A project management optimization method and system

Abstract

The present invention relates to the field of management, and in particular to an optimization method and system for project management. An optimization method for project management comprises the following steps: S1: constructing a project risk model using KD-Tree according to relevant historical data of the project, and using KNN algorithm and real-time project-related data to input into the project risk model to obtain a first project risk level set; S2: obtaining the project risk level after processing the first project risk level set; S3: obtaining relevant data of the project when a project changes, and inputting the relevant data of the project into the project risk model to obtain the project risk level, analyzing the project risk level, and obtaining a first analysis result or a second analysis result. The present invention constructs a project risk model using KD-Tree, obtains the project risk level using KNN and LRU algorithms, and adds trained personnel to the project to speed up the progress of the task.

Description

Technical Field

[0001] The present invention relates to the technical field of project management, and in particular to an optimization method and system for project management. Background Art

[0002] Project schedule management is a key component of project management. It requires that project tasks be completed efficiently and orderly within the scheduled time to ensure that the project can be delivered on time. The assessment of project schedule risk in project schedule management can assess potential impacts and formulate response strategies. The existing project schedule risk assessment relies more on digital equipment for remote communication and collaboration when managing company projects, and the management of projects relies more on the average progress of projects over a period of time. However, this method ignores the particularity or importance of individual tasks. At the same time, when the company's task progress does not meet the standard, it usually hires external personnel to assist in speeding up the task progress, which increases the company's expenses. Summary of the invention

[0003] In order to overcome the shortcomings of inaccurate project management, the present invention provides an optimization method and system for project management.

[0004] The technical solution is: an optimization method for project management, including the following steps:

[0005] S1: constructing a project risk model using KD-Tree based on relevant historical data of the project, obtaining relevant data of the real-time project, and inputting the relevant data of the real-time project into the project risk model and then using KNN algorithm to obtain a first project risk level set;

[0006] S2: Processing the first project risk level set to obtain a project risk level;

[0007] S3: when a project changes, obtaining relevant data of the changed project, and re-obtaining the project risk level based on the relevant data of the changed project, analyzing the project risk level, and obtaining a first analysis result or a second analysis result;

[0008] S4: Obtain the training cycle and progress of the training personnel, and adjust and plan the personnel of the project according to the second analysis result.

[0009] Preferably, the method of using KD-Tree to construct a project risk model based on relevant historical data of the project, obtaining relevant data of the real-time project, and inputting the relevant data of the real-time project into the project risk model before using the KNN algorithm to obtain the first project risk level set includes: collecting relevant historical data about the project, and performing data cleaning operations on the data to supplement missing values, remove outliers, and remove duplicate values, and after normalizing the data, constructing the project risk model according to the KD-Tree construction algorithm.

[0010] Preferably, the project risk model is constructed using KD-Tree based on relevant historical data of the project, relevant data of the real-time project is obtained, and the relevant data of the real-time project is input into the project risk model, and then the KNN algorithm is used to obtain the first project risk degree set, including: constructing a project risk model based on project-related data, obtaining N real-time project progress within a preset time period and inputting them into the project risk model respectively, obtaining N project risk degree frameworks, each of which contains K project risk degrees within a preset range, and then taking the intersection of the project risk degrees in the N project risk degree frameworks as the first project risk degree set.

[0011] Preferably, the project risk level is obtained after processing the first project risk level set, including: obtaining the project risk level with the largest number in the first project risk level set as the project risk level within this time period; if there is no single project risk level with the largest number in the first project risk level set, then sorting the elements in the N project wind direction level framework to obtain a first sorting sequence, using the LRU algorithm to obtain a second sorting sequence for the first sorting sequence, and obtaining the project risk level according to the project risk calculation formula.

[0012] Preferably, if there is no single project risk level with the largest number in the first project risk level concentration, then the elements in the N project risk level frameworks are sorted to obtain a first sorting sequence, and the first sorting sequence is obtained by using the LRU algorithm to obtain a second sorting sequence, and the project risk level is obtained according to the project risk calculation formula, including: comparing the elements in the N project risk level frameworks to obtain repeated elements in each framework, and sorting the repeated elements in advance according to the inherent size relationship to obtain N first sorting sequences, and according to the different times of inputting N real-time project progress in a preset time period into the project risk model, the N first sorting sequences are sorted by using the LRU algorithm to obtain a first preset number of second sorting sequences, and the project risk level is obtained by using the project risk calculation formula on the second sorting sequence, wherein the project risk calculation formula is:

[0013] Wherein K, B, and Q are confidence factors, and the total number of confidence factors is a first preset number, S(t) is the project risk level, and X, Y, and Z are the project risk levels in the second sorting sequence, respectively.

