Method for evaluating ability and quality coupling coordination of skill post personnel and application

Abstract

The invention relates to the technical field of human resource management, in particular to a skill post personnel ability and quality coupling coordination evaluation method and application, and aims to solve the problem that existing research neglects the coordination relationship and development degree between ability and quality constituent elements. The invention provides a skill post personnel ability and quality coupling coordination evaluation method and application. The method comprises the following steps: S1, constructing a skill post personnel ability and quality evaluation index system; s2, completing conversion from initial data information to a confidence distribution form; s3, constructing an initial evaluation model, and evaluating the secondary indexes; s4, performing preliminary evaluation on the ability and quality coupling coordination of the personnel; and S5, optimizing the preliminary evaluation result in the step S4. According to the skill post personnel ability and quality coupling coordination evaluation method and application, expert experience knowledge can be effectively embedded, evaluation of personnel ability and quality comprehensive level and coupling coordination is realized, and model evaluation precision can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of human resource management, and in particular to a method for evaluating the coupling and coordination of the capabilities and qualities of skilled personnel and its application. Background Art

[0002] With the development of the national economy and the optimization and upgrading of the industrial structure, technical and skilled personnel have gradually become the most dynamic and promising key factor in production activities and an increasingly important foundation for the formation and development of a company's core competitiveness. Practice has demonstrated that, due to a variety of factors such as individual subjective intentions, ability and quality characteristics, and the actual requirements of the position, the work potential and performance of the same employee in different positions and work environments vary significantly. Therefore, in modern human resource management, how to ensure the right person for the right position has become a major concern for many companies and scholars.

[0003] Assessing the competencies of personnel in key skill positions involves combining their specific competencies with societal needs, using appropriate means to make contemporary value judgments. Scientifically and efficiently assessing the competencies of personnel in skilled positions helps identify and guide the growth and application of talent, and is a fundamental prerequisite for both enterprises and society to support the alignment of talent and position. Existing research suggests that the comprehensive competencies of personnel in key skill positions can be reflected directly or indirectly through competency indicators. However, the components of personnel competencies are multi-source and complex, with varying approaches to description and quantification. Furthermore, the positive role of expert experience in the assessment process cannot be ignored. This makes competency assessment inherently subjective and ambiguous, as the information processed is multi-source, heterogeneous, and uncertain. Therefore, competency assessment can be categorized as a typical multi-level, multi-indicator, complex modeling problem with uncertainty.

[0004] Currently, four types of methods are commonly used to address assessment issues: 1) Model-driven assessment methods. These methods require a precise mathematical description of the modeling object when building the model. However, the object of comprehensive capability and quality assessment is not an actual physical system, and the relationship between capability and quality indicators and the comprehensive level cannot be described using precise mathematical expressions. 2) Knowledge-driven assessment methods, such as the Analytic Hierarchy Process (AHP) and the Delphi method, rely on extensive expert experience, can handle fuzzy information and uncertainties, and their processes are highly interpretable. However, due to the limitations and subjectivity of expert experience, the results produced by these assessment methods may not accurately reflect the comprehensive level of personnel capabilities and qualities. 3) Data-driven assessment methods, such as support vector machines and artificial neural networks, offer simple operation and high assessment accuracy, but their performance is heavily dependent on high-quality historical labeled data. Furthermore, these methods lack mechanisms for describing and handling uncertain information, cannot effectively utilize limited expert experience, and lack strong explanations for the feasibility and rationality of the assessment results. 4) Hybrid information-driven assessment methods. Compared with the aforementioned methods, the biggest advantage of this type of method is that it does not require a precise mathematical description of the evaluation object and can effectively utilize limited empirical knowledge and historical data to construct an evaluation model.

[0005] The Belief Rule Base (BRB) approach is a typical hybrid information-driven modeling method. Professor Yang Jianbo of the University of Manchester proposed this approach by introducing a confidence framework into traditional fuzzy rules based on Dempster-Shafer evidence theory, fuzzy theory, and decision theory. BRB describes semi-quantitative information with uncertainty using belief rules, reasoning about this uncertainty using an evidence reasoning algorithm (ER), and optimizing model parameters using optimization mechanisms and historical data samples. BRB features a transparent and interpretable modeling process and is widely used in multi-source information fusion evaluation, multi-attribute decision analysis, and pattern recognition. Traditional BRBs face the problem of a "combinatorial explosion" of rules when performing multi-level, multi-metric evaluations, which severely limits their modeling capabilities. Numerous researchers have conducted research to address this issue, with representative results including the union-hypothesis BRB model and the principal component analysis-based BRB model. Both models can effectively reduce the size of the rule base. However, the union hypothesis of the former may be inconsistent with the correlation between metrics, while the physical meaning of the latter model inputs is lost, reducing interpretability. Cao et al. optimized the multi-input confidence rules into a single-input confidence rule form and further proposed a naive belief rule base modeling method (NBRB). Compared with the aforementioned methods, NBRB can reduce the size of the rule base while ensuring good interpretability, and has obvious advantages in solving multi-level and multi-indicator evaluation problems.

