Method for setting optimal similarity value range of nuclear power information chunks

Abstract

The invention discloses a nuclear power information block optimal similar value range setting method, which comprises the following steps of: (1) setting the composition of a nuclear power information block, and defining the object characteristics and the operation process of the information block; step (2), aiming at the feature description of the nuclear power information chunks, introducing similarity feature setting elements, and extracting similarity feature elements of the chunks; step (3), obtaining a regression model calculation equation of the overall similarity value of the nuclear power block through the semantic difference questionnaire scale; and step (4), according to similar characteristics contained in a regression model calculation equation of the overall similar values of the nuclear power blocks, designing a plurality of sets of nuclear power interface blocks with different similar values as experimental materials, and performing a behavior experiment of a scene task normal form to obtain a similar value range of optimal performance with significant difference. According to the method, the similarity influence factors of the nuclear power information chunks can be quantified, and the optimal similarity range is determined, so that operation errors caused by sensing similarity deviation of nuclear power operators are reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of interface design and relates to a method for setting the optimal similarity value range of nuclear power information chunks. Background Art

[0002] The nuclear power monitoring system plays an important role in monitoring and controlling as the center of a nuclear power plant. In this system, icons and interfaces form nuclear power information chunks, which become the main visual targets. The nuclear power information chunks have relatively single relevance at the functional level, layout structure, and type of design elements, resulting in convergent similarity characteristics among them. On the one hand, a certain degree of similarity is necessary, which can make the nuclear power monitoring system have a unified style as a whole, thereby improving the coordination degree of the system; on the other hand, too high a similarity degree is not conducive to the operator quickly identifying the meaning conveyed by the information, which may lead to cognitive biases of the operator and delays or mistakes in performing operations. Summary of the Invention

[0003] To solve the above problems, the present invention provides a method for setting the optimal similarity value range of nuclear power information chunks, aiming to obtain the optimal similarity value range of specific information chunks and provide a guiding reference for the design practice of information chunks in the nuclear power and other industrial control industries.

[0004] The technical solution of the present invention is as follows:

[0005] A method for setting the optimal similarity value range of nuclear power information chunks includes the following steps: Step (1), set the composition of nuclear power information chunks, and clarify the object characteristics and operation processes of the information chunks: Define the nuclear power icons or interfaces composed of several visual forms as nuclear power information chunks. The information chunks of nuclear power icons are simply referred to as icon groups, and the information chunks of nuclear power interfaces are simply referred to as drawing groups. Select N nuclear power information chunk objects that need to be discriminated for similarity, describe the characteristics of the icon group or drawing group, and set the operation process of its corresponding task, where N≥3;

[0006] Step (2), for the characteristic description of nuclear power information chunks, introduce similarity characteristic setting elements, and extract the similarity characteristic elements of the chunks: Set 16 items including complexity, density, familiarity, hierarchy, interface hierarchy, layout structure, interaction mode, information flow, text, color, size, direction, icon, control, text passage, and navigation as the similarity characteristic element set of nuclear power information chunks. According to the selected characteristic description of nuclear power information chunks, combined with interviews and investigations of nuclear power experts, screen and extract the similarity characteristic elements of the nuclear power information chunks from the 16 items;

[0007] Step (3): Obtain the regression model calculation equation for the overall similarity value of nuclear power chunks through a semantic differential questionnaire scale: Conduct a semantic differential scale questionnaire survey to compare the pairwise similarity characteristics of the selected nuclear power information chunk objects; perform a secondary screening of the similarity characteristics through linear regression analysis and obtain the regression model calculation equation for the overall similarity value of nuclear power chunks.

[0008] Step (4): Conduct a scenario task experiment to determine the optimal similarity value range. Design multiple sets of nuclear power interface chunks with different similarity values as experimental materials according to the similarity characteristics included in the regression model calculation equation for the overall similarity value of nuclear power chunks, and conduct a behavioral experiment using the scenario task paradigm; by analyzing the objective data and subjective data of users, obtain the similarity value range with the optimal performance showing significant differences. The objective data includes the correct rate and reaction time, and the subjective data includes the SUS usability score and the post-experiment experience questionnaire.

[0009] Further, step (1) specifically includes:

[0010] Step (1-1): Identify the nuclear power information chunk objects for which similarity needs to be discriminated and ensure that they are of the same category, where the category includes nuclear power diagrams or nuclear power paintings.

[0011] Step (1-2): Describe the specific characteristics of the power information chunk objects, including visual characteristics and functional uses, and combine with the actual selection of specific nuclear power typical task processes, and simplify and concatenate the operations according to the regulatory documents, where the regulatory documents include procedures or alarm cards.

