Comprehensive evaluation method for safety and comfort of urban expressway maintenance operation area

Abstract

The invention discloses an urban expressway maintenance operation area safety comfort comprehensive evaluation method, which comprises the steps of S1, according to length setting of a warning area and an upstream transition area, constructing a simulation experiment scene of an urban expressway construction area; s2, carrying out a driving simulation experiment based on the simulation experiment scene, and obtaining driving feature data of different drivers in the driving process; s3, ANP-EWM combination weighting is adopted to solve the weight of each evaluation index, and a comprehensive weight vector W is constructed; s4, grading the evaluation indexes through a K-means clustering algorithm, and establishing a safety comfort evaluation standard; and S5, determining a membership value according to the evaluation index data and the established evaluation standard, constructing a membership matrix, and calculating a fuzzy comprehensive evaluation result vector and a comprehensive evaluation grade feature value by combining the comprehensive weight vector. The method can effectively evaluate the influence of length combinations of different warning areas and upstream transition areas on driving safety and comfort, and can provide reference for determining the length of an urban expressway construction operation area.

Description

Technical Field

[0001] The present invention relates to the field of highway traffic safety technology, and in particular to a comprehensive evaluation method for safety and comfort in a maintenance operation area of an urban expressway. Background Art

[0002] Urban expressways are a vital component of urban roads, playing a crucial role in improving urban traffic efficiency and shortening travel times. After decades of development, my country's expressway system has gradually improved, and urban expressway maintenance and construction operations have become a regular occurrence. During urban expressway construction, lane closures are necessary, complicating the driving environment, reducing road capacity, and ultimately degrading driving safety and comfort in construction areas. These maintenance and construction areas have become not only bottlenecks on urban expressways but also frequent accident sites.

[0003] According to the standard (GB 5768.4-2017) issued by the Standardization Administration of the People's Republic of China, a road work zone consists of six areas: a warning zone, an upstream transition zone, a buffer zone, a work zone, a downstream transition zone, and a termination zone. The warning zone is where drivers adjust their speed and find a gap to change lanes, and its length is crucial for traffic flow. The upstream transition zone is designed to prevent sudden changes in traffic flow during lane changes, ensuring smoother flow. Its length is particularly important for smooth merging. Research indicates that traffic accidents in the warning and upstream transition zones account for 80% of all accidents in the work zone. Therefore, properly designing the length of these zones is crucial for improving driving safety and comfort in construction zones.

[0004] To this end, the present invention provides a comprehensive evaluation method for the safety and comfort of urban expressway maintenance work areas, which can effectively evaluate the impact of different combinations of warning zone and upstream transition zone lengths on driving safety and comfort, and plays an important guiding role in determining the length of maintenance work areas in urban expressway sections. Summary of the Invention

[0005] In view of this, in order to improve the effectiveness and rationality of the safety and comfort evaluation of different length combinations of warning zones and upstream transition zones, the present invention developed a comprehensive safety and comfort evaluation method for maintenance operation areas consisting of three levels: visual effect, cognitive effect and driving effect by integrating eye movement, EEG and manipulation behavior indicators.

[0006] In order to achieve the above technical objectives, the technical solution adopted by the present invention is: a comprehensive evaluation method for the safety and comfort of urban expressway maintenance work areas, comprising:

[0007] S1. Setting the lengths of the warning zone and the upstream transition zone in the construction zone of the expressway section, and constructing a driving simulation experiment scenario in the construction zone of the urban expressway section based on the length settings of the warning zone and the upstream transition zone;

[0008] S2. Conducting a driving simulation experiment based on the constructed driving simulation experiment scenario to obtain driving characteristic data of different drivers during driving, wherein the driving characteristic data includes a plurality of evaluation index data;

[0009] S3, use ANP-EWM combined weighting to obtain the weight of each evaluation index and construct the comprehensive weight vector W;

[0010] S4. Use K-means clustering algorithm to classify the evaluation indicators and establish safety and comfort evaluation standards;

[0011] S5. Determine the membership value based on the evaluation index data and the established evaluation criteria, construct a membership matrix, calculate the fuzzy comprehensive evaluation result vector and the comprehensive evaluation grade eigenvalue based on the comprehensive weight vector, and evaluate the experimental scenario based on the comprehensive evaluation grade eigenvalue.

[0012] In some embodiments, the number of drivers (test samples) in the driving simulation experiment in S2 is determined based on the expected variance, target confidence, and error margin, and the calculation formula is shown in formula (1):

[0013]

[0014] Where: N is the experimental sample size; Z is the standard normal distribution statistic, σ is the standard deviation, and E is the maximum error.

[0015] In some embodiments, the evaluation index data in the driving characteristic data in S2 includes pupil area U 11 , scanning frequency U12, α wave absolute power U21, β wave absolute power U22, θ / β value U23, longitudinal acceleration U31, instantaneous speed entering the working area U32, lane change duration U33.

[0016] In some embodiments, S3 specifically includes the following steps:

[0017] S3.1. ANP determines subjective weights

[0018] Experts measure the relative importance of each evaluation indicator and score them on a scale of 1-9. The subjective weight of each evaluation indicator is calculated using the ANP method and Super Decision software. cj , and finally the subjective weight vector Wc=[w c1 , w c2 ,…,w cn ];

[0019] S3.2. Determine objective weights using entropy weight method

[0020] The original data obtained during the driver simulation experiment is normalized to obtain the normalized decision matrix B shown in formula (2):

[0021]

[0022] Where: b ij —The normalized value of the jth evaluation indicator of the i-th evaluation object;

[0023] According to the normalized decision matrix, the entropy value of the jth indicator in the evaluation index system is calculated:

[0024]

[0025] Where: m is the number of evaluation objects; n is the number of evaluation indicators; A is the translation amount;

[0026] Determine the objective weight w of each indicator pj :

[0027]

[0028] S3.3. Portfolio empowerment

[0029] The subjective and objective weights are combined and weighted using formula (6):

[0030]

[0031] Where: w j is the weight of the jth indicator; w cj is the subjective weight of the jth indicator; w pj is the objective weight of the jth indicator;

[0032] S3.4. Construct a comprehensive weight vector W, where W = (w1, w2, ..., w m ).

[0033] In some embodiments, the translation amount A is 0.00001.

[0034] In some embodiments, S4 specifically includes:

[0035] The original data obtained during the driver simulation experiment were normalized so that the data range fell within the interval [0,1].

