Block chain-based tourism scenic spot visitor flow peak prediction system

Abstract

The invention relates to the technical field of people flow peak prediction, in particular to a block chain-based tourism scenic spot people flow peak prediction system. The method comprises an interest point prediction unit, an attraction index calculation unit and a peak value prediction unit. The interest point judgment module analyzes interest point values according to historical behavior data of the scenic spot area and judges whether objects are interest points of the scenic spot area, when the analyzed interest point values are larger than set interest point values, the objects are the interest points of the scenic spot area, and then the number of the interest points of the scenic spot area is recorded; according to the method, scenic spot area interest points are obtained, feature analysis is carried out on the scenic spot area interest points, the preference of tourists for objects in different scenic spot areas can be known by analyzing interest point values, the objects with high interest point values in the scenic spot areas can be emphatically input and managed to ensure that the tourists obtain high-quality experience, and the objects with low interest point values in the scenic spot areas can be emphatically managed to ensure that the tourists obtain high-quality experience. The attraction can be increased by improving services and developing new activities, and the optimization efficiency of resources is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of peak flow prediction of human traffic, and in particular to a peak flow prediction system for cultural and tourist attractions based on blockchain. Background Art

[0002] In the existing management of cultural and tourism scenic spots, crowd flow prediction usually relies on simple statistical analysis, lacking comprehensive consideration of tourists' points of interest and real-time data, resulting in inaccurate prediction results. When tourists visit cultural and tourism scenic spots casually, they will have an interest in an object in the cultural and tourism scenic spots. Due to the inaccurate prediction of the peak flow of cultural and tourism scenic spots, some interest areas have high peak flow and some areas have low peak flow, which makes it impossible for the scenic area management department to make resource allocation in advance and unable to reasonably allocate human resources, facility resources and service resources, resulting in waste or shortage of resources and reducing the management efficiency of cultural and tourism scenic spots. Since some interest areas have high peak flow and some areas have low peak flow, when tourists visit the high peak flow area according to their points of interest, they cannot take timely measures to guide tourists to avoid crowded areas, which is prone to stampedes, resulting in reduced comfort and safety of tourists. Therefore, we provide a cultural and tourism scenic spot crowd flow peak prediction system based on blockchain. Summary of the invention

[0003] The purpose of the present invention is to provide a peak traffic prediction system for cultural and tourism attractions based on blockchain to solve the problems raised in the above-mentioned background technology.

[0004] To achieve the above-mentioned purpose, the present invention provides a peak prediction system for the flow of people at cultural and tourist attractions based on blockchain, including a prediction point of interest unit, an attraction index calculation unit and a peak prediction unit;

[0005] The prediction point of interest unit obtains tourist behavior data and historical data, divides the cultural and tourist attractions into multiple scenic area areas, analyzes the point of interest values ​​according to the historical data, and determines whether the object is a point of interest in the scenic area;

[0006] The unit for calculating attraction index is used to receive a command that an object in the prediction point of interest unit is a point of interest in a scenic area, and the unit for calculating attraction index obtains tourist behavior data and historical data from the prediction point of interest unit to calculate the maximum value of the dwell time, visits, photos, and interactions in front of the object, and performs a standardized combination of the dwell time, visits, photos, and interactions in front of the object, and calculates the attraction index of the scenic area according to the combined standardized data of the dwell time, visits, photos, and interactions in front of the object;

[0007] The peak prediction unit is used to receive the tourist behavior data and historical data in the prediction point of interest unit and the regional object attraction index in the calculation attraction index unit to determine whether the scenic spot area is attractive to tourists. When attraction is generated, the regional object attraction index, tourist behavior data and historical data are used to predict the peak flow of people in the scenic spot area and determine whether there is a peak flow period in the scenic spot area. When there is a peak flow period in the scenic spot area, the tourist preference characteristics are analyzed according to the tourist behavior data and the points of interest in the scenic spot area with low peak flow are matched, and then the route is planned according to the points of interest in the scenic spot area with low peak flow.