[0014] Preferably, the method of obtaining relevant data of the changed project when a project changes, and re-obtaining the project risk level based on the relevant data of the changed project, analyzing the project risk level, and obtaining a first analysis result or a second analysis result includes: obtaining relevant data of the changed project after modification at the current time, and re-obtaining the project risk level based on the relevant data of the changed project, and analyzing according to the project risk level, if it is the first analysis result, continuing to analyze and detect; if it is the second analysis result, obtaining the project predicted completion time according to the project progress prediction model, and comparing the project predicted completion time with the preset range of the deadline and then making adjustments.

[0015] Preferably, the training cycle and progress of the trainees are obtained, and the personnel are adjusted and planned for the project according to the second analysis result, including: if the predicted completion time of the project is less than the preset range of the deadline, the training completion time of each trainee is obtained by using the trainee progress prediction model, wherein the historical trainee progress data is trained using the Transformer model to obtain the trainee progress prediction model, and the corresponding strategy is adjusted according to the training completion time of each trainee; if the predicted completion time of the project is greater than or equal to the preset range of the deadline, the corresponding progress strategy is adjusted.

[0016] Preferably, the Transformer model is used to train the historical training personnel progress data to obtain a training personnel progress prediction model, and the corresponding strategy is adjusted according to the training completion time of each training personnel, including: after the training is completed, the personnel who have the ability to complete the training within the project completion time will be used to work together with the project personnel using a rotation scheduling algorithm, and the team collaboration ability of each training personnel and project personnel is obtained according to the project progress growth value when each training personnel and project personnel work together, and the strategy is adjusted according to the team collaboration ability.

[0017] Preferably, if the predicted completion time of the project is greater than or equal to the preset range of the deadline, the corresponding progress strategy is adjusted, including: adding all personnel who can complete the training within the project completion time to the project task after completing the training, and adjusting the date progress correspondence table.

[0018] Preferably, a project management optimization system further includes:

[0019] Data collection and processing module, used to collect historical project data, personnel training data and project progress data, and perform pre-processing such as data cleaning on the data;

[0020] The project risk model building module uses KD-Tree to build a project risk model based on the relevant data of the project;

[0021] The project risk degree acquisition module is used to use the project risk model and obtain the project risk degree by using the KNN algorithm and real-time project-related data;

[0022] The first judgment module is used to determine whether the project schedule can be completed on time when it is modified midway;

[0023] The personnel change module is used to adjust the corresponding strategy according to the training completion time of each training personnel and the predicted completion time of the project.

[0024] The beneficial effects are as follows: the present invention constructs a project risk model by using KD-Tree to detect the project risk level of a period of time. At the same time, in order to reduce the inaccuracy of project risk level detection caused by taking the average value, the KNN, LRU algorithm and project risk calculation formula are used to further detect the project risk, obtain a more accurate project risk level, and manage the project task progress according to the project risk level. The RR algorithm is used to calculate the team collaboration ability between training personnel and project personnel, and when the task progress is slow, the training personnel are added to the project for training while increasing the progress of the project task. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 The present invention is a flowchart of a project management optimization method and system. DETAILED DESCRIPTION

[0026] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0027] Example 1: A project management optimization method, such as Figure 1 As shown, the following steps are included:

[0028] S1: constructing a project risk model using KD-Tree based on relevant historical data of the project, obtaining relevant data of the real-time project, and inputting the relevant data of the real-time project into the project risk model and then using KNN algorithm to obtain a first project risk level set;

[0029] Collect relevant historical data about the project, and perform data cleaning operations to supplement missing values, remove outliers, and remove duplicate values. After normalizing the data, build a project risk model according to the KD-Tree construction algorithm.

