Subway trip chain satisfaction evaluation method based on three-stage configuration factor analysis
Through the three-stage configuration factor analysis method, the problem that the existing technology cannot fully capture the subway travel satisfaction factors is solved, the key travel stages are determined, and the subway sharing rate and passenger satisfaction are improved.
Patent Information
- Application Number
- CN202510132530.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-06
- Publication Date
- 2025-03-07
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing technology is difficult to fully capture all factors affecting subway travel satisfaction, and fails to establish a complete mechanism to evaluate the factors affecting subway satisfaction.
The three-stage configuration factor analysis method is adopted to determine the key travel stages of the entire subway travel process by obtaining the factors that affect the state, improve the evaluation process and mechanism, and guide the subway operation and planning rectification opinions.
It has achieved the acquisition of all passengers' overall needs during subway travel, identified the key travel stages, and improved the subway sharing rate and passenger satisfaction.
Smart Images
Figure CN119579242A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis, and belongs to the field of traffic engineering. Background Art
[0002] Travelers' satisfaction with subway travel determines whether they use the subway. Higher satisfaction with subway travel increases the frequency of use. The factors affecting subway travel satisfaction are complex and diverse, mainly involving the environment and services. The environment includes travel environment, travel facilities, built environment, etc., and services include service quality, service level, service capacity, etc. The existing methods mainly evaluate the influencing factors of a certain process in the travel, and fail to fully capture all the influencing factors of passengers' use of subway travel, and fail to establish a complete mechanism for evaluating the influencing factors of subway satisfaction. Summary of the invention
[0003] Purpose of the invention: In order to overcome the deficiencies in the prior art, the present invention provides a subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis. By obtaining the state influencing factors, all the needs of passengers in the overall subway travel process can be obtained, and the critical travel stages in the entire subway travel process can be determined. The existing evaluation process and mechanism are improved, and the results obtained are used to guide subway operations and planning to formulate rectification opinions and improve the subway share rate.
[0004] Technical solution: To solve the above technical problems, the present invention provides a method for evaluating subway travel chain satisfaction based on three-stage configuration factor analysis, comprising the following steps: S1: Divide the subway travel chain and formulate subway travel satisfaction rules; S2: Prepare questionnaires, distribute and collect questionnaires, and obtain data sets; S3: Obtain basic information of travelers and travel perception information and assign values; S4: Determine the perceived and observed variables involved in the analysis; S5: Transformation of measured variables into perceived variables; S6: Convert data into fuzzy sets; S7: Construct fuzzy set truth table; S8: Get the configuration factors that affect subway travel satisfaction; S9: Count the frequency of occurrence of individual factors in the configuration factors and perform statistical sorting.
[0005] Preferably, in step S1, the satisfaction evaluation method is as follows: the traveler directly selects the degree of travel satisfaction through the Likert 7-point scale.
[0006] Preferably, in step S3, the basic information includes: the entrance travel time T1, the subway ride time T2, the time from the exit subway station to the destination T3, the transportation tool F used for the entrance travel 1c , where c∈C, C={1, 2, ...4, 5}, 1 represents walking, 2 represents shared bicycles, 3 represents electric bicycles, 4 represents public transportation, and 5 represents online car-hailing. The transportation method used from the exit subway station to the destination is F 2e , e∈E, E={1, 2, 3, 4, 5}, 1 represents walking, 2 represents shared bicycle, 3 represents company car, 4 represents bus, 5 represents online car-hailing, the entrance subway station name is A n and exit subway station name E n , n represents the nth subway station of the entrance; The evaluation method of T1 is as follows: the acquired data is coded, the travel time within 10 minutes is coded as 1, the travel time between 10 and 20 minutes is coded as 2, the travel time between 20 and 30 minutes is coded as 3, and the travel time over 30 minutes is coded as 4. By calculating the proportion of a certain number of codes to the sample size of high travel satisfaction evaluation, they are ranked in descending order and assigned values. The code ranked as 1 is assigned a value of "6", followed by "5", "3", and "2" points. The T2 evaluation method is as follows: encode the acquired data, the travel time within 15 minutes is coded as 1, the travel time between 15 and 30 minutes is coded as 2, the travel time between 30 and 45 minutes is coded as 3, and the travel time over 45 minutes is coded as 4. By calculating the proportion of a certain number of codes to the sample size of high travel satisfaction evaluation, they are ranked in descending order and assigned values. The code ranked as 1 is assigned a value of "6", followed by "5", "3", and "2" points. The T3 evaluation method is as follows: encode the acquired data, the travel time within 10 minutes is coded as 1, the travel time between 10 and 20 minutes is coded as 2, the travel time between 20 and 30 minutes is coded as 3, and the travel time over 30 minutes is coded as 4. By calculating the proportion of a certain number of codes to the sample size of high travel satisfaction evaluation, they are ranked in descending order and assigned values. The code ranked as 1 is assigned a value of "6", followed by "5", "3", and "2" points. Among them, F 1c The evaluation method is as follows: encode the acquired data, F11 is coded as 1, F12 is coded as 2, F13 is coded as 3, F14 is coded as 4, and F15 is coded as 5. By calculating the proportion of a certain number of codes in the sample size of high travel satisfaction evaluation, they are ranked in descending order and assigned values. The code ranked as 1 is assigned a value of "6", followed by "5", "3", and "2". The last two codes with the lowest proportion are uniformly coded as "2"; Among them, F2e The evaluation method is as follows: encode the acquired data, F21 is coded as 1, F22 is coded as 2, F23 is coded as 3, F24 is coded as 4, and F25 is coded as 5. By calculating the proportion of a certain code number to the sample size of high travel satisfaction evaluation, they are ranked in descending order and assigned values. The code ranked as 1 is assigned a value of "6", followed by "5", "3", and "2". The last two codes with the lowest proportion are uniformly coded as "2"; Among them, A n and E n The function is to confirm the land use attributes of subway entrances and exits. The method is as follows: confirm the land use attribute type within the 800-meter buffer zone through the station name of the subway station, and assign a value to the land use degree within the range, specifically: (1); Where: Mixn is the land use degree of the 800-meter buffer zone of the nth subway entrance, p m is the ratio of the mth type of land use (the ratio of the number of a certain land use attribute within the 800-m buffer zone of a subway station to the total number of all land use types in the area), and k depends on the number of different land uses in the station buffer zone.