[0008] As a further improvement of the technical solution, the interest point prediction unit includes a prediction segmentation module and an interest point judgment module;

[0009] The prediction and segmentation module obtains tourist faces and tourist behavior data through monitoring equipment at the entrance of cultural and tourist attractions, and then obtains historical data, uses blockchain technology to verify the acquired facial data, and then performs a deduplication operation on the successfully verified facial data, and uses time series analysis to predict the initial flow of people based on the deduplication facial data, and then divides the cultural and tourist attractions into multiple scenic spots, and arbitrarily labels the multiple scenic spots;

[0010] The point of interest judgment module is used to receive the scenic area data and historical data labeled in the prediction and segmentation module, obtain the historical behavior data of the scenic area from the historical data, analyze the point of interest value according to the historical behavior data of the scenic area, and then use the analyzed point of interest value and the set point of interest value to judge whether the object is a point of interest in the scenic area. When the analyzed point of interest value is greater than the set point of interest value, it means that the object is a point of interest in the scenic area, and then record the number of points of interest in the scenic area, and perform feature analysis on the points of interest in the scenic area.

[0011] As a further improvement of the technical solution, the principle of analyzing the interest point value in the interest point judgment module is as follows:

[0012] The historical tourists’ stay time Ta in front of the objects in the scenic area, the number of historical tourists’ viewing times Na, and the historical tourists’ interactions with the objects in the scenic area Ne are collected and input into the linear regression model. The linear regression model normalizes the input data and then outputs the corresponding weight coefficients w1, w2, and w3. The interest point values ​​are calculated using the historical tourists’ stay time Ta in front of the objects in the scenic area, the number of historical tourists’ viewing times Na, and the historical tourists’ interactions with the objects in the scenic area Ne and the corresponding weight coefficients w1, w2, and w3. The calculated interest point values ​​are the analyzed interest point values ​​Bs. The specific algorithm formula is:

[0013] Bs=Ta×w1+Na×w2+Ne×w3.

[0014] As a further improvement of the technical solution, the unit for calculating the attraction index includes a data calculation module and an attraction index module;

[0015] The data calculation module is used to receive the command from the point of interest judgment module that the object is a point of interest in the scenic area, and the data calculation module obtains the tourist behavior data and historical data from the prediction segmentation module, calculates the maximum value of the time spent in front of the object, visits, photos, and interactions based on the tourist behavior data and historical data, and then calculates the average value of the time spent in front of the object, visits, photos, and interactions based on the tourist behavior data;

[0016] The attraction index module is used to receive the tourist behavior data in the prediction and segmentation module and the average value of the dwell time, visits, photos, and interactions in front of the object in the data calculation module and standardize them with the maximum value of the dwell time, visits, photos, and interactions in front of the object; calculate the standard deviation of the dwell time, visits, photos, and interactions in front of the object according to the tourist behavior data and the average value of the dwell time, visits, photos, and interactions in front of the object; then perform mean standardization according to the average value of the dwell time, visits, photos, and interactions in front of the object and the standard deviation of the tourist behavior data and the dwell time, visits, photos, and interactions in front of the object; use the maximum value standardization of the dwell time, visits, photos, and interactions in front of the object and the mean standardization of the dwell time, visits, photos, and interactions in front of the object to perform a standardized combination of the dwell time, visits, photos, and interactions in front of the object; input the combined standardized data of the dwell time, visits, photos, and interactions in front of the object into a linear regression model; the linear regression model outputs a weight coefficient; and calculates the attraction index of the scenic spot area according to the combined standardized data of the dwell time, visits, photos, and interactions in front of the object and the weight coefficient.

[0017] As a further improvement of the technical solution, the implementation principle of calculating the attraction index of regional objects in the attraction index module is as follows:

[0018] Collect the combined object dwell time, visit, photo, interaction standardization data and dwell time weight coefficient W T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I Calculate the attraction index A of the scenic area. The specific algorithm formula is:

[0019] A n =T ns ·W T +V ns ·W V +P ns ·W P +I ns ·W I ;

[0020] Among them, A nRefers to the attractiveness index of the nth scenic spot area to tourists.

