning signal is generated in time. This real-time dynamic comparison and warning mechanism can detect the deviation of the engineering cost from the expected target at the first time, so that project managers can quickly take countermeasures, reduce the risk of cost overruns, and ensure that the project is completed within the planned budget.

[0013] 3. This application introduces the man-hour utilization rate, a key factor affecting production efficiency, when estimating the number of people in each type of work, and presets the reasonable man-hour utilization rate of each type of work through a series of scientific steps such as reference sample analysis, identification of influencing factors, and utilization rate estimation, rather than subjective assumptions, so that the method is more rigorous and the results are closer to the actual situation of the project. By analyzing and estimating the man-hour utilization rate of each type of work from both horizontal and vertical dimensions, the present invention not only draws on the empirical data of other similar projects, but also fully considers the particularity of this project, and achieves "the stone from other mountains can be used to polish jade". The preset reasonable man-hour utilization rate reflects the expected production efficiency level of each type of work under the plan of this project. Using it to estimate the required number of people can maximize the efficiency of human resource utilization and control labor costs while meeting the construction period requirements. The higher the man-hour utilization rate, the greater the effective workload completed per unit time, and the fewer people are required; conversely, the lower the man-hour utilization rate, the less effective workload actually completed, and the more people are required. The introduction of labor time utilization rate enables the estimation of the number of people required for each type of work to be based on a more scientific and reasonable labor quota, which can better guide the optimal allocation of human resources and ultimately achieve "using the least number of people, doing the most work, and achieving the best results." BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A flowchart of a method for cost management of the entire process of project settlement provided in an embodiment of the present application; Figure 2 An architectural diagram of a cost management system for the entire process of project settlement provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0015] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0016] In the description of the embodiments of the present application, words such as "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "for example" or "for example" is intended to present related concepts in a specific way.

[0017] In the description of the embodiments of the present application, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0018] In order to facilitate understanding of the method and system provided by the embodiments of the present application, before introducing the embodiments of the present application, the background of the embodiments of the present application is first introduced.

[0019] Project settlement is a link that both the construction unit and the construction unit attach great importance to, and is directly related to the economic interests of all parties. In the process of editing and reviewing the project settlement, due to the different positions and purposes of the editors and reviewers of each party, as well as the differences in personal work ability, the editing and review results often have different degrees of deviation. Under normal circumstances, it is normal for the settlement results to have certain differences. However, if the difference is too great, there may be a situation of deliberately suppressing the cost or overestimating it.

[0020] In order to strengthen the management of project costs and avoid the problem of excessive settlement differences, some studies have proposed methods for dynamically managing project costs. For example, Chinese patent application CN201911160731.1 discloses a method for dynamic management of the cost of the entire project settlement process. This method compares the actual investment and planned investment for each process of the project. When deviations occur, the deviation data and reasons are promptly reported to the approval personnel to facilitate adjustment of the planned investment when necessary. The approval personnel will proceed to the next process only after completing the approval.

[0021] This method has strengthened the dynamic monitoring of project costs to a certain extent and helped control settlement differences. However, this method still has shortcomings. The main problem is that only when the current process is completed and the actual investment is inconsistent with the planned investment, the reasons will be analyzed and the corresponding approval will be carried out. This leads to a lag in the management's control over the cost of the entire project, making it difficult to achieve advance warning and timely adjustment.

[0022] After the background introduction of the above content, those skilled in the art can understand the problems existing in the prior art. The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments.

[0023] Based on the above background technology, please refer to Figure 1 , Figure 1 A flow chart of a method for cost management of the whole process of project settlement provided in an embodiment of the present application. The system can be implemented by a computer program or can be run as an independent tool application. Specifically, in an embodiment of the present application, the method can be applied on a server, but can also be applied on electronic devices such as a server. A method for cost management of the whole process of project settlement includes the following steps: S101, obtaining a project plan of the project, and determining an expected construction cost of the current process according to the project plan; Specifically, after obtaining the project plan, the expected project cost of the current process is calculated based on the relevant information in the project plan, such as the project quantity, material usage, labor requirements, etc., combined with the unit price and quota standards of various resources. The expected project cost refers to the amount of cost that should be consumed by the current process, estimated based on design drawings, construction plans, etc. when the project plan is formulated. It reflects the normal cost level of the current process under the planned state and is an important reference for subsequent dynamic cost management.

[0024] S102, obtaining the completed engineering quantity, the engineering cost of the completed engineering quantity and the unfinished engineering quantity of the current process, and determining the manpower data, engineering cycle and raw material data required for the unfinished engineering quantity according to the unfinished engineering quantity and the engineering project plan of the engineering project; Specifically, firstly, the amount of work completed in the current process is obtained through on-site measurement and progress reports, such as the area of ​​poured concrete and the weight of completed steel structures. These data reflect the actual progress of the project and are important inputs for subsequent cost calculations. At the same time, the actual project cost of the completed project is calculated based on the consumption of materials, labor, and machine hours, combined with the corresponding prices.

[0025] After obtaining the data of the completed part, the unfinished amount is obtained by subtracting the completed amount from the total amount. The unfinished amount is a key factor in predicting the completion cost, which determines the resource input and cost expenditure in the subsequent construction process. In order to further quantify the impact of the unfinished amount on the cost, it is necessary to determine the manpower data, engineering cycle and raw material data required to complete this part of the amount based on the unfinished amount and the project plan.

[0026] Based on the above embodiment, as an optional embodiment, the manpower data, engineering cycle and raw material data required for the unfinished engineering volume are determined according to the unfinished engineering volume and the engineering project plan of the engineering project, including: S201, calculating the man-hours required for each type of work according to the labor quota of the current process and the unfinished work quantity in the project plan; Specifically, labor quota refers to the number of workers required to complete a unit of engineering work, which is determined during the preparation of the project plan based on the engineering design drawings, construction plans, quota standards, etc. Labor quotas are usually measured in man-days or man-hours and are subdivided according to different types of work, such as masonry, reinforcement, and carpentry. Labor quota data reflects the human resource input standard for completing a specific process and is an important basis for calculating labor costs.

[0027] After obtaining the labor quota of the current process, it is necessary to multiply it by the unfinished work volume to obtain the man-hour data of each type of work required to complete the remaining work volume. This calculation process can be expressed as: the man-hour required for a certain type of work = the labor quota of the type of work × the unfinished work volume. Through this formula, the time invested in each type of work in the subsequent construction can be estimated.

[0028] S202, estimating the number of people required for each type of work based on the required working hours for each type of work and the preset reasonable working hour utilization rate; Specifically, the man-hour utilization rate refers to the ratio of the time actually used for productive work to the total attendance time per unit time, reflecting the actual utilization efficiency of human resources. In project management, the man-hour utilization rate is usually expressed as a percentage, such as 80%, 90%, etc. Setting a reasonable man-hour utilization rate is determined based on comprehensive consideration of factors such as the construction environment, technical difficulty, labor intensity, and worker skills, reflecting the expectations and requirements for the efficiency of human resource utilization.

[0029] After obtaining the required working hours and working time utilization rate of each type of work, the required working hours can be divided by the working time utilization rate to obtain the actual working hours required for each type of work. Then, based on the normal attendance days or hours per unit time, the number of people required for each type of work can be estimated. This process can be expressed by the formula: the number of people required for a certain type of work = the required working hours for the type of work ÷ the working time utilization rate ÷ the number of attendance per unit time. Among them, the number of attendance per unit time is determined according to the work calendar and attendance requirements in the project plan, such as 8 hours a day, 5 days a week, etc.

[0030] For example, assuming that the masonry workers need 3,000 man-days, the preset man-hour utilization rate is 90%, and the project requires masonry workers to work 8 hours a day, then the actual working time required for masonry workers is 3,333 man-days (i.e., 3,000÷90%), and the required number of workers is 104 (i.e., 3,333÷8÷40). This means that in order to complete the unfinished masonry tasks within the specified construction period, 104 masonry workers need to be arranged to participate in the construction. Similarly, the number of people required for other types of work can be calculated.