[0036] Then, SPSS software was used to classify the evaluation indicators using the K-means clustering algorithm to establish a safety and comfort evaluation standard.

[0037] In some embodiments, S5 specifically includes the following steps:

[0038] S5.1. Determine the factor set U and the evaluation set V: The factor set is a collection of factors that characterize the evaluation object and is represented by U: U = (u1, u2, ..., u m ); the evaluation set is a set of various possible results that the evaluator may make on the evaluation object, represented by V: V=(v1,v2,…,v n ), the evaluation set is generally divided into five levels, and each level is regarded as a fuzzy subset;

[0039] S5.2. Determine the degree of membership: The degree of membership of each factor indicator is determined by the membership function;

[0040] S5.3. Construct the membership matrix R: Make the evaluation indicators dimensionless and use min-max normalization to the interval [0,1];

[0041] Determine the membership of the factor set to the evaluation set and construct the membership matrix, which is represented by R:

[0042]

[0043] Where: r ij is the membership degree of the i-th element in the factor set U to the j-th level in the evaluation set V;

[0044] S5.4. Obtaining the fuzzy comprehensive evaluation result vector S: The comprehensive weight vector W and the membership matrix R are synthesized to obtain the fuzzy comprehensive evaluation result vector S of each evaluation object. The expression of S is shown in formula (8):

[0045]

[0046] Where: s i For the evaluation object v j The membership value of the level.

[0047] S5.5. Analysis of evaluation results: The membership degree is processed using the weighted average method to calculate the comprehensive evaluation grade characteristic value G. The expression of G is shown in formula (9):

[0048]

[0049] Where: k is the unknown coefficient, ε j It represents the value corresponding to the j-th evaluation level and is assigned a weight value according to the actual evaluation system.

[0050] In some embodiments, in S52 , a relatively small, medium, or large trapezoidal membership function is selected to determine the membership value in the single factor evaluation vector.

[0051] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art:

[0052] Existing research primarily focuses on highway construction zones. The lengths of warning zones and upstream transition zones are set solely based on road traffic characteristics, without considering the driver's driving characteristics or the impact of different section length combinations. While urban expressways are similar to highways, they differ in terms of traffic functions and characteristics, making it difficult to directly apply existing research findings. To this end, the present invention conducts indoor driving simulation experiments based on a simulation environment, selecting driver eye movement indicators, EEG indicators, and maneuvering behavior indicators as indicators for evaluating driving safety and comfort. This effectively evaluates the impact of different warning zone and upstream transition zone length combinations on driving safety and comfort, providing a reference for determining the length of urban expressway construction zones. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0054] Figure 1 It is a driving simulation scenario for a two-way six-lane urban expressway;

[0055] Figure 2 It is the DSR-1000TS2.0 driving simulation system;

[0056] Figure 3 is the distribution of pupil area of subjects in different experimental scenarios;

[0057] Figure 4 is the distribution of subjects’ scanning frequency in different experimental scenarios;

[0058] Figure 5 is the distribution of the absolute power values of the subjects' α waves in different experimental scenarios;

[0059] Figure 6 is the distribution of the absolute power values of the subjects' β waves in different experimental scenarios;

[0060] Figure 7 is the distribution of θ / β values of subjects in different experimental scenarios;

[0061] Figure 8 is the distribution of longitudinal acceleration of the subjects in different experimental scenarios;

[0062] Figure 9 is the instantaneous speed distribution of the subjects entering the work area under different experimental scenarios;

[0063] Figure 10 is the distribution of lane-changing duration of subjects in different experimental scenarios;

[0064] Figure 11 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0065] The present invention will be described in further detail below with reference to the accompanying drawings and examples. It is particularly noted that the following examples are intended only to illustrate the present invention and are not intended to limit the scope of the present invention. Similarly, the following examples are only some embodiments of the present invention and are not intended to be exhaustive. All other embodiments obtained by those of ordinary skill in the art without creative effort are intended to fall within the scope of protection of the present invention.

[0066] Refer to the attached Figure 11 As shown, the present invention provides a comprehensive evaluation method for safety and comfort in urban expressway maintenance work areas, comprising:

[0067] S1. Setting the lengths of the warning zone and the upstream transition zone in the construction zone of the expressway section, and constructing a driving simulation experiment scenario in the construction zone of the urban expressway section based on the length settings of the warning zone and the upstream transition zone;

[0068] S2. Conduct a driving simulation experiment based on the constructed driving simulation experiment scenario to obtain driving characteristic data of different drivers during the driving process. The number of drivers in the driving simulation experiment is determined based on the expected variance, target confidence, and error margin. The calculation formula is shown in formula (1):

[0069]

[0070] Where: N is the experimental sample size; Z is the standard normal distribution statistic, σ is the standard deviation, and E is the maximum error.

[0071] The driving characteristic data includes multiple evaluation index data, including:

[0072] Pupil area U11 (pixel): Changes in the pupil can reflect the driver's visual comfort level. When the visual effect is poor, the pupil will dilate, and when the visual effect is good, the pupil will shrink. At the same time, during the operation, the pupil area will also change, and the magnitude of the change is closely related to the difficulty of the operation.

[0073] Scan rate (U12 times / second): Rapid eye movements from one point to another, with a dwell time of less than 120 milliseconds, are considered scans. Scan rate is the number of scans per unit time. When the driver is familiar with the driving environment or the road is simple, the scan rate is low. When the driver is unfamiliar with the driving environment or the road is complex, the scan rate is high, indicating greater search difficulty.

[0074] α wave absolute power U21(μV 2 ): When the driver is physically relaxed and focused, the absolute power of the alpha waves will increase. The driver will be more relaxed, with a moderate cognitive load, and will be more able to calmly deal with sudden road situations, which is more beneficial to driving safety.

[0075] β wave absolute power U22(μV 2 Beta waves typically occur when a person is exposed to external stimuli, experiencing heightened alertness and a high cognitive load. An increase in the absolute power of beta waves often signals a transition from alertness to tension, leading to increased energy consumption and fatigue, which can reduce driving safety and comfort.

[0076] θ / β value U23: When the θ wave pattern increases, the brain is in a relaxed state, indicating that the person's mental stress is relieved. Therefore, the θ / β value can represent changes in cognitive level. The larger the θ / β value, the more relaxed the driver is, the more moderate the cognitive load is, and the driver is more able to calmly deal with sudden road situations, which is more beneficial to driving safety.