[0007] Preferably, in step S4, the travel perception information includes: the perception P1 of the ancillary facilities of the entrance subway station, using the observation variable U 1j Reflects the perception of the service in the subway station P2, using the observed variable U 2j Reflects the perception of convenience in the subway station P3, using the observed variable U 3j Reflection, perception P4 during subway ride, using observation variable U 4j Reflects the perception of the ancillary facilities of the exit subway station P5, using the observed variable U 5j Reflection, exit travel perception P6, using the observed variable U 6j Reflect, where j represents the Jth factor, j∈J, J={1, 2, ...j}; Among them U 1j , U 2j , U 3j , U 4j , U 5j , U 6j All of them are observed variables reflecting the perception variables P1, P2, P3, P4, P5, and P6. The evaluation of the observed variables by travelers is obtained through a 7-point Likert scale.
[0008] For the convenience of subsequent expression, define U ij is the i-th perception variable P i The observed variables under ijIt is the set of samples belonging to this observed variable, that is, each U ij All are composed of {u 11 ,u 12 ,...u ij}constitute; Preferably, the path coefficient of the observed variable to the perceived variable is used to determine whether the perceived variable can be explained. The path coefficient is calculated by the formula: ij Factors with coefficients less than 0.7 are considered to be unable to fully express the perception variable U ij Should be deleted.
[0009] (2); In the formula, Indicates U ij P i The path coefficient, SAT j The set of satisfaction is represented by {sat1, sat2, ...sat j}, i∈I, I=1, 2...6, j∈J, J=1, 2...J, coefficient Less than 0.7, delete P i The corresponding U ij , Var represents the function of variance; The test of the reflection system is carried out through the values of the indicators, CR, AVE, and VIF, and the calculation method is as follows: (3); (4); Where θ ij represents the variance of the observed variable error.
[0010] (5); In the formula, R is the coefficient of determination; The evaluation index requires CR greater than 0.7, AVE greater than 0.5, and VIF less than 5. i The square root of AVE is greater than the correlation coefficient with other perceptual variables. When it reflects the system inspection index and meets all the above index requirements, it indicates that it has a stable and accurate interpretation basis for subsequent evaluation content.
[0011] Preferably, in step S5, the formula for converting the measured variable into the perceived variable is as follows: (6); In the formula, i∈I, I={1, 2...6}, j∈J, J={1, 2...j}, after the effects of the observed variables Pij involved in all samples are aggregated to the perception variable Pi, a new data set Ω1 is obtained from the original data set.
[0012] Preferably, in step S6, the data is converted into fuzzy sets, and the data set Ω1 is converted again to obtain a data set Ω2 that can be used for qualitative evaluation analysis. The main function of the fuzzy set is to map the satisfaction value to a value in the range of (0, 1), wherein (0, 0.5) represents dissatisfaction with the factor, and a fuzzy set value between (0.5, 1) represents satisfaction with the factor, and (0.5) represents a neutral state that is neither affiliated with the former nor the latter, so that all factors can participate in the evaluation at the same order of magnitude; The transformation method of the fuzzy set is mainly to confirm the qualitative anchor point, using the quartile point calculation, where P i is the set of variables in Ω1, P i25% , P i75% Respectively represent the 25% quantile and 75% quantile of the set of the i-th perceptual variable P, m=i, i∈I, I={1, 2, 3, 4, 5, 6, 7}; (7); After obtaining the new data set Ω2, a necessity test needs to be performed on each perception variable to confirm whether it is set as a variable that must exist when the result is output in the truth table analysis, which will affect the analysis result; The necessity means that when condition X is considered to be a necessary condition for result Y, all high membership of X corresponds to high membership of result Y. It is evaluated by two key indicators: consistency and coverage. It is calculated by formulas (8) and (9); (8); (9); Where, X m represents the fuzzy set membership of the condition, that is, the i-th condition variable in the data set Ω2, SAT m Represents the fuzzy set membership set of the result, and each condition X in the sample corresponds to a result SAT.
