Multi-project personnel intelligent arrangement and dynamic balance method and device

Abstract

The invention relates to the technical field of human resource and project management crossing, and particularly provides a multi-project personnel intelligent arrangement and dynamic balance method and device, and the method comprises the following steps: S1, carrying out the quantification of employee capability portraits, and generating an operable capability vector; s2, project demand modeling is carried out, and a demand standard of a matching capability vector is generated; s3, calculating an employee-project matching degree, and quantifying a suitability level; s4, constructing a'employee-project-neutral position 'bipartite graph, and solving global optimal distribution by using a minimum cost flow algorithm; s5, performing '1 + N' focusing and teaching copying on the high-ability employees, and establishing an L1-L3 gradient task pool and cross-project scheduling on the low-ability employees; and S6, performing dynamic iterative optimization, real-time monitoring and periodic redisk. Compared with the prior art, the method can achieve a remarkable effect through quantitative matching, global optimization and dynamic balance.

Description

Technical Field

[0001] This invention relates to the field of human resources and project management, specifically providing a method and apparatus for intelligent staffing and dynamic balancing across multiple projects. Background Technology

[0002] In the current multi-project operation model of enterprises, traditional personnel allocation methods have long relied on the subjective experience of project managers or simple responsibility division tables (such as RACI), which have the following unavoidable core pain points, seriously restricting project delivery efficiency and maximizing human resource value:

[0003] Subjective matching of capabilities and needs leads to a high mismatch rate: There is a lack of quantitative assessment of employees' multidimensional capabilities (such as hard and soft technical skills) and project requirements. Personnel are assigned based on vague descriptions such as "experienced" and "good at communication", resulting in a mismatch problem where "high-capable employees undertake basic work (waste of resources)" and "low-capable employees participate in core tasks (project risks)". The project rework rate is generally higher than 15%.

[0004] Highly capable employees are overworked and at high risk of losing their jobs: Due to the scarcity of highly capable employees (such as technical experts and core business backbones), project leaders compete for them in a disorderly manner, resulting in them participating in 4-5 projects at the same time, working more than 50 hours per week, and often exceeding 120% of their workload, which leads to burnout. The turnover rate of core talents is 30%-40% higher than that of ordinary employees.

[0005] Low-skilled employees are idle and wasteful, and their growth is stagnant: Low-skilled employees (such as new employees and those whose skills need to be improved) are unable to take on complex tasks due to insufficient ability, and the company lacks a suitable basic task pool, resulting in their weekly working hours saturation being less than 60%, and the human resource waste rate exceeding 35%. At the same time, due to the lack of practical opportunities, their ability growth cycle is as long as 18 months or more.

[0006] The lack of a dynamic adjustment mechanism results in poor adaptability: Project requirements are often adjusted with market changes (such as early / late schedules or new requirements), but the existing methods lack real-time monitoring and cross-project scheduling capabilities. Resource adjustment response time exceeds 72 hours, leading to delays in high-priority projects or a continuous worsening of uneven workload among personnel.

[0007] In summary, there is an urgent need for a systematic solution that combines "quantitative matching, global optimization, dynamic balancing, and a closed-loop growth model" to address the core pain points of managing personnel across multiple projects. Summary of the Invention

[0008] This invention addresses the shortcomings of the prior art by providing a highly practical method for intelligent arrangement and dynamic balancing of personnel across multiple projects.

[0009] A further technical objective of this invention is to provide a reasonably designed, safe, and applicable intelligent personnel arrangement and dynamic balancing device for multiple projects.

[0010] The technical solution adopted by this invention to solve its technical problem is:

[0011] The intelligent deployment and dynamic balancing method for personnel across multiple projects involves the following steps:

[0012] S1. Quantify employee capability profiles and generate computable capability vectors;

[0013] S2. Project requirements modeling, generating requirement standards that match capability vectors;

[0014] S3. Calculate employee-project matching degree and quantify the suitability level;

[0015] S4. Construct a bipartite graph of "employee-project-vacancy" and use the minimum cost flow algorithm to solve for the global optimal allocation;

[0016] S5. For high-ability employees, adopt a "1+N" focus and mentorship replication approach; for low-ability employees, establish an L1-L3 graded task pool and cross-project scheduling.

[0017] S6. Dynamic iterative optimization, real-time monitoring and periodic review.

[0018] Furthermore, in step S1, the employee's capabilities are first transformed from a "qualitative description" into a "quantitative vector".

[0019] include:

[0020] S1-1, Definition of Competency Dimensions: Select 5 quantifiable dimensions that cover the core value of employees, taking into account both hard skills and soft qualities;

[0021] Dimension 1 is hard technical skills; proficiency in core job skills, quantified through skills assessment tools and historical project results, with an initial score range of 0-5 points.

[0022] Dimension 2 is project experience; the number of years and rounds of participation in similar projects and the role in the projects are quantified through file verification in the enterprise project management system, with an initial score range of 0-5 points;

[0023] Dimension 3 is soft skills; communication and collaboration skills and stress resistance, which are quantified through 360-degree feedback and behavioral event interviews, with an initial score range of 0-5.

[0024] Dimension 4 is certification and performance, including authoritative industry certifications and recent KPI completion rates, which are quantified through the company's HR performance system and certification database, with an initial score range of 0-5 points;

[0025] Dimension 5 is growth potential; including the speed of learning new tools or processes and the implementation rate of innovation proposals, which are quantified through training assessment results and the innovation proposal management system, with an initial score range of 0-5 points.

[0026] S1-2. Using the min-max standardization algorithm, the initial scores of each dimension are mapped to the [0, 1] interval (eliminating differences in units) to generate the ability vector for each employee: E i =(e i1 e i2 e i3 e i4 e i5 ); where e ik ∈[0,1], k=1~5, corresponding to the normalized scores of dimensions 1 to 5 above;

[0027] S1-3: By default, the capability vector is automatically refreshed quarterly based on employee project performance; if an employee participates in a major project milestone, an update is triggered immediately to ensure that the profile is synchronized with the employee's actual capabilities.

[0028] Furthermore, in step S2, project requirements are aligned with employee capabilities, and priority weights are assigned to ensure that high-performing projects receive priority access to quality resources.