[0031] Through the above steps, the introduction of man-hour utilization rate on the basis of man-hour demand and the combination of attendance system can more accurately and reasonably estimate the number of people required for each type of work. On the one hand, by using the man-hour utilization rate indicator, the actual utilization efficiency of human resources can be considered in the man-hour demand calculation, making the estimation result closer to the actual project; on the other hand, converting the required man-hours into the required number of people provides a direct basis for construction organization design and human resource allocation, making it easier for managers to formulate employment plans and optimize human resource allocation.

[0032] Based on the above embodiment, as an optional embodiment, before estimating the number of people required for each type of work according to the required working hours for each type of work and the preset reasonable working hour utilization rate, the method further includes: S301, collect actual man-hour utilization data of various types of work in similar engineering projects to form a reference sample; Specifically, the man-hour utilization rate is an indicator closely related to the type of project, construction conditions, management level, etc. The actual man-hour utilization rates of different projects may vary greatly. If the empirical value or industry average is directly used as the preset man-hour utilization rate, it may not be consistent with the actual situation of the project, resulting in deviations in the number of people estimated. Therefore, it is necessary to collect actual data from similar projects to form a reference sample to provide a basis for determining the preset man-hour utilization rate.

[0033] The process of collecting actual man-hour utilization data for similar engineering projects requires project managers to extensively collect data on similar projects that have been completed or are under construction, with a focus on their human resource management data. The original data such as actual attendance time and completed workload of each project and each type of work can be obtained by consulting project archives, communicating with relevant personnel, and conducting field visits, and then the actual man-hour utilization rate can be calculated. In this process, attention should be paid to selecting appropriate project samples that are as similar as possible to the target project in terms of scale, technology, management, etc., in order to improve the referenceability of the data.

[0034] For example, in a certain construction project, the project manager collected and collated data from 10 similar projects and obtained sample data on the actual man-hour utilization rate of masonry workers: 80%, 85%, 90%, 88%, 92%, 87%, 91%, 89%, 93%, and 90%. These data reflect the actual work efficiency level of masonry workers in similar projects and can serve as an important reference for determining the preset man-hour utilization rate. Similarly, the actual man-hour utilization rate data of other types of work can be collected to form a complete reference sample.

[0035] S302, analyzing the factors affecting the utilization rate of working hours of each type of work according to the reference sample, and obtaining the factors affecting the utilization rate; Specifically, the man-hour utilization rate is a comprehensive indicator that is affected by many factors, such as the difficulty of construction technology, working environment conditions, labor organization management, and worker skill levels. The mode and degree of action of different influencing factors may be different, and it is necessary to identify the key influencing factors through systematic analysis and reveal their mechanism of action. Only by accurately grasping the influencing factors of man-hour utilization rate can we make targeted adjustments when determining the preset value to better meet the actual situation of the project.

[0036] The process of analyzing the factors affecting the utilization rate of working hours in various types of work requires project managers to comprehensively use statistical analysis, causal analysis, expert experience and other methods to deeply explore the laws and relationships contained in the reference sample data. The influencing factors can be analyzed from different angles by combining qualitative and quantitative methods. Qualitative analysis is mainly based on theoretical knowledge and practical experience, and identifies key factors that may affect the utilization rate of working hours from aspects such as management, technology, and environment; quantitative analysis uses mathematical and statistical methods to examine the correlation and significance between various factors and the utilization rate of working hours, and identify the main influencing factors. In the analysis process, attention should be paid to the comprehensiveness and diversity of influencing factors, and both general laws should be revealed and the particularities of different types of work should be considered.

[0037] For example, through qualitative and quantitative analysis of sample data on masonry workers' time utilization, project managers found that the main factors affecting masonry workers' time utilization include: the height and complexity of the masonry site, the supply of materials and equipment, weather conditions, and the level of construction organization and management. Among them, the height and complexity of the masonry site show a significant negative correlation with the time utilization rate, that is, the greater the construction difficulty, the lower the time utilization rate; while the timeliness of material and equipment supply and the level of construction organization and management are positively correlated with the time utilization rate. Timely supply and good management help improve the time utilization rate. Similarly, the factors affecting the time utilization rate of other types of work can be analyzed to obtain a relatively systematic and complete set of utilization influencing factors.

[0038] Through the above steps, based on the reference sample data, the influencing factors of the man-hour utilization rate of each type of work are systematically analyzed to obtain a set of utilization influencing factors, revealing the influencing mechanism of man-hour utilization rate. This analysis process makes full use of the information contained in the reference sample, and extracts key influencing factors from the complex data through qualitative and quantitative analysis methods, making the influencing mechanism clearer and more specific. Clarifying the influencing factors of man-hour utilization rate will help project managers make targeted adjustments when determining the preset values ​​according to the actual situation of the target project to improve the accuracy of the estimate.

[0039] S303, estimating the man-hour utilization rate according to the engineering project plan and utilization rate influencing factors, and obtaining a preset reasonable man-hour utilization rate.

[0040] Specifically, the project plan is an important basis for carrying out the estimation of man-hour utilization. It provides various conditions and requirements for project implementation, such as construction plan, schedule plan, resource allocation, etc., which will have an impact on man-hour utilization. At the same time, the utilization influencing factors obtained from the previous analysis reveal the key factors affecting man-hour utilization and their mechanism of action, providing a basis for judgment and adjustment direction for estimation. In the estimation process, it is necessary to consider both together, and make appropriate adjustments to the man-hour utilization data of the reference sample according to the specific project situation and influencing factors to obtain a preset value that meets the actual situation of the target project.

[0041] The process of estimating the man-hour utilization rate is actually a process in which project managers use qualitative and quantitative analysis methods to make conditional corrections to the reference sample data. First, we need to compare the project plan and analyze the differences between the target project and the reference sample in terms of construction conditions, technical requirements, management level, etc., and determine the direction and degree of impact of these differences on the man-hour utilization rate. Then, combined with the utilization influencing factors determined earlier, the reference sample data is corrected. Analogy, interpolation, regression analysis and other methods can be used to obtain a preliminary estimate. Finally, sensitivity analysis and uncertainty analysis are also required to examine the sensitivity of the estimate to changes in key parameters and the possible range of changes to ensure the robustness of the estimate.

[0042] For example, when estimating the man-hour utilization rate of masonry workers, the project management personnel comprehensively considered the construction plan, schedule requirements, technical specifications and other factors in the project plan, and believed that the overall conditions of the masonry construction of this project were slightly more difficult than the reference sample, so the average man-hour utilization rate might be 3-5 percentage points lower. At the same time, according to the analysis of meteorological data, there may be more rainy days during the construction of this project, which will also lead to a decrease in the man-hour utilization rate. The project department plans to adopt advanced masonry technology and scientific management methods, which can improve the man-hour utilization rate to a certain extent. After comprehensive analysis and calculation, the project management personnel preliminarily estimated that the average man-hour utilization rate of masonry workers in this project was 87%. Through further sensitivity analysis, it was believed that the reasonable range of variation was 85%-90%, and the final preset value was determined to be 87%. Similarly, the preset values ​​of man-hour utilization rates of other types of work can be estimated.

[0043] S203, calculating raw material data according to the material consumption quota and unfinished project quantity of each sub-project in the engineering project plan; Specifically, the material consumption quota refers to the quantity of various materials required to complete a unit of engineering work, which is determined during the preparation of the project plan based on design drawings, construction plans, quota standards, etc. The material consumption quota is usually based on physical quantities such as weight, volume, and area, and is subdivided according to different types of materials, such as cement, sand and gravel, steel bars, wood, paint, etc. The material consumption quota data reflects the material input standards for completing a specific sub-project and is an important basis for preparing material demand plans and procurement plans.

[0044] After obtaining the material consumption quota of each sub-project, it is necessary to multiply it by the corresponding unfinished project volume to obtain the quantity of various materials required to complete the remaining workload, that is, the raw material data. This calculation process can be expressed as: the required quantity of a certain material = the material consumption quota × the unfinished project volume of the corresponding sub-project. Through this formula, the demand for major materials such as cement, sandstone, and steel bars in subsequent construction can be estimated.