[0006] With the deepening understanding of the Scientific Outlook on Development, achieving a comprehensive, scientific, and objective assessment of personnel competence requires considering both spatial and temporal domains. On the one hand, the overall level of competence should be evaluated, reflecting the overall state of competence and emphasizing competence for the position; on the other hand, the developmental nature of competence should be evaluated, reflecting the degree of development and coordination between its components, emphasizing the personal growth of skilled personnel. The overall level and developmental nature of competence are important indicators of skilled personnel's ability to perform well in their current positions and adapt to future job development. However, existing research has largely focused on the overall level of competence in skilled positions, neglecting the coordination and development between its components. Summary of the Invention

[0007] In order to solve the problem that most existing research focuses on the comprehensive level of ability and quality of skilled positions, but ignores the coordination relationship and development degree between the components of ability and quality, the present invention provides a method for evaluating the coupling coordination of ability and quality of skilled position personnel. On the one hand, a multi-layer ability and quality coupling coordination evaluation model for skilled position personnel is constructed based on the NBRB model, which can effectively embed expert experience knowledge to realize the evaluation of the comprehensive level and coupling coordination of personnel ability and quality; on the other hand, based on limited historical data, interpretable constraints are designed, and an optimization model with the accuracy of ability and quality coupling coordination evaluation as the goal is constructed to improve the accuracy of model evaluation.

[0008] To achieve the above object, the technical solution adopted by the present invention is:

[0009] The present application provides a method for evaluating the coupling coordination of the capabilities and qualities of skilled personnel, including the following steps: S1: constructing an evaluation index system for the capabilities and qualities of skilled personnel; S2: completing the conversion of initial data information corresponding to the evaluation index system into a confidence distribution form; S3: based on the evaluation index system established in step S1 and the naive confidence rule base modeling method, using the data converted in step S2 to construct an initial evaluation model, and evaluating the secondary indicators in the evaluation index system; S4: based on the evaluation results of the secondary indicators in step S3, conducting a preliminary evaluation of the coupling coordination of personnel capabilities and qualities; S5: optimizing the preliminary evaluation results in step S4 to obtain the final evaluation results.

[0010] Furthermore, the evaluation index system in S1 includes four first-level indicators: ideological and political quality, physical and mental quality, scientific and technological quality, and work ability; the second-level indicators corresponding to ideological and political quality include: political literacy, moral quality, compliance with laws and regulations, and professional spirit; the second-level indicators corresponding to physical and mental quality include: physical condition and psychological state; the second-level indicators corresponding to scientific and technological quality include: information literacy and application of technology; the second-level indicators corresponding to work ability include: professional skill level, professional knowledge level, organizational and coordination ability, innovative thinking ability, and independent learning ability.

[0011] Furthermore, the confidence distribution described in S2 is expressed as:

[0012] S(x i )={(A i,j ,α i,j ),i=1,2,...,M;j=1,2,...,J};where S() is the information conversion function,

[0013] {(A 1,1 ,α 1,1 ),...,(A M,J ,α M,J )} represents the transformed confidence distribution, A i,j is the index x i The jth reference level, α i,j is the index x i The matching degree of the jth reference level, j is the number of reference levels, M i (i=1,...,T) represents x i The number of reference levels is , and J represents the number of sub-indicators to be analyzed.

[0014] Furthermore, the specific operations of S3 include: S301: constructing an initial assessment model for the ability and quality of skilled position personnel based on S1 and S2; S302: solving the secondary indicators through information conversion, rule activation and rule fusion; S303: solving the primary indicators through secondary indicator information fusion.

[0015] Furthermore, the solution formula for the secondary index in S302 is:

[0016]

[0017] Where 0≤w i,m ≤1 represents the activation weight of the i-th indicator input information on the m-th rule, L=M i (i=1,...,T) is the total number of rules, It represents the confidence of the nth output conclusion in the output conclusion of the i-th model, and N is the number of confidence levels of the output conclusions. represents the unknown degree of the mth rule of the i-th NBRB model, β n,i,m (n=1,...,N) represents the confidence of the nth output conclusion.

[0018] Furthermore, the specific operations of S4 include:

[0019] S401: Based on S303, a quantitative model for coupling personnel capabilities and qualities is constructed as follows:

[0020]

[0021] Where Co∈[0,1] represents the coupling index, Y j (j=1, 2, 3, 4) represents the utility value of the comprehensive level evaluation results of skill position ability and quality;

[0022] S402: Calculating a coupling coordination index based on S401;

[0023]

[0024] Where De∈[0,1] represents the coupling coordination index, δ j ′ is the normalized indicator weight.

[0025] Furthermore, the specific operation in S5 includes the following steps:

[0026] S501: Constructing the optimization goal of coupling and coordination of the ability and quality of skilled post personnel;

[0027]

[0028] Where Ω represents the parameter vector to be optimized, Ξ(Ω) represents the error between the model output and the actual value, Represents the output utility value of the model, O represents the actual evaluation result utility value, δ represents the indicator weight, θ represents the rule weight, and β represents the rule confidence;

[0029] S502: Determine the constraints of the optimization target in step S501; S503: Use the covariance matrix adaptive evolutionary strategy with projection operation to solve the optimization target in step S501.

[0030] Furthermore, the constraint conditions in S502 are:

[0031] 0≤θ i,m ≤1;

[0032] 0≤δ i ≤1;

[0033] 0≤β n,i,m ≤1;

[0034]

[0035] A coupling coordination evaluation system for the ability and quality of skilled position personnel is characterized in that the evaluation system includes an index system module, an initial data conversion module, an index model solving module, a coupling coordination evaluation module and a result optimization module; the index system module is used to objectively measure the performance of skilled position personnel in terms of ideological and political quality, physical and mental quality, scientific and technological quality and work ability; the initial data conversion module is used to propose different conversion methods for information types, and complete the conversion of information collected in the index system into a confidence distribution form; the index model solving module is used to determine the reference value, confidence, rule weight, etc. of the index system, and evaluate the comprehensive level of skilled position personnel through reasoning fusion; the coupling coordination module is used to obtain the coupling coordination index of each indicator in the index model to describe the coupling coordination of the ability and quality of skilled position personnel; the index system module, initial data conversion module, index model solving module, coupling coordination evaluation module and result optimization module are all implemented using the evaluation method described above.

