Enterprise project profit and loss prediction method based on time sequence model
By using a time series model-based approach and combining multi-dimensional data from an enterprise project management system for data preprocessing and model training, the accuracy and reliability issues of profit and loss prediction in existing technologies have been resolved, achieving high-precision project profit and loss prediction.
Patent Information
- Application Number
- CN202511441758.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-10
- Publication Date
- 2026-01-20
AI Technical Summary
In existing enterprise project management systems, project profit and loss prediction lacks the exploration of the time dimension characteristics of data, and existing prediction methods rely on human experience or simple statistical analysis, making the prediction results susceptible to subjective factors and resulting in low accuracy and reliability.
A time series model-based approach is adopted, which collects multi-dimensional data from the project management system through a scheduled task executor, performs data preprocessing and model training, and combines the periodic patterns in the project operation process to predict profits and losses, including data cleaning, completion, formatting and model fitting.
It improves the accuracy and adaptability of project profit and loss forecasting, enabling more accurate prediction of profit and loss in the future, and is applicable to the full life cycle management of different types of projects.
Smart Images
Figure CN121365772A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application discloses an enterprise project profit and loss prediction method based on a time sequence model and relates to the technical field of enterprise project management and data analysis. BACKGROUND
[0002] In enterprise project management, project profit and loss prediction provides a basis for resource allocation, risk control and decision making. Although existing enterprise project management systems store a large amount of data such as project repayments, costs, personnel working hours and the like, the following limitations exist: first, data application is limited to the recording level and lacks exploration of the time dimension characteristics of data; and second, existing prediction methods mostly rely on artificial experience or simple statistical analysis and cannot effectively combine the mutual relationship among personnel, progress, cost, repayment and other variables, resulting in that the prediction result is easily affected by subjective factors and has low accuracy and reliability. SUMMARY
[0003] The application provides an enterprise project profit and loss prediction method based on a time sequence model, which fully utilizes multi-dimensional operation data in an enterprise project management system and combines the dynamic analysis capability of a time sequence model to accurately predict the profit and loss of a project in a future period of time and provide strong data support for enterprise project decision making.
[0004] The specific scheme provided by the application is as follows:
[0005] The application provides an enterprise project profit and loss prediction method based on a time sequence model, which fully utilizes multi-dimensional operation data in an enterprise project management system and combines the dynamic analysis capability of a time sequence model to accurately predict the profit and loss of a project in a future period of time and provide strong data support for enterprise project decision making.
[0006] Step 1: data acquisition: an operating data in a project management system is collected by a timing task executor through a data interface, multi-dimensional project data is collected in real time at a preset time interval, the multi-dimensional project data includes personnel input data, project node data, repayment condition data and cost data,
[0007] Step 2: data preprocessing: the collected data is preprocessed by the timing task executor to form a data set and guarantee data quality,
[0008] Step 3: total cost accounting: unit working hour cost in the preprocessed project personnel working hour data is converted into working hour cost by the timing task executor, the unit working hour cost is an enterprise historical average working hour unit price, the working hour cost is combined with direct cost and indirect cost to obtain total cost,
[0009] Step 4: model training and updating: personnel input, project node, repayment condition and total cost data in the data set are taken as input by the timing task executor, a prediction model is fitted, and the deployed prediction model is updated,
[0010] Step 5: Model prediction is performed: the model is called through an API interface, and the project number and related personnel input, project node, repayment situation and total cost data in the time period are input to perform profit and loss data prediction.
[0011] Further, the multi-dimensional project data collected in step 1 of the enterprise project profit and loss prediction method based on the time series model includes:
[0012] Personnel input data: number of personnel input in each project phase, post type, labor unit price, cumulative working hours,
[0013] Project node data: project start time, end time, progress completion rate, node delay time,
[0014] Repayment situation data: amount of repayment, repayment time, contract repayment node, estimated repayment amount and time,
[0015] Cost data: material procurement cost, equipment rental, procurement cost, operation and maintenance service cost.