[0045] For example, if the steel consumption quota of a reinforced concrete project is 100kg / m3 and the unfinished project volume is 2,000 m3, then the amount of steel bars required to complete the sub-project is 200,000kg, or 200 tons. Similarly, the demand for other materials such as cement and sand and gravel can be calculated. These raw material data provide a quantitative basis for subsequent material procurement, inventory management, and funding arrangements.

[0046] S204, determining the construction period according to the construction project plan and the unfinished construction volume; Specifically, the project plan is a comprehensive and systematic action plan compiled before the implementation of the project based on factors such as engineering tasks, resource conditions, and technical requirements, which clarifies key elements such as the project's goals, content, progress, and resource allocation. Among them, the construction period plan is an important part of the project plan, which stipulates the total construction period of the project and the time arrangement of each stage, and plays a leading and controlling role in the project implementation process.

[0047] After obtaining the project plan, it is necessary to extract the duration data related to the unfinished project, mainly including the total project duration, the planned duration of the sub-project, etc., as a reference for determining the project cycle. At the same time, it is also necessary to fully consider the specific conditions of the unfinished project, such as quantity, difficulty, external conditions, etc., as well as the actual progress and efficiency data of the completed project, and comprehensively analyze and determine the time required to complete the remaining workload. This process requires project managers to use their experience and professional judgment to propose a practical and feasible project cycle plan based on sufficient demonstration.

[0048] For example, suppose the total planned construction period of a project is one year, the current completed construction volume accounts for 30%, and the actual progress is basically consistent with the planned progress. Based on the specific situation of the unfinished construction volume, the project management personnel believe that the difficulty and complexity of the subsequent construction are comparable to those of the previous period, so it is determined that it will take about 210 days to complete the remaining 70% of the construction volume, that is, the construction cycle is 7 months. On this basis, it can be further refined into each sub-project and the completion time of each sub-project can be estimated.

[0049] S205, determining the required manpower data according to the project cycle and the number of people required for each type of work.

[0050] S103, obtaining a labor price trend curve and a raw material price trend curve, and determining a predicted cost of an unfinished project volume according to the labor price trend curve, the raw material price trend curve, the required labor data, the project cycle and the raw material data; Specifically, firstly, by collecting and analyzing historical price data, we draw the labor price trend curve and the raw material price trend curve. The labor price trend curve reflects the market price change trend of various types of workers, such as the fluctuation of the wage level of construction workers, installers, and machine operators over time. The raw material price trend curve reflects the market price change trend of various raw materials, components, and equipment, such as the phased characteristics of the prices of major materials such as cement, steel, sandstone, and formwork. By analyzing these trend curves, we can grasp the overall change rules of labor and raw material prices and provide a reference for predicting future price levels.

[0051] On the basis of obtaining the price trend curve, combined with the manpower data, engineering cycle and raw material data determined in the previous step, the predicted cost of the unfinished project can be estimated. Specifically, firstly, based on the manpower data and engineering cycle, the workload and working time of each type of worker required to complete the unfinished project are estimated. Then, based on the manpower price trend curve, the average wage level of each type of worker during the construction period is predicted. Multiply the workload and wage level to get the predicted value of labor cost.

[0052] Based on the above embodiment, as an optional embodiment, the predicted cost of the unfinished project volume is determined according to the labor price trend curve, the raw material price trend curve, the required labor data, the project cycle and the raw material data, including: S401, determining the predicted labor cost according to the labor price trend curve, the required labor data and the engineering cycle; Specifically, labor cost is an important part of the project cost and directly affects the economic benefits of the project. Since labor prices have a certain time trend and the labor demand in different periods is also different, the prediction of labor costs requires a comprehensive consideration of the dynamic changes in labor prices and the use of human resources. Only by combining the trend of labor prices, the use plan of human resources and the project duration can the labor cost be estimated more accurately, providing a reliable basis for the prediction of project costs.

[0053] The process of determining the predicted labor cost is actually the process of dynamic estimation of labor cost. Project managers need to superimpose and analyze the labor price trend curve, required labor data and engineering cycle obtained above, calculate the labor cost of each period, and accumulate the labor cost of the entire project. First, according to the human resource utilization plan, determine the labor volume of each type of work in different construction periods. Then, according to the labor price trend curve, estimate the unit price level of labor in each period. Then multiply the labor volume and the unit price of labor to obtain the labor cost of each type of work in each period. Finally, add up the labor cost of each type of work in each period to obtain the total predicted labor cost of the project. In this process, we should pay attention to the matching problem of time scale and choose the appropriate calculation time interval, which should reflect the dynamics of labor use and price changes, and take into account the operability of calculation. If necessary, we can also adjust the time value of money to improve the accuracy of estimation.

[0054] For example, when estimating the labor cost of a project, the project manager first divides the project cycle into several periods (such as months), and obtains the planned labor volume of each type of work in each period based on the required labor data, such as 30 masons, 15 carpenters, and 20 steel workers in the first month, 50 masons, 20 carpenters, and 25 steel workers in the second month, and so on. Then, based on the labor price trend curve, the labor unit price of each type of work in different periods is estimated, such as the labor unit price of masons in the first month is 200 yuan / man-day, and 210 yuan / man-day in the second month, the labor unit price of carpenters in the first month is 180 yuan / man-day, and 190 yuan / man-day in the second month, the labor unit price of steel workers in the first month is 220 yuan / man-day, and 230 yuan / man-day in the second month, and so on. Then multiply the planned labor volume and labor unit price of each type of work in each period to get the sub-item labor cost. For example, the labor cost of masonry workers in the first month is 200×30×22=132,000 yuan, and the second month is 210×50×22=231,000 yuan. The labor cost of carpentry workers in the first month is 180×15×22=59,400 yuan, and the second month is 190×20×22=83,600 yuan. The first month of steel workers is 220×20×22=96,800 yuan, and the second month is 230×25×22=126,500 yuan. The same applies to other months. Finally, the labor costs of each type of work in each period are accumulated and summarized to get the total predicted labor cost of the project.

[0055] S402, determining a predicted raw material cost based on raw material data and a raw material price trend curve; Specifically, the cost of raw materials for engineering projects usually accounts for a large proportion of the project cost and has an important impact on the economic benefits of the project. The cost of raw materials depends not only on the quantity and type of raw materials, but also on the market price of raw materials. Since the price of raw materials also has a certain time trend, the forecast of raw material costs requires a comprehensive consideration of the raw material usage plan and the trend of raw material prices. Only by combining the dynamics of the two can the raw material cost be estimated more accurately, providing a complete and reliable basis for the forecast of engineering cost.

[0056] The process of determining the forecasted raw material cost is actually a process of dynamically estimating the raw material cost. Project managers need to determine the planned usage of various raw materials based on the design drawings and construction plans, and then estimate the raw material costs in each period based on the raw material price trend curve, and accumulate and summarize to obtain the raw material cost of the entire project. First, the quantity of various raw materials required for the project should be calculated based on the engineering design and construction organization plan. Then, according to the raw material price trend curve, the price level of raw materials in each period is estimated. Then, the raw material quantity and price are multiplied to obtain the sub-item raw material cost. Finally, the sub-item raw material costs are accumulated and summarized to obtain the predicted total raw material cost. In this process, it should be noted that the time span of the raw material price trend curve should be adapted to the project cycle to ensure the time matching of the estimate. At the same time, the timeliness of raw material supply and quality stability should also be considered. If necessary, the raw material price forecast value needs to be adjusted to improve the reliability of the estimate.

[0057] For example, when estimating the cost of raw materials for a construction project, the project manager first calculated the quantity of various main raw materials required for the project based on the construction drawings and plans, such as 8,000 cubic meters of commercial concrete, 2,000 tons of steel, 600 tons of cement, 15,000 cubic meters of sand and gravel, and 100 cubic meters of wood. Then, based on the price monitoring data of the building materials market, the time trend curves of various raw material prices were drawn, such as the price trend of commercial concrete: 500 yuan / cubic meter in the first quarter, 520 yuan / cubic meter in the second quarter, and 530 yuan / cubic meter in the third quarter; the price trend of steel: 4,000 yuan / ton in the first quarter, 4,200 yuan / ton in the second quarter, and 4,100 yuan / ton in the third quarter; the price trends of other materials are similar. On this basis, according to the planned usage and price forecast of various raw materials, the sub-item raw material costs of each period are estimated. For example, the cost of commercial concrete in the first quarter is 500×3000=1500000 yuan, the second quarter is 520×3000=1560000 yuan, and the third quarter is 530×2000=1060000 yuan; the cost of steel in the first quarter is 4000×800=3200000 yuan, the second quarter is 4200×800=3360000 yuan, and the third quarter is 4100×400=1640000 yuan; the sub-item cost estimation method of other materials is similar. Finally, the sub-item costs of each material in each period are accumulated and summarized to obtain the predicted total raw material cost of the project.