[0036] An electronic device comprises at least one processor; and a memory communicatively connected to the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor to enable the processor to perform the evaluation method as described above.

[0037] The present invention has the following technical effects:

[0038] 1. This assessment method, through a series of rigorous design and technical means, accurately evaluates the coordinated and coordinated capabilities and qualities of skilled personnel. It not only provides a powerful tool for human resource management and development, but also promotes a positive interaction between employee personal growth and corporate development. Specifically, this method can help companies more effectively allocate human resources, improve work efficiency and service quality, and contribute to the continuous improvement of employee self-awareness and personal capabilities.

[0039] 2. By establishing a standardized description framework, the consistency and objectivity of the evaluation criteria are ensured, which not only helps to accurately reflect the ability and quality of employees, but also lays a solid foundation for subsequent data processing and analysis; it improves the comprehensiveness and accuracy of the evaluation.

[0040] 3. The NBRB assessment model combines multiple reference levels, output conclusions and their confidence levels, as well as weightings for rules and indicators, to form a complex decision-support system. It automatically derives the most likely outcome based on input information and assigns corresponding confidence levels, significantly enhancing the intelligence of the assessment process. This approach ensures that the assessment results not only account for individual differences but also comprehensively reflect overall trends, providing more accurate assessment conclusions. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a model diagram for evaluating the competence and quality of skilled personnel in the present invention;

[0042] Figure 2 This is a diagram showing the basic concept of parameter optimization for the capability and quality assessment model of the present invention;

[0043] Figure 3 This is a diagram showing the ability and quality coupling coordination evaluation results of the present invention;

[0044] Figure 4 This is a comparison chart of the capability and quality coupling coordination evaluation results of the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the technical solution of the present invention, the technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0046] Example 1

[0047] Combine Figure 1-4 As shown, the present invention provides a method for evaluating the coupling coordination of the ability and quality of skilled post personnel, comprising the following steps:

[0048] S1: Construct an evaluation index system for the competence and quality of skilled personnel;

[0049] Specifically, by summarizing existing relevant literature and consulting experts in the field, and adhering to the principles of comprehensiveness, independence, simplicity, and quantifiability, this paper selects 4 first-level indicators and 13 second-level indicators to form an indicator system for the ability and quality of skilled personnel, as shown in Table 1.

[0050] Ideological and political quality is primarily assessed in terms of political literacy, moral character, compliance with laws and regulations, and professionalism. These factors are difficult to quantify directly. To ensure scientificity and objectivity, they can be evaluated based on actual performance through a combination of self-assessment and peer-assessment. Physical and mental quality primarily considers physical and mental health. Physical health can be assessed through physical examinations, while mental health can be assessed through psychological assessments. Scientific and technological quality is primarily assessed through information literacy and technological application, both of which can be assessed through knowledge assessments. Work ability, the core foundation of skilled personnel, is primarily assessed in terms of professional skills, professional knowledge, organizational and coordination skills, innovative thinking skills, and autonomous learning ability. Professional skills and professional knowledge can be assessed through appropriate assessments, organizational and coordination skills can be assessed based on actual performance through a combination of self-assessment and peer-assessment, innovative thinking skills can be assessed through innovative thinking assessments, and autonomous learning ability can be assessed through peer-assessment. The weights of each indicator system shown in Table 1 are represented by δ, which can be assigned by experts using the Analytic Hierarchy Process (AHP). Reference can be made to existing research and will not be discussed in detail in this article.

[0051] Table 1 Ability and quality evaluation index system for skilled post personnel

[0052]

[0053]

[0054] S2: Complete the transformation of the initial data information corresponding to the evaluation index system into the form of confidence distribution;

[0055] Specifically, the confidence distribution form in the present invention is specifically expressed as:

[0056] S(x i )={(A i,j ,α i,j ),i=1,2,...,M; j=1,2,...,J} (1)

[0057] Among them, S() is the information conversion function, {(A 1,1 ,α 1,1 ),...,(A M,J ,α M,J )} represents the transformed confidence distribution. A i,j is the index x i The jth reference level, α i,j is the index x iThe match degree of the jth reference level, where j is the number of reference levels and J represents the number of sub-indicators to be analyzed. The reference levels and abbreviations for the different competency components are shown in Table 1. The reference levels for the four first-level indicators are "weak (W)", "qualified (S)", "average (N)", and "strong (ST)".

[0058] For example, when collecting data on an employee’s political literacy, it is set to three reference levels, namely “qualified”, “medium”, and “high”. If 90% of people believe that the employee’s political literacy is high, and 10% of people believe that his political literacy only reaches the medium standard, then the employee’s political literacy can be described as {(high, 0.9), (medium, 0.1), (qualified, 0)} using the confidence distribution form of formula (1).

[0059] More specifically, qualitative and quantitative information can be obtained through the data collection methods shown in Table 1. To complete the transformation of initial data information into a confidence distribution form, different transformation methods must be proposed for each type of information.