[0012] Further, the specific steps for screening and extracting the similarity characteristic elements of the nuclear power information chunk from 16 items in step (2) include:

[0013] Nuclear power experts analyze the nuclear power information chunk objects from the dimension of overall perception according to the characteristic descriptions of the chunks, and screen suitable similarity characteristic elements from the four characteristics of complexity, density, familiarity, and hierarchy.

[0014] Nuclear power experts analyze the nuclear power information chunk objects from the dimension of visual pattern according to the characteristic descriptions of the chunks, and screen suitable similarity characteristic elements from the four characteristics of interface level, layout structure, interaction method, and information flow.

[0015] Nuclear power experts analyze the nuclear power information chunk objects from the dimension of visual symbols according to the characteristic descriptions of the chunks, and screen suitable similarity characteristic elements from the eight characteristics of text, color, size, direction, icon, control, text paragraph, and navigation.

[0016] Further, the overall perception similarity of all the feature elements screened in step (2) and the nuclear power information chunk is collectively referred to as the nuclear power information chunk feature comparison dimension S for which the similarity needs to be discriminated:

[0017] S = {T0, T1, T2, T3, …, T n ; n ≤ 16}

[0018] where T0 represents the overall perception similarity of the nuclear power information chunk, and T1, T2, T3, …, T n represents the screened similarity feature elements.

[0019] Further, step (3) specifically includes:

[0020] Step (3-1): Invite no less than 15 nuclear power experts to compare the selected nuclear power information chunk objects in pairs, design a semantic questionnaire difference survey questionnaire, and the composed comparison items S' are as follows:

[0021] S' = {Zxy|x = 1, 2, …, N; y = 1, 2, …, N; x ≠ y}

[0022] where Zxy represents the comparison result between the xth nuclear power information chunk object and the yth nuclear power information chunk object, 1, 2, …, N refer to the serial numbers of the nuclear power information chunk objects, N is the total number of nuclear power information chunks, and after the nuclear power experts complete a questionnaire compare the chunks. Each pair of chunks conducts n + 1 feature comparisons around the feature comparison dimension S, n ≤ 16. The questionnaire is set as a seven-level scale, and the scores are 1, 2, 3, 4, 5, 6, 7 respectively. From 1 to 7, it indicates that the difference of the comparison items between the two nuclear power chunks gradually becomes smaller;

[0023] Step (3-2): Conduct a multiple linear regression analysis in statistics on no less than 15 questionnaire data. Use the overall similarity score of the nuclear power chunk as the dependent variable, and the n selected similarity features as the independent variables for multiple linear regression analysis. When the fitting situation passes, analyze and screen out the similarity feature elements T a , T b , T c , …, T m that can significantly affect the overall similarity, m < n ≤ 16. According to the unstandardized coefficients, the multiple linear regression model of the overall similarity of the nuclear power information chunk is expressed as the following formula (1):

[0024] T0 = b + αTa + βTb + γTc + … + ηTm, Ta, Tb, Tc, …, Tm ∈ [1, 7](1)

[0025] Among them, b, α, β, γ, …, η are the non-standardized coefficients of each characteristic element. The minimum value of the overall perceived similarity T0 of the nuclear power block is b + α + β + γ + … + η; the maximum value is b + 7(α + β + γ + … + η);

[0026] Step (3-3): To make the value range of the overall similarity of the block convenient for perceiving the similarity degree of the nuclear power block, the regression model (1) is normalized, and the calculation equation of the regression model for the overall similarity value of the nuclear power block is shown in Equation (2).

[0027]

[0028] At this time, the value range of T0’ Set the overall similarity degree to five levels, and the division criteria are as shown in Table 1 below:

[0029] Table 1 Division criteria for the overall similarity degree of nuclear power information blocks under the T0’ calculation formula

[0030]

[0031]

[0032] Furthermore, step (4) specifically includes:

[0033] Step (4-1): For the similarity characteristic elements T a , T b , T c , …, T m in the calculation equation of the regression model for the overall similarity value of the nuclear power block (2), the perceived similarity characteristic coding rules are formulated respectively, and the similarity coding design of m features is carried out. Finally, five sets of experimental materials of nuclear power information blocks with different similarity levels are designed, and each set of materials has N nuclear power information blocks;

[0034] Step (4-2): Invite 10 nuclear power experts, and conduct the difference questionnaire survey of pairwise comparison in 7 levels as in step (3-1) for these five sets of experimental materials of nuclear power information blocks respectively. The comparison items S” formed are as follows:

[0035] S” = {Qxy|x = 1, 2, …, N; y = 1, 2, …, N; x ≠ y}

[0036] Among them, Qxy represents the comparison result between the xth nuclear power information block object after similarity coding design and the yth nuclear power information block object after similarity coding design. 1, 2, …, N refer to the serial numbers of the nuclear power information block objects after similarity coding design. One questionnaire completes 5 For each pair of chunks, m feature comparisons are carried out around the feature comparison dimension S”. The questionnaire is set as a seven-point scale, and the scores of the feature dimension values are 1, 2, 3, 4, 5, 6, and 7 respectively; from 1 to 7, it indicates that the difference of the comparison items between the two nuclear power chunks gradually becomes smaller.

[0037] Substitute the average score of the m feature dimension values into formula (2) to obtain the average overall similarity value of the five sets of nuclear power chunks. The average overall similarity value of the nuclear power chunks represents the overall perceived similarity level, which are A, B, C, D, and E from small to large.

[0038] Step (4-3): Invite 25 nuclear power expert subjects to be divided into five groups for a behavioral experiment. The experimental task process is obtained according to the simplified operation logic in step (1-2). Each nuclear power expert subject completes one experiment. The experiment adopts a scenario task paradigm, which belongs to a between-subjects single-factor behavioral experiment. The independent variable is the five overall perceived similarity levels A, B, C, D, and E of the interface, and the dependent variables are cognitive performance indicators: task time, error rate, and System Usability Scale (SUS) score. The error rate is expressed as the number of incorrect clicks / number of clicks in the ideal situation. The task time and error rate are recorded by the background during the experiment; the SUS score is obtained by the nuclear power expert subjects filling out the product usability questionnaire after the experiment.

[0039] Step (4-4): Conduct a one-way repeated measures analysis of variance and pairwise comparisons on the reaction time data that is confirmed to be normally distributed to obtain the significance P.

[0040] Step (4-5): Determine the comparison groups with significance P ≤ 0.05 as significantly different. The significant differences include significant decrease and significant increase. Take the average overall similarity value R corresponding to the group with a significant increase in reaction time and the average overall similarity value R’ corresponding to the group with a significant decrease as the best range G of the nuclear power information chunk similarity value, where G ∈ [R, R’], R ∈ A / B / C / D / E; R’ ∈ A / B / C / D / E; R ≠ R’.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] The method for setting the best similarity value range of the nuclear power information chunks of the present invention is aimed at nuclear power monitoring systems and other similar industrial control industries. It quantifies the influencing factors of the similarity of nuclear power information chunks and determines the best similarity range, providing a similarity evaluation means for the existing nuclear power information chunk design scheme to reduce the operation errors caused by the operator's perception similarity deviation and providing a reference for the similarity evaluation in nuclear power information design. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a flowchart of the method for setting the best similarity value range of the nuclear power information chunks of the present invention.

[0044] Figure 2 It is a simplified task flow chart of the nuclear power module.

[0045] Figure 3 It is a set of 16 perceptual similarity feature elements of information blocks.

[0046] Figure 4 It is the color-coded result diagram of the five sets of experimental interfaces.

[0047] Figure 5 This is a diagram showing seven common layout methods in nuclear power interfaces.

[0048] Figure 6 It is the mean reaction time graph of five kinds of interface similarity.

[0049] Figure 7 It is a curve fitting diagram of the overall similarity value T0' of the information block and the reaction time. DETAILED DESCRIPTION

[0050] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention.

[0051] like Figure 1 As shown in FIG, a method for setting the optimal similarity value range of a nuclear power information block is applied to the similarity evaluation of a typical image group of a nuclear power alarm task. The specific steps are as follows:

[0052] (1-1) Based on the definition of nuclear power information blocks, seven typical interface images of nuclear power monitoring system alarm tasks were selected as similarity assessment targets. The specific image groups are: system main interface, alarm list, alarm card, control interface 1, operation window 1, control interface 2, and operation window 2.

[0053] The system's main interface consists of a set of key parameters related to the progress, execution, and results of long-term operational activities. This helps operators track the critical operating parameters that require monitoring as specified in the operating procedures and also assists in navigating to related auxiliary information and control interfaces. The alarm list integrates and categorizes alarm information from the nuclear power monitoring system using tables and color codes, allowing for rapid location of alarm faults. It provides auxiliary functions such as screening, confirmation, and troubleshooting, and links alarm interfaces and alarm cards. The control interface displays a simplified process flow based on the power plant system flow chart, sends commands to the process, and displays operational feedback information for executing operational tasks identified during the execution process. The operation window is an icon control panel beneath the control interface, presenting the icon's display information and operational options. This interface group is the lowest level where operators perform interface operations.

