Cooperative task efficiency evaluation method for manned ship and unmanned ship on water surface

Abstract

The invention discloses a cooperative task efficiency evaluation method for a manned ship and an unmanned ship on the water surface. The method comprises the following steps: determining an initial weight of each level index in an evaluation index system based on a fuzzy analytic hierarchy process; constructing a mutual influence relation matrix of the first-level indexes through a decision-making laboratory analysis method, and calculating an influence weight; fusing the initial weight and the influence weight in a combined weighting mode to obtain a comprehensive weight; and calculating a cooperative task efficiency evaluation value based on the comprehensive weight. According to the method, the relative importance and the mutual influence degree of each factor in the collaborative task can be accurately reflected, and the processing capability of the model on uncertainty and fuzziness is enhanced, so that the weight calculation result has higher practical adaptability and decision reference value.

Description

Technical Field

[0001] This invention belongs to the field of artificial intelligence technology, and in particular relates to a method for evaluating the collaborative task performance of manned and unmanned surface vessels. Background Technology

[0002] In recent years, the rapid advancement of artificial intelligence technology has endowed machines with human-like "intelligence." This development has driven the widespread application of various intelligent unmanned systems, particularly in modern warfare, providing more new tactical options and operational possibilities. With shifts in strategic orientation and technological progress, global competition for marine resources is intensifying. Maritime unmanned systems, including drones, unmanned surface vessels (USVs), wave gliders, and unmanned underwater vehicles (UUVs), are becoming key technologies with ever-expanding applications. USVs can perform specific tasks within designated sea areas by configuring different underwater acoustic and optical sensors, such as port patrols, maritime search and rescue, situational awareness, and underwater detection. Since the 21st century, countries worldwide have accelerated the research and development of small and medium-sized unmanned platforms, aiming to create advanced systems with high mobility, high efficiency, and the ability to handle high-risk missions. Although a single USV has limitations in independently executing missions, increasing the number of USVs and adopting swarm operations can fully leverage the advantages of individual USVs and swarm collaboration, thereby significantly improving combat effectiveness and even enabling them to achieve significant results with limited resources, such as performing complex anti-ship missions. In maritime operations and mission execution, the unpredictability of the environment and the complexity of the missions pose severe challenges to unmanned systems. Currently, while intelligent unmanned systems designed based on expert experience can undertake many tasks, they still have limitations in dealing with complex and ever-changing maritime situations and cannot completely replace the crucial role of operators on manned vessels. Therefore, how to deeply integrate the advantages of manned and unmanned vessels to achieve collaborative operations has become a key research direction for improving overall combat effectiveness and overcoming current limitations. Summary of the Invention

[0003] To address the aforementioned technical issues, this invention proposes a method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels. This method accurately reflects the relative importance and mutual influence of various factors during combat operations, enhances the model's ability to handle uncertainty and ambiguity, and thus makes the weight calculation results more realistically adaptable and valuable for decision-making.

[0004] To achieve the above objectives, the present invention provides a method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels, comprising:

[0005] The initial weights of each level of indicators in the evaluation index system are determined based on the fuzzy hierarchical analysis method.

[0006] A matrix of mutual influence relationships among primary indicators was constructed using the decision laboratory analysis method, and the influence weights were calculated.

[0007] The initial weights and the influence weights are combined using a combined weighting method to obtain a comprehensive weight;

[0008] The collaborative task effectiveness evaluation value is calculated based on the comprehensive weights.

[0009] Optionally, the evaluation index system includes a target layer, a criterion layer, and an index layer;

[0010] The target layer is used to display overall task metrics;

[0011] The criteria layer is used to decompose the overall task into stages and divide it into human-machine collaboration dimensions.

[0012] The indicator layer is used to establish key indicators under each criterion layer and to calculate key indicators from basic values.

[0013] Optionally, determining the initial weights of indicators at each level in the evaluation indicator system includes:

[0014] A fuzzy judgment matrix is ​​constructed using a fuzzy numerical scale, and the initial weights of each level of indicators are calculated based on the fuzzy judgment matrix through row sum normalization.