[0060] For qualitative information, such as qualitative judgments given in the process of self-evaluation or peer evaluation, in order to further reduce the subjectivity of qualitative information, this paper transforms the information through statistical analysis of multi-person evaluation results, which can be described as:

[0061]

[0062] Among them, n i,j To convert the index x i The number of people who are evaluated as the jth reference level. i,j is the index x i The matching degree of the j-th reference level.

[0063] For quantitative information, such as scores determined in skill assessment or knowledge assessment, this paper transforms them by designing a triangular membership function. First, determine the reference level A i,j and the reference value γ i,j ,Right now:

[0064] γ i,j meansA i,j ,i=1,2,...,M; j=1,2,...,J (3)

[0065] Then the corresponding matching degree is calculated as follows:

[0066]

[0067] Among them, α i,j For input x i The matching degree of the jth reference level, x iis the specific quantitative score of the corresponding indicator.

[0068] S3: Based on the evaluation index system established in step S1 and the naive belief rule base modeling method, the data converted in step S2 is used to build an initial evaluation model, and the secondary indicators in the evaluation index system are evaluated.

[0069] Specifically, based on the indicator system shown in Table 1, combined with the logical relationship between indicators and the NBRB model, the following Figure 1 The skill position personnel ability quality assessment model shown in the figure. Among them, the 13 secondary indicators are recorded as X1-X 13 , using the data from these secondary indicators as input, corresponding NBRB models are constructed. The four primary indicators, denoted as Y1-Y4, are the fusion results of the corresponding secondary indicator NBRB model outputs using the Evidential Reasoning (ER) algorithm. When evaluating the competence and quality of skilled personnel, fusing Y1-Y4 using the ER algorithm yields a comprehensive assessment of the competence and quality of the skilled position. The coupling and coordination of the four primary indicators is then analyzed and calculated.

[0070] For Figure 1 The evaluation model shown, the i-th indicator X i The mth rule corresponding to the NBRB model can be described as:

[0071]

[0072] Among them, M i (i=1,...,T) represents X i Number of reference levels; A i,m Indicates indicator X i The reference value corresponding to the mth reference level; D n (n=1,...,N) represents the nth output conclusion; β n,i,m (n=1,...,N) represents the confidence of the nth output conclusion, and N represents the total number of output conclusions; θ i,m Representation rule R i,m The weight of the rule represents the importance of the rule relative to other rules; the weight corresponding to the i-th indicator is δ i It represents the importance of this indicator relative to other indicators.

[0073] The initial assessment model for the competency and quality of skilled personnel is typically established by domain experts based on their experience. This primarily involves determining the relevant parameters within the aforementioned rules, such as the reference values, confidence levels, and rule weights of indicator information. Once the initial model is established, when corresponding indicator information is input into the model, multiple reasoning steps are required, including information conversion, rule activation, rule fusion, and indicator information fusion. These steps are as follows:

[0074] Step 1 (Information conversion): For the qualitative and quantitative information of the secondary indicators in Table 1, convert them into confidence distribution forms, namely {(A i,m ,α i,m ),i=1,...,T;m=1,...,M i}.

[0075] Step 2 (Rule Activation): Calculate the matching degree between the input information and the corresponding confidence rule:

[0076]

[0077] in, Indicates the matching degree between the input information of the ith indicator and the mth rule, which is between 0 and 1. The higher the matching degree, the better. The closer the value is to 1. After obtaining the matching degree of all rules of the i-th indicator, the activation weight of the corresponding rule is further calculated as follows:

[0078]

[0079] Where 0≤w i,m ≤1 represents the activation weight of the i-th indicator input information on the m-th rule. The activation weight represents the degree to which the rule is activated, w i,m =1 means that the rule is fully activated, w i,m =0 means not activated at all.

[0080] Step 3 (rule fusion): Use the ER iterative algorithm to fuse the activated rule consequences to obtain the output of the NBRB model. The specific process is as follows:

[0081]

[0082] Where L = M i (i=1,...,T) is the total number of rules, It represents the confidence of the nth output conclusion in the output conclusion of the i-th NBRB model. The unknown degree model output of the mth rule of the i-th NBRB model can be described in the form of confidence distribution:

[0083]

[0084] In the above formula, A * represents the actual input information of the NBRB model, and N is the number of confidence levels of the output conclusions. * ) can be calculated as follows, where u() represents the utility function:

[0085]

[0086] Step 4 (Secondary Index Information Fusion): Through steps 1 to 3, the output of the input information in the secondary index NBRB model can be obtained, that is, the evaluation result of the secondary index. In order to obtain the evaluation result of its corresponding primary index, it is necessary to combine Figure 1 The model characteristics shown are used to integrate the evaluation results of the secondary indicators corresponding to the primary indicators. For example, the output results of the four models NBRB1 to NBRB4 are integrated to obtain the evaluation results of ideological and political quality.

[0087] S4: Based on the evaluation results of the secondary indicators in step S3, conduct a preliminary evaluation of the coupling and coordination of personnel capabilities and qualities.

[0088] Specifically, based on steps 1 to 4 above, the evaluation results of the four first-level indicators of ideological and political quality, physical and mental quality, scientific and technological quality, and work ability can be obtained. Since the reference levels of the four first-level indicators are "weak (W)", "qualified (S)", "normal (N)" and "strong (ST)", the corresponding reference values ​​are 0, 0.5, 0.75 and 1 respectively. The confidence distribution form of the evaluation results can be expressed as in

[0089] If the reference level and reference level of the comprehensive level of skill position ability and quality are consistent with the four first-level indicators, the confidence level of the nth reference level can be obtained by the following formula:

[0090]

[0091] Among them, L=N=4 are the number of first-level indicators and the number of first-level indicator reference levels respectively; w m Take the standardized weight values ​​of the first-level indicators respectively, that is, Similarly, the utility value Y of the comprehensive level evaluation result of skill position ability quality is j (j=1,2,3,4) can be calculated by formula (11), and Y j ∈[0,1],Y j =0 means that the comprehensive level of this indicator is completely at the "weak" level, Y j =1 means that the comprehensive level of this indicator is completely at the "strong" level.