[0013] Among them, consistency means the sum of the memberships that achieve consistency in the antecedent conditions or combinations of antecedent conditions divided by the sum of the memberships in all antecedent conditions or combinations of antecedent conditions, which is used to verify the consistency of the results produced by samples with the same antecedent conditions. Coverage means the proportion of the overlapping parts of the two sets of antecedent condition combinations and results relative to the result set, verifying the proportion of samples that meet the consistency requirements in the total samples.
[0014] Among them, when the consistency is greater than 0.9 and the coverage is greater than 0.5, the condition is considered to be a necessary condition for the result, and only the combination with the condition of 1 is retained during truth table analysis; The purpose of achieving consistency and coverage is to ensure the accuracy of the results of factors affecting satisfaction; Preferably, the step S7 comprises the following specific steps: S71, constructing the original truth table: constructing an initial table through the data set Ω2, where each row represents the condition configuration of a case and each column represents the corresponding result value; S72. Identify the minimum conditional combination: Identify which conditional combinations are most common in the observed results, with the goal of finding the causal relationship between the conditional combination and the results. Use the following methods to achieve this goal: First, find all input combinations with an output of 1, find adjacent items to merge, and determine the minimum term. For example, (condition A=0, condition B=0, condition C=1) and (condition A=0, condition B=1, condition C=1), (condition A = 0, condition B = 1, condition C = 1), (condition A=0, condition B=1, condition C=1) can be merged because they differ only in B. After merging, they become A′C. (condition A=0, condition B=1, condition C=0) and (condition A=1, condition B=1, condition C=0)(condition A = 1, condition B = 1, condition C = 0)(condition A=1, condition B=1, condition C=0) can be merged. After merging, they become B′C, completing the determination of the minimum term.
[0015] S73. Remove redundant items: select the smallest items that can cover all combinations with output 1.
[0016] S74. Determine the final Boolean expression: The final Boolean expression is the union of the merged minimum terms. The result in the example is, f(A, B, C) = A′C + B′C′ + AC; Among them, for i binary conditions (for example, 1 represents "existence" and 0 represents "absence"), the total number of combinations is 2i. After the truth table analysis determines the final Boolean expression, the simplest causal relationship combination is found.
[0017] Among them, when performing truth table calculations, it is necessary to set the passing threshold according to the actual situation. It is necessary to pass the thresholds of two indicators, that is, the consistency of all cases involved in the calculation needs to be greater than 0.8, and the proportional consistency reduction (PRI) value needs to be greater than 0.50 to determine the occurrence of the result. Among them, consistency needs to ensure that when the truth table is calculated, the cases involved in the calculation have high explanatory power for the sample, and PRI is used to avoid the existence of the same subset relationship between the configurations in the result and the result negation.
[0018] The PRI is determined using formula (10): (10); Where, X m The set of fuzzy set memberships representing conditions, SAT m The set of fuzzy set memberships of the results; Among them, the membership threshold and PRI value determine which condition combinations are considered valid or met. By setting appropriate thresholds, the stability and accuracy of the results are guaranteed.
[0019] The closer the consistency value is to 1, the more consistent the relationship between the condition combination and the result is (i.e., high consistency). The closer the consistency value is to 0, the less consistent the relationship between the condition combination and the result is (i.e., low consistency).
[0020] A subway travel chain satisfaction evaluation system based on three-stage configuration factor analysis includes a questionnaire generation module, a questionnaire processing module, a questionnaire result generation module and a display terminal; The questionnaire generation module generates a questionnaire according to the subway travel satisfaction rule; The questionnaire processing module performs redundancy processing on the results of the questionnaires generated by the questionnaire generation module, converts the data into fuzzy sets, constructs a fuzzy set truth table, and obtains the configuration factors that affect the subway travel satisfaction. The questionnaire result push module pushes the single factors of the ranking obtained by the questionnaire processing module to the display terminal; The display terminal displays the key factors affecting subway travel satisfaction, which serve as a basic reference for subway planning and operation.
[0021] The steps S2-S3 are completed by the questionnaire generation module; The steps S4-S6 are completed by the questionnaire processing module; Steps S7-S8 are completed by the questionnaire result generation module; The step S9 is completed through the display terminal.