[0015] Optionally, constructing the mutual influence matrix of primary indicators includes:

[0016] The direct influence relationship between primary indicators is determined by expert scoring, and the degree of influence is quantified using a 0-3 scale. The direct influence matrix is ​​then normalized to generate the mutual influence matrix of primary indicators.

[0017] Optionally, the calculation of influence weights includes:

[0018]

[0019] Where, d i The influence degree can be obtained by summing the rows of each row in the mutual influence matrix; c i The degree of influence can be obtained by summing each column of the mutual influence matrix.

[0020] Optionally, obtaining the overall weight includes:

[0021]

[0022] Among them, u i g is the weight vector established between various performance indicators. i h is the weight vector used for verification in fuzzy hierarchical analysis. i The combined weight vector after weighting the combination.

[0023] Optionally, the rules for the 0-3 scale are as follows: 0 represents no effect; 1 represents a weak effect; 2 represents a moderate effect; and 3 represents a strong effect.

[0024] Optionally, the fuzzy judgment matrix is ​​constructed using fuzzy scaling rules: a scale of 0.5 indicates that the indicators are equally important; a scale of 0.6-0.9 indicates that the importance increases progressively; and a scale of 0.1-0.4 indicates that the importance is in reverse order and complementary.

[0025] Technical effects of this invention: This invention discloses a method for evaluating the collaborative task effectiveness of manned and unmanned surface vessels. The method uses the FAHP approach to determine the initial weights of each primary and secondary evaluation indicator, and clarifies the importance of different indicators within the indicator system based on expert scores. The Dematel method is used to calculate the weights of the mutual influence between primary indicators, taking into account the complexity of the interaction between multiple indicator factors. A combined weighting method is employed to integrate the weights of both, improving the scientific rigor and rationality of the weight calculation. Attached Figure Description

[0026] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0027] Figure 1 This is a flowchart illustrating a method for evaluating the collaborative task effectiveness of manned and unmanned surface vessels according to an embodiment of the present invention.

[0028] Figure 2 This is a schematic diagram of the evaluation index system for the collaborative combat effectiveness of manned and unmanned ships according to an embodiment of the present invention;

[0029] Figure 3 This is a directed graph showing the mutual influence of factors in an embodiment of the present invention. Detailed Implementation

[0030] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0031] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0032] like Figure 1 As shown, this embodiment provides a method for evaluating the collaborative task effectiveness of manned and unmanned surface vessels, including:

[0033] The DEMATEL method is used to analyze the causal relationships among various influencing factors.

[0034] The initial weights of each factor were determined by using the FAHP method to conduct hierarchical analysis on the evaluation indicators.

[0035] A causal relationship matrix of influencing factors was constructed using the DEMATEL method to analyze the degree of mutual influence among the factors;

[0036] By combining and assigning weights to the criteria layer, and integrating causal relationships with weights, a comprehensive evaluation value for the effectiveness of collaborative operations is derived.

[0037] Furthermore, the initial weights of the indicators are calculated based on FAHP:

[0038] Establish a multi-level hierarchical structure:

[0039] First, a multi-level evaluation structure is established. Generally, a hierarchical structure is divided into three layers. This invention uses a multi-level division to assess the effectiveness of manned / unmanned surface vessel collaborative operations. The first layer is the objective layer, i.e., the overall mission indicators; the second layer is the criterion layer, which decomposes the overall mission into stages and divides it into human-machine collaboration dimensions; the third layer is the indicator layer, which establishes the key indicators under each criterion layer and calculates these key indicators from basic numerical values. The second and third layer hierarchical structures of the mission stage decomposition include four criterion layers and 14 underlying indicators, which are summarized here as follows: Figure 2 As shown.