[0030] A project risk model is constructed based on project-related data, and N real-time project progress within a preset time period is obtained. The progress is input into the project risk model respectively to obtain N project risk degree frameworks. Each project risk degree framework contains K project risk degrees within a preset range. Then, the intersection of the project risk degrees within the N project risk degree frameworks is taken as the first project risk degree set.

[0031] It should be noted that accurate data can be obtained by cleaning relevant historical data, and relevant data such as project progress, available funds, and risk level can be used to construct a risk level that can reflect the project at this time. Obtain N real-time project progress within a preset time period, where N is greater than or equal to 1, obtain N project wind direction degree frameworks, use KNN to analyze the N real-time project progress to obtain N project wind direction degree frameworks, and the project risk degree framework contains K project risk levels within a preset range, where K is greater than or equal to 1, and use the intersection of the project risk levels within the N project wind direction degree frameworks as the first project risk level set. For example: when N is 2 and K is 3, the framework is: 5, 4, 3 at the first time; the second time is: 4, 3, 2, where the intersection is 4, 3.

[0032] S2: Processing the first project risk level set to obtain the project risk level;

[0033] The risk level of the project with the largest number in the first project risk level concentration is obtained as the project risk level within this time period. If there is no single risk level of the project with the largest number in the first project risk level concentration, the elements in the N project wind direction degree framework are sorted to obtain a first sorting sequence, and the LRU algorithm is used to obtain a second sorting sequence for the first sorting sequence, and the project risk level is obtained according to the project risk calculation formula.

[0034] The elements in the N project wind direction degree frameworks are compared to obtain the repeated elements in each framework, and the repeated elements are sorted in advance according to the inherent size relationship to obtain N first sorting sequences, and according to the different times of inputting the N real-time project progress in the preset time period into the project risk model, the N first sorting sequences are sorted using the LRU algorithm to obtain the first preset number of second sorting sequences, and the project risk degree is obtained by using the project risk calculation formula for the second sorting sequence, where the project risk calculation formula is:

[0035]

[0036] Wherein K, B, and Q are confidence factors, and the total number of confidence factors is a first preset number, S(t) is the project risk level, and X, Y, and Z are the project risk levels in the second sorting sequence, respectively.

[0037] It should be noted that the project risk level with the largest number in the first project risk level concentration is selected as the project risk level within this time period. For example, if the level of the project risk level with the largest number of repetitions in the intersection of the first time and the second time is 4, then 4 is taken as the project risk level. If there is no project risk degree with the largest number of repetitions in the intersection of the first time and the second time, the elements in the N project wind direction degree framework are sorted to obtain a first sorting sequence, and the LRU algorithm is used to obtain the second sorting sequence for the first sorting sequence, and the project risk degree is obtained according to the project risk calculation formula. For example, when N is 3 and K is 3, the framework is the first time: 5, 4, 3; the second time: 4, 6, 2; the third time: 3, 2, 1; the intersection is 4, 3, 2, and the number is 2. There is no project risk degree with the largest number, so 4, 3, 2 in each time is changed in advance to the first time: 5, 4, 3; the second time: 6, 4, 2; the third time: 1, 3, 2, and the first sorting sequence is obtained: 5, 4, 3, 6, 4, 2, 1, 3, 2, where the time is closer to the right; when the first preset number is 3, the second sorting sequence obtained by using the LRU algorithm is: 2, 3, 1, and the final project risk degree is obtained by using the project risk calculation formula for the second sorting sequence.

[0038] S3: when a project changes, obtaining relevant data of the changed project, and re-obtaining the project risk level based on the relevant data of the changed project, analyzing the project risk level, and obtaining a first analysis result or a second analysis result;

[0039] Obtain relevant data of the modified project at the current time, and re-obtain the project risk level based on the relevant data of the modified project, and perform analysis according to the project risk level. If it is the first analysis result, continue to analyze and detect; if it is the second analysis result, obtain the project predicted completion time according to the project progress prediction model, and compare the project predicted completion time with the preset range of the deadline and then make adjustments.

[0040] It should be noted that when the project goals are modified midway or part of the project tasks are terminated, resulting in changes in the project tasks, the project risk model is used to assess the project risks. If the project risk level is small, it is considered to have no impact, and analysis and testing will continue; if the project risk level is large, the project completion time is predicted to determine whether the completion time exceeds the acceptable range of the deadline, and corresponding adjustments will be made.