[0021] As a further improvement of the technical solution, the peak prediction unit includes a regional peak prediction module and a total peak prediction module;

[0022] The regional peak prediction module is used to receive the tourist behavior data, historical data, and numbered scenic spot area data in the prediction segmentation module, and the regional object attraction index and weight coefficient in the attraction index module, receive the interest point features of the scenic spot area analyzed by the interest point judgment module, and use the regional object attraction index to judge whether the scenic spot area is attractive to tourists. When the scenic spot area is attractive to tourists, the regional object attraction index, tourist behavior data, historical data and weight coefficient are used to predict the peak value of regional traffic at time t, and then use the predicted regional traffic peak at time t to judge whether there is a peak traffic period in the scenic spot area at time t;

[0023] The total peak prediction module is used to receive the initial passenger flow predicted in the prediction segmentation module, historical data and the regional passenger flow peak predicted at time t in the regional peak prediction module, and use the machine learning model to predict the total passenger flow peak of cultural and tourism attractions based on the predicted initial passenger flow, the predicted regional passenger flow peak at time t and historical data.

[0024] As a further improvement of the technical solution, the implementation principle of predicting the peak value of regional passenger flow at time t in the regional peak prediction module is as follows:

[0025] Collect the area object attraction index A, the dwell time T, the visit V, the photo P, the interaction I, the weather data Tq, the historical flow data Hs and the dwell time weight coefficient W of the scenic spot area T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I , attractiveness weight coefficient W A , weather weight coefficient W tq and the weight coefficient of historical flow of people in the scenic area W H Predict the peak flow rate of the area at time t, and predict the peak flow rate Fz of the area at time t. The specific algorithm formula is:

[0026] Fz=A n ·W A +T j ·W T +V j ·W V +P j ·W P +I j ·W I +Tq α ·Wtq +Hs·W H ;

[0027] Among them, Tq α Refers to the αth weather data.

[0028] As a further improvement of the technical solution, the specific judgment situation in the regional peak prediction module includes:

[0029] Case ①: When the predicted peak flow rate of the area at time t is greater than the set peak flow rate, it means that there is a peak flow rate in the scenic spot area at time t. The tourist behavior data is input into the linear regression model to analyze the tourist preference characteristics. The analyzed tourist preference characteristics are used to perform similarity matching with the analyzed points of interest characteristics of the scenic spot area to match the points of interest in the scenic spot area with low peak flow rate, and then the route is planned according to the points of interest in the scenic spot area with low peak flow rate;

[0030] Case ②: When the predicted peak flow rate of the area at time t is less than the set peak flow rate, it means that there is a low flow rate period in the scenic spot area at time t.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. In the peak flow prediction system of cultural and tourism scenic spots based on blockchain, the regional peak prediction module uses the regional object attraction index to determine whether the scenic spot area is attractive to tourists. When the scenic spot area is attractive to tourists, the regional object attraction index, tourist behavior data, historical data and weight coefficient are used to predict the regional peak flow at time t, and then the predicted regional peak flow at time t is used to determine whether there is a peak flow period in the scenic spot area at time t. When there is a peak flow period in the scenic spot area, the tourist preference characteristics are analyzed, and the analyzed tourist preference characteristics are used to perform similarity matching with the analyzed interest point characteristics of the scenic spot area to match the interest points of the scenic spot area with low peak flow, and then the route is planned according to the interest points of the scenic spot area with low peak flow. By integrating multi-dimensional data, the peak flow of people at specific times and regions can be predicted more accurately, which is helpful for the scenic spot management department to make resource allocation in advance, reasonably allocate human resources, facility resources and service resources, avoid waste or shortage of resources, and improve management efficiency. At the same time, by predicting the peak flow of people in advance, measures can be taken to guide tourists to avoid crowded areas and improve the comfort of the tour.