[0058] S403, taking the sum of the predicted labor cost and the predicted raw material cost as the predicted cost of the unfinished project volume.

[0059] Specifically, the project cost is composed of multiple cost items such as labor cost, material cost, machinery cost, and management fee. Among them, labor cost and raw material cost are usually the most important components, and the sum of the two largely determines the total cost level of the project. Therefore, when predicting the cost of unfinished engineering, adding the labor cost and raw material cost estimated previously can get a more accurate predicted cost value. This approach is based on the basic principle of cost composition and is reasonable and feasible.

[0060] The process of adding the predicted labor cost and the predicted raw material cost to get the predicted cost is actually a process of cost aggregation and calculation. Project managers only need to collect and sum up the labor cost and raw material cost data estimated previously according to time periods or cost items to get the predicted total cost. This process is mainly about data collation and calculation. The key is to do a good job of collecting and matching labor costs and raw material costs to ensure the consistency of the two in time and scope. If the estimated time periods of labor costs and raw material costs are inconsistent, it is necessary to convert and adjust the time scales first, and then add them up. At the same time, according to the specific situation, it is also necessary to consider whether it is necessary to convert labor costs and raw material costs at a certain ratio to eliminate the impact of price changes and the time value of money. In the process of summing up, you can first sum up the labor costs and raw material costs in different time periods, and then summarize the subtotals of each time period, or you can first summarize the total labor costs and the total raw material costs separately, and then add the two together, depending on the specific data situation and the convenience of calculation.

[0061] For example, based on the previous estimate of the labor cost and raw material cost of a construction project, the project manager collected and sorted the labor cost and raw material cost data on a monthly basis, and obtained the following comparison table: the labor cost of the first month is 288,200 yuan, and the raw material cost is 1,326,500 yuan; the labor cost of the second month is 441,100 yuan, and the raw material cost is 1,428,300 yuan; the labor cost of the third month is..., the raw material cost is...; and so on. Then, the labor cost of each month is added to the corresponding raw material cost to obtain the subtotal of the predicted cost for each month, such as 1,614,700 yuan in the first month, 1,869,400 yuan in the second month, and the third month..., and so on. Finally, the subtotal values ​​of the predicted cost of each month are summed up and added to obtain the predicted total cost of the unfinished project volume of the project. You can also first summarize the labor cost and raw material cost of each month to obtain the total predicted labor cost and the total predicted raw material cost, and then add the two together to obtain the predicted total cost.

[0062] Through the above steps, the predicted labor cost and predicted raw material cost are added together to obtain the predicted cost of the unfinished project. This estimation result comprehensively considers the main cost components of the project cost. By dynamically predicting and summarizing the labor cost and raw material cost, it overcomes the static limitations of the traditional quota pricing method and improves the accuracy and rationality of the project cost prediction. This "cost-based cost estimation" approach can not only predict the overall cost level of the project, but also predict the time distribution and composition structure of the cost, which has important guiding value for optimizing cost control, capital use, and progress management.

[0063] Based on the above embodiment, as an optional embodiment, before obtaining the labor price trend curve, the method further includes: S501, collecting historical data on raw material prices and data on factors affecting raw material prices; Specifically, the price of raw materials is an important part of the project cost and has a direct impact on the project cost and economic benefits. The price of raw materials is not static, but changes dynamically with the changes in time, market, policy and other factors. In order to accurately grasp the changing trend of raw material prices, it is necessary to collect and organize the historical data of raw material prices and analyze the various factors affecting the price of raw materials. Only on the basis of historical data analysis and combined with the changing trend of price influencing factors can the future trend of raw material prices be predicted more accurately. Therefore, collecting historical data on raw material prices and data on influencing factors is the premise and basis for obtaining the trend curve of raw material prices.

[0064] The process of collecting historical data on raw material prices and data on factors affecting raw material prices requires project managers to collect information extensively and organize data comprehensively. For historical data on raw material prices, it can be obtained from public channels such as statistical yearbooks, industry reports, and data released by price departments, or collected from internal channels such as the company's own procurement records and supplier quotations. The collected data should be as complete and continuous as possible, covering the main types of raw materials, and the time span should match the project cycle. For data on factors affecting raw material prices, information and data related to the prices of main raw materials can be collected from aspects such as macroeconomic policies, industry development trends, supply and demand conditions in the raw material market, and international trade situations. The collected data should be as comprehensive as possible, covering various factors that may affect prices. In the process of data collection, attention should be paid to the timeliness, reliability, and consistency of the data. For cases where the source is unknown, data is missing, and the indicator caliber is inconsistent, necessary screening, correction, and supplementation should be carried out.

[0065] For example, in the cost forecasting process of a certain construction project, project managers need to collect historical price data of major building materials such as cement, steel, sand and gravel, wood, and paint. They collected the monthly average price data of major materials in the past three years by consulting statistical yearbooks, building materials market information websites, and the owner's purchase ledgers, and made a price change table. At the same time, they also collected relevant data such as national macroeconomic policies, fixed asset investment conditions, construction market demand expectations, raw material production capacity and output, and international market price trends, and analyzed the direction and degree of influence of various factors on building material prices. In the process of data collection, they paid attention to identifying the authority of the data source, converted the data with inconsistent statistical calibers, and reasonably interpolated individual missing data to ensure the integrity and comparability of the data.

[0066] S502, constructing lagged price features according to the time series characteristics of historical data of raw material prices; Specifically, the historical data of raw material prices is a set of data series arranged in chronological order, reflecting the numerical changes of raw material prices at different points in time. This time series data contains the trend, periodicity, seasonality and other characteristics of price changes. By constructing lagged price features, the inertia effect and time lag effect of raw material price changes can be revealed, that is, the current price is not only related to current factors, but also to the price level of the past period. By introducing lagged price features, the "autocorrelation" of prices can be considered in the time series model, improving the ability to predict future price trends. Therefore, constructing lagged price features based on the time series characteristics of raw material prices is an important step in exploring the law of price changes and establishing a price prediction model.

[0067] The process of constructing lagged price features based on the time series characteristics of historical data of raw material prices is actually to perform time lag transformation and feature reconstruction on historical data of raw material prices. Project managers need to select appropriate lag orders according to the change cycle of raw material prices, the frequency of price data, etc., move and combine the original price series to form several lagged price series, and incorporate them into the price prediction model as new feature variables. First, according to the fluctuation characteristics of raw material prices, it is necessary to judge whether there is obvious periodicity and seasonality in price changes, and preliminarily determine the possible lag orders. Then, the original price time series is lagged for multiple orders such as 1 period, 2 periods, and 3 periods to form a series of lagged price series. Next, the correlation coefficient between the current price and the price of each lag period is calculated to examine the autocorrelation of the price series. Finally, according to the results of correlation analysis, the lagged price series with strong correlation is selected as the input feature of price prediction to construct a time series prediction model. In this process, it should be noted that price series with different lag periods may have multicollinearity, and necessary feature screening and dimensionality reduction processing are required. At the same time, it is also necessary to consider that the lag characteristics of different material prices may be different, and lag features need to be constructed separately.

[0068] For example, in the forecast of building material price trends, project managers found that steel prices have strong quarterly cyclical characteristics, so they decided to construct quarterly lagged price characteristics. Taking the month as the basic time unit, they lagged the original time series of steel prices by 1 quarter, 2 quarters, 3 quarters, and 4 quarters in turn, and obtained 4 sets of quarterly lagged price series. By calculating the Pearson correlation coefficient between each lagged price series and the current price, it was found that the correlation coefficient between the current price and the price of the previous quarter was as high as 0.8 or above, and the correlation coefficient with the price of the previous two quarters was also above 0.6, while the correlation with the price of earlier quarters was weaker. Based on this, they selected the price of the previous quarter and the price of the previous two quarters as the characteristic variables for predicting the trend of steel prices, and constructed a quarterly autoregressive model. Similarly, for other major raw materials such as cement and wood, corresponding lagged price characteristics were also constructed.