[0029] include:

[0030] S2-1. Based on the strategic value and execution complexity of the projects, the projects are divided into 4 categories, which are reviewed and adjusted by the enterprise's PMO over a period of time, and priority weights W are assigned accordingly. j The value range is [0, 1]. The higher the weight, the higher the priority of resource allocation.

[0031] S2-2. Referring to the five capability dimensions in step S1, define the demand intensity for each dimension for each type of project, on a scale of 0-1, where 0 represents no demand and 1 represents the highest demand. Generate the project demand vector: P j =(p j1 p j2 p j3 p j4 p j5 Among them, P jk ∈[0,1], k=1~5, corresponding one-to-one with the dimensions of the employee capability vector;

[0032] S2-3. The enterprise PMO can adjust the project requirement vector in real time through the visual interface of the supporting software system, so that the requirements are synchronized with the actual situation of the project.

[0033] Furthermore, in step S2-1, the four categories of items are as follows:

[0034] Strategic core projects: These projects impact the company's core annual goals and require breakthroughs in technology or business operations. Priority weight: W j =1.0;

[0035] Emergency support projects: Must be delivered in a short period of time, with a clear process but requiring efficient execution, and a priority weight of W. j =0.9;

[0036] Routine execution projects: daily operational tasks, standardized processes, priority and weight W j =0.7;

[0037] Reserve growth projects: long-term foundational tasks, open to trial and error, with priority and weight W. j =0.5.

[0038] Furthermore, in step S3, a weighted cosine similarity algorithm is used to calculate the matching degree between the employee's ability vector and the project requirement vector, and to determine the suitability level.

[0039] include:

[0040] S3-1, Matching degree formula: score(E i P j )=cos(E i P j )*W j ;

[0041] Among them, cos(E i P j W represents the cosine similarity between the employee capability vector and the project requirement vector, measuring the degree of dimensional overlap. Its value range is [-1, 1], with values ​​closer to 1 indicating a better dimensional match. j As a weight for project priority;

[0042] S3-2. Adaptability Level Classification: Based on the matching score, the employee-project adaptation relationship is divided into 3 levels;

[0043] High match: score(E) i P j A score of ≥0.85 indicates that the employee's capabilities fully cover the project requirements and they are capable of playing core roles.

[0044] Capable of: 0.7 ≤ score (E) i P j The score is less than 0.85, indicating that the staff's capabilities basically meet the project requirements, but some guidance is needed.

[0045] To be adapted: score(E) i P j If the score is less than 0.7, the employee's ability is far from meeting the project requirements, and they need to be mentored or only participate in basic work.

[0046] Furthermore, step S4 includes:

[0047] S4-1. Construct an undirected bipartite graph G = (V, E), with the nodes and edges defined as follows.

[0048] The node set V includes the source node S and the employee node set {E1, E2, ..., E}. n}, Project node set {P1, P2, ..., P m Virtual void node V0, sink node T;

[0049] The edge set E is used to connect the nodes and define resource flow rules, clarifying the capacity limits and cost coefficients between the nodes. Specifically, it includes the following edges:

[0050] Source node S → Employee node E i Capacity = 1, Cost = 0;

[0051] Employee Node E i →Project Node P j Capacity = 1, Cost = 1 - score(E) i P j The higher the matching degree, the lower the cost, ensuring that highly matched employees are prioritized for assignment to corresponding projects;

[0052] Employee Node E i →Virtual unused node V0: Capacity = 1, Cost = 0.1;

[0053] Project Node P j →Sink node T: Capacity = Project P j Required personnel, cost = 0;

[0054] S4-2. To avoid overloading of personnel or insufficient project resources, the following hard constraints are set:

[0055] Working hours constraints: High-ability employees, weekly working hours ≤ 40h; Medium-to-low-ability employees, weekly working hours ≤ 35h;

[0056] Project quantity constraints: High-ability employees can participate in ≤3 projects simultaneously; low-to-medium ability employees can participate in ≤2 projects simultaneously.

[0057] Demand constraint: The sum of the percentages of total man-hours for each project must equal the total manpower required for the project;

[0058] S4-3. With the goal of "lowest total matching cost + lowest idle cost", the formula is as follows:

[0059] min(∑ n i＝1 ∑ m j＝1(1-score(E i P j ))*X ij +0.1*Z i );

[0060] Among them, X ij The percentage of working hours allocated from employee i to project j, Z i For employee i's idle working hours, Z i =Standard working hours - ∑ m j＝1 X ij ;

[0061] The minimum cost maximum flow algorithm is used to solve the above model, and the optimal assignment matrix X is output. ij This establishes the correspondence between employees, projects, and working hours, and generates a visual assignment plan.

[0062] Furthermore, in step S5, for highly capable employees, a "1+N" core focus model, unified management of a high-capability employee resource pool, or a capability replication and talent pipeline development system are adopted.

[0063] The "1+N" core focus model assigns one "core project" to each highly capable employee, dedicating 70%-80% of their weekly work hours to ensure deep involvement; at the same time, it supports other projects with "N lightweight roles" (N≤2).

[0064] The high-ability employee resource pool is managed uniformly. The enterprise's HR or resource management department establishes the "high-ability employee resource pool" to record each high-ability employee's professional field, current project, available time, and expertise.

[0065] When applying for highly capable employees for a project, a "requirement specification" must be submitted. The resource pool administrator will approve applications based on "project priority + employee suitability" and reject "resources with no clear value".

[0066] The capability replication and talent pipeline development involves assigning multiple mid-level employees as mentors to each high-ability employee, with a mentoring cycle set for a certain period.

[0067] Furthermore, in step S5, for employees with low skill levels, a basic task pool is constructed in a hierarchical manner, and a tiered task allocation and promotion mechanism and a cross-project scheduling mechanism are adopted.

[0068] The basic task pool is structured hierarchically as follows: it is divided into three levels based on ability requirements, forming a "basic task pool" for employees with lower abilities to choose from.