[0041] S4: Obtain the training cycle and progress of the training personnel, and adjust and plan the personnel of the project according to the second analysis results.

[0042] If the project predicted completion time is less than the preset range of the deadline, the training personnel progress prediction model is used to obtain the training completion time of each training personnel. The Transformer model is used to train the historical training personnel progress data to obtain the training personnel progress prediction model, and the corresponding strategy is adjusted according to the training completion time of each training personnel; if the project predicted completion time is greater than or equal to the preset range of the deadline, the corresponding progress strategy is adjusted.

[0043] After completing the training, the personnel who have the ability to complete the training within the project completion time will use the rotation scheduling algorithm to work together with the project personnel. According to the project progress growth value when each training personnel and project personnel work together, the team collaboration ability of each training personnel and project personnel is obtained, and the strategy is adjusted according to the team collaboration ability.

[0044] All personnel who can complete the training within the project completion time will be added to the project tasks after completing the training, and the date progress correspondence table will be adjusted.

[0045] It should be noted that the Transformer model is used to train a project progress prediction model that can predict the completion time of the project progress, and a training personnel progress prediction model that can predict the completion time of the training personnel. When the project prediction completion time is less than the preset range of the final deadline, the personnel who have the ability to complete the training within the project completion time will use the round-robin scheduling algorithm to work together with the project personnel after completing the training. That is, the training personnel and the project personnel are allowed to work together on the project, and the time for all training personnel to collaborate is the same. The project progress growth value of all training personnel is recorded to obtain the team collaboration ability of the training personnel and the project personnel, and the strategy is adjusted according to the team collaboration ability. If the project prediction completion time is greater than or equal to the preset range of the final deadline, the personnel who can complete the training within the project completion time will be added to the project task after completing the training to speed up the task progress, and the date progress correspondence table is adjusted to make the daily progress increase reasonably to ensure that all project tasks are completed within the specified time limit.

[0046] Embodiment 2: Based on Embodiment 1, a project management optimization system is also included:

[0047] Data collection and processing module, used to collect historical project data, personnel training data and project progress data, and perform pre-processing such as data cleaning on the data;

[0048] The project risk model building module uses KD-Tree to build a project risk model based on the relevant data of the project;

[0049] The project risk degree acquisition module is used to use the project risk model and obtain the project risk degree by using the KNN algorithm and real-time project-related data;

[0050] The first judgment module is used to determine whether the project schedule can be completed on time when it is modified midway;

[0051] The personnel change module is used to adjust the corresponding strategy according to the training completion time of each training personnel and the predicted completion time of the project.

[0052] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A project management optimization method, characterized in that: The following steps are involved: S1: constructing a project risk model using KD-Tree based on relevant historical data of the project, obtaining relevant data of the real-time project, and inputting the relevant data of the real-time project into the project risk model and then using KNN algorithm to obtain a first project risk level set; S2: Processing the first project risk level set to obtain a project risk level; S3: when a project changes, obtaining relevant data of the changed project, and re-obtaining the project risk level based on the relevant data of the changed project, analyzing the project risk level, and obtaining a first analysis result or a second analysis result; S4: obtaining the training cycle and progress of the training personnel, and adjusting and planning the personnel of the project according to the second analysis result; The project risk model is constructed using KD-Tree according to the relevant historical data of the project, the relevant data of the real-time project is obtained, and the relevant data of the real-time project is input into the project risk model, and then the first project risk degree set is obtained using the KNN algorithm, including: constructing the project risk model according to the project-related data, obtaining N real-time project progress within a preset time period and inputting them into the project risk model respectively, obtaining N project risk degree frameworks, each of which contains K project risk degrees within a preset range, and then taking the intersection of the project risk degrees in the N project risk degree frameworks as the first project risk degree set; The obtaining of the project risk degree after processing the first project risk degree set includes: obtaining the risk degree of the project with the largest number in the first project risk degree set as the project risk degree in this time period, if there is no single risk degree of the project with the largest number in the first project risk degree set, then sorting the elements in the N project risk direction degree framework to obtain a first sorting sequence, using the LRU algorithm to obtain a second sorting sequence for the first sorting sequence, and obtaining the project risk degree according to the project risk calculation formula; If there is no single project risk level with the largest number in the first project risk level concentration, the elements in the N project risk level frameworks are sorted to obtain a first sorting sequence, and the first sorting sequence is obtained by using the LRU algorithm to obtain a second sorting sequence, and the project risk level is obtained according to the project risk calculation formula, including: comparing the elements in the N project risk level frameworks to obtain repeated elements in each framework, and sorting the repeated elements in advance according to the inherent size relationship to obtain N first sorting sequences, and according to the different times of inputting N real-time project progress in a preset time period into the project risk model, the N first sorting sequences are sorted by using the LRU algorithm to obtain a first preset number of second sorting sequences, and the project risk level is obtained by using the project risk calculation formula on the second sorting sequence, wherein the project risk calculation formula is: Wherein K, B, and Q are confidence factors, and the total number of confidence factors is a first preset number, S(t) is the project risk level, and X, Y, and Z are the project risk levels in the second sorting sequence, respectively.