[0033] 2. In the peak flow prediction system of cultural and tourism scenic spots based on blockchain, the point of interest judgment module obtains the historical behavior data of the scenic spot area from the historical data, analyzes the point of interest value according to the historical behavior data of the scenic spot area, and then uses the analyzed point of interest value and the set point of interest value to judge whether the object is a point of interest in the scenic spot area. When the analyzed point of interest value is greater than the set point of interest value, it means that the object is a point of interest in the scenic spot area. Then the number of points of interest in the scenic spot area is recorded, and the characteristics of the points of interest in the scenic spot area are analyzed. By analyzing the point of interest values, it is possible to understand tourists' preferences for objects in different scenic spot areas. For objects in scenic spot areas with high point of interest values, they can be invested in and managed with emphasis to ensure that tourists have a high-quality experience. For objects in scenic spot areas with low point of interest values, they can increase their attractiveness and improve the optimization efficiency of resources by improving services and developing new activities. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is an overall block diagram of the present invention;

[0035] Figure 2 It is a block diagram of the real-time detection unit of the present invention.

[0036] The meaning of each number in the figure is:

[0037] 1. Prediction of interest point unit; 11. Prediction segmentation module; 12. Interest point judgment module;

[0038] 2. Calculation unit for attraction index; 21. Data calculation module; 22. Attraction index module;

[0039] 3. Peak prediction unit; 31. Regional peak prediction module; 32. Total peak prediction module. DETAILED DESCRIPTION

[0040] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] Example 1

[0042] The present invention provides a peak flow prediction system for cultural and tourist attractions based on blockchain. Figure 1-Figure 2 , including an interest point prediction unit 1, an attraction index calculation unit 2 and a peak prediction unit 3;

[0043] The prediction point of interest unit 1 obtains tourist behavior data and historical data, divides the cultural and tourism attractions into multiple scenic areas, analyzes the point of interest values ​​according to the historical data, and determines whether the object is a point of interest in the scenic area; the calculation attraction index unit 2 is used to receive the command that the object is a point of interest in the scenic area in the prediction point of interest unit 1, and the calculation attraction index unit 2 obtains the tourist behavior data and historical data from the prediction point of interest unit 1 to calculate the maximum value of the residence time, visits, photos, and interactions in front of the object and perform a standardized combination of the residence time, visits, photos, and interactions in front of the object, and determines whether the object is a point of interest in the scenic area according to the combination of the standardized values ​​of the residence time, visits, photos, and interactions in front of the object. The peak prediction unit 3 is used to receive the tourist behavior data and historical data in the prediction point of interest unit 1 and the regional object attraction index in the calculation attraction index unit 2 to determine whether the scenic spot area is attractive to tourists. When attraction is generated, the regional object attraction index, tourist behavior data and historical data are used to predict the peak flow of people in the scenic spot area and determine whether there is a peak flow period in the scenic spot area. When there is a peak flow period in the scenic spot area, the tourist preference characteristics are analyzed according to the tourist behavior data and the points of interest in the scenic spot area with low peak flow are matched, and then the route is planned according to the points of interest in the scenic spot area with low peak flow.

[0044] The prediction interest point unit 1 includes a prediction segmentation module 11 and an interest point judgment module 12;

[0045] The prediction and segmentation module 11 obtains the facial and behavioral data of tourists through the monitoring equipment at the entrance of the cultural and tourist attractions, and then obtains the historical data, uses the blockchain technology to verify the acquired facial data, and then performs a deduplication operation on the successfully verified facial data, and uses time series analysis to predict the initial flow of people based on the deduplication of facial data, and then divides the cultural and tourist attractions into multiple scenic area, and arbitrarily labels the multiple scenic area, including scenic area Q1, scenic area Q2, ..., scenic area Q n , the numbered scenic area data is passed to the point of interest determination module 12;

[0046] Historical data includes historical traffic data of cultural and tourism attractions, historical traffic data of scenic spots, historical behavior data of scenic spots, historical tourists’ stay time in front of objects in scenic spots Ta, historical tourists’ viewing times of objects in scenic spots Na, historical tourists’ interactions with objects in scenic spots Ne, and the maximum stay time in front of objects in scenic spots T max , the maximum value V visited before the object in the scenic area max , the maximum value P of taking photos in front of objects in the scenic area max , the maximum value of interaction in front of objects in the scenic area I max , weather data;