[0092] More specifically, the development of human capabilities and qualities is a complex and dynamic process, characterized by a multi-faceted coupling relationship between their capabilities and qualities, characterized by mutual promotion, mutual restriction, and interaction. Coupling coordination assessment is based on a coupling coordination model, describing the interactions between several sub-indicators through a coupling coordination index. Therefore, based on reference to relevant research findings and in light of the specifics of this study, a quantitative model for the coupling of human capabilities and qualities was constructed, focusing on primary indicators as the analysis object:

[0093]

[0094] Among them, J represents the number of sub-indicators to be analyzed, Co∈[0,1] represents the coupling index, and Y j (j=1,2,3,4) represents the utility value of the comprehensive level evaluation results of skill position ability and quality; combined with Figure 1 It can be seen that there are four indicators in total; Co∈[0,1] represents the coupling index, and its different values ​​from 0 to 1 represent different degrees of coupling. The closer it is to 1, the stronger the interaction between the indicators.

[0095] Although the coupling index can reflect the degree of interaction between indicators, it is difficult to characterize whether the indicators promote each other at a high level or restrict each other at a low level. Therefore, it is necessary to further combine the comprehensive level of ability and quality on the basis of the coupling index to obtain the coupling coordination index. The specific calculation formula is:

[0096]

[0097] Among them, De∈[0,1] represents the coupling coordination index. The closer its value is to 1, the stronger the coordination between indicators. j ′ is the normalized indicator weight, which satisfies δ1′+δ2′+δ3′+δ4′=1. Referring to relevant research, the types of coupled coordination of personnel capabilities and qualities can be divided as shown in Table 2.

[0098] Table 2 Coupling and coordination types of competence and quality of skilled post personnel

[0099]

[0100]

[0101] The initial parameters of the above-mentioned skill position personnel ability and quality assessment model need to be given in combination with expert experience. However, due to the limitations and subjectivity of expert knowledge, the initial model is usually difficult to accurately implement the assessment. It is necessary to adjust the relevant parameters in combination with historical data to improve performance.

[0102] S5: Optimize the preliminary evaluation result in step S4 to obtain the final evaluation result.

[0103] Specifically, the goal of optimizing the skill position personnel competency assessment model is to minimize the error between the assessed coupling coordination and the actual value. Given that the coupling coordination index can comprehensively reflect the coupling coordination of competency and quality, the following optimization goal can be constructed:

[0104] minΞ(Ω)(15)

[0105] Where Ω represents the parameter vector to be optimized, which mainly includes the indicator weight δ, the rule weight θ, and the rule confidence β; Ξ(Ω) represents the error between the model output and the actual value. These errors can be measured based on the mean square error (MSE) between the two, that is:

[0106]

[0107] in, represents the output utility value of the model, and O represents the actual evaluation result utility value. Therefore, the optimization objective can be further expressed as;

[0108] Furthermore, when optimizing the capability and quality model, some parameters must also meet the following constraints to ensure their interpretability: all rule weights and attribute weights should be between 0 and 1; the confidence of each rule should be between 0 and 1, and the sum of the built-in confidence of each rule should be 1.

[0109] 0≤θ i,m ≤1 (17)

[0110] 0≤δ i ≤1 (18)

[0111] 0≤β n,i,m ≤1 (19)

[0112]

[0113] Therefore, the optimization objective function of the aforementioned skill position personnel ability and quality assessment model should take formula (15) as the target and formulas (17)-(20) as the constraints.

[0114] The basic idea of ​​optimizing the parameters of the capability and quality assessment model is as follows: Figure 2 As shown, the model parameters are adjusted to minimize the error between the evaluation results output by the model and the actual results obtained through the complex process of "evaluation-feedback-group decision-making".

[0115] Due to the complex input-output relationship of the capability and quality assessment model, the large scale of model parameters, and the difficulty in obtaining gradients, traditional gradient descent optimization methods are difficult to effectively adjust model parameters. In view of this, it is necessary to adopt an intelligent optimization algorithm. The covariance matrix adaptive evolutionary strategy (CMA-ES) is an efficient intelligent optimization algorithm that can effectively combine search intensity to generate new solutions. CMA-ES is only suitable for solving unconstrained nonlinear, non-convex real-valued continuous optimization problems. In order to solve the optimization problem with constraints, Hu Guanyu et al. proposed a covariance matrix adaptive evolutionary strategy with projection operation (P-CMA-ES) by introducing a projection operation. This paper will also optimize the model parameters based on this method.