[0054] (1-2) For the task process of the nuclear power alarm "upper bearing temperature H2", simplify and concatenate the operations according to regulatory documents such as procedures or alarm cards, and finally form Figure 2 the task process. One task needs to complete five icon visual search tasks separately. After finding a specific icon in the control interface, open the operation window to complete the corresponding operation, and then return to the alarm card for secondary confirmation.

[0055] (2) Invite 15 nuclear power interface engineers to screen the similarity features for typical nuclear power alarm task interfaces according to the information chunk perception similarity feature set (such as Figure 3 ). Due to the strong professionalism in the industrial control industry, operators need to undergo strict training and learning on the nuclear power monitoring system and are required to be fully familiar with and master the interface. Familiarity features are not applicable in similarity assessment; there is little difference between the "hierarchy" in the perception dimension and the "information hierarchy" in the visual pattern, so one of them is discarded; nuclear power regulations require that all interface interactions are mouse click operations, and the interaction method is meaningless as a visual similarity variable in the nuclear power field; the "direction" feature in visual symbols is not prominent in the interface. The current interfaces are all 16:9 landscape interfaces, and even if the direction of the main body such as icon text changes, it should be classified into variables such as "icon" and "text"; finally, "text paragraph" and "navigation" generally appear in the form of "controls" in industrial control interfaces, so they are classified into the "control" variable. After screening, 10 nuclear power interface similarity perception features are retained: T1 complexity, T2 density, T3 hierarchy, T4 layout structure, T5 information flow, T6 text, T7 color, T8 size, T9 icon, T 10 control, as shown in Table 1 specifically.

[0056] Table 1 Screening results of nuclear power information chunk similarity perception characteristics

[0057]

[0058]

[0059] Combine all the screened feature elements with the overall perception similarity of the chunks to form the comparison dimension S of the nuclear power information chunk features whose similarity needs to be discriminated:

[0060] S = {T0, T1, T2, T3, …, T 10}

[0061] (3-1) Invite 15 nuclear power interface engineers to compare the selected nuclear power interfaces in pairs. Design a semantic questionnaire difference questionnaire, and the comparison items formed are as follows:

[0062] S’ = {Zxy|x = 1, 2, …, 7; y = 1, 2, …, 7; x ≠ y}

[0063] Nuclear power experts need to complete a questionnaire The blocks were compared, with each pair of blocks compared across 11 features along the feature comparison dimension S. The questionnaire used a seven-point scale, with scores of 1, 2, 3, 4, 5, 6, and 7. A score of 1 indicated that the two blocks differed significantly in the comparison items, while a score of 7 indicated that the two blocks were essentially identical in the comparison items.

[0064] (3-2) A multiple linear regression analysis was performed on the data from the 15 questionnaires. The overall similarity score of the nuclear power block was used as the dependent variable, and the 10 similarity features selected were used as independent variables for the multiple linear regression analysis. The fitting results are shown in Table 2. The R-squared value is 0.63, and the explanatory power of the independent variable on the dependent variable is 63% (good effect); the Durbin-Watson coefficient is 1.571, and the sample independence is passed. Therefore, the analysis results have good explanatory significance. The specific results are shown in Table 3. The significance of T1, T4 and T7 is less than 0.001, and the significance of the other feature variables is greater than 0.05. Therefore, the three similarity features of interface complexity, layout structure and color have a significant impact on the overall similarity score, and the VIF values are all less than 5, indicating that the three similarity features have no collinear relationship with other features.

[0065] Table 2 Multiple linear regression fitting

[0066]

[0067] Table 3 Multivariate linear regression model coefficients

[0068]

[0069] Based on the unstandardized coefficients, the multivariate linear regression model of the overall interface similarity can be expressed as follows (1):

[0070] T0=-0.709+0.324T7+0.424T4+0.225T1 (1)

[0071] (3-3) Since T0∈[0.264,6.102], the value range is not convenient for perceiving the similarity, so it is normalized to T0'(2):

[0072] T0'=0.531T7+0.695T4+0.369T1-1.162 (2)

[0073] Where T X The scores are 1, 2, 3, 4, 5, 6, and 7, and the overall similarity value is normalized to score T0'∈(0.433, 10). The overall similarity is set to five levels, and the division criteria are shown in Table 4.