[0069] Based on the above embodiment, as an optional embodiment, a lagged price feature is constructed according to the time series characteristics of the historical data of raw material prices, including: S601, performing a stationary test on the time series of the historical data of the raw material price to obtain a test result, and performing a difference process on the time series when the test result is not unstable, until the test result is stable; Specifically, historical data on raw material prices often present certain time series characteristics, such as trend, periodicity, randomness, etc. These characteristics may cause the statistical properties of raw material price time series to change over time, that is, "non-stationarity". Statistics such as the mean, variance, and autocorrelation coefficient of non-stationary time series will change over time, which violates the basic assumptions of most statistical forecasting models. Direct modeling with non-stationary sequences is prone to problems such as "pseudo-regression", affecting the effectiveness of parameter estimation and the reliability of forecast results. Therefore, before modeling, the stationarity of historical data on raw material prices must be tested. When the sequence is found to be non-stationary, appropriate processing should be taken to convert it into a stationary sequence. Through stationarity testing and processing, non-stationary factors such as trend terms and seasonal terms in the time series can be removed to obtain a new sequence with stable statistical properties, thereby creating favorable conditions for subsequent feature construction and model training.

[0070] The stationarity test of the time series of the historical data of raw material prices is carried out to obtain the test results. When the test results are not stationary, the time series is subjected to difference processing until the test results are stationary. In fact, it is a process of repeatedly exploring the statistical characteristics of the time series. Project managers should first select one or several appropriate test methods to judge the stationarity of the sequence. Commonly used methods include ADF test, PP test, KPSS test, etc. Taking the ADF test as an example, its principle is to test whether there is a unit root in the first-order difference sequence of the sequence. If the null hypothesis of the existence of the unit root cannot be rejected, it means that the original sequence is non-stationary and needs further processing. When the test finds that the sequence is not stationary, the common processing method is to differentiate the sequence, that is, to subtract the previous value from the current value of the sequence to generate a new difference sequence. The first-order difference can effectively eliminate the linear trend in the sequence, while the second-order difference can further eliminate the acceleration trend. The differenced sequence should be re-tested for stationarity. If it is still not stationary, continue to differentiate until the test result is stationary. In practice, it is generally rare to do more than two orders of difference. Once the sequence becomes stationary, the lagged price features can be constructed on this basis.

[0071] For example, in the cement price forecast of a construction project, the project managers tried to use the monthly cement price data of the past two years to construct lagged price features. They first used the ADF test to test the stability of the price series and found that the p-value was 0.32, which was greater than the significance level of 0.05, indicating that the series was non-stationary. To this end, they performed first-order difference processing on the series, that is, using the price of this month minus the price of last month to obtain a price difference series. After testing, the ADF test p-value of the first-order difference series was 0.02, which was less than the significance level of 0.05, indicating that the series after difference has achieved stability. On this basis, they used the first-order difference series as a new price series and constructed a series of lagged price features, such as the differential price values ​​of 1 period, 3 periods, and 6 periods. These features characterize the short-term volatility characteristics of the price series on different time scales and can be used as an important input for subsequent prediction models.

[0072] Based on the above embodiment, as an optional embodiment, a stationary test is performed on the time series of historical data of raw material prices to obtain the test results, including: A time series graph is drawn based on the time series of historical data of raw material prices, and the time series graph is tested through the ADF test to obtain the test results.

[0073] S602, when the test result is stable, the lag order is calculated according to the time series; Specifically, the lagged price feature is an important feature constructed using the autocorrelation of the raw material price time series. The so-called autocorrelation of a time series refers to the correlation between the series at different lag periods. For the time series of raw material prices, the price of this period may not only be related to the price of the previous period, but may also have a certain correlation with the price of the previous period, the previous period, and the previous period. Introducing the price value of the appropriate lag order into the prediction model can make full use of the historical information of the price series, thereby improving the accuracy of price prediction. However, if the lag order is not properly selected, invalid information or redundant information will be introduced, reducing the prediction performance. If the lag order is too small, the long-term relevant information of the price series may be omitted; if the lag order is too large, too much noise interference may be introduced. Therefore, it is necessary to objectively determine the optimal lag order based on the autocorrelation characteristics of the series, and use it to guide the construction of the lagged price feature. By calculating the optimal lag order, the structure and quantity of the lagged price feature can be optimized, and the autocorrelation information of the price series can be extracted to the maximum extent, thereby providing more accurate and stable feature input for the subsequent price prediction model.

[0074] The process of calculating the lag order based on the time series is actually the process of examining the autocorrelation coefficient of the time series at different lags and conducting a significance test. Project managers usually use the partial autocorrelation coefficient to determine the best lag order. The partial autocorrelation coefficient measures the strength of the partial correlation between the series and a certain lag under the condition of controlling other lags. The larger its absolute value, the stronger the explanatory power of the lag for the current period. In a stationary time series, in theory, the partial autocorrelation coefficient will decay and tend to zero as the lag increases. Therefore, if the partial autocorrelation coefficient of a certain lag order is statistically significantly different from zero, while the partial autocorrelation coefficient of the next lag order is not significantly different from zero, then the lag order is usually regarded as the best lag order. In practice, the partial autocorrelation coefficients of each lag order can be jointly tested for significance using methods such as the Q test and the LM test, and the p-value, BIC information criterion and other indicators can be comprehensively examined to objectively determine the best lag order, and based on this, the lagged price characteristics can be constructed in a targeted manner.

[0075] For example, in the asphalt price forecast of a municipal engineering project, the project managers wanted to construct a lag feature based on the asphalt weekly price series of the past three years. After the stationarity test, the price series is a stationary series, which meets the conditions for calculating the lag order. In order to determine the best lag order, they calculated the partial autocorrelation coefficients of the price series from the 1st to the 15th order, and found that the partial autocorrelation coefficients of the first 5 orders were significantly different from zero at a confidence level of 5%, while the partial autocorrelation coefficient of the 6th order was not significantly different from zero. Furthermore, they conducted a Q test on the partial autocorrelation coefficients of the first 15 orders and found that the p-values ​​of the first 5 orders were all less than 0.05, while the p-value of the 6th order was greater than 0.05, and the BIC information criterion also achieved the minimum value at the 5th order. Based on the above results, they determined that the 5th order was the best lag order for the asphalt price series, that is, this week's price was not only related to last week's price, but also significantly related to the price in the previous 5 weeks. Based on this, they constructed five lagged price features of the first, second, third, fourth, and fifth order, and more fully extracted the autocorrelated information contained in the price series.

[0076] S603, extracting features according to the lag order to obtain lagged price features.

[0077] Specifically, the lag order determines the most critical autocorrelation structure in the raw material price time series, that is, there is a significant correlation between the current price and the previous prices. Although this information reveals some inherent laws of the price series, it cannot be directly used to build a price prediction model. In order to give the lag order computability and usability, it is necessary to further extract the lagged price values ​​corresponding to each lag order from the original price series to form a series of quantifiable lagged price features to characterize the short-term fluctuation trend of the price. These lagged features are based on the original price series and make full use of the autocorrelation of the series, which can provide the prediction model with prior information about the short-term price change trend. Through the construction of lagged features, not only the abstract data structure is concretized into a computable feature vector, but also the information related to the prediction problem is extracted from the original data to the greatest extent. The introduction of lagged price features can not only improve the prediction model's ability to characterize short-term price fluctuations, but also control the feature dimension and prevent the dimensional disaster caused by insufficient sample size. Through feature extraction, the autocorrelation of the time series can be fully utilized, and the short-term factors affecting the price can be incorporated into the model to improve the timeliness and accuracy of the prediction.