[0077] Longitudinal acceleration U31(m / s 2 ) refers to the change in vehicle speed per unit time during braking, which can intuitively reflect the driver's operational stability during the speed change process. Sudden acceleration or deceleration is considered to be a relatively unsafe driving behavior. The greater the fluctuation in acceleration, the greater the driving risk.

[0078] The instantaneous speed U32 (km / h) upon entering the work zone refers to the vehicle speed in the normal direction when entering the construction work zone. To a certain extent, it reflects the rationality of the length setting of the warning zone and the upstream transition zone, as well as the speed control effect of signs and markings.

[0079] Lane Change Duration (U33s): This is the difference between the time it takes the vehicle to reach the centerline of the target lane and the time it first crosses the lane line. This duration can, to a certain extent, reflect the difficulty of the lane change.

[0080] S3. Use ANP-EWM combined weighting to obtain the weights of each evaluation indicator and construct a comprehensive weight vector W, which specifically includes:

[0081] S3.1. ANP determines subjective weights

[0082] ANP (Analytical Network Process) is a subjective weighting method that uses experts to measure the relative importance of each evaluation indicator and score them on a scale of 1-9. Using the ANP method, the subjective weight of each evaluation indicator is calculated using Super Decision software. cj , and finally the subjective weight vector Wc=[w c1 , w c2 ,…,w cn ];

[0083] S3.2. Determine objective weights using entropy weight method

[0084] The entropy weight method is used to determine the weights of the driver reaction evaluation index system, which can objectively and truly reflect the evaluation effect. In order to eliminate the inconsistency of units and dimensions between the index data, the raw data obtained during the driver simulation experiment are normalized. The present invention specifically uses the minimum-maximum normalization method to normalize the data and obtain the normalized decision matrix B shown in formula (2):

[0085]

[0086] Where: b ij —The normalized value of the jth evaluation indicator of the i-th evaluation object;

[0087] According to the normalized decision matrix, the entropy value of the jth indicator in the evaluation index system is calculated:

[0088]

[0089] Where: m is the number of evaluation objects; n is the number of evaluation indicators; A is the translation (A is 0.00001). In order to ensure the integrity and validity of the data and make the logarithm in formula (3) meaningful, it is necessary to add a translation to the value of 0 after normalization so that it can become a small positive value.

[0090] Determine the objective weight w of each indicator pj :

[0091]

[0092] S3.3. Portfolio empowerment

[0093] The subjective and objective weights are combined and weighted using formula (6):

[0094]

[0095] Where: w j is the weight of the jth indicator; w cj is the subjective weight of the jth indicator; w pj is the objective weight of the j-th indicator.

[0096] S3.4. Construct a comprehensive weight vector W, where W = (w1, w2, ..., w m ).

[0097] S4. Use the K-means clustering algorithm to classify the evaluation indicators and establish a safety and comfort evaluation standard, which includes:

[0098] In order to eliminate the differences between the index data (units and dimensions are not uniform), the original data obtained during the driver simulation experiment were normalized so that the data range falls within the interval [0,1].

[0099] Then, SPSS software was used to classify the evaluation indicators using the K-means clustering algorithm to establish a safety and comfort evaluation standard.

[0100] S5. Determine the membership value based on the evaluation index data and the established evaluation criteria, construct a membership matrix, calculate the fuzzy comprehensive evaluation result vector and the comprehensive evaluation grade eigenvalue based on the comprehensive weight vector, and evaluate the experimental scenario based on the comprehensive evaluation grade eigenvalue. Specifically, the steps include:

[0101] S5.1. Determine the factor set U and the evaluation set V: The factor set is a collection of factors that characterize the evaluation object and is represented by U: U = (u1, u2, ..., u m ); the evaluation set is a set of various possible results that the evaluator may make on the evaluation object, represented by V: V=(v1,v2,…,v n ), the evaluation set is generally divided into five levels, and each level is regarded as a fuzzy subset;

[0102] S5.2. Determine the degree of membership: The degree of membership of each factor indicator is determined by a membership function. Small, medium, and large trapezoidal membership functions are selected to determine the degree of membership in the single factor evaluation vector.

[0103] S5.3. Construct the membership matrix R: Make the evaluation indicators dimensionless and use min-max normalization to the interval [0,1]. Determine the membership of the factor set to the evaluation set and construct the membership matrix, which is represented by R:

[0104]

[0105] Where: r ij is the membership degree of the i-th element in the factor set U to the j-th level in the evaluation set V;

[0106] S5.4. Obtaining the fuzzy comprehensive evaluation result vector S: The comprehensive weight vector W and the membership matrix R are synthesized to obtain the fuzzy comprehensive evaluation result vector S of each evaluation object. The expression of S is shown in formula (8):

[0107]

[0108] Where: s i For the evaluation object v j The membership value of the level indicates the degree of membership of level j.

[0109] S5.5. Analysis of evaluation results: The membership degree is processed using the weighted average method to calculate the comprehensive evaluation grade characteristic value G. The expression of G is shown in formula (9):

[0110]

[0111] Where: k is the unknown coefficient, ε j It represents the numerical value corresponding to the j-th evaluation level, and is generally assigned a weight value based on the actual evaluation system.

[0112] On the basis of effectively solving the problem of failure in determining the maximum membership degree in traditional fuzzy comprehensive evaluation, the present invention makes the comprehensive evaluation grade characteristic value G more intuitively reflect the specific results of the evaluation object.

[0113] Example 1

[0114] This embodiment provides a specific implementation case based on the solution of the present invention, which is as follows:

[0115] 1. Experimental scenario

[0116] Since existing regulations do not specify the lengths of various sections in urban expressway construction zones, the standards for Class I highways in the "Highway Maintenance Safety Operations Regulations" were used as a reference. The warning zone, while meeting the standard minimum length of 1,600 meters, was gradually increased by 200 meters, for a total of six warning zone lengths. The upstream transition zone, while meeting the standard minimum length of 100 meters, was gradually increased by 20 meters, for a total of six upstream transition zone lengths. Furthermore, based on the principle of least favorable conditions, an 80-meter buffer zone, a 1,000-meter work zone, a 30-meter downstream transition zone, and a 30-meter termination zone were established.

[0117] Using the control variable method, VISSIM simulation models of six different warning zone lengths and upstream transition zone lengths were constructed respectively, and five evaluation indicators, including average queue length, number of traffic conflicts, average travel delay, average travel time and number of stops, were output. Based on the entropy weight method, a comprehensive evaluation of the warning zone length and upstream transition zone length was conducted respectively. The driving simulation experiment control variables of 1800m, 2000m, and 2200m warning zone lengths and 120m, 140m, and 160m upstream transition zone lengths were selected for research.