[0040] Constructing a fuzzy judgment matrix:

[0041] After constructing the hierarchical analysis model, pairwise comparisons of factors at each level of the model are necessary based on the feedback from the expert questionnaire. The core of this process lies in assessing the importance of each factor relative to its superior factor, thereby determining their relative priority. Experts, by judging the importance of each factor and combining this with pre-set scoring criteria (as shown in Table 1), perform fuzzy quantification on each pair of factors, thus constructing a fuzzy judgment matrix A. The elements a of fuzzy matrix A... ij This represents the fuzzy evaluation value of the importance of factor i relative to factor j by experts. If 0 ≤ a ij If A ≤ 1 (i,j = 1,2,...,n), then A is called a fuzzy matrix. And for a fuzzy consistent matrix:

[0042] ①For There is a ij +a ji =1.

[0043] ②For There is a ii =0.5.

[0044] Table 1

[0045]

[0046] Based on the table above, after rating the importance of each factor, the judgment matrix A can be represented as:

[0047]

[0048] Calculate the weight vector based on the judgment matrix:

[0049] The weight vector W = (W1, W2, ..., Wn) of the judgment matrix A is calculated using the row sum normalization method. n ) T W i Vectors satisfy:

[0050]

[0051] in Let be the sum of the values ​​of the elements in the i-th row, and satisfy . W i ≥0, (i=1,2,...,n).

[0052] Calculate the characteristic matrix W * :

[0053] After calculating each weight vector, the feature matrix needs to be calculated for subsequent consistency checks. Let W be the feature judgment matrix of A. * Then W * The calculation formula is shown in equation (3):

[0054] W * =(W ij ) n×n (3);

[0055] Among them W ij The calculation formula is:

[0056]

[0057] Consistency check:

[0058] The fuzzy judgment matrix A and its characteristic matrix W are compared. * Calculate the compatibility index I(A,W) * If I(A,W) * If α ≤ 0, the consistency test is considered passed. The value of α indicates the degree of consistency required by the decision-maker for the fuzzy judgment matrix. Usually, α is taken as 0.1. The calculation of the compatibility index I is shown in equation (5).

[0059]

[0060] Analysis of experimental results:

[0061] This embodiment uses a questionnaire survey to evaluate the importance of each indicator relative to the previous level. The average of the expert evaluations is used to establish a fuzzy judgment matrix based on the questionnaire data. In this embodiment, after calculating the results at each level, each value is rounded to three decimal places. The subjective weights of FAHP are calculated according to the weight calculation formula (2) as follows:

[0062] Determining the weights of primary indicators:

[0063] like Figure 1 As shown in Table 2, the primary indicators “reconnaissance and search”, “tracking and surveillance”, “strike and encirclement” and “human-machine collaboration” under the evaluation system of manned / unmanned ship collaborative combat effectiveness are represented by A1, A2, A3 and A4 respectively. The results calculated based on the questionnaire are shown in Table 2.

[0064] Table 2

[0065]

[0066] Based on the weight vectors of these indicators, the feature matrix can be calculated using equation (3). As shown in equation (6),

[0067]

[0068] Furthermore, according to formula (5), the compatibility index I1 is calculated to be 0.03. This value is less than the α value of 0.1 for consistency test, so the weight value is valid and the consistency test is passed.

[0069] Determining the weights of secondary indicators:

[0070] In the reconnaissance and search assessment, the four indicators "instantaneous coverage rate," "cycle coverage time," "search time interval," and "blind spot time" are defined as secondary indicators, denoted by B1, B2, B3, and B4, respectively. The judgment matrix for these indicators is represented based on the questionnaire results, and the calculation results are shown in Table 4.3. Simultaneously, the weights of each indicator are calculated based on the judgment matrix, as shown in the weight column of Table 3.

[0071] Table 3

[0072]

[0073] Based on the index weights in Table 3 and the calculation using equation (3), the feature matrix can be obtained. for:

[0074]

[0075] Further calculation using equation (5) shows that the compatibility index I2 is 0.0193 < 0.1, so the consistency test of this weight is passed.

[0076] In the assessment at the tracking and monitoring level, the four indicators "target detection rate", "target identification rate", "successful tracking efficiency" and "effective tracking time" are defined as secondary indicators, represented by B5, B6, B7 and B8 respectively; the expert scores are displayed and their weights are calculated using formula (2) as shown in Table 4.