[0022] Beneficial effects: The subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis of the present invention extracts observed variables (influencing factors) and screens the data after processing, and deletes the observed variables that cannot fully explain the perceived variables; then, an analyzable reflection system is established, and the reflection system is tested by using indicators. After the conditions are met, the observed variables of the data are merged into perceived variables to obtain a new data set; then, a fuzzy set for qualitative analysis is constructed on the basis of the new data set, and the necessary conditions are tested by Boolean logic and a truth table is constructed to perform qualitative analysis; then, the configuration solution that affects travel satisfaction is obtained, and the configuration factors that affect travel satisfaction are obtained from it; then, according to the concentrated distribution of the configuration factors, the travel stage (key stage) that has the greatest impact on travel satisfaction is determined; finally, through the obtained configuration influencing factors and key stages, the planning and operation departments are guided to improve the traffic conditions during the subway travel process, improve the passengers' subway travel satisfaction, and improve the subway share rate. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 It is a schematic diagram of the process of the present invention.
[0024] Figure 2 It is a system schematic diagram of the present invention.
[0025] Figure 3 This is the scatter heat map of SQ and SAT.
[0026] Figure 4 This is the scatter heat map of ETP and SAT. DETAILED DESCRIPTION
[0027] The present invention will be further described below in conjunction with the accompanying drawings.
[0028] like Figure 1 and Figure 2 As shown, a subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis includes the following steps: Step S1, dividing the subway trip into three travel chains, namely, the entry travel process (from home to the entrance subway station), the subway riding process, and the exit travel process (from the subway station to the destination); Step S2, information acquisition: obtaining the traveler's satisfaction (SAT) evaluation of the entire subway trip. The satisfaction evaluation method is as follows: the traveler directly selects the degree of satisfaction with the trip through the Likert 7-point scale; Step S3, obtaining basic information of travelers and travel perception information, the observation variables involved, and the evaluation scores of the factors involved are consistent with the evaluation method of satisfaction scores; The basic information includes: entrance travel time T 1、 Subway ride time 2、 Time T from the exit subway station to the destination3， Means of transportation used for entrance F 1c， Transportation from the exit subway station to the destination F 2e， Among them, c∈C, C={1, 2, ..., 5}, 1 represents walking, 2 represents shared bicycles, 3 represents electric bicycles, 4 represents public transportation, and 5 represents online car-hailing; e∈E, E={1, 2, ..., 5}, 1 represents walking, 2 represents shared bicycles, 3 represents company cars, 4 represents public transportation, and 5 represents online car-hailing.
[0029] The travel perception information includes: the perception of the ancillary facilities of the entrance subway station U 1j , Perception of services in subway stations 2j , Perception of convenience in subway stations 3j , Perception during subway ride 4j , Perception of the auxiliary facilities of the exit subway station 5j , Exit travel perception U 6j , Built Environment Perception 7j . Where j represents the jth factor, j∈J, j={1, 2, ..., J}.
[0030] The evaluation method of T1 is as follows: the acquired data is coded, and the code for travel time within 10 minutes is 1, the code for travel time between 10 and 20 minutes is 2, the code for travel time between 20 and 30 minutes is 3, and the code for travel time over 30 minutes is 4. The proportion of codes in the high travel satisfaction evaluation is calculated and assigned in descending order. The code ranked 1 is assigned "6" points, followed by "5", "3", and "2".
[0031] The T2 evaluation method is as follows: the acquired data is coded, and the travel time within 15 minutes is coded as 1, the travel time between 15 and 30 minutes is coded as 2, the travel time between 30 and 45 minutes is coded as 3, and the travel time over 45 minutes is coded as 4. The proportion of codes in the high travel satisfaction evaluation is calculated and assigned in descending order. The code ranked 1 is assigned "6" points, followed by "5", "3", and "2".
[0032] The T3 evaluation method is as follows: the acquired data is coded, and the travel time within 10 minutes is coded as 1, the travel time between 10 and 20 minutes is coded as 2, the travel time between 20 and 30 minutes is coded as 3, and the travel time over 30 minutes is coded as 4. The proportion of codes in the high travel satisfaction evaluation is calculated and assigned in descending order. The code ranked 1 is assigned "6" points, followed by "5", "3" and "2".
[0033] Among them, F 1iThe evaluation method is as follows: encode the acquired data, F 11 Coded as 1, F 12 Coded as 2, F 13 Coded as 3, F 14 Coded as 4, F 15 The code is 5. By calculating the proportion of codes in the high travel satisfaction evaluation, the code is assigned in descending order. The code ranked 1 is assigned "6", followed by "5", "3", and "2". Since this variable includes 5 objects, the last two digits with the lowest proportion are uniformly coded as "2".
[0034] Among them, F 2j The evaluation method is as follows: encode the acquired data, F 21 Coded as 1, F 22 Coded as 2, F 23 Coded as 3, F 24 Coded as 4, F 25 The code is 5. By calculating the proportion of codes in the high travel satisfaction evaluation, the code is assigned in descending order. The code ranked 1 is assigned "6", followed by "5", "3", and "2". Since this variable includes 5 objects, the last two digits with the lowest proportion are uniformly coded as "2".