[0047] Tourist behavior data includes the time T of staying in front of the object and the number of stops Ttn , the object's previous visit V and the number of visits V vn , P photos taken in front of the object and the number of photos taken P pn , interaction before the object I and the number of interactions I in ;

[0048] The point of interest judgment module 12 is used to receive the scenic area data and historical data marked in the prediction and segmentation module 11, obtain the historical behavior data of the scenic area from the historical data, and then obtain the historical tourists' stay time Ta in front of the scenic area object, the number of historical tourists' viewing of the scenic area object Na, and the historical tourists' interaction with the scenic area object Ne from the historical behavior data of the scenic area, analyze the point of interest value according to the historical tourists' stay time Ta in front of the scenic area object, the number of historical tourists' viewing of the scenic area object Na, and the historical tourists' interaction with the scenic area object Ne, and then use the analyzed point of interest value and the set point of interest value to judge whether the object is a point of interest in the scenic area, and when the analyzed point of interest value is greater than the set point of interest value, the point of interest value is determined. When the point of interest value is set, it indicates that the object is a point of interest in the scenic area. Then the number of points of interest in the scenic area N is recorded, and the characteristics of the points of interest in the scenic area are analyzed. The characteristics of the points of interest in the scenic area analyzed include ancient buildings, museums, historical sites, water landscapes, zoos and botanical gardens, nature reserves, amusement projects, performances, and entertainment activities. By comprehensively considering the length of stay, the number of visits, and the interaction, it is possible to more comprehensively and accurately identify the points of interest of tourists and understand the preferences of tourists for different attractions. For scenic areas with higher point of interest values, they can be invested and managed with emphasis to ensure that tourists have a high-quality experience. For areas with lower point of interest values, they can increase their attractiveness by improving services and developing new activities to improve the optimization efficiency of resources.

[0049] The implementation principle of analyzing the interest point value in the interest point judgment module 12 is as follows:

[0050] The historical tourists’ stay time Ta in front of the objects in the scenic area, the number of historical tourists’ viewing times Na, and the historical tourists’ interactions with the objects in the scenic area Ne are collected and input into the linear regression model. The linear regression model normalizes the input data and then outputs the corresponding weight coefficients w1, w2, and w3. The interest point values ​​are calculated using the historical tourists’ stay time Ta in front of the objects in the scenic area, the number of historical tourists’ viewing times Na, and the historical tourists’ interactions with the objects in the scenic area Ne and the corresponding weight coefficients w1, w2, and w3. The calculated interest point values ​​are the analyzed interest point values ​​Bs. The specific algorithm formula is:

[0051] Bs=Ta×w1+Na×w2+Ne×w3;

[0052] Among them, the formula is used to calculate the point of interest value. By comprehensively considering the length of stay, number of visits and interaction, it can more comprehensively and accurately identify tourists' points of interest and understand tourists' preferences for different attractions;

[0053] The attraction index calculation unit 2 includes a data calculation module 21 and an attraction index module 22;

[0054] The data calculation module 21 is used to receive the command from the point of interest judgment module 12 that the object is a point of interest in the scenic area. The data calculation module 21 obtains the tourist behavior data and historical data from the prediction segmentation module 11, and extracts the stay time T and the number of stays T before the object from the tourist behavior data. tn , the object's previous visit V and the number of visits V vn , the number of photos taken in front of the object P and the number of photos taken P pn , interaction before the object I and the number of interactions I in , and then extract the maximum value T of the time spent in front of the object in the scenic area from the historical data max , the maximum value V visited before the object in the scenic area max , the maximum value P of taking photos in front of objects in the scenic area max , the maximum value of interaction in front of objects in the scenic area I max , according to the dwell time T in front of the object, the visit V in front of the object, the photo taken in front of the object P, the interaction I in front of the object and the maximum dwell time T in front of the object in the scenic area max , the maximum value V visited before the object in the scenic area max , the maximum value P of taking photos in front of objects in the scenic area max , the maximum value of interaction in front of objects in the scenic area I max The maximum values ​​of the dwell time, visits, photos and interactions in front of the object are calculated and standardized, and then the maximum values ​​of the dwell time T and the number of dwell times T are calculated. tn , the object's previous visit V and the number of visits V vn , the number of photos taken in front of the object P and the number of photos taken P pn , interaction before the object I and the number of interactions I in Calculate the average values ​​of the time spent in front of the object, visits in front of the object, photos taken in front of the object, and interactions in front of the object;