[0077] Table 4

[0078]

[0079] Based on the weight vectors of these indicators, the feature matrix can be calculated using equation (3). As shown in equation (8),

[0080]

[0081] Furthermore, the compatibility index I3 calculated according to formula (5) is 0.0246, and this value is less than 0.1. Therefore, the consistency test shows that the weight has good consistency.

[0082] The secondary indicators for combating encirclement and capture are "formation formation efficiency", "escape prevention efficiency", "hit efficiency" and "damage to the fleet", denoted by B9, B10, B11 and B12. The expert scores and weight calculation results are shown in Table 5.

[0083] Table 5

[0084]

[0085] Based on the weight vector values ​​in the table and formula (3), the feature matrix can be calculated. for:

[0086]

[0087] The compatibility index I4 is calculated to be 0.0133, which is less than 0.1. Therefore, the consistency test of the weight values ​​in the weight table is passed, and the weight values ​​are valid.

[0088] The secondary indicators of human-machine collaboration are identified as “abnormal handling time” and “number of personnel”, denoted as B13 and B14 respectively. The results of the weight calculation of the expert scores and Equation (2) are shown in Table 6.

[0089] Table 6

[0090]

[0091] Based on the weight vector table and equation (3), the feature matrix can be calculated. for:

[0092]

[0093] Finally, according to formula (5), the compatibility index is calculated to be I5 = 0.04 < 0.1, so the consistency test is passed.

[0094] Furthermore, the influence weights between indicators are calculated based on Dematel:

[0095] Decision-making laboratory method:

[0096] like Figure 3 The diagram shows a directed graph structure illustrating the interrelationships of factors, reflecting the causal flow and degree of influence among elements in the system. In the DEMATEL method, the entire causal modeling process can be divided into the following steps:

[0097] Define the elements that influence the relationship:

[0098] The Dematel method primarily employs graph theory to determine the influence relationships between indicators, such as... Figure 3 As shown, arrows represent the mutual influence relationships between indicators. If indicator i influences indicator j, a directed arrow from i to j is used. Furthermore, to more clearly represent the magnitude of the influence between indicators, a 0-3 scale is used to represent the influence relationships between elements in the system, where 0 represents no influence between elements, 1 represents a slight influence, 2 represents a significant influence, and 3 represents a very large influence. For example... Figure 3 The numbers indicated by the middle arrows are summarized in Table 7.

[0099] Table 7

[0100]

[0101] Determine the influence matrix:

[0102] Experts determine the pairwise influence relationships between all elements in the system, forming a direct relationship matrix denoted by P. If a system has n indicators, the direct relationship matrix is ​​an n×n matrix, with the diagonal representing the degree of influence of each indicator on itself, denoted by 0, as shown in equation (11). ij This indicates the degree of influence of indicator i on indicator j;

[0103]

[0104] Normalized influence matrix:

[0105] After completing the influence relationship matrix, in order to ensure that the influence relationships between the factors are compared under a unified dimension, the direct relationship matrix P needs to be normalized. That is, the normalized relationship matrix P′ can be obtained by multiplying the elements in the influence relationship matrix by the parameter γ, as shown in equation (12). The parameter setting formula is (13).

[0106] P′=γP(12);

[0107]

[0108] Establish the mutual influence matrix T:

[0109] After completing the normalized matrix, in order to further reflect the direct and indirect influence relationships between various factors in the system, a comprehensive influence matrix T needs to be constructed, the specific formula of which is as follows:

[0110] T = P′(EP′) -1 (14);

[0111] Where E is the identity matrix, as shown in equation (15).

[0112]

[0113] Calculate the degree of influence and the degree of being influenced:

[0114] Influence is typically represented by a matrix, where rows represent the influenced factors and columns represent the influencing factors. The influence d is obtained by summing the rows of matrix T. i Summing each column of matrix T yields the degree of influence c. i As shown in equations (16) and (17);

[0115]

[0116] Calculate centrality and causality:

[0117] Centrality M i The centrality value is calculated as shown in equation (18). It is used to measure the importance of a factor in a system and the significance of its effect. The higher the centrality value, the more significant the influence of the factor on other factors in the system, and it can usually be considered a major factor. Causality R i The calculation formula is shown in equation (19). A causal degree greater than 0 indicates that the factor has a strong influence on other factors and is a causal factor; a causal degree less than 0 indicates that the factor is more easily affected by other factors and is an outcome factor.