[0069] Level 1, pure execution: no professional skills required, only basic tool skills;

[0070] Level 2, light judgment: requires basic professional skills;

[0071] Level 3, Assisted Collaboration: Requires basic collaboration skills;

[0072] The tiered task allocation and promotion system: Low-ability employees are assigned tasks in the order of "L1→L2→L3". After completing each level of task with an accuracy rate of ≥90%, they are automatically promoted to the next level, ensuring that they are competent and can grow.

[0073] The cross-project scheduling mechanism monitors the workload saturation of low-skilled employees in real time through the "project-employee matching dashboard" of the supporting software system. If the basic tasks of Project A are saturated, the low-skilled employees will be immediately scheduled to the basic task pool of Project B to avoid uneven staffing between projects.

[0074] Furthermore, step S6 includes:

[0075] S6-1 Real-time Monitoring Dashboard: A visual dashboard is built using the accompanying software system. Core fields include:

[0076] Project-level information: Project type, priority, current progress, staffing gaps, and required skill types;

[0077] Employee side: Ability level, current project, work saturation, available time and areas of expertise;

[0078] Warning information: When the saturation of high-capacity employees is greater than or equal to a certain percentage, an "overload warning" is triggered; when the saturation of low-capacity employees is less than a certain percentage, an "idle warning" is triggered; when the project schedule is delayed by more than or equal to a certain percentage, a "resource replenishment warning" is triggered.

[0079] S6-2, Weekly Resource Coordination Meeting: A weekly resource coordination meeting is held by the PMO, project managers, and HR specialists to handle alerts according to the following process:

[0080] (1) Overload handling: If the saturation of high-capacity employees is greater than or equal to a certain percentage, medium-capacity employees should be dispatched from "projects that are closed ahead of schedule" to supplement the workforce, or high-capacity employees should be removed from "non-core modules".

[0081] (2) Idle handling: If the saturation of low-skilled employees is less than a certain percentage, tasks will be allocated from "reserve projects" or "basic modules of regular projects";

[0082] (3) Emergency Queueing: In case of an emergency project, low- to medium-capability employees will be temporarily dispatched from the "lowest priority reserve projects", while high-capability employees will only provide "short-term support";

[0083] S6-3, Quarterly Review and Calibration:

[0084] Data review: Statistics on overload rate of high-skilled employees, idle rate of low-skilled employees, on-time project delivery rate, and promotion rate of low-skilled employees; to analyze the shortcomings of the current strategy.

[0085] Parameter optimization: Adjust project priority weights and task pool level standards;

[0086] Profile Update: Based on the employee's quarterly project performance, update the employee's capability vector to ensure that subsequent matching is based on the latest capability data.

[0087] A multi-project personnel intelligent deployment and dynamic balancing device, comprising: at least one memory and at least one processor;

[0088] The at least one memory is used to store a machine-readable program;

[0089] The at least one processor is used to call the machine-readable program to execute a multi-project personnel intelligent arrangement and dynamic balancing method.

[0090] Compared with the prior art, the intelligent arrangement and dynamic balancing method and apparatus for multi-project personnel of the present invention has the following outstanding advantages:

[0091] (1) Compared with existing technologies, the accuracy of employee-project matching is improved by 55%, the overload rate of high-ability employees is reduced by 91%, the idle rate of low-ability employees is reduced by 82%, the on-time delivery rate of projects is improved by 28%, the promotion rate of low-ability employees is increased by 120%, the growth cycle of low-ability employees is reduced by 33%, and the turnover rate of core talents is reduced by 68%.

[0092] This invention achieves significant results through quantitative matching, global optimization, and dynamic balancing.

[0093] (2) In addition, the present invention can bring long-term value to enterprises: by replicating capabilities and building talent pipelines, the dependence on external core talents is reduced, and recruitment costs are reduced by 20%; by dynamic scheduling, the overall utilization rate of human resources is increased from 65% to 92%, indirectly increasing enterprise revenue by 15%-20%. Attached Figure Description

[0094] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0095] Figure 1 This is a flowchart illustrating a method for intelligent staffing and dynamic balancing across multiple projects.

[0096] Figure 2This is a schematic diagram illustrating the generation of employee capability vectors and project demand vectors in a multi-project intelligent personnel deployment and dynamic balancing method.

[0097] Figure 3 This is a schematic diagram of a bipartite graph model in a multi-project intelligent personnel arrangement and dynamic balancing method.

[0098] Figure 4 This is a schematic diagram of a real-time monitoring dashboard interface in a multi-project intelligent personnel arrangement and dynamic balancing method.

[0099] Figure 5 This is a schematic diagram of the "1+N" work hour allocation for highly capable employees in a multi-project intelligent personnel arrangement and dynamic balancing method.

[0100] Figure 6 This is a schematic diagram illustrating the graded task promotion of low-ability employees in a multi-project intelligent personnel arrangement and dynamic balancing method. Detailed Implementation

[0101] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to specific embodiments. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0102] The following is a preferred embodiment:

[0103] like Figure 1-6 As shown in this embodiment, a method for intelligent deployment and dynamic balancing of personnel across multiple projects includes the following steps:

[0104] S1. Quantify employee capability profiles and generate computable capability vectors;

[0105] Transforming employee capabilities from "qualitative descriptions" into "quantitative vectors" provides an objective basis for subsequent matching:

[0106] include:

[0107] S1-1, Definition of Capability Dimensions: Select 5 quantifiable dimensions that cover the core value of employees, taking into account both hard skills and soft qualities, and avoiding redundancy or lack of dimensions.

[0108] Dimension 1 (Hard Technical Skills): Proficiency in core job skills (such as programming ability, mastery of design software, and equipment operation skills), quantified through skills assessment tools (such as iMocha and Nowcoder) and historical project results (code reuse rate, task accuracy rate), with an initial score range of 0-5 points;

[0109] Dimension 2 (Project Experience): Years / rounds of participation in similar projects (e.g., "3 years of e-commerce system development experience" or "2 large-scale project leadership experiences") and project roles (leader / core executor / supporter), quantified through file verification in the enterprise project management system (e.g., Jira, Lark projects), with an initial score range of 0-5 points;

[0110] Dimension 3 (Soft Skills): Communication and Collaboration Skills (efficiency of cross-departmental requirement coordination and conflict resolution effectiveness), and Stress Resistance (quality of high-pressure project delivery). These are quantified through 360-degree feedback (evaluation from superiors / colleagues / subordinates) and Behavioral Event Interviews (BEI), with an initial score range of 0-5.