[0074] Table 4 Classification criteria for overall interface similarity under T0' calculation formula

[0075]

[0076] (4-1) In combination with the design standards of the nuclear power industry, for the similarity characteristic elements T1, T4, and T7 in the overall similarity calculation equation (2), the perceptual similarity characteristic coding rules are formulated respectively, and the similarity coding design of the three characteristics is carried out.

[0077] Color coding: In the design of nuclear power monitoring interfaces, the selection of colors needs to comply with established safety standards and regulatory requirements. For directional colors (such as red for warnings, green for safety, yellow for reminders, etc.), the general industry specifications must be strictly observed to ensure that operators can quickly identify and respond to relevant interface information. Interface color coding refers to the overall main color tone. According to relevant regulations and design guidelines, in the case of not involving directional meanings, dark-colored systems without special meanings are preferably selected. The application of these dark colors can not only reduce visual fatigue but also reduce the impact of ambient light reflection on the screen clarity, thereby improving the operation stability and safety of the interface.

[0078] Select the seven colors of red, orange, yellow, green, blue, purple, and gray in the dark gray color system in the interface design as the standard reference color values (the visual difference between cyan and green is not significant, so gray is used instead). The saturation S + 100 is used as the highlight color, the lightness L - 20 and the transparency O - 50 are used as the sinking colors, and the transparency O - 50 is used as the background color. In common interface designs, most of the main colors are three, so it is set as the number of main colors in the general similarity experiment interface; the greater the color change between the less similar experimental interfaces, the more main colors. Color coding is carried out according to the actual task situation. The main colors of the extremely dissimilar interfaces are 7, and the main colors of the extremely similar interfaces are 1. The color characteristic coding results of the five sets of interfaces are as Figure 4 shown.

[0079] Layout structure coding: It is found through observation that the current nuclear power monitoring interface designs mainly include grid layout, sidebar layout, column layout, and panel layout. On this basis, combined with the common layout methods of Internet B-side systems, seven layout structures are finally refined and sorted out as Figure 5 shown. Different layout methods are used for different design occasions. For example, the T-shaped layout is usually used in scenarios where information priorities are clear; the Chuan-shaped and Gong-shaped layouts are suitable for the parallel display of multiple functions; the grid layout can handle a large amount of data information in rows and columns; the whole picture + marker layout is the most widely used in nuclear power interfaces and is mostly used for visualizing the overall graphics such as process flows. When coding nuclear power interfaces, it should be adjusted as much as possible according to the actual situation and based on the application experience of the layout method to ensure clear interface logic, distinct levels, and improved information accessibility.

[0080] Among several interfaces under a set of nuclear power operation task processes, there are 3 common layout types, which are used as the number of layout categories for general similar experimental interfaces; the greater the difference between less similar experimental interfaces, the greater the layout changes and the more layout types. Layout coding is carried out according to the experimental task situation. The number of layout methods for extremely dissimilar interfaces is 7, and the number of layout methods for extremely similar interfaces is 1. The layout characteristic coding results of five sets of interfaces are shown in Table 5.

[0081] Table 5 Encoding Results of the Layout Structures of Five Sets of Experimental Interfaces

[0082]

[0083] Complexity coding: Most current methods for quantifying the complexity of interface information refer to the interface complexity theory proposed by Shannon and improved by Comber. The complexity of nuclear power blocks increases with the increase in the number and types of components. Through interface research and observation, the 8 most common component types and 26 component quantities in nuclear power interfaces are set as "generally complex" interfaces, and they are evenly extended to both poles from this intermediate value. Finally, the complexity settings of nuclear power blocks are shown in Table 6.

[0084] In terms of the similarity characteristic coding of the complexity of experimental interfaces, the 3 common complexities are used as the number of complexities for general similar interfaces. The greater the difference between less similar experimental interfaces, the greater the complexity changes and the more complexity quantities. Complexity coding is carried out according to the actual task situation. The number of complexities for extremely dissimilar interfaces is 5, and the number of complexities for extremely similar interfaces is 1. The final complexity coding results are shown in Table 7.

[0085] Table 6 Complexity Settings in Experimental Interfaces

[0086]

[0087] Table 7 Complexity Coding Results of Five Sets of Experimental Interfaces

[0088]

[0089] Summarize the above coding results as shown in Table 8, and thus design five sets of interface experimental materials.

[0090] Table 8 Coding Results of the Three Similar Characteristics of Experimental Interfaces

[0091]

[0092] (4-2) Conduct a 7-level pairwise comparison difference questionnaire survey on these five sets of interfaces from three dimensions: color, layout structure, and complexity. Invite 20 relevant practitioners to conduct it in three groups of questions, and each group needs to Compare the items. There are 105 multiple-choice questions in total. The scores from 1 to 7 represent very large differences - basically the same. The average values of the three dimensions are calculated and then substituted into the formula T0', and the average scores of the interface similarity values shown in Table 9 are obtained.

