A method and system for predicting the delay of a scientific research project

Abstract

The application relates to the technical field of project early warning, and discloses a scientific research project delay prediction method and a prediction system. The method comprises the following steps: obtaining natural month progress data and technical maturity of a completed project, obtaining statistical month progress data based on the natural month progress data, dividing the statistical month progress data into several parts according to progress prediction requirements, and constructing a project progress prediction model of different progress links based on the statistical month progress data and the technical maturity; obtaining statistical month progress data and technical maturity of a to-be-predicted project, inputting the project progress prediction model of a corresponding progress link based on the progress link of the to-be-predicted project, and obtaining an early warning result of the to-be-predicted project; obtaining early warning information and index information of the to-be-predicted project, and judging early warning to different relevant personnel based on the early warning result, the index information, the technical maturity and the early warning information in combination with a preset early warning threshold. The method solves the problem that the existing project management early warning method cannot effectively predict project delay.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of project early warning, and in particular to a scientific research project delay prediction method and system. BACKGROUND

[0002] In project management early warning, cost performance index (CPI) and schedule performance index (SPI) are usually used to calculate and evaluate the cost and schedule of the project. Taking the schedule performance index as an example, it reflects the ratio of the actual completed work to the planned work. SPI greater than 1 indicates that the project schedule is ahead of schedule, and less than 1 indicates that the schedule is behind schedule, so it is used to monitor the actual performance of the project and provide decision support for project adjustment.

[0003] SPI and CPI are based on linear growth ratio, which is usually true in engineering projects because engineering projects are usually highly mature projects. However, in technology enterprises and universities, projects are mostly scientific research projects, involving scientific research time, new equipment development cost, repeated experiments and other uncertainties, so the relationship between cost and schedule is not a linear ratio growth relationship, and therefore it is not sufficient as a direct basis for management decisions.

[0004] The current scientific research project management information system database table provides an index of "current completion progress" or similar name, which means the total progress percentage completed in the current natural month, which is used by project managers and project leaders to control the overall progress of the project. However, this value is a cumulative value, and the task planning for this month is based on the previous task, which cannot reflect the full life cycle execution of the project, nor can it control the progress of the full life cycle of the project, i.e. it cannot solve the problem of delay. SUMMARY

[0005] The present application provides a scientific research project delay prediction method and system to solve the problem of project delay caused by the inability of existing project management early warning methods to effectively predict projects.

[0006] In order to achieve the above purpose, the present application realizes the following technical solutions:

[0007] In the first aspect, the present application provides a scientific research project delay prediction method, comprising the following steps:

[0008] Step 1: Obtain the natural month progress data and technical maturity of the completed project, obtain statistical month progress data based on the natural month progress data, construct a plurality of statistical month progress data of different progress links according to the progress prediction requirements, and construct a plurality of project progress prediction models corresponding to the progress links based on the plurality of statistical month progress data and the technical maturity;

[0009] Step 2: Obtain the monthly progress data and technology maturity of the project to be predicted. Based on the progress stage of the project according to the monthly progress data, input the monthly progress data and technology maturity into the project progress prediction model of the corresponding progress stage to obtain the early warning result of the project to be predicted.

[0010] Step 3: Obtain early warning information and index information for the project to be predicted. Based on the early warning results, index information, technology maturity, and early warning information combined with preset early warning thresholds, determine to issue early warnings to different relevant personnel.

[0011] Furthermore, the monthly progress data is obtained by subtracting the current completion progress of each natural month from that of the corresponding previous natural month in the total execution time of the project.

[0012] The process of obtaining statistical monthly progress data based on natural month progress data includes: dividing the total execution time of the project into 12 statistical months, and converting the corresponding natural month progress data into statistical monthly progress data according to the progress conversion formula;

[0013] The progress conversion formula is expressed by the following formula:

[0014] ;

[0015] ;

[0016] ;

[0017] in, Represents each calendar month The corresponding monthly statistical index, This indicates the total number of calendar months the project was executed. Indicates the first i The progress data for each statistical month. Indicates the first Progress data for each natural month. Indicates the natural month Corresponding statistical monthly index and i When they are equal, the value is 1.

[0018] Furthermore, the progress prediction requirement involves setting nodes for different progress stages in 12 statistical months, and selecting nodes based on the progress stage of the project to be predicted.

[0019] The monthly progress data for each of the different progress stages all include monthly progress data for at least one statistical month.