[0111] Dimension 4 (Certification and Performance): Industry authoritative certifications (such as PMP, AWS certification, skill level certificates), KPI completion rate in the past 3 months (on-time delivery rate of tasks, pass rate of results), quantified through the enterprise HR performance system and certification certificate database, with an initial score range of 0-5 points;

[0112] Dimension 5 (Growth Potential): Learning speed of new tools / processes (such as the time to get started with a new framework, proficiency in operating a new system), and implementation rate of innovation proposals (such as the number of project optimization suggestions adopted, and the effectiveness of cost reduction and efficiency improvement solutions). These are quantified through training assessment results and the innovation proposal management system, with an initial score range of 0-5 points.

[0113] S1-2, Normalization and Vector Generation: Using the min-max normalization algorithm, the initial scores of each dimension are mapped to the [0, 1] interval (eliminating differences in units) to generate the ability vector for each employee: E i =(e i1 e i2 e i3 e i4 e i5 );

[0114] Among them, e ik ∈[0,1](k=1~5), corresponding to the normalized scores of the above 5 dimensions respectively.

[0115] For example, the ability vector of senior engineer A is E. A = (0.9, 0.8, 0.7, 0.8, 0.9), the ability vector of new employee B is E. B = (0.4, 0.5, 0.6, 0.3, 0.4).

[0116] S1-3. Dynamic Profile Update: By default, the capability vector is automatically updated quarterly based on employee project performance (task completion rate, mentoring results); if an employee participates in a major project milestone (such as core module delivery, implementation of innovative solutions), an update is triggered immediately to ensure that the profile is synchronized with the employee's actual capabilities.

[0117] S2. Project requirements modeling, generating requirement standards that match capability vectors;

[0118] This step aligns project requirements with employee capabilities, while also assigning priority weights to ensure that high-priority projects receive the best resources first.

[0119] include:

[0120] S2-1. Project Classification and Priority Assignment: Based on the project's strategic value (its impact on the company's annual goals) and execution complexity (technical / business challenge difficulty), projects are classified into 4 categories. These categories are reviewed and adjusted monthly by the company's PMO (Project Management Office), and a priority weight W is assigned accordingly. j (Value range [0, 1], the higher the weight, the higher the priority of resource allocation):

[0121] The four types of projects are as follows:

[0122] Strategic core projects: These projects impact the company's core annual goals and require significant technological or business breakthroughs (e.g., next-generation product development, customized systems for major clients). Priority weight: W j =1.0;

[0123] Emergency support projects: Must be delivered within a short period (e.g., 1-2 weeks), with clearly defined processes but requiring efficient execution (e.g., system vulnerability patching, quarterly financial report audit support), with priority weight W. j =0.9;

[0124] Routine execution projects: daily operational tasks, standardized processes (such as monthly user operations, routine customer maintenance), priority and weight W j =0.7;

[0125] Reserve growth projects: long-term preparatory tasks, allowing for trial and error (such as new technology research and development, new employee training), with priority and weight W. j =0.5.

[0126] S2-2, Project Requirements Vector Generation: Referring to the 5 capability dimensions in step S1, define the requirement intensity for each dimension for each type of project, on a scale of 0-1, where 0 represents no requirement and 1 represents the highest requirement. Generate the project requirements vector: P j =(p j1 p j2 p j3 p j4 p j5 Among them, P jk ∈[0,1](k=1~5), which corresponds one-to-one with the dimensions of the employee capability vector.

[0127] For example, the demand vector for the strategic core project (new APP development) is P1 = (0.9, 0.8, 0.7, 0.8, 0.9), and the demand vector for the reserve growth project (new employee training) is P4 = (0.3, 0.2, 0.4, 0.3, 0.7).

[0128] S2-3, Demand Vector Adjustment: Supports PMOs to adjust project demand vectors in real time through the visual interface of the supporting software system (such as adding "resilience" demand intensity to emergency projects) to ensure that the demand is synchronized with the actual situation of the project.

[0129] S3. Calculate employee-project matching degree and quantify the suitability level;

[0130] This step uses a weighted cosine similarity algorithm to calculate the matching degree between the employee's ability vector and the project's requirement vector, and to determine the suitability level.

[0131] include:

[0132] S3-1, Matching degree formula: score(E i P j )=cos(E i P j )*W j ;

[0133] Among them, cos(E i P j W represents the cosine similarity between the employee capability vector and the project requirement vector (measuring the degree of dimensional overlap, with a value range of [-1, 1], where the closer to 1, the better the dimensional match). j This represents the project priority weight; the formula considers both the fit between capabilities and needs and the impact of project priority on the matching results.

[0134] S3-2. Adaptability Level Classification: Based on the matching score, the employee-project adaptation relationship is divided into 3 levels to provide a basis for subsequent allocation;

[0135] High match: score(E) i P j )≥0.85, the employee's capabilities fully cover the project requirements and can serve in core roles (such as project leader, core module leader);

[0136] Capable of: 0.7 ≤ score (E) i P j The score is less than 0.85, indicating that the employees' abilities basically meet the project requirements, but a small amount of guidance is needed (e.g., employees with medium abilities are responsible for executing key modules).

[0137] To be adapted: score(E) i P jIf the score is less than 0.7, the employee's ability is far from meeting the project requirements, and they need to be paired with a mentor or only participate in basic work (such as low-ability employees undertaking data entry).

[0138] Example calculation: Senior Engineer A (E A The matching degree between the value of E = (0.9, 0.8, 0.7, 0.8, 0.9) and the strategic core project P1 (P1 = (0.9, 0.8, 0.7, 0.8, 0.9), W1 = 1.0) is cos(E) A P1)*1.0=0.98, which is a "high match"; new employee B(E) B =(0.4, 0.5, 0.6, 0.3, 0.4)) has a matching degree of 0.64*1.0=0.64 with P1, which belongs to "to be adapted".