[0118] The driving simulation experiment scene was set up using the control variable method to reduce the interference of non-experimental factors. Except for the difference in the length of the warning area and the upstream transition area, the other elements remained consistent. The experiment selected a two-way six-lane urban expressway as the experimental scene. The total length is about 8km, the width of a single lane is 3.5m, the design speed is 100km / h, and the speed limit in the work area is 60km / h. The work area is arranged in accordance with "Road Traffic Signs and Markings Part 4: Work Area", and the working condition is set to half-closed outer lane operation ( Figure 1 ).

[0119] Considering the impact of traffic flow, a steady flow of 2000 pcu / h was selected as the traffic load for this experiment. For three warning zone lengths and three upstream transition zone lengths, a total of 3 × 3 = 9 experimental scenarios were set, as shown in Table 1.

[0120] Table 1 Experimental scene settings

[0121]

[0122] 2. Number of subjects

[0123] The sample size of the driving simulation experiment is determined based on the expected variance, target confidence level, and error margin. The calculation formula is shown in formula (1). Where: N is the experimental sample size; Z is the standard normal distribution statistic. The confidence level of this experiment is 95%, so Z = 1.96; σ is the standard deviation, which ranges from 0.25 to 0.5

[96] and is 0.30 in this experiment; E is the maximum error, which is 10%. The minimum sample size calculated by the formula is 35.

[0124] To ensure a sufficient sample size after eliminating invalid data, this experiment recruited 45 drivers (25 males and 20 females) aged 20 to 28 years (mean M = 23.94 years, standard deviation SD = 1.78) with 1 to 6 years of driving experience (mean M = 1.56 years, standard deviation SD = 1.87). All participants held a Class 1 driver's license, had normal visual function, and had a visual acuity of 1.0 or higher, either uncorrected or corrected. All participants had experience driving on urban expressways and were unfamiliar with the driving simulations. All participants were aware of the experimental procedures and signed informed consent. For 24 hours prior to the experiment, participants were ensured to eat and rest normally, refrain from alcohol or medication, and avoid strenuous exercise.

[0125] 3. Experimental equipment

[0126] The experiment collected data on drivers' eye movements, EEG, and driving behavior while driving in a construction zone on a two-way six-lane urban expressway. The experiment was conducted indoors to control the effects of climate, light, and noise. Figure 2 The instruments required for the experiment are:

[0127] Driving simulation system

[0128] This experiment uses the DSR-1000TS2.0 driving simulation system, which consists of a cockpit system, a vehicle dynamics system, a scene display system, and an independent console. It can record and export driving behavior parameters and vehicle dynamics parameters, such as vehicle speed, vehicle acceleration and deceleration, etc.

[0129] Dikablis Glasses Eye Tracker

[0130] A Dikablis Glasses eye tracker was used to track and measure the driver's eye movement behavior on a simulated road section. The device, working in conjunction with D-Lab analysis software, recorded real-time eye activity data, such as pupil area, fixation duration, and scan frequency. The eye tracker collected samples at a 60Hz frequency, with an accuracy range of 0.1° to 0.3°.

[0131] 32-channel NE wireless EEG system

[0132] This experiment used a 32-channel NE wireless EEG device to collect EEG data from drivers during a driving simulation. This device, with 32 experimental channels, a 500 SPS sampling rate, and 24-bit resolution, uses wireless technology to transmit EEG data and obtain EEG signals, such as alpha and beta waves, providing strong support for subsequent data processing and analysis.

[0133] 4. Evaluation indicators

[0134] The driver's psychophysiological characteristics and driving behavior are the most important factors affecting driving safety. The driver's eye movement, EEG, and driving behavior data can reflect their safety and comfort levels throughout the driving process. This experiment collects the following eight indicators to assess the driver's driving safety and comfort.

[0135] Visual effect (eye movement indicator U1)

[0136] Pupil area U11 (pixel): Changes in the pupil can reflect the driver's visual comfort level. When the visual effect is poor, the pupil will dilate, and when the visual effect is good, the pupil will shrink. At the same time, during the operation, the pupil area will also change, and the magnitude of the change is closely related to the difficulty of the operation.

[0137] Scan rate (U12 times / second): Rapid eye movements from one point to another, with a dwell time of less than 120 milliseconds, are considered scans. Scan rate is the number of scans per unit time. When the driver is familiar with the driving environment or the road is simple, the scan rate is low. When the driver is unfamiliar with the driving environment or the road is complex, the scan rate is high, indicating greater search difficulty.

[0138] Cognitive effect (EEG index U2)

[0139] α wave absolute power U21 (μV 2 ): When the driver is physically relaxed and focused, the absolute power of the alpha waves will increase. The driver will be more relaxed, with a moderate cognitive load, and will be more able to calmly deal with sudden road situations, which is more beneficial to driving safety.

[0140] β wave absolute power U22(μV 2 Beta waves typically occur when a person is exposed to external stimuli, experiencing heightened alertness and a high cognitive load. An increase in the absolute power of beta waves often signals a transition from alertness to tension, leading to increased energy consumption and fatigue, which can reduce driving safety and comfort.

[0141] θ / β value U23: When the θ wave pattern increases, the brain is in a relaxed state, indicating that the person's mental stress is relieved. Therefore, the θ / β value can represent changes in cognitive level. The larger the θ / β value, the more relaxed the driver is, the more moderate the cognitive load is, and the driver is more able to calmly deal with sudden road situations, which is more beneficial to driving safety.

[0142] Driving performance (control index U3)

[0143] Longitudinal acceleration U31(m / s 2) refers to the change in vehicle speed per unit time during braking, which can intuitively reflect the driver's operational stability during the speed change process. Sudden acceleration or deceleration is considered to be a relatively unsafe driving behavior. The greater the fluctuation in acceleration, the greater the driving risk.

[0144] The instantaneous speed U32 (km / h) upon entering the work zone refers to the vehicle speed in the normal direction when entering the construction work zone. To a certain extent, it reflects the rationality of the length setting of the warning zone and the upstream transition zone, as well as the speed control effect of signs and markings.

[0145] Lane Change Duration (U33s): This is the difference between the time it takes the vehicle to reach the centerline of the target lane and the time it first crosses the lane line. This duration can, to a certain extent, reflect the difficulty of the lane change.