[0020] If the progress prediction requirement is to set nodes at 25% and 50% in 12 statistical months, namely, to predict three progress links (stages) of 0% to 25%, 25% to 50%, and 50% to 100%, 12 statistical months are divided into three statistical month data, the first statistical month data includes statistical month data of 1, 2, and 3 statistical months, the second statistical month data includes statistical month data of 1 to 6 statistical months, and the third statistical month data includes statistical month data of 1 to 12 statistical months. Based on this, the project progress can be more flexibly monitored.

[0021] Further, the project progress prediction model is constructed by the following steps:

[0022] Step 101: Select one statistical month progress data from several statistical month progress data, obtain the statistical month cumulative data corresponding to the progress prediction requirement by combining node division, obtain the average progress of the statistical month according to the statistical month cumulative data, and calculate the standard deviation based on the average progress;

[0023] The statistical month cumulative data is the cumulative value of the statistical month progress data between nodes in one statistical month progress data;

[0024] If the progress prediction requirement is to set nodes at 25% and 50% in 12 statistical months, the third statistical month data, namely, the statistical month data of 1 to 12 statistical months, is selected, and the corresponding statistical month cumulative data is three, the first statistical month cumulative data of the statistical month data of 1 to 3 statistical months, the second statistical month cumulative data of the statistical month data of 4 to 6 statistical months, and the third statistical month cumulative data of the statistical month data of 7 to 12 statistical months.

[0025] Step 102: Construct a target variable based on the technology maturity, statistical month cumulative data, average progress, standard deviation, and statistical month progress data;

[0026] Step 103: Construct a project progress prediction model of the corresponding progress link of the one statistical month progress data based on the target variable.

[0027] Further, the target variable is constructed based on a characteristic variable and a sigmoid function;

[0028] The characteristic variable is constructed by the technology maturity, statistical month cumulative data, average progress, standard deviation, and statistical month progress data;

[0029] The characteristic variable is represented by the following formula:

[0030] ;

[0031] wherein, indicates the characteristic variable corresponding to a statistical monthly progress; indicates the total number of statistical monthly progress data in a statistical monthly progress data based on progress demand division indicates the statistical monthly progress data of the th statistical month; indicates the cumulative data of the th statistical month; indicates the average progress; indicates the standard deviation; indicates the technology maturity;

[0032] The target variable is represented by the following formula:

[0033]

[0034] Wherein, indicates the target variable; all indicate training coefficients, indicates a sigmoid function.

[0035] Further, step 103 comprises: constructing a model data set based on the data structure of the target variable, dividing the model data set into a training subset and a test subset according to a predetermined ratio, training the model using the training subset in a loop, testing the model using the test subset, and selecting the model with the lowest error rate as the project progress prediction model after repeating a predetermined number of times.

[0036] Further, the preset warning threshold includes a warning threshold and an index threshold;

[0037] The warning based on the warning result, the index information, the technology maturity, and the warning information in combination with the preset warning threshold includes: if the warning result exceeds the warning threshold, warning different relevant personnel based on the comparison result of the index information, the technology maturity, and the index threshold in combination with the warning information.

[0038] Further, the warning information includes a project leader, a direct supervisor, a financial manager, a project name, and a contact method, the index information includes a cost performance index and a progress performance index, and the index threshold includes a first preset threshold and a second preset threshold;

[0039] The warning based on the comparison result of the index information and the index threshold in combination with the warning information includes:

[0040] If the cost performance index and the progress performance index do not exceed the first preset threshold, the warning result is sent to the project leader;

[0041] ​​​​If the cost performance index and the progress performance index exceed the first preset threshold value and the technology maturity exceeds the second preset threshold value, the early warning result is sent to the project manager and the financial manager;

[0042] If the cost performance index and the progress performance index exceed the first preset threshold value and the technology maturity does not exceed the second preset threshold value, the early warning result is sent to the project manager, the financial manager and the direct supervisor.

[0043] In a second aspect, the present application also provides a scientific research project delay prediction system, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to realize the steps of the method of any one of the above aspects.

[0044] Advantages:

[0045] The scientific research project delay prediction method and system provided by the present application comprehensively consider the multivariate prediction method of the whole life cycle of the project, can realize the probability prediction of the possibility of the delay of the scientific research project, can realize the linkage alarm according to the strategy, and provides an effective monitoring and prediction tool for project management, so that the project manager or the management personnel can take measures in advance, re-plan the progress and resources, reduce the possibility of the delay of the project and improve the completion quality of the project. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 A flowchart of the scientific research project delay prediction method of the present application;

[0047] Figure 2 An information data transmission schematic diagram of the scientific research project delay prediction method of the present application. DETAILED DESCRIPTION

[0048] The technical solutions of the present application will be described clearly and completely below. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0049] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms "an" or "a" and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms "connected" or "linked" and similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up," "down," "left," "right," etc., are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship also changes accordingly.