[0139] S4. Construct a bipartite graph of "employee-project-vacancy" and use the minimum cost flow algorithm to solve for the global optimal allocation;

[0140] include:

[0141] S4-1. Bipartite Graph Model Construction: Construct an undirected bipartite graph G = (V, E), with the nodes and edges defined as follows, ensuring coverage of the entire "employee-project-idle" scenario:

[0142] Node set V: includes source node S, employee node set {E1, E2, ..., E...} n}, Project node set {P1, P2, ..., P m Virtual idle node V0 (handles temporary idleness of low-skilled employees), sink node T;

[0143] The edge set E is used to connect the nodes and define resource flow rules, clarifying the capacity limits and cost coefficients between the nodes. Specifically, it includes the following edges:

[0144] Source node S → Employee node E i Capacity = 1 (1 employee can participate in multiple projects but work hours need to be split), Cost = 0 (no additional cost);

[0145] Employee Node E i →Project Node P j Capacity = 1 (1 employee can be assigned to this project), Cost = 1 - score(E) i P j (The higher the matching degree, the lower the cost, ensuring that highly matched employees are prioritized for assignment to corresponding projects);

[0146] Employee Node E i →Virtual idle node V0: Capacity = 1, Cost = 0.1 (Idle nodes incur a slight penalty, forcing companies to explore basic tasks and reduce manpower waste);

[0147] Project Node P j →Sink node T: Capacity = Project P j Required number of personnel (e.g., capacity required for a 5-person project = 5), cost = 0 (no additional cost).

[0148] S4-2. Constraint Setting: To avoid personnel overload or insufficient project resources, the following hard constraints are set:

[0149] Working hours constraints: High-ability employees (employees with 3 or more values ​​greater than 0.8 in the ability vector are considered high-ability) have weekly working hours ≤ 40h; Medium-to-low-ability employees (employees with one value greater than 0.6 or one value less than 0.4 in the ability vector are considered low-ability) have weekly working hours ≤ 35h.

[0150] Project number constraints: High-ability employees can participate in ≤3 projects simultaneously (1 core project + 2 light roles, such as reviewer and mentor); low-to-medium ability employees can participate in ≤2 projects simultaneously.

[0151] Demand constraint: The sum of the percentage of total working hours for each project must equal the number of people required for the project (e.g., for a 5-person project, the sum of the percentage of total working hours must be 500%, meaning that each of the 5 employees contributes 100% of their working hours, or each of the 10 employees contributes 50% of their working hours).

[0152] S4-3. Objective Function and Algorithm Solution:

[0153] Objective function: The objective is to minimize the total matching cost plus the idle cost, as shown in the following formula:

[0154] min(∑ n i＝1 ∑ m j＝1 (1-score(E i P j ))*X ij +0.1*Z i )

[0155] Among them, X ij The percentage of working hours allocated from employee i to project j, Z i Idle working hours of employee i (Z) i =Standard working hours - ∑ m j＝1 X ij );

[0156] Algorithm Implementation: The minimum cost maximum flow algorithm (implemented based on the Google OR-Tools open-source framework, balancing solution efficiency and accuracy) is used to solve the above model and output the optimal assignment matrix X. ij(i.e., the correspondence between employees, projects, and working hours), and at the same time generate a visual assignment scheme (such as a Gantt chart or personnel allocation table).

[0157] S5. For high-ability employees, adopt a "1+N" focus and mentorship replication approach; for low-ability employees, establish an L1-L3 graded task pool and cross-project scheduling.

[0158] To address the core issue of overworked high-skilled employees and underutilized low-skilled employees, a differentiated management strategy is designed to balance resource utilization and talent development.

[0159] For highly capable employees, a "1+N" core focus model is adopted, along with a unified management of a high-capability employee resource pool or a capability replication and talent pipeline development system.

[0160] The "1+N" core focus model assigns each highly capable employee one "core project" (strategic core or emergency support project), investing 70%-80% of their weekly work hours to ensure deep involvement; at the same time, it supports other projects with "N lightweight roles" (N≤2), such as technical review (≤2 hours per week) and new employee mentoring (≤4 hours every two weeks), to avoid overload caused by being fully committed to multiple projects.

[0161] Example: The working hours of technical expert A (highly capable) are allocated as follows: 70% to strategic project P1 (core module development) + 15% to emergency project P2 (technical review) + 15% to reserve project P4 (new employee mentoring), with a total load of 90% (no overload).

[0162] Unified management of the high-ability employee resource pool: The company's HR or resource management department establishes a "high-ability employee resource pool" to record each high-ability employee's professional field (such as Java development, marketing planning), current project, available time, and modules of expertise;

[0163] When applying for highly capable employees for a project, a "requirement specification" must be submitted (clearly defining the core problem to be solved, the estimated working hours, and the impact on the project). The resource pool administrator approves applications based on "project priority + employee suitability" and rejects "resources with no clear value" (such as assigning basic data statistics to technical experts).

[0164] Capability replication and talent pipeline development: Assign 1-2 mid-level employees (selected from the capability vector with "growth potential ≥ 0.7 + core competencies ≥ 0.6") to each high-performing employee as mentors, with a 3-month mentoring cycle.

[0165] Month 1: Highly capable employees demonstrate core tasks (such as building the solution framework), while mid-capable employees assist in collecting data and refining details;

[0166] Months 2-3: Mid-level employees independently take charge of non-core modules (such as the development of the APP log module), while high-level employees review and correct issues regularly.

[0167] Mentoring incentives: If the mentor can independently complete the non-core work of the original high-ability employee within 3 months, the high-ability employee will receive an additional 10% monthly performance score and their corresponding working hours (e.g., 15%) will be released for core projects or new lightweight support tasks, reducing reliance on a single high-ability employee.

[0168] For employees with low skill levels, a tiered system of basic task pools, tiered task allocation and promotion, and cross-project scheduling is adopted.

[0169] A tiered basic task pool is constructed: Basic tasks that are "standardized, low-complexity, and highly forgiving" are identified from all projects and categorized into three levels based on ability requirements, forming a "basic task pool" for employees with lower skill levels to choose from.