[0055] Steps to calculate the maximum value of the dwell time, visits, photos, and interactions in front of an object:

[0056] Step 1: According to the dwell time T in front of the object and the maximum dwell time T in front of the object in the scenic area max Calculate the maximum normalized time of staying in front of the object, and obtain the maximum normalized time of staying in front of the object T norm , the specific algorithm formula is:

[0057]

[0058] Among them, this formula is used to calculate the maximum value of the dwell time in front of the object, which is the basis for the subsequent standardized combination of the dwell time in front of the object;

[0059] Step 2: According to the previous visit V of the object and the maximum value V of the previous visit of the object in the scenic area max Calculate the maximum value of the object before the visit to normalize, get the maximum value of the object before the visit normalization V norm , the specific algorithm formula is:

[0060]

[0061] Among them, this formula is used to calculate the maximum value of the pre-object visit normalization, which is the basis for the subsequent pre-object visit normalization combination;

[0062] Step 3: Based on the maximum value P of the photos taken in front of the object and the photos taken in front of the object in the scenic area max Calculate the maximum value of the normalized image taken in front of the object, and obtain the maximum value of the normalized image taken in front of the object P norm , the specific algorithm formula is:

[0063]

[0064] Among them, this formula is used to calculate the maximum value of the normalization of the photos taken in front of the object, which serves as the basis for the subsequent standardized combination of photos taken in front of the object;

[0065] Step ④: According to the interaction before the object I and the maximum value of the interaction before the object in the scenic area I max Calculate the maximum normalized value of the interaction before the object and obtain the maximum normalized value of the interaction before the object I norm , the specific algorithm formula is:

[0066]

[0067] Among them, this formula is used to calculate the maximum value standardization of the interaction before the object, which serves as the basis for the subsequent standardized combination of the interaction before the object;

[0068] Steps to calculate the average of the time spent in front of an object, visits, photos, and interactions:

[0069] Step 1: Collect the dwell time T and the number of dwell times T before the object tn The average dwell time in front of the object is calculated to obtain the average dwell time in front of the object μT. The specific algorithm formula is:

[0070]

[0071] Among them, Tj Refers to the j-times of residence time in front of an object. This formula is used to calculate the average residence time in front of an object, which serves as the basis for the subsequent calculation of the standard deviation of the residence time in front of an object.

[0072] Step 2: Collect the number of visits V and V before the object is accessed vn Calculate the average value of the object's front access and obtain the average value of the object's front access μV. The specific algorithm formula is:

[0073]

[0074] Among them, V j Refers to the j visits before the object. This formula is used to calculate the average of the visits before the object, which serves as the basis for the subsequent calculation of the standard deviation of the visits before the object.

[0075] Step 3: Collect the number of photos P and photos taken before the object pn Calculate the average value of the photos taken in front of the object to obtain the average value μP of the photos taken in front of the object. The specific algorithm formula is:

[0076]

[0077] Among them, P j Refers to the jth number of photos taken in front of the object. This formula is used to calculate the average value of photos taken in front of the object, which serves as the basis for the subsequent calculation of the standard deviation of photos taken in front of the object.

[0078] Step 4: Collect the interactions before the object I and the number of interactions I in The average value of interaction before the object is calculated to obtain the average value of interaction before the object μI. The specific algorithm formula is:

[0079]

[0080] Among them, I j Refers to the j interactions before the object. This formula is used to calculate the average value of interactions before the object, which serves as the basis for the subsequent calculation of the standard deviation of interactions before the object.