[0016] Further, the pre-processing in step 2 of the enterprise project profit and loss prediction method based on the time series model includes:
[0017] Data cleaning: abnormal value detection algorithm is used to identify and correct abnormal values in the data, and repeated records are deleted,
[0018] Data completion: default value is used to complete the missing non-key data, and manual recording is prompted for key data,
[0019] Time formatting: node time and repayment time are formatted into standard date format, and project phase duration and repayment interval features are extracted.
[0020] Further, in step 4 of the enterprise project profit and loss prediction method based on the time series model, the data set is divided into training set and validation set in the ratio of 7:3 by the timing task executor, and the personnel input, project node, repayment situation and total cost data in the training set are used as input to fit the prediction model.
[0021] The application also provides an enterprise project profit and loss prediction system based on a time series model, which comprises a collection module, a preprocessing module, an accounting module, a training and updating module and a prediction module,
[0022] The collection module collects data: the timing task executor collects the operating data in the project management system through the data interface, and collects multi-dimensional project data in real time at preset time intervals, including personnel input data, project node data, repayment situation data and cost data,
[0023] The preprocessing module performs data preprocessing: the collected data is preprocessed by a timing task executor to form a data set, and the data quality is guaranteed,
[0024] The accounting module accounts for the total cost: the unit labor cost in the preprocessed project personnel working hour data is converted into labor cost by the timing task executor, the unit labor cost is the historical average labor cost per hour of the enterprise, and the labor cost is combined with the direct cost and the indirect cost to obtain the total cost,
[0025] The training and updating module trains and updates the model: the personnel input, project node, repayment situation and total cost data in the data set are input to fit the prediction model, and the deployed prediction model is updated,
[0026] The prediction module performs model prediction: through the API interface call, the project number and related personnel input, project node, repayment situation and total cost data in the time period are input to predict the profit and loss data.
[0027] Further, the multi-dimensional project data collected by the collection module of the enterprise project profit and loss prediction system based on the time series model includes:
[0028] Personnel input data: number of personnel input in each project phase, post type, labor unit price, cumulative working hours,
[0029] Project node data: project start time, end time, progress completion rate, node delay time,
[0030] Repayment situation data: amount of repayment, repayment time, contract repayment node, estimated repayment amount and time,
[0031] Cost data: material procurement cost, equipment rental, procurement cost, operation and maintenance service cost.
[0032] Further, the preprocessing module of the enterprise project profit and loss prediction system based on the time series model includes:
[0033] Data cleaning: identify and correct outliers in the data using outlier detection algorithms, and delete duplicate records,
[0034] Data completion: use default values to complete missing non-critical data, and manually record critical data with alarm prompts,
[0035] Time formatting: format node time and repayment time into standard date format, extract project phase duration and repayment interval features.
[0036] Further, the training update module of the enterprise project profit and loss prediction system based on a time series model divides the data set into a training set and a validation set by a timing task executor in a 7:3 ratio, and takes the personnel input, project node, repayment condition and total cost data in the training set as input to fit the prediction model.
[0037] The present application has the advantages of:
[0038] High prediction accuracy: through the time series model, the periodicity of project data during project operation can be combined for prediction, and the prediction accuracy is higher than that of traditional statistical methods;
[0039] Strong adaptability: the method can adapt to the business characteristics of different types of projects such as software development projects and engineering construction projects, and meet the prediction needs of the whole project life cycle. BRIEF DESCRIPTION OF DRAWINGS
[0040] Figure 1 is a method flowchart of the present application. DETAILED DESCRIPTION
[0041] The present application will be further described below in conjunction with the drawings and specific embodiments, so that those skilled in the art can better understand the present application and implement it, but the embodiments are not limiting to the present application.
[0042] Example 1
[0043] The present application provides a kind of enterprise project profit and loss prediction method based on time series model, comprising:
[0044] Step 1: data acquisition: through timing task executor, data interface is used to collect operating data in project management system, and multi-dimensional project data are collected in real time according to preset time interval, and multi-dimensional project data include personnel input data, project node data, repayment condition data and cost data.