[0118] M i =d i +c i ,i=1,2,3,...,n(18);

[0119] R i =d i -c i ,i=1,2,3,...,n(19).

[0120] Calculate the overall impact weight:

[0121] In DEMATE1, there are generally two methods for calculating the overall weight. One method is to calculate the overall influence weight of each weight indicator based on centrality, and the other method is to calculate the overall weight based on influence and affectedness.

[0122] The centrality is used to calculate the comprehensive weight W2. By calculating the sum of the influence degree and the degree of being influenced, the influence degree and the degree of being influenced are considered together to measure the importance of the factor, reflecting the core position of a factor in the whole system. The importance of the comprehensive influence is considered in a more balanced way. The calculation formula (20) is as follows:

[0123]

[0124] The comprehensive weight W2 is calculated using the degree of influence and the degree of being influenced. By multiplying the degree of influence and the degree of being influenced, the comprehensive role of each factor in the system is reflected from both the degree of influence and the degree of being influenced, highlighting the interaction between factors. The calculation formula (21) is as follows:

[0125]

[0126] In this embodiment, since the FAHP method has already assessed and judged the importance of each factor, while the DEMATEL method focuses more on the degree of mutual influence between each factor, the present invention uses the product of influence degree and affected degree to calculate the comprehensive influence weight, that is, uses formula (21) to calculate the comprehensive weight.

[0127] Analysis of experimental results:

[0128] This embodiment analyzes the influence between various indicators in the indicator system. Since the mutual influence between all secondary indicators is weak or the relationships are complex and difficult to determine, and the importance of primary indicators in influencing decision-making is greater than that of secondary indicators, this invention only judges the mutual influence between primary indicators and combines it with the weights of the fuzzy hierarchical analysis method. A questionnaire was still distributed to 15 experts for survey, and the average value was used to construct the direct influence matrix and calculate subsequent formulas. The final result for each stage was taken to three decimal places. The results show that the direct influence matrix is ​​as follows:

[0129]

[0130] Based on the direct influence matrix, the row sum and maximum value are calculated, and the direct influence matrix is ​​normalized according to equation (12). The results are as follows:

[0131]

[0132] And according to equation (14), the mutual influence matrix T is calculated as follows:

[0133]

[0134] The influence degree d between each performance index is calculated based on the comprehensive influence matrix T and equations (16) and (17). i With the degree of influence c i Furthermore, the centrality M is calculated based on the degree of influence and the degree of being influenced among various performance indicators. i With causal degree R i The calculation results are shown in Table 8. According to Table 8 and Equation (21), the comprehensive influence weights of each indicator in the criterion layer are calculated as [0.248, 0.278, 0.272, 0.202].

[0135] Table 8

[0136]

[0137] Furthermore, the combined weighting method:

[0138] The combination of FAHP and DEMATEL methods aims to achieve more comprehensive, accurate, and effective results in decision analysis. FAHP excels at handling fuzzy and uncertain information, revealing the relative importance of each factor through a fuzzy comparison matrix. However, FAHP relies on the independence of influencing factors, which can be difficult to accurately represent in complex systems, such as manned / unmanned surface vessel collaborative combat effectiveness assessment systems, potentially leading to biases. DEMATEL, on the other hand, reflects the degree of mutual influence between factors, incorporating the degree of influence into the weight calculation to arrive at a more comprehensive overall weight. This method, to some extent, compensates for FAHP's shortcomings in handling inter-factor correlations and reduces the influence of expert subjectivity on key indicators. Therefore, the combination of FAHP and DEMATEL integrates the advantages of both methods, more comprehensively and accurately addressing complex, fuzzy, and interrelated decision problems, thus improving the effectiveness and credibility of decisions.