[0050] Please see Figure 1 This application provides a method for predicting the delay of scientific research projects, including the following steps:

[0051] Step 1: Obtain the natural month progress data and technology maturity of completed projects. Based on the natural month progress data, obtain the statistical month progress data. Construct several sets of statistical month progress data for different progress stages according to the progress prediction requirements. Based on the several sets of statistical month progress data and technology maturity, construct several project progress prediction models corresponding to the progress stages.

[0052] For details, please see Figure 2 The system uses SQL queries to retrieve monthly progress data and technology maturity of completed projects from the scientific research project management information system. Monthly progress data is obtained by subtracting the current progress of each month from the current progress of the preceding month within the total project execution time, and then querying a historical table of SQL query relationships. For a two-year project in this embodiment, progress data for 24 months can be obtained.

[0053] After obtaining the progress data for each natural month, the total execution time of the project is divided into 12 statistical months, and the corresponding natural month progress data is converted into statistical month progress data according to the progress conversion formula.

[0054] The schedule transition formula is expressed as follows:

[0055] ;

[0056] ;

[0057] ;

[0058] in, Represents each calendar month The corresponding monthly statistical index, denotes the total number of natural months of project execution, denotes the statistical month progress data of the i statistical month, denotes the natural month progress data of the natural month, denotes 1 when the corresponding statistical month index is equal to . i

[0059] Based on the above, statistical month progress data can be obtained, wherein each statistical month progress data contains three natural month progress data;

[0060] The progress prediction requirement is to set nodes for different progress links in 12 statistical months, and to select nodes based on the progress link of the project to be predicted;

[0061] Each statistical month data contains statistical month data of at least one statistical month.

[0062] In this embodiment, the accuracy prediction requirement of the project to be detected is 3 nodes, which are 25%, 50% and 75% respectively. The statistical month progress data corresponding to the completed project is divided into 4 statistical month data. The first one is the statistical month progress data of 1 to 3 statistical months, the second one is the statistical month progress data of 1 to 6 statistical months, the third one is the statistical month progress data of 1 to 9 statistical months, and the fourth one is the statistical month progress data of 1 to 12 statistical months.

[0063] For the construction of the project progress prediction model, the following steps are included:

[0064] Step 101: Select one statistical month progress data from several statistical month progress data, obtain the statistical month cumulative data corresponding to the progress prediction requirement by combining node division, obtain the average progress of the statistical month according to the statistical month cumulative data, and calculate the standard deviation based on the average progress;

[0065] The statistical month cumulative data is the cumulative value of the statistical month progress data between nodes in one statistical month progress data;

[0066] In this embodiment, four statistical month cumulative data are obtained. The first one corresponds to the first one, which is the statistical month progress data of 1 to 3 statistical months, the second one corresponds to the statistical month progress data of 4 to 6 statistical months, the third one corresponds to the statistical month progress data of 7 to 9 statistical months, and the fourth one corresponds to the statistical month progress data of 10 to 12 statistical months.

[0067] Step 102: Construct the target variable according to the technology maturity, the statistical month cumulative data, the average progress, the standard deviation and the statistical month progress data;

[0068] ​The target variable is constructed based on the feature variables combined with the sigmoid function;

[0069] The feature variables were constructed using technology maturity, cumulative monthly statistical data, average progress, standard deviation, and monthly statistical progress data.

[0070] The characteristic variable is represented by the following formula:

[0071] ;

[0072] in, These represent the characteristic variables corresponding to several monthly statistical progress reports; This indicates that a statistical monthly progress data set based on schedule requirements contains a total of [number] items. Monthly progress data, ; Indicates the first Cumulative data for each statistical month; Indicates average progress; Indicates standard deviation; Indicates the level of technological maturity;

[0073] The target variable is represented by the following formula:

[0074] ;

[0075] in, Represent the target variable; to All represent training coefficients. This represents the sigmoid function.

[0076] Step 103: Construct a project progress prediction model based on the target variable and the corresponding progress stages of the monthly progress data.

[0077] Specifically, based on the target variable The model dataset is constructed using a data structure. Based on the model dataset and a predetermined ratio, it is divided into a training subset and a test subset. The training subset is used to train the model, and the test subset is used to test the model. After repeating this process a predetermined number of times, the model with the lowest error rate is selected as the project progress prediction model.