[0035] Among them, F 1c The evaluation method is as follows: the score of the traveler's evaluation of the i-th factor is obtained, which is an integer in the range of [1, 7]. 2c ...F 6c The evaluation method is the same.
[0036] Obtain the land use attribute information of the traveler's starting subway station and exit subway station.
[0037] Among them, A n and E n The function is to confirm the land use attributes of subway entrances and exits. The method is as follows: confirm the land use attribute type within the 800-meter buffer zone through the names of the subway stations entering and exiting, and assign a value to the land use degree within the range. Specifically, collect the name of the starting subway station of the traveler as An, and the name of the exit subway station as En. Among them, n represents n. Download the land use attribute information around the subway station through OpeanMap, use ArcGIS to extract the land use attribute type and quantity within the 800-meter buffer zone around the subway station, and use formula (1) to evaluate the land use attributes; (1); Where: Mix is the subway station n, the land use type within the 800-m range is assigned, and p m is the ratio of the mth type of land use, and k depends on the number of different land uses in the station buffer.
[0038] Step S4, select research factors and construct a reflection system. (1) Use the PLS-SEM reflection system to select factors with analytical value, The observed variable (factor) U is calculated by formula (2): ij P i The path coefficient of the perception variable, the factor with a coefficient less than 0.7 is considered to be unable to fully express the perception variable P i Should be deleted; (2); In the formula, Indicates U ij Yes i The path coefficient, SAT j The set of satisfaction is represented by {sat1, sat2, ...sat j}, i∈I, I={1, 2...6}, j∈J, J={1, 2...j}.
[0039] (2) Use formulas (3) and (4) to determine whether the reflection system has reliability and validity. (3); (4); In the formula, express right The path coefficient is, i∈I, I=1, 2...6, j∈J, J=1, 2...J, represents the variance of the observed variable error.
[0040] Use formula (5) to determine the correlation between observed variables; (5); In the formula, R is the coefficient of determination.
[0041] Wherein, in step (5), CR is required to be greater than 0.7 and AVE is required to be greater than 0.5, indicating that the reflection system has reliability. When the square root of AVE of the perception variable is greater than the correlation coefficient with other perception variables, it indicates that the reflection system has validity.
[0042] When the VIF in step (6) is less than 5, it indicates that there is no collinearity between the observed variables. When the reflection system meets all the above index requirements, the next step of configuration research can be carried out.
[0043] S5: Transformation of measured variables into perceived variables; Use formula (6) to calculate the value of Pi (perception variable); (6); Wherein, i∈I, I={1, 2...6}, j∈J, J={1, 2...j} (integer).
[0044] S6: Convert data into fuzzy sets (this step should explain how to convert data into fuzzy sets), Use formula (7) to calculate P i The fuzzy set membership corresponding to each case is used to establish the fuzzy set.
[0045] (7); When condition X is considered to be a necessary condition for the result SAT, all high membership cases of X correspond to high membership of the result SAT. It is evaluated by two key indicators: consistency (Consistency for Necessity) and coverage (Coverage for Necessity). Calculated by formulas (8) and (9); (8); (9); In the formula, X m The fuzzy set membership of the condition, SAT m The resulting fuzzy set membership set.
[0046] Among them, when the consistency in step (9) is greater than 0.9 and the coverage is greater than 0.5, the condition is considered to be a necessary condition for the result to be generated, and only the combination with the condition being 1 is retained during truth table analysis.
[0047] S7: Construct the truth table and obtain the configuration solution. A truth table lists all possible combinations of causal conditions and observes whether these combinations lead to the occurrence of the result. For i binary conditions (for example, 1 represents "presence" and 0 represents "absence"), the total number of combinations is 2 i For each combination, the occurrence of the result is determined based on the actual data passing the threshold (fuzzy set membership greater than 0.8, proportional consistency reduction PRI greater than 0.50). PRI is determined using formula (10); (10); In the formula, X m The set of fuzzy set memberships representing conditions, SAT m The resulting fuzzy set membership is a collection of
[0048] S8: Get the configuration factors that affect subway travel satisfaction; S9: Count the frequency of occurrence of individual factors in the configuration factors and perform statistical sorting.
[0049] According to the content of the invention, the application is explained through actual cases.
[0050] Based on the collected 414 Nanjing subway travel satisfaction questionnaires, the required information was extracted, and the method proposed in the present invention was used to explore the influence of configuration factors on satisfaction. The collected data are shown in Table 1.
[0051] In step 1, the obtained satisfaction score index is defined as SAT, among which there are 319 samples that are satisfied with the trip (score greater than 4 points), 41 samples that are dissatisfied (score less than 4 points), and the rest are neutral. Only some samples are selected as examples (see Table 1).
[0052] Table 1. Original questionnaire display (part)
[0053] To T g , g={1, 2, 3} is encoded and assigned, and the results are shown in Table 2.
[0054] Table 2 Time T assignment results
[0055] F 1e / F 2c , e, c = {1, 2, 3, 4, 5}, the results after encoding and assignment are shown in Table 3.