[0078] The process of extracting features according to the lag order to obtain lagged price features is actually the process of restructuring the original price time series based on the lag order. After determining the optimal lag order p, the project manager reorganizes the original price series into a sample matrix of p+1 columns based on p, where the first column is the predicted current price, the second column to the p+1 column are the price values ​​from lag 1 to lag p, and each row corresponds to a complete price sample. In practice, the above sample matrix can be quickly constructed using methods such as circular shift, that is, the price series is shifted back by 1 to p positions as a whole to obtain p lagged price series, and then the current price series is combined with the p lagged price series by column to obtain the target sample matrix. Each column of the matrix corresponds to a lagged price feature, which contains the fluctuation information of the original price series under a specific time lag. On this basis, more derivative features can be further constructed around each lagged feature, such as month-on-month, year-on-year, moving average, etc., in order to characterize the short-term price change pattern from more angles.

[0079] For example, in the steel price forecast of a certain industrial park project, the project managers determined that the optimal lag order of the monthly steel price series was 3 based on the previous analysis, that is, the price of the current month was not only related to the price of the previous month, but also significantly related to the price in the previous three months. Therefore, they reorganized the steel price series based on 3 and constructed a sample matrix containing 4 columns of features, namely the price of the current month, the price of the previous month, and the price of the previous month. At the same time, they also calculated the first-order difference values ​​of these three lag features, that is, the month-on-month change rate, and included them in the sample matrix as derivative features reflecting the short-term price fluctuation trend. In the end, they obtained a lag price feature set containing 7 features, which more comprehensively characterized the short-term factors affecting the steel price forecast. By inputting these features into price forecasting models such as support vector machines, the trend of steel prices in the next 1 to 2 months can be better predicted, and the average forecast error is controlled within 5%. Through the above steps, the project managers extracted features from the historical data of raw material prices according to the lag order and constructed a series of lag price features.

[0080] S503, screening the data of factors affecting the price of raw materials by a feature selection method to determine the data of significant factors affecting the price of raw materials; Specifically, the data on factors affecting raw material prices usually include a large number of macroeconomic indicators, industry development indicators, market supply and demand indicators, etc. The degree of influence of these indicators on raw material prices varies. Some indicators may be closely related to prices, while some indicators may have a weak relationship with prices. If all influencing factor data are included in the price prediction model, on the one hand, the data dimension and computational complexity will increase, affecting the training efficiency of the model; on the other hand, redundant information and noise data may also be introduced, reducing the prediction accuracy of the model. Therefore, it is necessary to screen the influencing factor data through feature selection methods before modeling, remove redundant factors and non-significant factors, and retain only the key factors that have the most significant impact on prices, which can both reduce the complexity of the model and improve the prediction effect. This practice of determining significant influencing factors through feature selection is an important means of effective dimensionality reduction and data optimization.

[0081] The process of screening the data of factors influencing raw material prices through feature selection and determining the data of significant factors is actually to examine the correlation and statistical significance between each factor and the raw material price. Project managers can use common feature selection methods such as Filter, Wrapper, and Embedded to evaluate the importance of each factor from the perspectives of statistical correlation, prediction contribution, model fitting effect, etc., and screen out significant factors. Taking the Filter method as an example, the Pearson correlation coefficient and Spearman rank correlation coefficient of each factor index and the raw material price can be calculated to examine the strength and direction of the correlation between the two. At the same time, hypothesis testing is also required to determine whether the correlation coefficient passes the test of statistical significance level. Comprehensively analyzing the results of correlation and significance, sorting by the absolute value of the correlation coefficient, and selecting a group of indicators with the strongest correlation and the most significant as key factors. Of course, you can also use Wrapper methods and Embedded methods such as Lasso regression and decision tree to automatically screen out a group of factors that contribute the most to price prediction through iterative training and optimization of the model.

[0082] For example, in the steel price forecast of a certain construction project, the project managers initially collected dozens of influencing factor indicators, including GDP growth rate, fixed asset investment growth rate, real estate development investment growth rate, infrastructure investment growth rate, steel output, steel inventory, iron ore price, coke price, etc. In order to screen out significant influencing factors, they used the Filter feature selection method. First, each indicator was matched with the steel price series to form an indicator-price data pair. Then, the Pearson correlation coefficient between each indicator and the price was calculated, and a significance level of 0.05 was set for a two-sided test. The results showed that the absolute value of the correlation coefficient of five indicators, such as steel output, steel inventory, and iron ore price, was greater than 0.6, and passed the significance test. The correlation coefficient of three indicators, such as infrastructure investment growth rate, was between 0.4-0.6, and also passed the significance test, while the correlation of other indicators was weak or did not pass the significance test. Based on this, they selected the above eight indicators as significant influencing factors of steel prices.

[0083] S504, determining the exchange relationship of the significant influencing factor data to obtain the interaction characteristics between the significant influencing factor data; Specifically, the significant influencing factors of raw material prices are not independent of each other, but have certain internal connections and interactions. For example, there may be a certain positive correlation between the growth rate of GDP and the growth rate of fixed asset investment, because economic growth is often accompanied by an increase in investment demand. For another example, there may be a certain negative correlation between steel production and steel inventory, because an increase in production often leads to a decrease in inventory. The interaction between such influencing factors may produce a "leverage effect" or "buffer effect" on raw material prices, causing the price fluctuation range or fluctuation direction to change. If the price prediction model ignores this interaction effect, it may underestimate or overestimate the role of certain influencing factors, affecting the prediction accuracy. Therefore, it is necessary to reveal the interaction mechanism between significant influencing factors through interactive relationship determination, construct interactive features, and provide more comprehensive and accurate information for the price prediction model. The introduction of interactive features can improve the model's ability to characterize complex influencing mechanisms, thereby improving the accuracy of predictions.

[0084] The process of determining the interactive relationship of the data of significant influencing factors and obtaining the interactive characteristics between the data of significant influencing factors is actually to examine the effect of the combination of each influencing factor on the price of raw materials. Project managers need to match several significant influencing factors in pairs, construct interaction items, introduce them as new features into the price prediction model, and judge whether there is a significant interaction between the factors by evaluating the significance of the regression coefficient of the interaction item. A common practice is to first assume that there may be interactions between certain influencing factors based on business understanding and expert experience, and to construct several interaction items in a targeted manner. For example, it is believed that there may be interactions between macroeconomic factors and industry development factors, and between supply-side factors and demand-side factors, so as to construct corresponding product items as interactive features. Another approach is to use exploratory combination strategies, such as factor analysis and principal component analysis, to automatically find the most likely combination of factors with interactive effects from the perspective of statistical correlation, and then construct interaction items. Regardless of which approach is adopted, after the interaction items are formed, they must be included in the statistical model together with the original influencing factors for regression analysis, and the significance test is used to determine which interaction items have a significant impact on the price of raw materials, and finally screen out effective interaction features.

[0085] For example, in the steel price forecast of a certain construction project, after selecting eight significant influencing factors such as steel production, steel inventory, iron ore price, and infrastructure investment growth rate, the project managers further considered the possible interaction effects between these factors. Based on their professional knowledge, they speculated that there might be strong interactions between steel production and steel inventory, between iron ore price and coke price, and between the growth rate of real estate development investment and the growth rate of infrastructure investment. They then constructed interaction terms such as "steel production × steel inventory", "iron ore price × coke price", and "real estate development investment growth rate × infrastructure investment growth rate", and included them in the multivariate linear regression model together with the original eight influencing factors. Through stepwise regression and F test, it was found that the regression coefficients of the two interaction terms "steel production × steel inventory" and "iron ore price × coke price" reached a significance level of 0.05, so they were identified as effective interaction features for subsequent price prediction model training.

[0086] S505, constructing a sample set based on the lagged price characteristics, significant influencing factor data, interaction characteristics and historical data of raw material prices; Specifically, raw material price prediction is a supervised learning task, which requires the use of known historical data to train the prediction model so that it has the ability to predict future prices. Before training the prediction model, a training sample set must be constructed based on the existing feature data. A complete training sample consists of two parts: a feature vector and a target value, wherein the feature vector contains various feature data that affect the target value, and the target value is the raw material price value to be predicted. In the present invention, the feature vector should include three types of data: lagged price features, significant influencing factor data, and interactive features, and the target value is the actual price in the historical data of raw material prices. Matching the three types of features with the actual price can form a complete training sample. The quality of the training sample directly determines the performance of the prediction model, so the construction of the sample set must be highly valued. By scientifically constructing a sample set containing multidimensional features, the inherent laws of raw material price fluctuations can be fully characterized, laying a data foundation for training a high-precision price prediction model.