[0170] Level L1 (Pure Execution): No professional skills are required, only basic tool skills (such as Excel / Word usage), such as customer information entry and project document archiving. The acceptance criterion is a task accuracy rate of ≥95%.

[0171] Level L2 (Light Judgment): Requires basic professional skills (core dimension ≥ 0.4), such as comparing sales data with CRM records and generating templated weekly reports. The acceptance criteria are a task accuracy rate ≥ 92% and efficiency meeting the standard (e.g., data verification ≥ 100 records / day).

[0172] Level L3 (Assisted Collaboration): Requires basic collaboration skills (soft skills ≥ 0.5), such as cross-departmental request communication and task progress tracking (updating Jira status according to templates). The acceptance criteria are a collaboration feedback timeliness rate ≥ 100% and unambiguous request communication.

[0173] Tiered task allocation and promotion: Low-ability employees are assigned tasks in the order of "L1→L2→L3". After completing each level of task with an accuracy rate of ≥90%, they are automatically promoted to the next level, ensuring that they are "competent and have opportunities for growth".

[0174] Example: Task path for new employee B (low ability, EB = (0.4, 0.5, 0.6, 0.3, 0.4): Month 1: Assigned L1 task (customer information entry, saturation 70%, accuracy 96%) → Month 2: Promoted to L2 task (sales data verification, saturation 80%, accuracy 93%) → Month 3: Promoted to L3 task (progress tracking, saturation 90%, timely collaboration feedback 100%).

[0175] Cross-project scheduling mechanism: Through the "project-employee matching dashboard" of the supporting software system, the workload saturation of low-skilled employees can be monitored in real time.

[0176] If the basic tasks of Project A are saturated (such as a monthly operation project entering its final stage), low-skilled employees should be immediately reassigned to the basic task pool of Project B (such as L1-L2 tasks of a new customer maintenance project) to achieve "people follow tasks" and avoid uneven staffing between projects.

[0177] S6. Dynamic iterative optimization, real-time monitoring and periodic review;

[0178] To address changing project requirements and employee skill development, a dynamic iterative mechanism of "real-time monitoring - weekly scheduling - quarterly review" has been established.

[0179] Real-time monitoring dashboard: A visual dashboard is built using the accompanying software system. Core fields include:

[0180] Project-side information: Project type, priority, current progress, staffing shortage, and required skill types;

[0181] Employee side: Ability level, current project, work saturation (%), available time, and areas of expertise;

[0182] Warning information: When the saturation of high-capacity employees is ≥120%, an "overload warning" is triggered; when the saturation of low-capacity employees is <60%, an "idle warning" is triggered; when the project schedule is delayed by ≥10%, a "resource replenishment warning" is triggered.

[0183] Weekly Resource Coordination Meeting: A resource coordination meeting is held weekly by the PMO, project managers, and HR specialists to handle alerts according to the following process:

[0184] Overload handling: If the saturation of high-capacity employees is ≥120% (e.g., technical expert A is simultaneously undertaking 3 core projects), prioritize the allocation of medium-capacity employees from "projects that have been completed ahead of schedule" to supplement the workforce, or allow high-capacity employees to be removed from "non-core modules" (e.g., stop participating in regular project reviews).

[0185] Idle tasks: If the saturation of low-skilled employees is less than 60% (e.g., new employee B is only responsible for L1 tasks of one project), tasks will be allocated from "Reserve Projects" or "Regular Project Basic Modules" (e.g., document archiving work for a new P3 project).

[0186] Emergency Priority: In the event of an emergency project (such as a customer complaint handling project), low- to medium-skilled staff will be temporarily dispatched from the "lowest priority reserve projects". High-skilled staff will only provide "short-term support" (≤8 hours per week) to avoid delays in core projects.

[0187] Quarterly review and calibration:

[0188] Data review: Statistically analyze the overload rate of high-ability employees (target ≤ 5%), the idle rate of low-ability employees (target ≤ 8%), the on-time delivery rate of projects (target ≥ 98%), and the promotion rate of low-ability employees (target ≥ 30%), and analyze the shortcomings of the current strategy (such as the lack of certain basic tasks leading to the idleness of low-ability employees).

[0189] Parameter optimization: Adjust project priority weights (e.g., market changes cause a regular project to be upgraded to an emergency project), and task pool level standards (e.g., increase the professional ability requirements for L2 tasks);

[0190] Profile Update: Based on the employee's quarterly project performance (task completion rate, accuracy rate, mentoring results), update the employee's competency vector to ensure that subsequent matching is based on the latest competency data.

[0191] To support the implementation of the above six-step technical solution, this invention designs a supporting software system, comprising five core modules. These modules work together to automate the entire process of "data acquisition - modeling - matching - scheduling - review."

[0192] (1) Capability profiling engine:

[0193] Functions: Integrates with enterprise HR systems, skills assessment tools (iMocha), and project management systems (Jira) to collect employee competency data; executes normalization algorithms to generate competency vectors; and automatically updates profiles periodically.

[0194] Output: Employee competency vector library, competency level classification table (high / medium / low competency).

[0195] (2) Project Requirements Manager:

[0196] Features: Supports PMO to input project information (type, cycle, requirement dimensions); generates project requirement vectors; allows for visual adjustment of priority weights;

[0197] Outputs: Project requirement vector library, project priority list.

[0198] (3) Intelligent matching algorithm:

[0199] Functionality: Load employee capability vectors and project requirement vectors, calculate the matching degree; construct a bipartite graph model; call the minimum cost flow algorithm to solve; generate assignment schemes;

[0200] Outputs: Optimal assignment matrix, personnel-project matching Gantt chart, resource allocation report.

[0201] (4) Dynamic balance controller:

[0202] Functions: Real-time monitoring of staff saturation and project progress, triggering alerts; generating weekly scheduling suggestions; recording mentoring logs and employee promotion information;

[0203] Outputs: Overload / idle warning list, scheduling suggestion report, and teaching results statistics.

[0204] (5) Visualized review dashboard:

[0205] Features: Displays project delivery data, staff workload data, and employee growth data; supports multi-dimensional filtering (such as by project type and competency level); generates quarterly review reports;

[0206] Outputs: Data visualization charts (radar charts, bar charts), quarterly review reports, and a list of optimization suggestions.