[0056] Table 3 Connection mode F assignment results
[0057] Land use attribute U 71 with U 72 Directly use formula (1) to solve and obtain the results, as shown in Table 4.
[0058] Table 4 Land use attribute assignment (partial)
[0059] After solving formula (2), as shown in Table 5, Uij less than 0.7 needs to be deleted, and the factor selection is completed.
[0060] Table 5 Factor loadings
[0061] Note: The bold ones are observed variables that meet the load requirements; CR, AVE, validity test value, and VIF were calculated using relevant formulas, and the results are shown in Tables 6 to 8. The diagonal in Table 7 is the square root value of AVE, and the corresponding value of Pi is greater than the value other than the diagonal value.
[0062] Table 6 CR and AVE values
[0063] Table 7 Validity test
[0064] Table 8 VIF test
[0065] Step 3: Use formula (7) to complete data transformation and fuzzy set establishment. Thirty cases of fuzzy sets were randomly selected, as shown in Table 9.
[0066] Table 9 Fuzzy sets
[0067] The results of the necessity analysis are shown in Table 10. Figure 3 and Figure 4 The darker the color, the more points there are in that area. It can be seen that most of the data are not above the Y=X line, which indicates that the variables in this study do not have the necessary conditions to produce high satisfaction.
[0068] Table 10 Necessity test results
[0069] Note: The bold data are necessary conditions to be tested, and the judgment conditions are consistency>0.9 and coverage>0.5. “~” represents no, that is, the condition is missing.
[0070] The results of constructing the truth table are shown in Table 11.
[0071] Table 11 Truth table (partial)
[0072] After truth table analysis and deletion of configurations with consistency less than 0.8 and coverage less than 0.2, there are 7 groups of intermediate solutions that can be used to explain the results (Table 12).
[0073] Table 12 Configuration solution
[0074] Note:" " indicates that the condition exists, " " indicates missing conditions, and dark colors indicate core conditions. The original coverage rate refers to the proportion of cases with the configuration result; the unique coverage rate refers to the coverage rate obtained by eliminating the repeated parts with other combinations; the overall coverage rate and overall consistency indicate the explanatory power and overall consistency of all condition configurations for the results.
[0075] Among them, configuration 1 indicates that entrance facilities, subway services and exit travel experience are the dominant factors that determine the generation of high satisfaction, while the perception of taking the subway is missing as an auxiliary condition, indicating that it is not the main aspect that travelers pay attention to. Configuration 2 is dominated by entrance facilities, subway services and exit facility experience, and the exit travel process exists as an auxiliary condition for the generation of high satisfaction results. Configuration 3 is similar to configuration 1, but the requirements are slightly higher than configuration 1, showing that convenience in subway travel is equally important. Configuration 4 has a higher order than configuration 2, and exit facilities are also the core conditions that dominate the generation of high satisfaction. Configuration 5 expresses some groups that do not care about the entrance environment. Subway services, convenience, exit facilities and exit travel experience are the core conditions that dominate the generation of high satisfaction. The generation of high satisfaction in configurations 6 and 7 is both generated under the dominance of entrance facilities and exit facilities as core conditions, and non-core factors assist the generation of high satisfaction. Through the 7 groups of configurations, it can be concluded that the most important and least important factors for high satisfaction are entrance facilities and subway perception.
[0076] P1 and P5 appear most frequently in all configurations. P1 is used to measure the entrance travel stage and appears 6 times, while P5 is used to measure the exit travel stage and appears 5 times. This indicates that the entrance travel stage should be the first stage to be improved in the entire subway travel process, followed by the exit travel stage.
[0077] Among them, in the operation planning stage, it is necessary to find the key to improvement by reflecting the specific observation variables of P1 and P5. Since P1 is reflected by U12 and U14, among which U12 is observed by traffic signs and U14 is observed by road greening, it shows that the setting of traffic signs and greening coverage are the key contents that need to be paid attention to. P5 is reflected by U52 and U53. Among them, U52 is observed by the exit road conditions, and U53 is observed by the number of parking spaces at the terminal. It shows that in the process from the exit subway station to the terminal, the road conditions (flatness, non-motorized vehicle separation facilities, road width, etc.) and the parking spaces for non-motorized vehicles at the terminal directly affect the travel experience of subway travelers.
[0078] A subway travel chain satisfaction evaluation system based on three-stage configuration factor analysis includes a questionnaire generation module, a questionnaire processing module, a questionnaire result generation module and a display terminal; The questionnaire generation module generates a questionnaire according to the subway travel satisfaction rule; The questionnaire processing module performs redundancy processing on the results of the questionnaires generated by the questionnaire generation module, converts the data into fuzzy sets, constructs a fuzzy set truth table, and obtains the configuration factors that affect the subway travel satisfaction. The questionnaire result push module pushes the single factors of the ranking obtained by the questionnaire processing module to the display terminal; The display terminal displays the key factors affecting subway travel satisfaction, which serve as a basic reference for subway planning and operation.