[0087] The process of constructing a sample set based on lagged price features, significant influencing factor data, interactive features, and historical data on raw material prices is actually a process of matching and integrating multi-source heterogeneous data and converting them into structured training samples. Project managers need to time-align various feature data with price data to form a synchronized time series data set. For lagged price features, it is necessary to misalign the price series of different lag periods with the current price series, with the current price as the target value and the lagged price as the feature value. For significant influencing factor data, it is necessary to match each influencing factor indicator with the price series in time and match them with a specific time lag period. For example, if there is a certain lag effect in the influencing factor, it may be necessary to match the current period's price with the previous period's influencing factor indicator value. For interactive features, it is necessary to cross-combine the interactive factors and calculate the interactive item value based on the matching influencing factor data. After integrating multi-source data to form a structured data set, it is also necessary to divide the training set, validation set, and test set, and perform necessary preprocessing such as data enhancement and data cleaning to finally form a high-quality training sample set.

[0088] For example, in the concrete price forecast of a certain construction project, the project managers constructed the price characteristics of three time points with a lag of 1, 2, and 3 periods, screened out four significant influencing factor indicators, namely cement price, sand and gravel price, labor cost, and transportation cost, and determined the interactive feature of "cement price × sand and gravel price". At the same time, they collected the monthly price data of concrete in the past five years and formed a price series of 60 time nodes. On this basis, they matched the three types of feature data with the price series, as follows: for the lagged price characteristics, the price series with a lag of 1, 2, and 3 periods were shifted back by 1, 2, and 3 months, respectively, and aligned with the price series of the current month; for the significant influencing factor indicators, considering that these indicators have a stronger correlation with prices in the same period, they were synchronized with the price series; for the interactive characteristics, the cement price series and the sand and gravel price series were first synchronized, and then the product of the two series was calculated month by month to form the interactive term series. Through the above data alignment operation, a data set containing 60 samples is finally formed. Each sample contains 1 current price target value and 8 feature values ​​(3 lagged price features + 4 significant influencing factors + 1 interactive feature), and is randomly divided into training set, validation set and test set.

[0089] S506, constructing an initial raw material price trend prediction model, and training the initial raw material price trend prediction model based on the sample set to obtain a target raw material price trend prediction model.

[0090] Specifically, the raw material price trend prediction model is an intelligent algorithm model that can learn the price fluctuation law from historical price data and influencing factor data, and predict the price trend in the future period based on current information. It is generally composed of a specific network structure and parameter learning mechanism. To build a raw material price trend prediction model, it is necessary to select a suitable model type based on the statistical characteristics of the raw material price time series, such as periodicity, trend, and randomness, combined with the requirements of actual scenarios for prediction accuracy and prediction timeliness. In practice, the commonly used models for price prediction include moving average method, exponential smoothing method, autoregressive moving average model, support vector machine, neural network, etc. Project managers can select one or several models for initialization based on data conditions and task requirements. During initialization, the hyperparameters of the model should be reasonably set based on professional experience, and the parameters of the model should be randomly initialized. After the initial model is built, it is also necessary to train the model using the constructed training sample set. By repeatedly iterating the model parameters, the model can fit the sample data to the maximum extent, thereby mastering the general law of price prediction. Model training is the key link in the evolution of the initial model to the target model. It transforms the model from a general template into a practical tool with strong prediction ability for specific raw material prices. Through model construction and model training, we can obtain a target model that can more accurately predict the future trend of raw material prices, thereby providing strong support for related engineering cost management decisions.

[0091] The process of constructing an initial raw material price trend forecasting model and training the initial raw material price trend forecasting model based on the sample set to obtain the target raw material price trend forecasting model is actually a process of model development, debugging and optimization. After selecting the initial model type, the project manager should design the network structure of the model appropriately according to the scale of the sample data, the number and type of features, etc. For example, for time series models, the time span and frequency of the samples should be considered to select the appropriate number of lags; for machine learning models, the dimension and distribution form of the features should be considered to select the appropriate number of network layers and neurons. After determining the network structure, the hyperparameters of the model, such as the loss function, optimization algorithm, and learning rate, should also be initialized. After initialization, the sample set can be used to train the initial model. The training process usually adopts the supervised learning paradigm, that is, the feature data in the sample set is input into the model, the model is allowed to make predictions, the prediction results are compared with the true target values ​​of the samples, the prediction errors are calculated, and the errors are back-propagated to the model to guide the model to adjust the parameters. Through repeated iterations of the "forward prediction-backward propagation" process, the model will gradually master the inherent laws of price prediction. During the training process, we need to use cross-validation and other strategies to evaluate the model performance, and dynamically adjust the hyperparameters based on the evaluation results. This cycle is repeated until the prediction error of the model on the validation set is sufficiently small or does not decrease significantly. At this point, the optimized model can be determined as the final target price prediction model.

[0092] For example, in the steel bar price forecasting of a certain construction project, project managers selected the long short-term memory neural network (LSTM) as the initial forecasting model based on the constructed training sample set. They comprehensively considered the sequence length and feature dimension of the monthly steel bar price data, and designed a three-layer network structure consisting of 2 LSTM hidden layers (64 neurons in each layer) and 1 fully connected output layer. At the same time, they selected the mean square error as the loss function, the Adam optimizer, and the initial learning rate of 0.01 for parameter learning. During the training process, they adopted the 5-fold cross-validation method, randomly dividing the sample set into 5 mutually exclusive subsets, and each time using 4 of the subsets to train the model, and the remaining 1 subset to evaluate the model, and repeated this 5 times. By monitoring the training error and validation error of each iteration, they found that the error of the model slowed down after the 50th iteration, so the parameter state of the 50th iteration was taken as the optimal state, and the model at this time was determined as the final target forecasting model. After evaluation, the model has a mean absolute percentage error (MAPE) of 3.5% on the test set, which can well predict the price trend of steel bars in the next 1 to 3 months.

[0093] S104, determine the predicted cost of the current process based on the predicted cost of the unfinished project volume and the project cost of the completed project volume, and compare the predicted cost of the current process with the expected project cost in real time to obtain a comparison difference. When the comparison difference exceeds a preset difference threshold, a risk warning signal is generated.

[0094] Specifically, after obtaining the predicted cost of the unfinished project and the project cost of the completed project, the two are added together to obtain the predicted cost of the current process. The predicted cost of the current process reflects the total cost level of the process predicted based on the current actual completion status and future construction plans, and is an important indicator for dynamically evaluating the project cost status.

[0095] In order to reveal the difference between the predicted cost of the current process and the expected level, it is necessary to compare it with the expected project cost in real time. The expected project cost refers to the normal cost level of the current process estimated when the project plan is formulated, reflecting the cost control target of the project under ideal conditions. Comparing the predicted cost with the expected cost can intuitively show whether the current process meets expectations in terms of cost control and how big the deviation is.

[0096] When making a specific comparison, the predicted cost of the current process is subtracted from the expected project cost to obtain a difference, which is called the comparison difference. The positive or negative and size of the comparison difference reflect the direction and degree of deviation of the predicted cost from the expected cost. If the comparison difference is a positive number, it means that the predicted cost is higher than the expected cost, and there is a risk of cost overrun; if the comparison difference is a negative number, it means that the predicted cost is lower than the expected cost, and the cost control is good; if the comparison difference is close to zero, it means that the predicted cost is basically consistent with the expected cost, and the cost control is in line with expectations.

[0097] When the absolute value of the comparison difference exceeds a preset difference threshold, a risk warning signal will be generated. The difference threshold is a reasonable interval set according to factors such as the scale, type, and management requirements of the project. When the comparison difference falls within the threshold interval, although there is a certain deviation, it is within the acceptable range and does not require special treatment; when the comparison difference exceeds the threshold, the deviation is already large, indicating that there are obvious problems in cost control, and timely measures need to be taken to correct and adjust.