[0207] Based on the above method, the multi-project personnel intelligent arrangement and dynamic balancing device in this embodiment includes: at least one memory and at least one processor;

[0208] The at least one memory is used to store a machine-readable program;

[0209] The at least one processor is used to call the machine-readable program to execute a multi-project personnel intelligent arrangement and dynamic balancing method.

[0210] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor can be a microprocessor or any conventional processor.

[0211] Memory is used to store computer programs and / or modules. The processor implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and by accessing data stored in the memory. Memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, at least one application program required for a function, etc.; the data storage area can store data created based on the use of the terminal, etc. In addition, memory can also include high-speed random access memory, and can also include non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart memory cards (SMC), secure digital cards (SD cards), flash memory cards, at least one disk storage device, flash memory devices, or other volatile solid-state storage devices.

[0212] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent personnel deployment and dynamic balancing across multiple projects, characterized in that, It has the following steps: S1. Quantify employee capability profiles and generate computable capability vectors; S2. Project requirements modeling, generating requirement standards that match capability vectors; S3. Calculate employee-project matching degree and quantify the suitability level; S4. Construct a bipartite graph of "employee-project-vacancy" and use the minimum cost flow algorithm to solve for the global optimal allocation; S5. For high-ability employees, adopt a "1+N" focus and mentorship replication approach; for low-ability employees, establish an L1-L3 graded task pool and cross-project scheduling. S6. Dynamic iterative optimization, real-time monitoring and periodic review.

2. The method for intelligent deployment and dynamic balancing of personnel across multiple projects according to claim 1, characterized in that, In step S1, firstly, the employee's capabilities are transformed from "qualitative descriptions" into "quantitative vectors". include: S1-1, Definition of Competency Dimensions: Select 5 quantifiable dimensions that cover the core value of employees, taking into account both hard skills and soft qualities; Dimension 1 is hard technical skills; proficiency in core job skills, quantified through skills assessment tools and historical project results, with an initial score range of 0-5 points. Dimension 2 is project experience; the number of years and rounds of participation in similar projects and the role in the projects are quantified through file verification in the enterprise project management system, with an initial score range of 0-5 points; Dimension 3 is soft skills; communication and collaboration skills and stress resistance, which are quantified through 360-degree feedback and behavioral event interviews, with an initial score range of 0-5. Dimension 4 is certification and performance, including authoritative industry certifications and recent KPI completion rates, which are quantified through the company's HR performance system and certification database, with an initial score range of 0-5 points; Dimension 5 is growth potential; including the speed of learning new tools or processes and the implementation rate of innovation proposals, which are quantified through training assessment results and the innovation proposal management system, with an initial score range of 0-5 points. S1-2. Using the min-max standardization algorithm, the initial scores of each dimension are mapped to the [0, 1] interval (eliminating differences in units) to generate the ability vector for each employee: E i =(e i1 e i2 e i3 e i4 e i5 ); where e ik ∈[0,1], k=1~5, corresponding to the normalized scores of dimensions 1 to 5 above; S1-3: By default, the capability vector is automatically refreshed quarterly based on employee project performance; if an employee participates in a major project milestone, an update is triggered immediately to ensure that the profile is synchronized with the employee's actual capabilities.

3. The method for intelligent deployment and dynamic balancing of personnel across multiple projects according to claim 2, characterized in that, In step S2, project requirements are aligned with employee capabilities, and priority weights are assigned to ensure that high-performing projects receive priority access to quality resources. include: S2-1. Based on the strategic value and execution complexity of the projects, the projects are divided into 4 categories, which are reviewed and adjusted by the enterprise's PMO over a period of time, and priority weights W are assigned accordingly. j The value range is [0, 1]. The higher the weight, the higher the priority of resource allocation. S2-2. Referring to the five capability dimensions in step S1, define the demand intensity for each dimension for each type of project, on a scale of 0-1, where 0 represents no demand and 1 represents the highest demand. Generate the project demand vector: P j =(p j1 p j2 p j3 p j4 p j5 Among them, P jk ∈[0,1], k=1~5, corresponding one-to-one with the dimensions of the employee capability vector; S2-3. The enterprise PMO can adjust the project requirement vector in real time through the visual interface of the supporting software system, so that the requirements are synchronized with the actual situation of the project.

4. The method for intelligent deployment and dynamic balancing of personnel across multiple projects according to claim 3, characterized in that, In step S2-1, the four types of items are as follows: Strategic core projects: These projects impact the company's core annual goals and require breakthroughs in technology or business operations. Priority weight: W j =1.0; Emergency support projects: Must be delivered in a short period of time, with a clear process but requiring efficient execution, and a priority weight of W. j =0.9; Routine execution projects: daily operational tasks, standardized processes, priority and weight W j =0.7; Reserve growth projects: long-term foundational tasks, open to trial and error, with priority and weight W. j =0.

5.

5. The method for intelligent deployment and dynamic balancing of personnel across multiple projects according to claim 4, characterized in that, In step S3, the weighted cosine similarity algorithm is used to calculate the matching degree between the employee's ability vector and the project requirement vector, and to determine the suitability level. include: S3-1, Matching degree formula: score(E i P j )=cos(E i P j )*W j ; Among them, cos(E i P j W represents the cosine similarity between the employee capability vector and the project requirement vector, measuring the degree of dimensional overlap. Its value range is [-1, 1], with values ​​closer to 1 indicating a better dimensional match. j As a weight for project priority; S3-2. Adaptability Level Classification: Based on the matching score, the employee-project adaptation relationship is divided into 3 levels; High match: score(E) i P j A score of ≥0.85 indicates that the employee's capabilities fully cover the project requirements and they are capable of playing core roles. Capable of: 0.7 ≤ score (E) i P j The score is less than 0.85, indicating that the staff's capabilities basically meet the project requirements, but some guidance is needed. To be adapted: score(E) i P j If the score is less than 0.7, the employee's ability is far from meeting the project requirements, and they need to be mentored or only participate in basic work.