[0079] The steps S2-S3 are completed by the questionnaire generation module; The steps S4-S6 are completed by the questionnaire processing module; Steps S7-S8 are completed by the questionnaire result generation module; The step S9 is completed through the display terminal.
[0080] Through the above factors, targeted improvements were made, including: 1. Ensure that traffic signs and markings are clear and visible, maintain them regularly, and improve and update them according to actual road conditions.
[0081] 2. Secondly, ensure that the green trees on the roads are pruned regularly, and that the ornamental green plants are cleaned and replaced promptly after their flowering period, to ensure that there are ornamental green vegetation on both sides of the roads in all seasons.
[0082] 3. Secondly, it is necessary to provide convenient shared bicycle return points (non-motor vehicle parking spaces) so that travelers can enjoy convenient transportation facilities without any other worries.
[0083] 4. Ensure that the road is safe for travelers from the exit subway station to the destination, that is, install non-motorized vehicle separation facilities when the road width permits, to ensure the safety of travelers and enable them to reach the destination smoothly and quickly in the event of lateness or other emergencies.
[0084] Through the above improvement measures, the travel experience of subway passengers can be directly and effectively improved, thereby attracting potential travelers to use the subway.
[0085] In addition, the questionnaire collection module can be used to regularly collect passenger satisfaction questionnaires. Through the questionnaire processing module and the result generation module, different measurement variables can be used for analysis and comparison to identify whether the factors that passengers pay attention to when traveling at different times are different. Finally, the planning operator can use the display terminal to discover changes in passenger needs and make improvements. This will form a virtuous cycle and continuously improve the service quality of subway travel, so that more potential travelers (private car commuters) can travel green and achieve green and sustainable transportation development.
[0086] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.
Claims
1. A subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis, characterized in that: The following steps are involved: S1: Divide the subway travel chain and formulate subway travel satisfaction rules; S2: Prepare questionnaires, distribute and collect questionnaires, and obtain data sets; S3: Obtain basic information of travelers and travel perception information and assign values; S4: Determine the perceived and observed variables involved in the analysis; S5: Transformation of measured variables into perceived variables; S6: Convert data into fuzzy sets; S7: Construct fuzzy set truth table; S8: Get the configuration factors that affect subway travel satisfaction; S9: Count the frequency of occurrence of individual factors in the configuration factors and perform statistical sorting.
2. The subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis according to claim 1 is characterized by: In step S1, the satisfaction evaluation method is as follows: the traveler directly selects the travel satisfaction level through the Likert 7-point scale.
3. The subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis according to claim 1 is characterized by: In the step S3, the basic information includes: the entrance travel time T1, the subway ride time T2, the time from the exit subway station to the destination T3, the transportation tool F1c used for the entrance travel, where c∈C, c=1, 2, ...4, 5, 1 represents walking, 2 represents shared bicycles, 3 represents electric bicycles, 4 represents buses, and 5 represents online car-hailing, the transportation tool F2e used from the exit subway station to the destination, e∈E, e=1, 2, 3, 4, 5, 1 represents walking, 2 represents shared bicycles, 3 represents company cars, 4 represents buses, 5 represents online car-hailing, the entrance subway station name An and the exit subway station name En, and n represents the nth subway station at the entrance; The evaluation method of T1 is as follows: the acquired data is coded, the travel time within 10 minutes is coded as 1, the travel time between 10 and 20 minutes is coded as 2, the travel time between 20 and 30 minutes is coded as 3, and the travel time over 30 minutes is coded as 4. The proportion of a certain number of codes to the number of samples with high travel satisfaction evaluation is calculated, and they are ranked in descending order and assigned values. The code ranked as 1 is assigned a value of "6", followed by "5", "3", and "2" points. The T2 evaluation method is as follows: encode the acquired data, with the travel time within 15 minutes coded as 1, the travel time between 15 and 30 minutes coded as 2, the travel time between 30 and 45 minutes coded as 3, and the travel time over 45 minutes coded as 4. Calculate the proportion of a certain number of codes to the sample size of high travel satisfaction evaluations, sort them in descending order, and assign values. The code ranked 1 is assigned a value of "6", followed by "5", "3", and "2" points. The T3 evaluation method is as follows: encode the acquired data, the travel time within 10 minutes is coded as 1, the travel time between 10 and 20 minutes is coded as 2, the travel time between 20 and 30 minutes is coded as 3, and the travel time over 30 minutes is coded as 4. By calculating the proportion of a certain number of codes to the number of samples with high travel satisfaction evaluation, they are ranked in descending order and assigned values. The code ranked as 1 is assigned a value of "6", followed by "5", "3", and "2" points. Among them, the evaluation method of F1c is as follows: encode the acquired data, F11 is coded as 1, F12 is coded as 2, F13 is coded as 3, F14 is coded as 4, and F15 is coded as 5. By calculating the proportion of a certain code number to the sample size of high travel satisfaction evaluation, they are sorted in descending order and assigned values. The code ranked as 1 is assigned a value of "6", followed by "5", "3", and "2". The last two codes with the lowest proportion are uniformly coded as "2"; Among them, the F2e evaluation method is as follows: encode the acquired data, F21 is coded as 1, F22 is coded as 2, F23 is coded as 3, F24 is coded as 4, and F25 is coded as 5. By calculating the proportion of a certain code number to the sample size of high travel satisfaction evaluation, they are sorted in descending order and assigned values. The code ranked as 1 is assigned a value of "6", followed by "5", "3", and "2". The last two codes with the lowest proportion are uniformly coded as "2"; Mixn is the land use degree of the 800-meter buffer zone of the nth subway entrance, specifically: （1）； Where: p m is the ratio of the mth type of land use, and k is the number of different land uses in the station buffer.