[0078] Based on the above steps and the requirements for progress forecasting, four project progress forecasting models are obtained, which are used to predict the progress of the project to be forecasted at 0% to 25%, 25% to 50%, 25% to 75%, and 75% to 100%, respectively. It should be noted that when constructing project progress forecasting models for different progress stages, the target variable should also be changed according to the statistical monthly data of the completed projects selected. The above steps can be repeated to obtain the target variable, which will not be elaborated here.

[0079] Step 2: Obtain the statistical monthly progress data and the technical maturity of the project to be predicted, input the statistical monthly progress data and the technical maturity into the project progress prediction model corresponding to the progress stage of the statistical monthly progress data of the project to be predicted, and obtain the early warning result of the project to be predicted;

[0080] Obtain the statistical monthly progress data and the technical maturity of the project to be predicted, which is consistent with the statistical monthly progress data of the completed project. Based on the natural monthly progress data of the project to be predicted, based on the progress stage of the project to be predicted, select the corresponding project progress prediction model for prediction, and obtain the early warning result of the project to be predicted.

[0081] For example, the execution period of a certain project to be predicted is 2 years, and the current execution period is 17 months, the current natural month is between 50%-75%, and the characteristic variable and the target variable can be further represented as:

[0082] ;

[0083] ;

[0084] The prediction model for predicting the project should be a model trained by a data set based on the target variable ;

[0085] Step 3: Obtain the early warning information and index information of the project to be predicted, and judge the early warning to different relevant personnel based on the early warning result, index information, technical maturity, and early warning information combined with the preset early warning threshold.

[0086] Obtain the early warning information, including the project manager, direct supervisor, financial manager, project name, and contact information. The index information includes the cost performance index and the progress performance index. The preset early warning threshold includes the early warning threshold and the index threshold, and the index threshold includes the first preset threshold and the second preset threshold.

[0087] If the early warning result does not exceed the early warning threshold, no early warning is given, and if the early warning result exceeds the early warning threshold, the early warning is given to different relevant personnel based on the comparison result of the index information and the index threshold combined with the early warning information.

[0088] The specific circumstances of exceeding the early warning threshold are as follows:

[0089] If the cost performance index and the progress performance index do not exceed the first preset threshold, the early warning result is sent to the project manager;

[0090] If the cost performance index and the progress performance index exceed the first preset threshold and the technical maturity exceeds the second preset threshold, the early warning result is sent to the project manager and the financial manager;

[0091] If the cost performance index and the progress performance index exceed the first preset threshold value and the technology maturity does not exceed the second preset threshold value, an early warning result is sent to the project manager, the financial manager and the direct superior.

[0092] The model trained using the dataset constructed by the target variable The project delay probability prediction results of the model trained using the dataset constructed by the target variable

[0093] Table 1: Project progress prediction model project delay probability prediction results.

[0094]

[0095] In Table 1, taking project P6 as an example, if the traditional “current completion progress” index is used for judgment, the completion progress of P6 is greater than 75%, the compound expectation is good, and the cumulative execution is good, but the future situation cannot be reflected; if SPI>1 is used for judgment, the progress of the project is currently ahead of schedule, the execution is good, and the future situation cannot be reflected; and according to the prediction output of the model, the project delay probability is 0.91, because the model captures the key information that the project execution is faster at the beginning and slower at the later stage, indicating that the progress is difficult to advance, and the technology maturity of the project is relatively low, and there is uncontrollable risk at the later stage, so the output delay probability is high. Combined with the setting of the early warning threshold, the first preset threshold value and the second preset threshold value, an early warning result can be sent to the project manager, or the project manager and the financial manager, or the project manager, the financial manager and the direct superior, so as to provide a basis for management decision.

[0096] The above describes the preferred embodiments of the present application in detail. It should be understood that those skilled in the art can make many modifications and changes without creative labor according to the concept of the present application. Therefore, any technical solution obtained by logical analysis, reasoning or limited experiment on the basis of the prior art according to the concept of the present application should be within the protection scope determined by the claims.