[0146] 5. Experimental process

[0147] First, the subjects were asked to sign an experimental consent form and fill out a personal information questionnaire, which included name, gender, age, years of driving experience, and mental state on that day.

[0148] Before the formal experiment began, participants read the experimental instructions and were briefed on the experimental tasks. They then practiced driving in a non-experimental setting to familiarize themselves with the simulator. After becoming familiar with the controls, participants were fitted with an EEG cap and eye tracker and calibrated. Participants were free to leave the experiment at any time if they felt uncomfortable.

[0149] The experimenter created experimental number information for the subjects in D-lab and started the formal experiment. In order to eliminate the interference of the scene order effect on the experimental results, the experimental scene was randomly loaded for each subject. According to the experimental requirements, the subjects entered the experimental section at a speed of 100km / h and needed to change their driving behavior and control the driving speed according to the prompts on the signboards. During the experiment, the experimenter needed to confirm whether the subject's EEG signals and eye movement images were normal. If there was any gaze point deviation, the instrument needed to be recalibrated before continuing the experiment. After each scene was completed, the subject had a 5-minute break until the nine scene experiments were completed. The above experimental process was repeated until all 45 test drivers completed the experiment.

[0150] Because EEG signals are extremely weak and highly sensitive to interference from various environmental factors, the experimental environment must be kept quiet throughout the experiment to prevent external factors from interfering with the subject's EEG. If a subject experiences an accident at the experimental site, such as hitting another car or running off the road, they will be asked to stop and repeat the experiment.

[0151] 6. Data Preprocessing

[0152] The raw eye movement data was imported into D-LAB software, and eye movement metrics such as pupil area and saccade frequency were extracted based on the determined AOI. Finally, the driving behavior data was exported from the vehicle dynamics module, and the collected EEG data was exported to MATLAB software for preliminary analysis.

[0153] 7 Experimental results

[0154] Eye movement data, EEG data, and vehicle manipulation data were collected from participants in each driving simulation. After eliminating invalid data (e.g., uncalibrated Dikablis glasses, missing data, or incomplete experimental tasks), data from 40 participants were retained. A two-way repeated-measures analysis of variance (ANOVA) was used to investigate the effects of three warning zone lengths (1800 m, 2000 m, and 2200 m), three upstream transition zone lengths (120 m, 140 m, and 160 m), and their interactions on participants' visual, cognitive, and driving performance at a significance level (α) of 0.05. Prior to the ANOVA, the Shapiro-Wilk method was used to test the data for normality. The results showed that each data set passed the normal distribution test and contained no outliers.

[0155] This experiment used IBM SPSS Statistics software to conduct a two-way repeated-measures ANOVA. If the data did not conform to the Mauchly test of sphericity (P < 0.05), correction was performed using the Greenhouse-Geisser method (1959). The results of the tests for visual, cognitive, and driving effects are shown in Tables 2, 3, and 4. When main effects were present, post hoc multiple comparisons were performed. When interaction effects were present, simple effects analysis was performed.

[0156] Table 2 Two-factor repeated measures ANOVA of visual effects

[0157]

[0158]

[0159] Table 3 Two-factor repeated measures ANOVA of cognitive effects

[0160]

[0161] Table 4 Two-factor repeated measures ANOVA of driving performance

[0162]

[0163] 7.1 Visual Effects

[0164] Pupil area

[0165] The pupil area data of 40 subjects were collected in 9 experimental scenes. The pupil area distribution of the subjects in different experimental scenes is as follows: Figure 3 As shown in the figure, as the length of the warning zone increases, the pupil area of the subjects gradually decreases and shows a convergence trend; as the length of the upstream transition zone increases, the pupil area of the subjects gradually decreases and shows a convergence trend. This shows that appropriate redundancy in the length of the warning zone and upstream transition zone can improve driver visual comfort.

[0166] As shown in Table 2, there was a statistically significant difference in the effects of the warning zone length and upstream transition zone length on pupil area, but no statistically significant difference in the effect of the interaction term on pupil area. Therefore, post hoc multiple comparisons were performed for each of the warning zone length and upstream transition zone length (Table 5).

[0167] Table 5 Post hoc multiple comparisons of pupil area of subjects

[0168]

[0169] Table 5 shows that pupil area showed significant differences between different warning zone lengths, with the largest pupil area at a warning zone length of 1800 m, indicating the worst driver visual comfort. Significant differences in pupil area were observed between upstream transition zone lengths of 120 m, 140 m, and 160 m, respectively. However, no significant difference existed between upstream transition zone lengths of 140 m and 160 m, indicating that these two upstream transition zone lengths provided comparable visual comfort to the subjects.

[0170] Saccade frequency

[0171] The scanning frequency data of 40 subjects in 9 experimental scenes were collected. The distribution of the scanning frequency of the subjects in different experimental scenes is as follows: Figure 4 As shown in the figure, the subjects' scanning frequency showed a downward trend as the length of the warning zone increased, and the changing trends were similar across the different upstream transition zones. As the length of the upstream transition zone increased, the subjects' scanning frequency showed a downward trend, and the changing trends were similar across the different warning zones. When the upstream transition zone length was set to 120m, the subjects showed frequent scanning behavior, and the scanning frequency was highest when the warning zone length was 1800m.

[0172] As shown in Table 2, there was a statistically significant difference in the effects of the warning zone length and upstream transition zone length on the subjects' scanning frequency, but there was no statistically significant difference in the effect of the interaction term on the subjects' scanning frequency. Therefore, post hoc multiple comparisons were performed for each of the warning zone length and upstream transition zone length (Table 6).

[0173] Table 6 Post hoc multiple comparisons of subjects' scanning rate (times / s)

[0174]

[0175] Table 6 shows that the scanning frequency is highest when the warning zone length is 1800m, indicating that the driver has a greater need to search the surrounding road environment and poor visual stability. There is no significant difference in scanning frequency between warning zone lengths of 2000m and 2200m. The scanning frequency shows significant differences between different upstream transition zone lengths, and the scanning frequency is highest when the upstream transition zone length is 120m, indicating that the driver's visual stability is the worst.

[0176] 7.2 Cognitive Effects (Eeg Analysis)

[0177] Beta value

[0178] The absolute power values of α waves of 40 subjects were collected in 9 experimental scenes. The distribution of the absolute power values of α waves of subjects in different experimental scenes is as follows: Figure 5 As shown in Figure 2, the absolute power of the alpha wave gradually decreases as the warning zone length increases. This suggests that as the warning zone length increases, the driver's cognitive load increases, gradually experiencing fatigue, which is detrimental to driving safety. However, as the warning zone length increases, the difference in the absolute power of the alpha wave across different upstream transition zone lengths decreases.