[0116] Specific case analysis:

[0117] In 2014, a certain industrial group established a high-voltage electromechanical product operation and maintenance department. In the past 10 years, it has introduced and trained 286 operation and maintenance engineers. In order to ensure the scientific and efficient management of the engineer group, it is necessary to reasonably, reliably and efficiently evaluate their capabilities and qualities in order to formulate corresponding management mechanisms to stimulate motivation and vitality. At present, during operation, the industrial group regularly organizes expert groups to evaluate the coupling coordination of the capabilities and qualities of 286 operation and maintenance engineers through the "assessment-feedback-group decision-making" process. The evaluation results have been gradually used to optimize the management mechanism. However, the above organizational process is complex, time-consuming, consumes a lot of financial resources, and has subjective preferences. Therefore, it is necessary to propose new methods to improve the quality and efficiency of personnel management. During the operation of the operation and maintenance department, some data were collected based on the group's human resources management system, including regular psychological tests, knowledge and skills tests for qualification assessments and promotions, annual physical examination reports, annual self-evaluation and mutual evaluation reports, etc., which can support the completion of Figure 1 To verify the effectiveness of the proposed method, 148 typical research subjects were selected according to the principles of data completeness, coverage, and typicality, including 41 junior engineers, 65 intermediate engineers, and 42 senior engineers.

[0118] Construct an initial model. The data obtained by the human resources management system are in qualitative and quantitative forms. The data collection method of qualitative indicators such as political literacy, moral quality, compliance with laws and regulations, professional spirit, organizational coordination ability, and autonomous learning ability in Table 1 is a realistic performance. The data is obtained by statistically analyzing the frequency ratio of the corresponding mutual evaluation report results using formula (2); for physical condition, the health level given in the physical examination report is used to determine it. It belongs to a type of qualitative classification data, which can be described as {(B, α 2,1 ),(S,α 2,2 ),(H,α 2,3 )}, where α 2,m= 1 indicates that the physical condition belongs to the mth level. For psychological status, similar to a physical examination report, it primarily examines related abnormalities such as depression. Considering the robustness and validity of the assessment results, the results of the past six months are used as the basis. For quantitative indicators such as information literacy, technology application, professional skills, professional knowledge, and innovative thinking, their reference values ​​are shown in Table 3 below. Furthermore, the reference values ​​for the first-level indicators of "weak (W)", "qualified (S)", "average (N)", and "strong (ST)" are given as 0, 0.4, 0.7, and 1, respectively. The initial weight of each level of indicator is given as 1.

[0119] Table 3 Reference values ​​of secondary indicators

[0120]

[0121]

[0122] To ensure the scientific nature of the research, the aforementioned 148 data sets were divided and uniformly randomized to produce 74 training and 74 test sets. Based on the training data, a statistical analysis was conducted on the personnel capability and quality coupling coordination index of the secondary indicator scores at different reference levels to determine their frequencies at different levels. Taking into account the bias of historical data samples and further fine-tuning the above frequency distribution based on expert experience, the final frequency distribution was used as the initial parameter for the rule conclusion, as shown in Table 4. Since the relative importance of each initial rule was determined to be consistent during the determination process, the initial weight of each rule was set to 1.

[0123] Table 4 Initial model parameters

[0124]

[0125]

[0126] Optimizing initial model parameters. The initial evaluation model described above was constructed based on a combination of expert experience and historical data. However, due to the limitations of the historical data analysis process and expert judgment, the constructed initial model is insufficient to provide reliable evaluation results. Therefore, further optimization and adjustment based on training data samples is necessary.

[0127] The optimization objective function is constructed based on formulas (16)-(20), where the model output is the capability-quality coupling coordination index. The P-CMA-ES algorithm is used for parameter optimization. The maximum number of iterations is set to 100, the initial step size is set to 0.27, and the population size is set to 23. The optimized model parameters are shown in Table 5. The optimized first-level indicator weights δ1, δ2, δ3, and δ4 are equal to 0.79, 0.52, 0.50, and 0.96, respectively. The weights of the 13 second-level indicators δ 1,1, δ 1,2 , δ 1,3 , δ 1,4 , δ 2,1 , δ 2,2 , δ 3,1 , δ 3,2 , δ 4,1 , δ 4,2 , δ 4,3 , δ 4,4 and δ 4,5 They are 0.89, 0.93, 0.74, 0.93, 0.96, 0.87, 0.78, 0.76, 0.78, 0.83, 0.72, 0.56 and 0.91 respectively.

[0128] Table 5 Optimized model parameters

[0129]

[0130]

[0131] Ability and quality coupling coordination evaluation. To facilitate subsequent discussion, the initial evaluation model constructed is denoted as NBRB-CC0, and the optimized model is denoted as NBRB-CC1. The above evaluation models are used to evaluate the ability and quality coupling coordination of the test set data. The evaluation results are shown in the figure below. Figure 3 As shown. Combined Figure 3 As can be seen from Table 2, the coupling coordination index of the test set samples is all above 0.3, that is, they are all above the slight imbalance range. This is because the industrial group has strict standards in the recruitment and internship placement stages, and has carried out early diversion of personnel with serious imbalances in ability and quality development, and retained and trained some personnel with slight imbalances in ability and quality development.

[0132] The modeling errors of the initial model NBRB-CC0 and the optimized model NBRB-CC1 are 0.0052 and 0.0004 respectively. It can be seen that the modeling performance of NBRB-CC1 has been significantly improved compared with the initial model. Figure 3 Analysis shows that the Coupling Coordination Index for data samples 19 and 20 exceeds 0.9, indicating that their competencies are highly coordinated. This indicates that the various competency elements within the employee are mutually reinforcing and highly balanced. Both NBRB-CC0 and NBRB-CC1 perform well in assessing highly coordinated competencies. This is because highly coordinated individuals are more typical and meet employer requirements, and the expert experience embedded in the initial model is more accurate. Furthermore, a comparison of the evaluation results for data samples 19 and 20 using the initial and optimized models shows that the model optimization process supplements expert experience and can correct the accuracy of the evaluation results.