[0045] Among them, the collected multi-dimensional project data includes:
[0046] Personnel input data: the number of personnel input in each project phase, post type, labor unit price, cumulative working hours,
[0047] Project node data: project start time, end time, progress completion rate, node delay time,
[0048] Repayment condition data: amount of repayment, repayment time, contract repayment node, estimated repayment amount and time,
[0049] Cost data: material procurement cost, equipment rental, procurement cost, operation and maintenance service cost.
[0050] Step 2: data preprocessing: the collected data is preprocessed by a timing task executor to form a data set and ensure data quality. The preprocessing includes:
[0051] Data cleaning: abnormal values in the data are identified and corrected by an abnormal value detection algorithm, and duplicate records are deleted,
[0052] Data completion: non-key data missing is completed using default values, and key data is manually recorded with an alarm prompt,
[0053] Time formatting: node time and repayment time are formatted into standard date format, and project phase duration and repayment interval features are extracted.
[0054] Step 3: total cost calculation: the unit labor cost in the preprocessed project personnel labor hour data is converted into labor cost by a timing task executor, the unit labor cost is the historical average labor hour unit price of the enterprise, and the labor cost is combined with the direct cost and indirect cost to obtain the total cost.
[0055] Step 4: model training and updating: the personnel input, project node, repayment condition and total cost data in the data set are input to fit the prediction model, and the deployed prediction model is updated.
[0056] The data set is divided into training set and validation set by a timing task executor in the ratio of 7:3, and the personnel input, project node, repayment condition and total cost data in the training set are input to fit the prediction model.
[0057] Step 5: model prediction: the project number and related personnel input, project node, repayment condition and total cost data in a certain time period are input to the model to predict the profit and loss data through API interface.
[0058] Embodiment 2
[0059] The application also provides an enterprise project profit and loss prediction system based on a time series model, which comprises a collection module, a preprocessing module, a calculation module, a training and updating module and a prediction module,
[0060] The collection module collects data: the timing task executor collects operating data in the project management system through a data interface, and real-time multi-dimensional project data is collected at a preset time interval, including personnel input data, project node data, repayment condition data and cost data,
[0061] The preprocessing module preprocesses the data: the collected data is preprocessed by a timing task executor to form a data set and ensure data quality,
[0062] The accounting module accounts for the total cost: the unit labor cost in the preprocessed project personnel labor hour data is converted into labor cost by the timing task executor, the unit labor cost is the historical average labor hour unit price of the enterprise, the labor cost is combined with the direct cost and the indirect cost to obtain the total cost,
[0063] The training update module trains and updates the model: the personnel input, project node, repayment situation and total cost data in the data set are input by the timing task executor, the prediction model is fitted, and the deployed prediction model is updated,
[0064] The prediction module performs model prediction: the project number and related personnel input, project node, repayment situation and total cost data in a time period are input into the model through API interface to predict profit and loss data.
[0065] The information interaction and execution process between the modules in the above system are based on the same concept as the method embodiments of the present application, and the specific content can be referred to the description in the method embodiments of the present application, which will not be repeated here.
[0066] Similarly, the advantages of the system of the present application are: high prediction accuracy: through the time series model, the periodicity of project data in the project operation process can be combined for prediction, and the prediction accuracy is higher than that of traditional statistical methods;
[0067] Strong adaptability: the method can adapt to the business characteristics of different types of projects such as software development projects and engineering construction projects, and meet the prediction needs of the whole life cycle of the project.
[0068] It should be noted that not all steps and modules in the above processes and system structures are necessary, and some steps or modules can be ignored according to actual needs. The execution order of each step is not fixed and can be adjusted as needed. The system structure described in the above embodiments can be a physical structure or a logical structure, that is, some modules can be implemented by the same physical entity, or some modules can be implemented by multiple physical entities, or they can be implemented by some components in multiple independent devices.
[0069] The above-described embodiments are only preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Equivalent substitutions or transformations made by those skilled in the art based on the present application are within the protection scope of the present application. The protection scope of the present application is subject to the claims.