[0179] As shown in Table 3, the length of the warning zone, the length of the transition zone, and their interaction have significant effects on the absolute power value of the subjects' α waves. Therefore, a simple effect analysis was performed on the length of the warning zone and the length of the transition section one by one.

[0180] Pairwise comparisons showed that when the warning zone length was 2000 m, there was no significant difference in the absolute alpha wave power between the upstream transition zone lengths of 120 m and 140 m [M(120-140)=-38.947, SE=31.974, P=0.692]. However, pairwise comparisons between the other upstream transition zone lengths showed significant differences. When the upstream transition zone length was 160 m, there was no significant difference in the absolute alpha wave power between the warning zone lengths of 2000 m and 2200 m [M(2000-2200)=16.397, SE=9.642, P=0.291]. However, pairwise comparisons between the other warning zone lengths showed significant differences. When the warning zone length was 1800 m, the driver's absolute alpha wave power was the highest, and the driver's cognitive load was the lowest.

[0181] Beta waves

[0182] The absolute power values of β waves of 40 subjects were collected in 9 experimental scenes. The distribution of the absolute power values of β waves of subjects in different experimental scenes is as follows: Figure 6As shown, the absolute power of the beta wave increases with the length of the warning zone. This suggests that increasing the length of the warning zone increases the driver's cognitive load, leading to increased tension and fatigue, which is detrimental to driving safety. The absolute power of the beta wave was maximum when the warning zone was 2200 meters long and the upstream transition zone was 120 meters long.

[0183] As shown in Table 3, the length of the warning zone, the length of the transition section, and their interaction all have significant effects on the absolute power value of the subjects' β waves. Therefore, a simple effect analysis was performed on the length of the warning zone and the length of the transition section one by one.

[0184] Pairwise comparisons showed that when the warning zone length was 1800 m and 2200 m, there was a significant difference in the absolute power values of β waves between the upstream transition zone lengths of 120 m and 160 m [M(120-160)=-85.444, SE=28.048, P=0.012; M(120-160)=174.706, SE=59.965, P=0.018]. When the warning zone length was 2000 m, there was a significant difference in the absolute power values of β waves between the upstream transition zone lengths of 140 m and 160 m [M(140-160)=-114.461, SE=41.052, P=0.024]. When the upstream transition zone length was 140 m, there was no significant difference in the absolute power values of β waves between the warning zone lengths of 1800 m and 2000 m [M(1800-2000)=-37.154, SE=25.897, P=0.478]; when the upstream transition zone length was 160 m, there was no significant difference in the absolute power values of β waves between the warning zone lengths of 2000 m and 2200 m [M(2000-2200)=-17.738, SE=50.765, P=1.000]; pairwise comparisons between the other warning zone lengths all showed significant differences.

[0185] θ / β value

[0186] The θ / β values of 40 subjects in 9 experimental scenarios were calculated. The distribution of θ / β values of subjects in different experimental scenarios is as follows: Figure 7 As shown in the figure, the θ / β value gradually decreases as the warning zone length increases. This indicates that as the warning zone length increases, the driver's cognitive load increases, gradually causing fatigue, which is detrimental to driving safety. However, as the warning zone length increases, the difference in the θ / β value for different upstream transition zone lengths decreases. When the warning zone length reaches 2200m, the θ / β values for different upstream transition zone lengths are roughly the same, indicating that the driver's cognitive load and fatigue are comparable.

[0187] As shown in Table 3, the effects of the warning zone length, transition zone length, and interaction on the subjects' θ / β values are significant. Therefore, a simple effect analysis was conducted on the warning zone length and transition zone length one by one.

[0188] Pairwise comparisons showed that when the warning zone length was 2200 m, there was no significant difference in the θ / β values between the upstream transition zone length of 120 m and other upstream transition zone lengths [M(120-140)=0.035, SE=0.018, P=0.160; M(120-160)=-0.042, SE=0.022, P=0.186]. Pairwise comparisons between the other upstream transition zone lengths showed significant differences. When the upstream transition zone length was 160 m, there was no significant difference in the θ / β values between the warning zone lengths of 2000 m and 2200 m [M(2000-2200)=-0.034, SE=0.029, P=0.737]. Pairwise comparisons between the other warning zone lengths showed significant differences. When the warning zone length was 1800 m, the driver's θ / β value was the largest, and the driver's cognitive load was the lowest.

[0189] 7.3 Driving behavior analysis (U3)

[0190] Longitudinal acceleration

[0191] The longitudinal acceleration data of 40 subjects were collected in 9 experimental scenes. The longitudinal acceleration distribution of the subjects in different experimental scenes is as follows: Figure 8 As shown in the figure, as the length of the warning zone increases, the longitudinal acceleration shows an upward trend and gradually converges; as the length of the upstream transition zone increases, the longitudinal acceleration shows an upward trend. When the length of the warning zone is 1800m, the absolute value of the longitudinal acceleration of the subject is large, indicating that the subject's operating load is large and the subject is eager to brake. This may be due to the shorter length of the warning zone, which shortens the deceleration distance.

[0192] As shown in Table 4, the effects of the warning zone length, transition section length, and their interaction on the longitudinal acceleration of the subjects are significant. Therefore, a simple effect analysis was performed on the warning zone length and transition section length one by one.

[0193] Pairwise comparisons showed that when the warning zone length was 1800 m, there was no significant difference in longitudinal deceleration between upstream transition zone lengths of 120 m and 140 m [M(120-140)=-0.072, SE=0.05, P=0.464]. Pairwise comparisons between the remaining upstream transition zone lengths all showed significant differences. When the transition zone length was constant, significant differences in longitudinal acceleration were observed between different warning zone lengths. When the warning zone length was 2200 m, the absolute value of the driver's longitudinal acceleration was smaller, indicating a gentle deceleration.

[0194] Instantaneous speed when entering the work area (Steering wheel angle)

[0195] The instantaneous speed data of 40 subjects entering the work area in 9 experimental scenarios were collected. The instantaneous speed distribution of subjects entering the work area in different experimental scenarios is as follows: Figure 9 As shown in the figure, the speed limit in the construction zone was set at 60 km / h. All participants took deceleration measures before entering the work zone, resulting in an average speed between 40 km / h and 50 km / h, meeting the speed limit in the construction zone. As the length of the warning zone increased, the participants' instantaneous speed upon entering the work zone gradually decreased; similarly, as the length of the upstream transition zone increased, the participants' instantaneous speed upon entering the work zone gradually decreased. This suggests that longer warning and upstream transition zones provide participants with a longer deceleration zone, reducing the pressure on them to decelerate.