[0133] For data samples in the moderate and basic coordination ranges, NBRB-CC1's assessment results are significantly more accurate than NBRB-CC0's. For example, for data sample number 51, the NBRB-CC0 assessment result is close to 0.65, which is in the basic coordination zone, while the NBRB-CC1 and actual results are both 0.75, which is in the moderate coordination zone. Because the assessment results of the ability-quality coupling coordination are an important basis for the group to formulate talent management mechanisms, errors in such assessment results may reduce the efficiency of resource allocation and utilization. For data sample number 40, the NBRB-CC1 assessment result is 0.5, which is right on the edge of the slight misalignment range. This is due to the lack of accumulated expert experience and data samples for such cases. As experience and data increase, the accuracy of NBRB-CC1 can be further improved.

[0134] Comparative Study. To further demonstrate the effectiveness of the proposed method, this section conducts a comparative study using a backpropagation neural network (BPNN), an extreme learning machine (ELM), and a fuzzy rule system (FRS). Qualitative indicators of mental and physical states are labeled by category. These models utilize the same training and test sets and optimization algorithms as the previous study. Ten rounds of repeated experiments were conducted based on each of these models. The mean and variance of the mean squared error (MSE) were used to characterize the modeling performance, as shown in Table 6.

[0135] Table 6 Comparison of modeling performance of each model

[0136] Model Name Minimum error Maximum error Mean error Error variance NBRB-CC1 0.0003 0.0006 0.0005 1.46E-08 BPNN 0.0009 0.0041 0.0027 7.80E-07 ELM 0.0016 0.0029 0.0022 2.44E-07 FRS 0.0021 0.0033 0.0026 1.26E-07

[0137] Combined with Table 6, it can be seen that the mean and variance of the modeling error of NBRB-CC1 are smaller than those of the other three models. On the one hand, this shows that the proposed method has better modeling performance in an average sense. On the other hand, since the initial model of NBRB-CC1 is given by expert experience, while the initial parameters of the other three models are randomly generated, NBRB-CC1 has better robustness during the optimization process.

[0138] In the aforementioned comparative study, the training set accounted for 50% of the entire dataset. To illustrate the relationship between varying the training set proportion and model performance, repeated experiments were conducted with training set proportions of 60%, 30%, and 10%. The mean model output errors are shown in Table 7. The analysis shows that when the training set proportion is low, the data-driven model parameters are not fully trained. However, the NBRB-CC model achieves lower modeling errors because its initial parameters are determined through expert experience, a process similar to parameter pre-training. When the training set proportion is high, i.e., the training data available is more abundant, the data-driven model achieves similar modeling performance to the proposed model. This analysis demonstrates that the proposed model has better engineering adaptability.

[0139] Table 7 Comparison of modeling performance of each model with different proportions of training sets

[0140] Model Name 10% 30% 50% 60% NBRB-CC1 0.0042 0.0021 0.0005 0.0004 BPNN 0.0746 0.0081 0.0027 0.0005 ELM 0.0933 0.0098 0.0022 0.0009 FRS 0.0990 0.0087 0.0026 0.0011

[0141] Conclusion. This paper addresses the challenge of assessing the competence and quality of skilled personnel, constructs an evaluation index system and an index conversion method for the competence and quality of skilled personnel, and proposes a competence and quality coupling coordination evaluation method based on the NBRB model, which can effectively utilize expert experience and limited historical data. Case studies show that the proposed method has better modeling accuracy and optimization robustness than traditional data-driven methods under the condition of limited data samples. The assessment of personnel competence and quality is a complex system engineering problem, and its ultimate goal is to assist in optimizing management. Therefore, it is necessary to explore the connection between employer policy formulation and the coupling coordination of personnel competence and quality in further research to improve the quality and effectiveness of policy formulation and application.

[0142] Example 2

[0143] Embodiment 2 provides a coupling coordination evaluation system for the ability and quality of skilled post personnel, which includes an indicator system module, an initial data conversion module, an indicator model solution module, a coupling coordination evaluation module and a result optimization module.

[0144] The indicator system module is used to objectively measure the performance of personnel in skilled positions in terms of ideological and political quality, physical and mental quality, scientific and technological quality, and work ability.

[0145] The initial data conversion module is used to propose different conversion methods for different types of information and complete the conversion of the information collected in the indicator system into the form of confidence distribution.

[0146] The indicator model solving module is used to determine the reference value, confidence level, rule weight, etc. of the indicator system, and evaluate the comprehensive level of skilled personnel through reasoning fusion.

[0147] The coupling coordination module is used to obtain the coupling coordination index of each indicator in the indicator model to describe the coupling coordination of the ability and quality of skilled position personnel.

[0148] The result optimization module is used to construct optimization objectives and constraints, and to minimize the error between the evaluation results output by the model and the actual results obtained through the complex process of "evaluation-feedback-group decision-making" by adjusting the model parameters.

[0149] It should be noted that the indicator system module, initial data conversion module, indicator model solution module, coupling coordination evaluation module and result optimization module in the present invention are all implemented using the method in Example 1.

[0150] Example 3

[0151] Embodiment 3 provides an electronic device, which includes at least one processor; and a memory communicatively connected to the processor; wherein the memory stores instructions that can be executed by the processor, and the instructions are executed by the processor to enable the processor to execute the evaluation method described in embodiment 1.