[0098] See also Figure 2 , Figure 2 The present invention provides an overall cost management system for project settlement, which may include: Data acquisition module 1, used to acquire the engineering project plan of the engineering project and determine the expected engineering cost of the current process according to the engineering project plan; The unfinished workload determination module 2 is used to obtain the completed workload of the current process, the engineering cost of the completed workload and the unfinished workload, and determine the manpower data, engineering cycle and raw material data required for the unfinished workload based on the unfinished workload and the engineering project plan; The cost prediction module 3 is used to obtain the labor price trend curve and the raw material price trend curve, and determine the predicted cost of the unfinished project volume according to the labor price trend curve, the raw material price trend curve, the required labor data, the project cycle and the raw material data; The early warning module 4 is used to determine the predicted cost of the current process based on the predicted cost of the unfinished project volume and the project cost of the completed project volume, and to compare the predicted cost of the current process with the expected project cost in real time to obtain a comparison difference. When the comparison difference exceeds a preset difference threshold, a risk early warning signal is generated.

[0099] It should be noted that: when the system provided in the above embodiment realizes its functions, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0100] Please refer to Figure 3 The application also discloses an electronic device. Figure 3 The electronic device 300 may include: at least one processor 301, at least one network interface 304, a user interface 303, a memory 305, and at least one communication bus 302 or end-to-end wireless communication.

[0101] The communication bus 302 is used to realize the connection and communication between these components.

[0102] The user interface 303 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0103] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0104] Among them, the processor 301 may include one or more processing cores. The processor 301 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the processor 301 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field~Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 301 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 301, and it can be implemented separately through a chip.

[0105] Among them, the memory 305 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read~Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may optionally be at least one storage system located away from the aforementioned processor 301. Refer to Figure 3 , the memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module and an application program of a method for cost management of the entire process of project settlement.

[0106] exist Figure 3In the electronic device 300 shown, the user interface 303 is mainly used to provide an input interface for the user to obtain the data input by the user; and the processor 301 can be used to call the application program of the whole process cost management method of the project settlement stored in the memory 305, and when executed by one or more processors 301, the electronic device 300 executes one or more methods in the above-mentioned embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited by the described action sequence, because according to the present application, some steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application. In the above embodiments, the description of each embodiment has its own emphasis, and the part that is not described in detail in a certain embodiment can refer to the relevant description of other embodiments.

[0107] In the several implementations provided in this application, it should be understood that the disclosed system can be implemented in other ways. For example, the system embodiments described above are only schematic, such as the division of modules, which is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, indirect coupling or communication connection of systems or modules, which can be electrical or other forms.

[0108] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0109] The present application also provides a computer storage medium that can store multiple instructions, which are suitable for being loaded and executed by a processor as described above. Figure 1 The whole process cost management method of project settlement in the embodiment shown in the figure can be found in the specific implementation process. Figure 1 The specific description of the illustrated embodiment will not be repeated here.

[0110] In addition, each functional module in each embodiment of the present application can be integrated into a processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The above integrated modules can be implemented in the form of hardware or software functional modules.

[0111] If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0112] The above are only exemplary embodiments of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, those skilled in the art will easily think of other embodiments of the present disclosure.

[0113] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A method for cost management in the whole process of project settlement, characterized in that: The method comprises: Obtaining a project plan for the project and determining an expected construction cost for a current process based on the project plan; Obtaining the completed engineering quantity, the engineering cost of the completed engineering quantity and the unfinished engineering quantity of the current process, and determining the manpower data, engineering cycle and raw material data required for the unfinished engineering quantity according to the unfinished engineering quantity and the engineering project plan of the engineering project; Obtaining a labor price trend curve and a raw material price trend curve, and determining a predicted cost of an unfinished project volume based on the labor price trend curve, the raw material price trend curve, the required labor data, the project cycle, and the raw material data; Based on the predicted cost of the unfinished project volume and the project cost of the completed project volume, the predicted cost of the current process is determined, and the predicted cost of the current process is compared with the expected project cost in real time to obtain a comparison difference. When the comparison difference exceeds a preset difference threshold, a risk warning signal is generated.

2. The method according to claim 1, characterized in that The manpower data, engineering cycle and raw material data required for determining the unfinished engineering quantity according to the unfinished engineering quantity and the engineering project plan of the engineering project include: Calculate the working hours required for each type of work based on the labor quota of the current process in the project plan and the unfinished work volume; Estimate the number of people required for each type of work based on the required working hours for each type of work and the preset reasonable working hours utilization rate; Calculate the raw material data according to the material consumption quota of each sub-project in the project plan and the unfinished project quantity; Determining the engineering period according to the engineering project plan and the unfinished engineering quantity; The required manpower data is determined based on the project cycle and the number of people required for each type of work.

3. The method according to claim 1, characterized in that The step of determining the predicted cost of the unfinished project volume according to the labor price trend curve, the raw material price trend curve, the required labor data, the project cycle and the raw material data includes: Determine the predicted labor cost according to the labor price trend curve, the required labor data and the project cycle; Determine the predicted raw material cost according to the raw material data and the raw material price trend curve; The sum of the predicted labor cost and the predicted raw material cost is used as the predicted cost of the unfinished project volume.

4. The method according to claim 1, characterized in that: Before obtaining the human price trend curve, the method further includes: Collect historical data on raw material prices and data on factors affecting raw material prices; constructing a lagged price feature according to the time series characteristics of the historical data of raw material prices; Screening the data of factors influencing the price of raw materials by a feature selection method to determine the data of significant influencing factors; Determining the exchange relationship of the significant influencing factor data to obtain the interaction characteristics between the significant influencing factor data; Constructing a sample set according to the lagged price feature, the significant influencing factor data, the interaction feature and the raw material price historical data; An initial raw material price trend prediction model is constructed, and the initial raw material price trend prediction model is trained based on the sample set to obtain a target raw material price trend prediction model.

5. The method according to claim 2, characterized in that: The method of estimating the number of people required for each type of work based on the required working hours for each type of work and the preset reasonable working hours utilization rate also includes: Collect actual man-hour utilization data of various types of work in similar engineering projects to form a reference sample; Analyze the factors affecting the utilization rate of working hours of each type of work according to the reference sample to obtain the factors affecting the utilization rate; The man-hour utilization rate is estimated according to the engineering project plan and the utilization rate influencing factors to obtain the preset reasonable man-hour utilization rate.

6. The method according to claim 4, characterized in that The step of constructing the lagged price feature according to the time series characteristics of the historical data of raw material prices includes: Performing a stationary test on the time series of the raw material price historical data to obtain a test result, and performing differential processing on the time series when the test result is not unstable until the test result is stable; When the test result is stable, calculating the lag order according to the time series; The lagged price feature is obtained by performing feature extraction according to the lag order.

7. The method according to claim 6, characterized in that The stationary test is performed on the time series of the historical data of the raw material prices to obtain the test results, including: A time series graph is drawn according to the time series of the raw material price historical data, and the time series graph is tested by an ADF test to obtain the test result.

8. A cost management system for the entire process of project settlement, characterized in that: The system comprises: A data acquisition module, used to acquire a project plan of a project and determine the expected construction cost of a current process according to the project plan; The unfinished workload determination module is used to obtain the completed workload of the current process, the engineering cost of the completed workload and the unfinished workload, and determine the manpower data, engineering cycle and raw material data required for the unfinished workload based on the unfinished workload and the engineering project plan of the engineering project; A cost prediction module, used to obtain a labor price trend curve and a raw material price trend curve, and determine the predicted cost of the unfinished project volume according to the labor price trend curve, the raw material price trend curve, the required manpower data, the project cycle and the raw material data; The early warning module is used to determine the predicted cost of the current process based on the predicted cost of the unfinished project volume and the project cost of the completed project volume, and to compare the predicted cost of the current process with the expected project cost in real time to obtain a comparison difference. When the comparison difference exceeds a preset difference threshold, a risk early warning signal is generated.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: It includes a processor, a memory and a transceiver, the memory is used to store instructions, the transceiver is used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the method according to any one of claims 1 to 7.

Images