[0196] As shown in Table 4, the length of the warning zone, the length of the upstream transition zone, and their interaction have significant effects on the instantaneous speed of the subjects entering the work area. Therefore, a simple effect analysis was conducted on the length of the warning zone and the length of the transition section one by one.

[0197] Paired comparisons showed that when the warning zone length was 1800 m, there was no significant difference in instantaneous speed between upstream transition zone lengths of 120 m and 140 m [M(120-140)=1.212, SE=0.788, P=0.396]; when the warning zone length was 2200 m, there was no significant difference in instantaneous speed between upstream transition zone lengths of 140 m and 160 m [M(120-140)=1.872, SE=0.762, P=0.056], and pairwise comparisons between other upstream transition zone lengths showed significant differences; when the transition zone length was fixed, there were significant differences in instantaneous speed between different warning zone lengths.

[0198] Saccade frequency

[0199] The lane-changing duration data of 40 subjects were collected in 9 experimental scenarios. The distribution of lane-changing duration of subjects in different experimental scenarios is as follows: Figure 10 As shown in the figure, as the warning zone length increases, the lane change duration of the subjects gradually decreases, and the changing trends are similar across different upstream transition zones. As the upstream transition section length increases, the lane change duration of the subjects gradually decreases, and the changing trends are similar across different warning zones. This may be because the shorter warning zone or upstream transition zone length leads to increased cross-sectional traffic volume, which compresses the subjects' available lane change maneuvers and increases the difficulty of lane changes.

[0200] Table 4 shows that there is a statistically significant difference in the effects of the warning zone length and upstream transition zone length on the lane change duration of the subjects, but there is no statistically significant difference in the effect of the interaction term on lane change duration. Therefore, post hoc multiple comparisons were conducted for each of the warning zone length and upstream transition zone length (Table 7).

[0201] Table 7 Post hoc multiple comparisons of lane-changing duration of subjects (s)

[0202]

[0203] Table 7 shows that the lane-changing durations for different warning zone lengths show significant differences, and the lane-changing duration is greatest when the warning zone length is 1800 m, indicating that the lane-changing difficulty is the greatest and the driving safety is the worst. The lane-changing durations for different upstream transition section lengths also show significant differences, and the lane-changing duration is greatest when the upstream transition section length is 120 m, indicating that the lane-changing difficulty is the greatest and the driving safety is the worst.

[0204] 7.4 Determining weights and ranking quality

[0205] Determining the comprehensive weight of indicators

[0206] Seven experts in related fields were invited to compare the importance of the indicators from 1 to 9. The subjective weight of each evaluation indicator was calculated using the ANP method and Super Decision software. The mean of each evaluation indicator obtained from the driving simulation experiment was calculated, and the objective weight was calculated using the entropy weight method according to equations (2) to (5). The comprehensive weight of the indicators was solved using equation (6) and is shown in Table 8.

[0207] Table 8 Comprehensive weights of evaluation indicators

[0208]

[0209] Determine the indicator grading standards

[0210] To eliminate discrepancies between indicator data (inconsistent units and dimensions), the raw data were normalized to fall within the range [0, 1]. SPSS software was then used to categorize the evaluation indicators using the K-means clustering algorithm. The grading criteria for the evaluation indicators obtained using the K-means clustering algorithm are shown in Table 9.

[0211] Table 9 Evaluation index grading standards

[0212]

[0213] Fuzzy comprehensive evaluation results and analysis

[0214] Based on the experimental data and the established evaluation criteria, the membership function values were determined, and a fuzzy judgment matrix was constructed. Combined with the comprehensive weight vector obtained by ANP-EWM, the fuzzy comprehensive evaluation result vector was calculated according to Equation (8). The comprehensive evaluation grade eigenvalue G was calculated according to Equation (9). In this embodiment, Equation (9) takes the undetermined coefficient k = 2, and the experimental scenarios were evaluated based on the comprehensive evaluation grade eigenvalue. The experimental scenarios were ranked according to the comprehensive evaluation grade eigenvalue. The results are shown in Table 10.

[0215] Table 10 Fuzzy comprehensive evaluation results and comprehensive ranking of each scene

[0216]

[0217] Table 10 shows that the comprehensive ranking of safety and comfort for the last nine experimental scenarios is I > F > H > E > G > C > D > A > B. The higher the comprehensive evaluation grade characteristic value, the higher the driver's safety and comfort in that scenario. Scenario I (2200m warning zone × 160m upstream transition zone) has the highest comprehensive evaluation grade characteristic value, indicating the best driver safety and comfort. Scenario B (1800m warning zone × 140m upstream transition zone) has the lowest comprehensive evaluation grade characteristic value, indicating the worst driver safety and comfort.

[0218] Based on the evaluation results in Table 10, it can be concluded that when the upstream transition zone length is constant, the comprehensive evaluation grade characteristic value increases with the increase in the warning zone length, indicating an improvement in driver safety and comfort. When the warning zone length is 2000m and 2200m, the comprehensive evaluation grade characteristic value increases with the increase in the upstream transition zone length, indicating an improvement in driver safety and comfort. Therefore, in actual engineering applications, when road space conditions permit, it is recommended to set a "2200m warning zone × 160m upstream transition zone."

[0219] Scenario A (1800m warning zone × 120m upstream transition zone) has a higher safety and comfort level than Scenario B (1800m warning zone × 140m upstream transition zone), Scenario C (1800m warning zone × 160m upstream transition zone) has a higher safety and comfort level than Scenario D (2000m warning zone × 120m upstream transition zone), Scenario E (2000m warning zone × 140m upstream transition zone) has a higher safety and comfort level than Scenario G (2200m warning zone × 120m upstream transition zone), and Scenario F (2000m warning zone × 160m upstream transition zone) has a higher safety and comfort level than Scenario H (2200m warning zone × 140m upstream transition zone). Therefore, when road space conditions are limited, a reasonable length combination can also achieve a high level of driving safety and comfort.