[0093] Table 9 Mean values of three similarity characteristics and average scores of overall similarity value T0' of five sets of experimental interfaces

[0094]

[0095] (4 - 3) Invite 50 nuclear power experts, operators or related personnel and divide them into five groups for behavioral experiments. The experimental task process is derived from the simplified operation logic in step (1), and each subject completes one experiment. The experiment adopts the scenario task paradigm, which belongs to the between-subjects single-factor behavioral experiment. The independent variable is the overall perceived similarity level of the five interfaces, and the dependent variables are cognitive performance indicators: task time consumption, error rate (number of incorrect clicks / number of clicks in the ideal situation), and experience scale score (SUS). The task time consumption and error rate are recorded in the background during the experiment; the SUS score is obtained from the subjective product usability questionnaire filled out by the subjects after the experiment.

[0096] (4 - 4) Conduct descriptive analysis of the number of errors and reaction times, as shown in Table 10 and Table 11. The data of the number of errors do not conform to the Shapiro-Wilk normal distribution (<0.05), and the data of reaction times conform to the normal distribution except for the extremely similar group. Perform a base-10 logarithmic transformation on the reaction times, and the five groups of data obtained all conform to the normal distribution. Therefore, the logarithmically transformed reaction times are used as the main data for subsequent processing.

[0097] Table 10 Descriptive statistics of the number of errors for five similarity interfaces

[0098]

[0099] Table 11 Descriptive statistics of the reaction times for five similarity interfaces

[0100]

[0101] Make a line graph of the mean reaction times as shown in Figure 6 shown. Conduct a one-way repeated measures analysis of variance on the transformed reaction times. Under the condition of satisfying Mauchly's sphericity assumption (significance P = 0.405 > 0.05), the main effect is significant (F = 22.708, P < 0.001). Conduct pairwise comparisons of the logarithms of the five groups of reaction times as shown in Table 12. It can be seen from the table that there is no significant difference between the extremely dissimilar interface and the dissimilar interface, and there are significant differences with the generally similar, similar, and extremely similar interfaces (P < 0.05); there are significant differences between the generally similar group and the other three groups except the similar group (P < 0.05). It can be seen that there are significant differences in cognitive performance for different interface similarities. When the overall interface is generally similar or similar, the user's cognitive performance is better.

[0102] Table 12 Pairwise Comparisons of Logarithms of Response Times for Five Interface Similarities

[0103]

[0104] Note: Groups 1 - 5 represent the extremely dissimilar group - the extremely similar group (4 - 5) according to the mean response time Figure 6 , and in combination with the overall interface similarity value T0' in Table 9, subjectively speculate and curve fit the relationship between the overall interface similarity and cognitive performance, as Figure 7 shown. The first speculation (blue line) is that the inflection point of cognitive performance appears between the dissimilar group (similarity value 3.517) and the generally similar group (similarity value 4.624); the second speculation (red line) is that the inflection point of cognitive performance appears between the generally similar group (similarity value 4.624) and the similar group (similarity value 5.538). Since it is impossible to determine which speculation is more accurate, it is preliminarily determined that: the closer the overall similarity value is to 4.624, the better the cognitive performance. Divide the overall similarity value corresponding to the generally similar group with a significant increase in reaction time into R (similarity value 4.624), and divide the overall similarity value corresponding to the similar group with a significant decrease into R' (similarity value 5.538) as the best range G of the similarity value of nuclear power information chunks, G ∈ [4.624, 5.538]. Generally speaking, when the interface similarity value is determined to be "generally similar" (similarity value 4 - 6), the cognitive performance is better.

[0105] The present invention can quantify the similarity influencing factors of nuclear power information chunks and determine the best similarity range, providing an evaluation method based on similarity measurement for the existing nuclear power information chunk design scheme to reduce the operation errors caused by the operator's perception similarity deviation.

[0106] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and deformations can still be made, and these improvements and deformations should also be regarded as the protection scope of the present invention.