Claims
1. A time series model-based enterprise project profit and loss prediction method, characterized by Comprise: Step 1: data collection: through the data interface by the timing task executor to collect the project management system within the business data, real-time acquisition of multi-dimensional project data in preset time interval, multi-dimensional project data including personnel input data, project node data, repayment data and cost data, Step 2: data preprocessing: through the timing task executor to preprocess the collected data, form a data set, guarantee data quality, Step 3: accounting total cost: through the timing task executor to convert the unit time cost in the preprocessed project personnel working hours data into working hours cost, the unit time cost is the historical average working hours unit price of the enterprise, and the working hours cost is combined with the direct cost and indirect cost to obtain the total cost, Step 4: model training and updating: through the timing task executor, the personnel input, project node, repayment and total cost data in the data set are used as input to fit the prediction model, and the deployed prediction model is updated, Step 5: model prediction: through the API interface call, the project number and related personnel input, project node, repayment and total cost data in the time period are input into the model to predict the profit and loss data.
2. The enterprise project profit and loss prediction method based on a time series model according to claim 1, characterized in that The multi-dimensional project data collected in step 1 includes: Personnel input data: the number of personnel input in each project phase, post type, labor unit price, cumulative working hours, Project node data: project start time, end time, progress completion rate, node delay time, Repayment data: the amount of repayment, repayment time, contract repayment node, estimated repayment amount and time, Cost data: material procurement cost, equipment rental, procurement cost, operation and maintenance service cost.
3. The enterprise project profit and loss prediction method based on time series model according to claim 1, characterized in that In step 2, preprocessing includes: Data cleaning: identify and correct outliers in the data through outlier detection algorithm, delete duplicate records, Data completion: use default value to complete the missing non-key data, and manually record the key data, Time formatting: format the node time and repayment time into standard date format, extract the project phase length and repayment interval features.
4. The enterprise project profit and loss prediction method based on time series model according to claim 1, characterized in that In step 4, the timing task executor divides the data set into training set and validation set in the ratio of 7:3, and then uses the personnel input, project node, repayment and total cost data in the training set as input to fit the prediction model.
5. A time series model-based enterprise project profit and loss prediction system, characterized by Comprise collection module, preprocessing module, accounting module, training and updating module and prediction module, The collection module collects data: through the data interface by the timing task executor to collect the project management system within the business data, real-time acquisition of multi-dimensional project data in preset time interval, multi-dimensional project data including personnel input data, project node data, repayment data and cost data, The preprocessing module preprocesses data: through the timing task executor to preprocess the collected data, form a data set, guarantee data quality, The accounting module accounts for the total cost: through the timing task executor to convert the unit time cost in the preprocessed project personnel working hours data into working hours cost, the unit time cost is the historical average working hours unit price of the enterprise, and the working hours cost is combined with the direct cost and indirect cost to obtain the total cost, Model training and updating: The data set of personnel input, project node, repayment situation and total cost data are input through the timing task executor to fit the prediction model and update the deployed prediction model, Model prediction: The model is called through the API interface to predict the profit and loss data of the project number and related personnel input, project node, repayment situation and total cost data in the time period.
6. The time series model based enterprise project profit and loss prediction system according to claim 5, characterized in that The multi-dimensional project data collected by the acquisition module includes: Personnel input data: number of personnel input in each project stage, post type, labor unit price, cumulative working hours, Project node data: project start time, end time, progress completion rate, node delay time, Repayment situation data: amount of repayment, repayment time, contract repayment node, estimated repayment amount and time, Cost data: material procurement cost, equipment rental, procurement cost, operation and maintenance service cost.
7. The time series model based enterprise project profit and loss prediction system of claim 5, wherein The preprocessing module performs preprocessing, including: Data cleaning: identify and correct outliers in the data using outlier detection algorithms, delete duplicate records, Data completion: use default values to complete missing non-critical data, and manually record critical data with alarm prompts, Time formatting: format node time and repayment time into standard date format, extract project stage duration and repayment interval features.
8. The time series model based enterprise project profit and loss prediction system of claim 5, wherein The training update module divides the data set into training set and validation set in the ratio of 7:3 through the timing task executor, and then inputs the personnel input, project node, repayment situation and total cost data in the training set to fit the prediction model.