[0220] In summary, in order to improve the scientific validity and rationality of the safety and comfort evaluation of different length combinations of warning zones and upstream transition zones, the present invention has developed a comprehensive safety and comfort evaluation method for maintenance operation zones consisting of three levels: visual effect, cognitive effect, and driving effect, by integrating eye movement, EEG, and manipulation behavior indicators. The present invention adopts ANP-EWM combined weighting to obtain the weights of evaluation indicators, which reduces the influence of subjective factors to a certain extent, objectively reflects the importance relationship between each evaluation indicator, enhances the scientific nature of the results, and provides a new idea for the evaluation of driving safety and comfort. In response to the uncertainty problem of the combined evaluation system of the length of the warning zone and the upstream transition zone, fuzzy theory is introduced on the basis of the combined weighting method, and the K-means clustering algorithm is used to grade the evaluation indicators. The trapezoidal membership function is selected to calculate the membership matrix values of the indicators at all levels. Combined with the comprehensive weight vector, the fuzzy comprehensive evaluation grade eigenvalue of each scene is calculated, and the experimental scene is evaluated based on the comprehensive evaluation grade eigenvalue G, which can provide a reference for determining the length of the construction operation zone of the urban expressway.

[0221] The above descriptions are only some embodiments of the present invention and do not limit the scope of protection of the present invention. Any equivalent device or equivalent process transformation made by using the contents of the description and drawings of the present invention, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A comprehensive safety and comfort evaluation method for urban expressway maintenance work areas, characterized by: include: S1. Setting the lengths of the warning zone and the upstream transition zone in the construction zone of the expressway section, and constructing a driving simulation experiment scenario in the construction zone of the urban expressway section based on the length settings of the warning zone and the upstream transition zone; S2. Conducting a driving simulation experiment based on the constructed driving simulation experiment scenario to obtain driving characteristic data of different drivers during driving, wherein the driving characteristic data includes a plurality of evaluation index data; S3, use ANP-EWM combined weighting to obtain the weight of each evaluation index and construct the comprehensive weight vector W; S4. Use K-means clustering algorithm to classify the evaluation indicators and establish safety and comfort evaluation standards; S5. Determine the membership value based on the evaluation index data and the established evaluation criteria, construct a membership matrix, calculate the fuzzy comprehensive evaluation result vector and the comprehensive evaluation grade eigenvalue based on the comprehensive weight vector, and evaluate the experimental scenario based on the comprehensive evaluation grade eigenvalue.

2. The comprehensive safety and comfort evaluation method for urban expressway maintenance work area according to claim 1 is characterized in that: The number of drivers in the driving simulation experiment in S2 is determined based on the expected variance, target confidence, and error margin. The calculation formula is shown in formula (1): Where: N is the experimental sample size; Z is the standard normal distribution statistic, σ is the standard deviation, and E is the maximum error.

3. The comprehensive safety and comfort evaluation method for urban expressway maintenance work area according to claim 1 is characterized in that: The evaluation index data in the driving characteristic data in S2 includes pupil area U 11 , scanning frequency U12, α wave absolute power U21, β wave absolute power U22, θ / β value U23, longitudinal acceleration U31, instantaneous speed entering the working area U32, lane change duration U33.

4. The comprehensive safety and comfort evaluation method for urban expressway maintenance work area according to claim 1 is characterized in that: S3 specifically includes the following steps: S3.

1. ANP determines subjective weights Experts measure the relative importance of each evaluation indicator and score them on a scale of 1-9. The subjective weight of each evaluation indicator is calculated using the ANP method and Super Decision software. cj , and finally the subjective weight vector Wc=[w c1 , w c2 ,…,w cn ]; S3.

2. Determine objective weights using entropy weight method The original data obtained during the driver simulation experiment is normalized to obtain the normalized decision matrix B shown in formula (2): Where: b ij —The normalized value of the jth evaluation indicator of the i-th evaluation object; According to the normalized decision matrix, the entropy value of the jth indicator in the evaluation index system is calculated: Where: m is the number of evaluation objects; n is the number of evaluation indicators; A is the translation amount; Determine the objective weight w of each indicator pj : S3.

3. Portfolio empowerment The subjective and objective weights are combined and weighted using formula (6): Where: w j is the weight of the jth indicator; w cj is the subjective weight of the jth indicator; w pj is the objective weight of the jth indicator; S3.

4. Construct a comprehensive weight vector W, where W = (w1, w2, ..., w m ).

5. The comprehensive safety and comfort evaluation method for urban expressway maintenance work area according to claim 4 is characterized in that: The translation amount A is 0.00001.

6. The comprehensive safety and comfort evaluation method for urban expressway maintenance work area according to claim 1 is characterized in that: S4 specifically includes: The original data obtained during the driver simulation experiment were normalized so that the data range fell within the interval [0,1]. Then, SPSS software was used to classify the evaluation indicators using the K-means clustering algorithm to establish a safety and comfort evaluation standard.

7. The comprehensive safety and comfort evaluation method for urban expressway maintenance work area according to claim 1 is characterized in that: S5 specifically includes the following steps: S5.

1. Determine the factor set U and the evaluation set V: The factor set is a collection of factors that characterize the evaluation object and is represented by U: U = (u1, u2, ..., u m ); the evaluation set is a set of various possible results that the evaluator may make on the evaluation object, represented by V: V=(v1,v2,…,v n ), the evaluation set is divided into five levels, and each level is regarded as a fuzzy subset; S5.

2. Determine the degree of membership: The degree of membership of each factor indicator is determined by the membership function; S5.

3. Construct the membership matrix R: Make the evaluation indicators dimensionless and use min-max normalization to the interval [0,1]; Determine the membership of the factor set to the evaluation set and construct the membership matrix, which is represented by R: Where: r ij is the membership degree of the i-th element in the factor set U to the j-th level in the evaluation set V; S5.

4. Obtaining the fuzzy comprehensive evaluation result vector S: The comprehensive weight vector W and the membership matrix R are synthesized to obtain the fuzzy comprehensive evaluation result vector S of each evaluation object. The expression of S is shown in formula (8): Where: s i For the evaluation object v j The membership value of the level. S5.

5. Analysis of evaluation results: The membership degree is processed using the weighted average method to calculate the comprehensive evaluation grade characteristic value G. The expression of G is shown in formula (9): Where: k is the unknown coefficient, ε j It represents the value corresponding to the j-th evaluation level and is assigned a weight value according to the actual evaluation system.

8. The comprehensive safety and comfort evaluation method for urban expressway maintenance work area according to claim 1 is characterized in that: In S52, a relatively small, medium, or large trapezoidal membership function is selected to determine the membership value in the single factor evaluation vector.

Images