Claims

1. A method for setting the optimal similarity value range of nuclear power information chunks, characterized in that It includes the following steps: Step (1): Set the composition of the nuclear power information chunks, and clarify the object characteristics and operation processes of the information chunks: Define the nuclear power graphic symbols or interfaces composed of several visual forms as the nuclear power information chunks. The information chunks of nuclear power graphic symbols are simply called graphic groups, and the information chunks of nuclear power interfaces are simply called drawing groups. Select N nuclear power information chunk objects that need to be judged for similarity, describe the characteristics of the graphic group or drawing group, and set the operation process of its corresponding task, where N≥3; Step (2): For the feature description of the nuclear power information chunks, introduce the similarity feature setting elements, and extract the similarity feature elements of the chunks: Set 16 items including complexity, density, familiarity, hierarchy, interface hierarchy, layout structure, interaction mode, information flow, text, color, size, direction, icon, control, text passage, and navigation as the similarity feature element set of the nuclear power information chunks. According to the selected feature description of the nuclear power information chunks, combined with the nuclear power expert interview and research, screen and extract the similarity feature elements of the nuclear power information chunks from the 16 items; Step (3): Obtain the regression model calculation equation for the overall similarity value of the nuclear power chunks through the semantic differential questionnaire scale: Conduct a semantic differential scale questionnaire survey to compare the pairwise similarity features of the selected nuclear power information chunk objects; Perform secondary screening of the similarity features through the linear regression analysis method, and obtain the regression model calculation equation for the overall similarity value of the nuclear power chunks; Step (4): Conduct a scenario task experiment to determine the optimal similarity value range. According to the similarity characteristics included in the regression model calculation equation for the overall similarity value of the nuclear power chunks, design multiple sets of nuclear power interface chunks with different similarity values as experimental materials, and conduct a behavioral experiment in the scenario task paradigm; By analyzing the objective data and subjective data of the users, obtain the similarity value range of the optimal performance with significant differences. The objective data includes the correct rate and reaction time, and the subjective data includes the SUS usability score and the post-experiment experience questionnaire.

2. The method for setting the optimal similarity value range of nuclear power information chunks according to claim 1, characterized in that, Step (1) specifically includes: Step (1-1): Clarify the nuclear power information chunk objects that need to be judged for similarity, and ensure that their categories are the same. The categories include nuclear power graphic groups or nuclear power drawing groups; Step (1-2): Describe the specific characteristics of the electrical information chunk objects, including visual characteristics and functional uses, and combine the actual situation to select a specific nuclear power typical task process, and simplify and concatenate the operations according to the regulatory documents. The regulatory documents include procedures or alarm cards.

3. The method for setting the optimal similarity value range of nuclear power information chunks according to claim 2, characterized in that, The specific process of screening and extracting the similarity feature elements of the nuclear power information chunks from the 16 items in Step (2) includes: Nuclear power experts analyze the nuclear power information chunk objects from the dimension of overall perception according to the feature description of the chunks, and screen appropriate similarity feature elements from the four characteristics of complexity, density, familiarity, and hierarchy; Nuclear power experts analyze the nuclear power information chunk objects from the dimension of visual mode according to the feature description of the chunks, and screen appropriate similarity feature elements from the four characteristics of interface hierarchy, layout structure, interaction mode, and information flow; Nuclear power experts analyze the nuclear power information chunk objects from the dimension of visual symbols according to the feature descriptions of the chunks, and screen out appropriate similarity feature elements from eight features including text, color, size, direction, icon, control, text passage, and navigation.

4. The method for setting the optimal similarity value range of nuclear power information chunks according to claim 3, characterized in that, All the feature elements screened out in step (2), together with the overall perception similarity of the nuclear power information chunks, are collectively referred to as the comparison dimension S of the nuclear power information chunk features for which similarity needs to be discriminated: S = {T0, T1, T2, T3, …, Tn; n ≤ 16} Among them, T0 represents the overall perception similarity of the nuclear power information chunk, and T1, T2, T3, …, T n represent the selected similarity feature elements.