2. A project management optimization method according to claim 1, characterized in that: The method uses KD-Tree to build a project risk model based on relevant historical data of the project, obtains relevant data of the real-time project, and inputs the relevant data of the real-time project into the project risk model before using the KNN algorithm to obtain the first project risk level set, including: collecting relevant historical data about the project, and performing data cleaning operations on the data to supplement missing values, remove outliers, and remove duplicate values, and after normalizing the data, build a project risk model according to the KD-Tree construction algorithm.

3. The project management optimization method according to claim 1, characterized in that: The method of obtaining relevant data of the changed project when a project changes, and re-obtaining the project risk level based on the relevant data of the changed project, analyzing the project risk level, and obtaining a first analysis result or a second analysis result includes: obtaining relevant data of the changed project after modification at the current time, and re-obtaining the project risk level based on the relevant data of the changed project, analyzing according to the project risk level, and if it is the first analysis result, continuing to analyze and detect; if it is the second analysis result, obtaining the project predicted completion time according to the project progress prediction model, and comparing the project predicted completion time with the preset range of the deadline and then making adjustments.

4. The project management optimization method according to claim 1, characterized in that: The obtaining of the training cycle and progress of the training personnel, and the adjustment and planning of the project personnel according to the second analysis result, include: if the predicted completion time of the project is less than the preset range of the deadline, using the training personnel progress prediction model to obtain the training completion time of each training personnel, wherein the training personnel progress prediction model is obtained by training the historical training personnel progress data using the Transformer model, and the corresponding strategy is adjusted according to the training completion time of each training personnel; if the predicted completion time of the project is greater than or equal to the preset range of the deadline, the corresponding progress strategy is adjusted.

5. A project management optimization method according to claim 4, characterized in that: The Transformer model is used to train the historical training personnel progress data to obtain a training personnel progress prediction model, and the corresponding strategy is adjusted according to the training completion time of each training personnel, including: after the training of personnel who have the ability to complete the training within the project completion time is completed, a rotation scheduling algorithm is used to work together with the project personnel, and the team collaboration ability of each training personnel and project personnel is obtained according to the project progress growth value when each training personnel and project personnel work together, and the strategy is adjusted according to the team collaboration ability.

6. A project management optimization method according to claim 4, characterized in that: If the predicted completion time of the project is greater than or equal to the preset range of the deadline, the corresponding progress strategy is adjusted, including: adding all personnel who can complete the training within the project completion time to the project tasks after completing the training, and adjusting the date progress correspondence table.

7. A project management optimization system, according to the project management optimization method according to claim 1, characterized in that: Included are: Data collection and processing module, used to collect historical project data, personnel training data and project progress data, and perform pre-processing such as data cleaning on the data; The project risk model building module uses KD-Tree to build a project risk model based on the relevant data of the project; The project risk degree acquisition module is used to use the project risk model and obtain the project risk degree by using the KNN algorithm and real-time project-related data; The first judgment module is used to determine whether the project schedule can be completed on time when it is modified midway; The personnel change module is used to adjust the corresponding strategy according to the training completion time of each training personnel and the predicted completion time of the project.

Images