6. The method for intelligent deployment and dynamic balancing of personnel across multiple projects according to claim 5, characterized in that, Step S4 includes: S4-1. Construct an undirected bipartite graph G = (V, E), with the nodes and edges defined as follows. The node set V includes the source node S and the employee node set {E1, E2, ..., E}. n }, Project node set {P1, P2, ..., P m Virtual void node V0, sink node T; The edge set E is used to connect the nodes and define resource flow rules, clarifying the capacity limits and cost coefficients between the nodes. Specifically, it includes the following edges: Source node S → Employee node E i Capacity = 1, Cost = 0; Employee Node E i →Project Node P j Capacity = 1, Cost = 1 - score(E) i P j The higher the matching degree, the lower the cost, ensuring that highly matched employees are prioritized for assignment to corresponding projects; Employee Node E i →Virtual unused node V0: Capacity = 1, Cost = 0.1; Project Node P j →Sink node T: Capacity = Project P j Required personnel, cost = 0; S4-2. To avoid overloading of personnel or insufficient project resources, the following hard constraints are set: Working hours constraints: High-ability employees, weekly working hours ≤ 40h; Medium-to-low-ability employees, weekly working hours ≤ 35h; Project quantity constraints: High-ability employees can participate in ≤3 projects simultaneously; low-to-medium ability employees can participate in ≤2 projects simultaneously. Demand constraint: The sum of the percentages of total man-hours for each project must equal the total manpower required for the project; S4-3. With the goal of "minimizing total matching cost + minimizing idle cost", the formula is as follows: min(∑ n i＝1 ∑ m j＝1 (1-score(E i ,P j ))*X ij +0.1*Z i ); Among them, X ij The percentage of working hours allocated from employee i to project j, Z i For employee i's idle working hours, Z i =Standard working hours - ∑ m j＝ 1X ij ; The minimum cost maximum flow algorithm is used to solve the above model, and the optimal assignment matrix X is output. ij This establishes the correspondence between employees, projects, and working hours, and generates a visual assignment plan.

7. The method for intelligent deployment and dynamic balancing of personnel across multiple projects according to claim 6, characterized in that, In step S5, for highly capable employees, a "1+N" core focus model and a unified management of the highly capable employee resource pool or a capability replication and talent pipeline development system are adopted. The "1+N" core focus model assigns one "core project" to each highly capable employee, dedicating 70%-80% of their weekly work hours to ensure deep involvement; at the same time, it supports other projects with "N lightweight roles" (N≤2). The high-ability employee resource pool is managed uniformly. The "high-ability employee resource pool" is established by the company's HR or resource management department, which records the professional field, current project, available time and modules of expertise of each high-ability employee. When applying for highly capable employees for a project, a "requirement specification" must be submitted. The resource pool administrator will approve applications based on "project priority + employee suitability" and reject "resources with no clear value". The capability replication and talent pipeline development involves assigning multiple mid-level employees as mentors to each high-ability employee, with a mentoring cycle set for a certain period.

8. The method for intelligent deployment and dynamic balancing of personnel across multiple projects according to claim 7, characterized in that, In step S5, for employees with low skills, a basic task pool is built in a hierarchical manner, and a tiered task allocation and promotion mechanism and a cross-project scheduling mechanism are adopted. The basic task pool is structured hierarchically as follows: it is divided into three levels based on ability requirements, forming a "basic task pool" for employees with lower abilities to choose from. Level 1, pure execution: no professional skills required, only basic tool skills; Level 2, light judgment: requires basic professional skills; Level 3, Assisted Collaboration: Requires basic collaboration skills; The tiered task allocation and promotion system: Low-ability employees are assigned tasks in the order of "L1→L2→L3". After completing each level of task with an accuracy rate of ≥90%, they are automatically promoted to the next level, ensuring that they are competent and can grow. The cross-project scheduling mechanism monitors the workload saturation of low-skilled employees in real time through the "project-employee matching dashboard" of the supporting software system. If the basic tasks of Project A are saturated, the low-skilled employees will be immediately scheduled to the basic task pool of Project B to avoid uneven staffing between projects.

9. The method for intelligent deployment and dynamic balancing of personnel across multiple projects according to claim 8, characterized in that, Step S6 includes: S6-1 Real-time Monitoring Dashboard: A visual dashboard is built using the accompanying software system. Core fields include: Project-level information: Project type, priority, current progress, staffing gaps, and required skill types; Employee side: Ability level, current project, work saturation, available time and areas of expertise; Warning information: When the saturation of high-capacity employees is greater than or equal to a certain percentage, an "overload warning" is triggered; when the saturation of low-capacity employees is less than a certain percentage, an "idle warning" is triggered; when the project schedule is delayed by more than or equal to a certain percentage, a "resource replenishment warning" is triggered. S6-2, Weekly Resource Coordination Meeting: A weekly resource coordination meeting is held by the PMO, project managers, and HR specialists to handle alerts according to the following process: (1) Overload handling: If the saturation of high-capability employees is greater than or equal to a certain percentage, medium-capability employees should be dispatched from "projects that are closed ahead of schedule" to supplement the workforce, or high-capability employees should be removed from "non-core modules". (2) Handling of idle employees: If the saturation of low-skilled employees is less than a certain percentage, tasks will be allocated from "reserve projects" or "basic modules of regular projects". (3) Emergency Queueing: In case of an emergency project, low- to medium-capability employees will be temporarily dispatched from the "lowest priority reserve projects", while high-capability employees will only provide "short-term support"; S6-3, Quarterly Review and Calibration: Data review: Statistics on overload rate of high-skilled employees, idle rate of low-skilled employees, on-time project delivery rate, and promotion rate of low-skilled employees; to analyze the shortcomings of the current strategy. Parameter optimization: Adjust project priority weights and task pool level standards; Profile Update: Based on the employee's quarterly project performance, update the employee's capability vector to ensure that subsequent matching is based on the latest capability data.

10. A multi-project personnel intelligent deployment and dynamic balancing device, characterized in that, include: At least one memory and at least one processor; The at least one memory is used to store a machine-readable program; The at least one processor is configured to invoke the machine-readable program to perform the method according to any one of claims 1 to 9.

Images