5. The method for setting the optimal similarity value range of nuclear power information chunks according to claim 4, characterized in that, Step (3) specifically includes: Step (3-1): Invite no less than 15 nuclear power experts to compare the selected nuclear power information chunk objects in pairs, design a semantic questionnaire difference questionnaire, and the formed comparison items S’ are as follows: S’ = {Zxy|x = 1, 2, …, N; y = 1, 2, …, N; x ≠ y} Among them, Zxy represents the comparison result between the x-th nuclear power information chunk object and the y-th nuclear power information chunk object. 1, 2, …, N refer to the serial numbers of the nuclear power information chunk objects, and N is the total number of nuclear power information chunks. After a nuclear power expert completes a questionnaire, compare the chunks. Each pair of chunks conducts n + 1 feature comparisons around the feature comparison dimension S, where n ≤ 16. The questionnaire is set as a seven-point scale, with scores of 1, 2, 3, 4, 5, 6, and 7 respectively. From 1 to 7, it indicates that the difference in the comparison items between the two nuclear power chunks gradually decreases; Step (3-2): Conduct a multiple linear regression analysis in statistics on no less than 15 sets of questionnaire data. Use the overall similarity score of the nuclear power block as the dependent variable and the n selected similarity features as the independent variables for the multiple linear regression analysis. When the fitting condition is passed, analyze and screen out the similarity feature elements T that can significantly affect the overall similarity a , T b , T c , …, T m , where m < n ≤ 16. According to the unstandardized coefficients, the multiple linear regression model for the overall similarity of the nuclear power information block is expressed as the following formula (1): T0 = b + αT a + βT b + γT c + … + ηT m , T a , T b , T c , …, T m ∈ [1, 7](1) Where b, α, β, γ, …, η are the non-standardized coefficients of each feature element, and the minimum value of the overall perception similarity T0 of the nuclear power chunk is b + α + β + γ + … + η; the maximum value is b + 7(α + β + γ + … + η); Step (3-3): Normalize the regression model (1) to obtain the calculation equation of the regression model for the overall similarity value of the nuclear power chunk as shown in formula (2). The value range of T0' at this time Set the overall similarity level to five levels, and the classification criteria are as shown in Table 1 below: Table 1 Classification criteria for the overall similarity degree of nuclear power information chunks under the calculation formula of T0’ 6. The method for setting the optimal similarity value range of nuclear power information chunks according to claim 5, characterized in that Step (4) specifically includes: Step (4-1): Calculate the similarity feature element T in Equation (2) for the regression model of the overall similarity value of the nuclear power block a ,T b ,T c ,…,T m , respectively formulate the coding rules for the perceived similarity features, conduct the similarity coding design for m features, and finally design five sets of experimental materials for the nuclear power information blocks with different similarity levels. Each set of materials has N nuclear power information blocks; Step (4-2): Invite 10 nuclear power experts to conduct a difference questionnaire survey on the five sets of nuclear power information chunk experimental materials respectively in a pairwise comparison of 7 levels as in step (3-1), and the formed comparison items S” are as follows: S” = {Qxy|x = 1, 2, …, N; y = 1, 2, …, N; x ≠ y} Among them, Qxy represents the comparison result between the x-th nuclear power information chunk object designed by similarity coding and the y-th nuclear power information chunk object designed by similarity coding. 1, 2, …, N refer to the serial numbers of the nuclear power information chunk objects designed by similarity coding, and one questionnaire is completed comparison items. Each pair of chunks conducts m feature comparisons around the feature comparison dimension S”. The questionnaire is set as a seven-point scale, and the score values of the feature dimension are 1, 2, 3, 4, 5, 6, and 7 respectively; from 1 to 7, it indicates that the difference of the comparison items between the two nuclear power chunks gradually becomes smaller Calculate the average value of the scores of the m feature dimension values and substitute them into formula (2) to obtain the average score of the overall similarity values of the five sets of nuclear power chunks. The average score of the overall similarity values of the nuclear power chunks represents the overall perception similarity degree level, which are A, B, C, D, and E from small to large; Step (4-3): Invite 25 nuclear power expert subjects to be divided into five groups for a behavioral experiment. The experimental task process is derived from the simplified operation logic in step (1-2). Each nuclear power expert subject completes one experiment. The experiment adopts a scenario task paradigm, which belongs to a between-subjects single-factor behavioral experiment. The independent variable is the five interface overall perception similarity degree levels A, B, C, D, and E, and the dependent variables are cognitive performance indicators: task time consumption, error rate, and experience scale score (SUS). The error rate represents the number of incorrect clicks / the number of clicks in the ideal situation. The task time consumption and error rate are recorded in the background during the experiment; the experience scale score (SUS) is obtained by the nuclear power expert subjects filling out a product usability questionnaire after the experiment; Step (4-4): Conduct a one-way repeated measures analysis of variance and pairwise comparison on the reaction time data confirmed to be normally distributed to obtain the significance P; Step (4-5): Determine the comparison groups with a significance P ≤ 0.05 as significantly different. The significant differences include significant decreases and significant increases. Divide the overall similarity value corresponding to the group with a significant increase in reaction time equally by R, and divide the overall similarity value corresponding to the group with a significant decrease equally by R'. These values are used as the optimal range G of the nuclear power information chunk similarity value, where G ∈ [R, R'], R ∈ A / B / C / D / E; R' ∈ A / B / C / D / E; R ≠ R'.

Images