[0152] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the coupling coordination of the ability and quality of skilled post personnel, characterized by: The following steps are involved: S1: Construct an evaluation index system for the competence and quality of skilled personnel; S2: Complete the transformation of the initial data information corresponding to the evaluation index system into the form of confidence distribution; S3: Based on the evaluation index system established in step S1 and the naive belief rule base modeling method, the data converted in step S2 is used to build an initial evaluation model, and the secondary indicators in the evaluation index system are evaluated; S4: Based on the evaluation results of the secondary indicators in step S3, conduct a preliminary evaluation of the coupling coordination of personnel capabilities and qualities; S5: Optimize the preliminary evaluation result in step S4 to obtain the final evaluation result.

2. The method for evaluating the coupling coordination of the ability and quality of skilled post personnel according to claim 1 is characterized in that: The evaluation index system in S1 includes four first-level indicators: ideological and political quality, physical and mental quality, scientific and technological quality, and work ability; The secondary indicators corresponding to ideological and political qualities include: political literacy, moral quality, compliance with laws and regulations, and professional spirit; the secondary indicators corresponding to physical and mental qualities include: physical condition and psychological state; the secondary indicators corresponding to scientific and technological qualities include: information literacy and application of science and technology; the secondary indicators corresponding to work ability include: professional skills level, professional knowledge level, organizational and coordination ability, innovative thinking ability, and independent learning ability.

3. The method for evaluating the coupling coordination of the ability and quality of skilled post personnel according to claim 2 is characterized in that: The confidence distribution form described in S2 is expressed as: S(x i )={(A i,j ,α i,j ),i=1,2,...,M;j=1,2,...,J}; Among them, S() is the information conversion function, {(A 1,1 ,α 1,1 ),...,(A M,J ,α M,J )} represents the transformed confidence distribution, A i,j is the index x i The jth reference level, α i,j is the index x i The matching degree of the jth reference level, j is the number of reference levels, M i (i=1,...,T) represents x i The number of reference levels is , and J represents the number of sub-indicators to be analyzed.

4. The method for evaluating the coupling coordination of the ability and quality of skilled personnel according to claim 3 is characterized in that: The specific operations of S3 include: S301: Construct an initial assessment model for the competence and quality of skilled personnel based on S1 and S2; S302: solving the secondary indicators through information conversion, rule activation and rule fusion; S303: Solve the primary indicator by fusing the secondary indicator information.

5. The method for evaluating the coupling coordination of the ability and quality of skilled post personnel according to claim 4 is characterized in that: The solution formula for the secondary index in S302 is: Where 0≤w i,m ≤1 represents the activation weight of the i-th indicator input information on the m-th rule, L=M i (i=1,...,T) is the total number of rules, Indicates the confidence of the nth output conclusion in the output conclusion of the i-th model, N is the number of confidence levels of the output conclusions, represents the unknown degree of the mth rule of the i-th NBRB model, β n,i,m (n=1,...,N) represents the confidence of the nth output conclusion.

6. The method for evaluating the coupling coordination of the ability and quality of skilled post personnel according to claim 5 is characterized in that: The specific operations of S4 include: S401: Based on S303, a quantitative model for coupling personnel capabilities and qualities is constructed as follows: Where Co∈[0,1] represents the coupling index, Y j (j=1, 2, 3, 4) represents the utility value of the comprehensive level evaluation results of skill position ability and quality; S402: Calculating a coupling coordination index based on S401; Where De∈[0,1] represents the coupling coordination index, δ j ′ is the normalized indicator weight.

7. The method for evaluating the coupling coordination of the ability and quality of skilled personnel according to claim 6 is characterized in that: The specific operation in S5 includes the following steps: S501: Constructing the optimization goal of coupling and coordination of the ability and quality of skilled post personnel; Where Ω represents the parameter vector to be optimized, Ξ(Ω) represents the error between the model output and the actual value, Represents the output utility value of the model, O represents the actual evaluation result utility value, δ represents the indicator weight, θ represents the rule weight, and β represents the rule confidence; S502: Determine the constraints of the optimization target in step S501; S503: Solve the optimization objective in step S501 using a covariance matrix adaptive evolutionary strategy with a projection operation.

8. The method for evaluating the coupling coordination of the ability and quality of skilled post personnel according to claim 7 is characterized in that: The constraints in S502 are: 0≤θ i,m ≤1; 0≤δ i ≤1; 0≤β n,i,m ≤1； 9. A system for evaluating the coupling and coordination of the competence and quality of skilled personnel, characterized by: The evaluation system includes an indicator system module, an initial data conversion module, an indicator model solution module, a coupling coordination evaluation module and a result optimization module; The indicator system module is used to objectively measure the performance of skilled personnel in terms of ideological and political quality, physical and mental quality, scientific and technological quality, and work ability; The initial data conversion module is used to propose different conversion methods for different types of information and complete the conversion of the information collected in the indicator system into a confidence distribution form; The indicator model solving module is used to determine the reference value, confidence level, rule weight, etc. of the indicator system, and evaluate the comprehensive level of skilled personnel through reasoning fusion; The coupling coordination module is used to obtain the coupling coordination index of each indicator in the indicator model to describe the coupling coordination of the ability and quality of skilled position personnel; The indicator system module, initial data conversion module, indicator model solution module, coupling coordination evaluation module and result optimization module are all implemented using the evaluation method described in any one of claims 1-8.

10. An electronic device, characterized in that: The electronic device includes at least one processor; and a memory communicatively connected to the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor to enable the processor to perform the evaluation method described in any one of claims 1 to 8.

Images