[0139] In this embodiment, the fuzzy hierarchical analysis method is used to determine the initial weight values ​​of the indicators at each level. Then, the DEMATEL method is used to modify and optimize the weight values ​​between the first-level indicators, thereby ensuring the objective, scientific, and accurate assignment of weights. Let u be the weight vector established by the FAHP method between the various performance indicators. iThe weight vector g confirmed by the DEMATEL method i The combined weight vector after weighting is h. i The formula for calculating the overall weight is as follows:

[0140]

[0141] As shown in Table 2, the initial weight values ​​determined using the fuzzy analytic hierarchy process are:

[0142] u i = [0.274, 0.245, 0.239, 0.242](26);

[0143] As shown in Table 8, the weight values ​​determined using the decision laboratory method are:

[0144] g i = [0.248, 0.278, 0.272, 0.202](27);

[0145] Combining equation (25) to comprehensively calculate the effectiveness weight values ​​of the two evaluation methods, the comprehensive weight vector h of the primary indicator is then obtained. i The value is [0.272, 0.272, 0.26, 0.196].

[0146] This invention discloses a method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels. The method uses the FAHP method to determine the initial weights of each primary and secondary evaluation indicator, and uses expert scores to clarify the importance of different indicators in the indicator system. The method uses the DEMATEL method to calculate the weights of the mutual influence between the primary indicators, taking into account the complexity of the interaction between multiple indicator factors. The method uses a combined weighting approach to integrate the weights of both, which improves the scientificity and rationality of the weight calculation.

[0147] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for evaluating the collaborative task effectiveness of manned and unmanned surface vessels, characterized in that, include: The initial weights of each level of indicators in the evaluation index system are determined based on the fuzzy hierarchical analysis method. A matrix of mutual influence relationships among primary indicators was constructed using the decision laboratory analysis method, and the influence weights were calculated. The initial weights and the influence weights are combined using a combined weighting method to obtain a comprehensive weight; The collaborative task effectiveness evaluation value is calculated based on the comprehensive weights.

2. The method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels as described in claim 1, characterized in that, The evaluation index system includes an objective layer, a criterion layer, and an indicator layer; The target layer is used to display overall task metrics; The criteria layer is used to decompose the overall task into stages and divide it into human-machine collaboration dimensions. The indicator layer is used to establish key indicators under each criterion layer and to calculate key indicators from basic values.

3. The method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels as described in claim 1, characterized in that, Determining the initial weights of indicators at each level in the evaluation indicator system includes: A fuzzy judgment matrix is ​​constructed using a fuzzy numerical scale, and the initial weights of each level of indicators are calculated based on the fuzzy judgment matrix through row sum normalization.

4. The method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels as described in claim 1, characterized in that, Constructing the mutual influence matrix of primary indicators includes: The direct influence relationship between primary indicators is determined by expert scoring, and the degree of influence is quantified using a 0-3 scale. The direct influence matrix is ​​then normalized to generate the mutual influence matrix of primary indicators.

5. The method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels as described in claim 1, characterized in that, The calculation of influence weights includes: Where, d i The influence degree can be obtained by summing the rows of each row in the mutual influence matrix; c i The degree of influence can be obtained by summing each column of the mutual influence matrix.

6. The method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels as described in claim 1, characterized in that, The overall weighting includes: Among them, u i g is the weight vector established between various performance indicators. i h is the weight vector used for verification in fuzzy hierarchical analysis. i The combined weight vector after weighting the combination.

7. The method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels as described in claim 4, characterized in that, The rules for the 0-3 scale are as follows: 0 represents no effect; 1 represents a weak effect; 2 represents a moderate effect; and 3 represents a strong effect.

8. The method for evaluating the collaborative mission effectiveness of manned and unmanned surface vessels as described in claim 3, characterized in that, The fuzzy judgment matrix is ​​constructed using fuzzy scaling rules: a scale of 0.5 indicates that the indicators are equally important; a scale of 0.6-0.9 indicates that the importance increases progressively; and a scale of 0.1-0.4 indicates that the importance is in reverse order and complementary.

Images