4. The subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis according to claim 1 is characterized by: In step S4, the travel perception information includes: the perception P1 of the ancillary facilities of the entrance subway station, using the observation variable U 1j Reflects the perception of the service in the subway station P2, using the observed variable U 2j Reflects the perception of convenience in the subway station P3, using the observed variable U 3j Reflects the perception P4 during the subway ride, using the observed variable U 4j Reflects the perception of the ancillary facilities of the exit subway station P5, using the observed variable U 5j Reflect, exit travel perception P6, using observed variables U 6j Reflect, where j represents the jth factor, j∈J, J={1, 2, ...j}, where U 1j , U 2j , U 3j , U 4j , U 5j , U 6j All of them reflect the perception variables P1, P2, P 3， The observed variables of P4, P5, and P6 and the evaluation of the observed variables of travelers were obtained using a 7-point Likert scale.
5. The subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis according to claim 4 is characterized by: Calculate the path coefficient of the observed variable to the perceived variable ; (2)； In the formula, Indicates U ij P i The path coefficient of SAT j The set of satisfaction is represented by {sat1, sat2, ...sat j }, i∈I, I=1, 2...6, j∈J, J=1, 2...J, coefficient When it is less than 0.7, delete the corresponding Uij under Pi, Var represents the function of variance; the values of CR, AVE, and VIF are calculated as follows: （3）； （4）； In the formula, θ ij represents the variance of the observed variable error, （5）； In the formula, R is the coefficient of determination; The evaluation index requires that CR is greater than 0.7, AVE is greater than 0.5, and VIF is less than 5, wherein the square root of the AVE of Pi obtained is greater than the correlation coefficient with other perceptual variables.
6. The subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis according to claim 1 is characterized by: In step S5, the formula for converting the measured variable into the perceived variable is as follows: （6）； In the formula, i∈I, I=1, 2..., 7, j∈J, J=1, 2..., J, all the observed variables U involved in the samples are ij The effect is attributed to the perception variable P i , that is, multiple U ij The set of P is transformed into a single i The set of P i New dataset Ω1.
7. The subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis according to claim 6 is characterized by: In step S6, the data is converted into a fuzzy set, the upper quartile is defined as Q3, the median is defined as Q2, and the lower quartile is defined as Q1. The fuzzy set membership degree of each case corresponding to Pi is calculated using formula (7) to establish a fuzzy set: （7）； Where P i is the set of variables in Ω1, P i 25% , P i 75% Respectively represent the 25% quantile and 75% quantile of the set of the i-th perceptual variable P, m=i, i∈I, I={1, 2, 3, 4, 5, 6, 7}; After the data set is converted to data set Ω2 by formula (7), the original values in the Pi set are converted to membership values between (0, 1); The set of SAT is converted into membership using formula (7). The converted SAT is a number between (0, 1). SAT is {sat1, sat2, ...sat m }, m∈M, M={1, 2, ...i}; Next, we need to confirm whether the necessary conditions have an impact on the subsequent analysis. This can be evaluated using two key indicators: consistency for necessity and coverage for necessity. These indicators can be calculated using formulas (8) and (9): （8）； （9）； Where, X m represents the fuzzy set membership of the condition, that is, the i-th condition variable in the data set Ω2, SAT m The fuzzy set membership set representing the result, each condition X in the sample corresponds to a result SAT; Among them, when the consistency is greater than 0.9 and the coverage is greater than 0.5, the condition is considered to be a necessary condition for the result, and only the combination with the condition of 1 is retained during truth table analysis.
8. The subway travel chain satisfaction evaluation method based on three-stage configuration factor analysis according to claim 1 is characterized by: The specific steps of step S7 are as follows: S71, constructing the original truth table: constructing an initial table through the data set Ω2, where each row represents the condition configuration of a case and each column represents the corresponding result value; S72. Identify the minimum conditional combinations: Identify which conditional combinations are most common in the observed results; S73, remove redundant items: select the smallest item that can cover all combinations with output 1; S74. Determine the final Boolean expression: the final Boolean expression is the union of the merged minimum terms.