Claims

1. A method for predicting a delay of a scientific research project, characterized by, The method comprises the following steps: Step 1: obtaining natural month progress data and technical maturity of a completed project, obtaining statistical month progress data based on the natural month progress data, constructing a plurality of statistical month progress data of different progress links according to progress prediction requirements, and constructing a plurality of project progress prediction models of corresponding progress links based on the plurality of statistical month progress data and the technical maturity; The natural month progress data is obtained by subtracting the current completion progress of each natural month from the corresponding previous natural month in the total execution time of the project; The statistical month progress data is obtained based on the natural month progress data, which comprises the following steps: dividing the total execution time of the project into 12 statistical months, and converting the corresponding natural month progress data into statistical month progress data according to a progress conversion formula; The progress conversion formula is represented by the following formula: ； ； ； wherein, represents the number of natural months, corresponding to the statistical month, represents the total number of natural months of project execution, represents the statistical month progress data of the i th statistical month, represents the natural month progress data of the th natural month, represents 1 when the corresponding statistical month index is equal to ; and i 1. The progress prediction requirement is to set nodes for different progress links in the 12 statistical months, and to select nodes based on the progress link of the project to be predicted; The statistical month progress data of the different progress links each comprises statistical month progress data of at least one statistical month; The project progress prediction model is constructed by the following steps: Step 101: selecting one statistical month progress data from the plurality of statistical month progress data, obtaining statistical month cumulative data of the corresponding progress prediction requirement in combination with node division, obtaining the average progress of the statistical month according to the statistical month cumulative data, and calculating the standard deviation based on the average progress; The statistical month cumulative data is the cumulative value of the statistical month progress data between nodes in one statistical month progress data; Step 102: constructing a target variable according to the technical maturity, the statistical month cumulative data, the average progress, the standard deviation and the statistical month progress data; Step 103: constructing a project progress prediction model of the corresponding progress link of the one statistical month progress data based on the target variable; Step 2: obtaining statistical month progress data and technical maturity of a project to be predicted, inputting the statistical month progress data and the technical maturity into the project progress prediction model of the corresponding progress link based on the progress link of the statistical month progress data of the project to be predicted, and obtaining an early warning result of the project to be predicted; Step 3: obtaining early warning information and index information of the project to be predicted, and judging early warning to different relevant personnel based on the early warning result, the index information, the technical maturity and the early warning information in combination with a preset early warning threshold.

2. The method of claim 1, wherein, The target variable is constructed based on a characteristic variable in combination with a sigmoid function; The characteristic variable is constructed by the technical maturity, the statistical month cumulative data, the average progress, the standard deviation and the statistical month progress data; The characteristic variable is represented by the following formula: ； in, These represent the characteristic variables corresponding to several monthly statistical progress reports; This indicates that a statistical monthly progress data set based on schedule requirements contains a total of [number] items. Monthly progress data, That is to say, the first Monthly progress data for each statistical month; ; Indicates the first Cumulative data for each statistical month; Indicates average progress; Indicates standard deviation; Indicates the level of technological maturity; The target variable is represented by the following formula: ； wherein, denotes the target variable; denotes the target variable; denote the training coefficients, denotes the sigmoid function.

3. The method of claim 1, wherein, Step 103 comprises the following steps: constructing a model data set based on the data structure of the target variable, dividing the model data set into a training subset and a test subset according to a predetermined ratio, training the model by using the training subset and testing the model by using the test subset in a loop, and selecting the model with the lowest error rate as the project progress prediction model after repeating a predetermined number of times.

4. The method of claim 1, wherein, The preset early warning threshold comprises an early warning threshold and an index threshold. The early warning to different relevant personnel based on the early warning result, the index information, the technology maturity and the early warning information combined with the preset early warning threshold value comprises: if the early warning result exceeds the early warning threshold value, early warning to different relevant personnel based on the comparison result of the index information and the index threshold value combined with the early warning information.

5. The method of claim 4, wherein, The early warning information comprises a project leader, a direct supervisor, a financial manager, a project name and a contact method, the index information comprises a cost performance index and a progress performance index, and the index threshold value comprises a first preset threshold value and a second preset threshold value. The early warning to different relevant personnel based on the comparison result of the index information and the index threshold value combined with the early warning information comprises: if the cost performance index and the progress performance index do not exceed the first preset threshold value, sending the early warning result to the project leader; if the cost performance index and the progress performance index exceed the first preset threshold value and the technology maturity exceeds the second preset threshold value, sending the early warning result to the project leader and the financial manager; if the cost performance index and the progress performance index exceed the first preset threshold value and the technology maturity does not exceed the second preset threshold value, sending the early warning result to the project leader, the financial manager and the direct supervisor.

6. A system for predicting delays in scientific research projects, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor implements the steps of the method of any one of claims 1 to 5 when executing the computer program.

Images