[0081] The attraction index module 22 is used to receive the tourist behavior data in the prediction segmentation module 11 and the average values ​​of the object dwell time, visits, photos, and interactions in the data calculation module 21 and normalize them with the maximum values ​​of the object dwell time, visits, photos, and interactions, and extract the object dwell time T and the number of dwells T from the tourist behavior data. tn , the object's previous visit V and the number of visits V vn , the number of photos taken in front of the object P and the number of photos taken P pn , interaction before the object I and the number of interactions I in , according to the time T in front of the object and the number of stops T tn, the object's previous visit V and the number of visits V vn , the number of photos taken in front of the object P and the number of photos taken P pn , interaction before the object I and the number of interactions I in The standard deviation of the dwell time, visits, photos, and interactions in front of the object is calculated based on the average value of the dwell time, visits, photos, and interactions in front of the object. Then, the mean is standardized based on the average value of the dwell time, visits, photos, and interactions in front of the object and the standard deviation of the dwell time, visits, photos, and interactions in front of the object. The standardized combination of the dwell time in front of the object is performed by using the maximum value standardization of the dwell time, visits, photos, and interactions in front of the object and the mean standardization of the dwell time, visits, photos, and interactions in front of the object. Standardized combination of pre-object visits Standardized combination for taking photos in front of an object Standardized set of interactions in front of objects The combined object dwell time, visit, photo, and interaction standardized data are input into the linear regression model, and the linear regression model outputs the dwell time weight coefficient W T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I , based on the combined object dwell time, visit, photo, interaction normalization data and dwell time weight coefficient W T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I Calculate the attraction index of scenic spots and use it to warn of possible congestion in advance, take measures to avoid safety accidents, deploy emergency forces in advance, ensure rapid response and handling of emergencies during peak traffic, and improve safety management;

[0082] Steps to calculate the standard deviation of dwell time, visits, photos, and interactions in front of an object:

[0083] Step 1: Collect the time T and number of stops before the object is collected tn The standard deviation σT of the object's residence time is calculated by using the average value μT of the object's residence time. The specific algorithm formula is:

[0084]

[0085] Among them, this formula is used to calculate the standard deviation of the dwell time in front of the object, which serves as the basis for the subsequent calculation of the mean standardization of the dwell time in front of the object;

[0086] Step 2: Collect the number of visits V and V before the object is collected vnThe standard deviation σV of the object's front visit is calculated by the average value μV of the object's front visit. The specific algorithm formula is:

[0087]

[0088] Among them, this formula is used to calculate the standard deviation of the object's pre-visit, which serves as the basis for the subsequent calculation of the standardization of the object's pre-visit mean;

[0089] Step 3: Take photos in front of the object P, and the number of photos P pn The standard deviation σP of taking photos in front of the object is calculated by the average value μP of taking photos in front of the object. The specific algorithm formula is:

[0090]

[0091] Among them, this formula is used to calculate the standard deviation of the photos taken in front of the object, which serves as the basis for the subsequent calculation of the mean standardization of the photos taken in front of the object;

[0092] Step 4: Collect the interaction before the object I and the number of interactions I in The standard deviation of interaction before the object σI is calculated by using the average value of interaction before the object μI. The specific algorithm formula is:

[0093]

[0094] Among them, this formula is used to calculate the standard deviation of pre-object interactions, which serves as the basis for the subsequent calculation of the mean standardization of pre-object interactions;

[0095] Steps to implement mean standardization:

[0096] Step ①, collect the dwell time T, the average dwell time μT and the standard deviation σT of the dwell time in front of the object, calculate the mean standardized dwell time in front of the object, and obtain the mean standardized dwell time T std , the specific algorithm formula is:

[0097]

[0098] Among them, this formula is used to calculate the mean standardization of the dwell time in front of the object, which serves as the basis for the subsequent standardized combination of the dwell time in front of the object;

[0099] Step ②, collect the object access V, the object access mean μV and the object access standard deviation σV to calculate the object access mean standardization, and obtain the object access mean standardization V std , the specific algorithm formula is:

[0100]

[0101] Among them, this formula is used to calculate the standardized mean of the pre-object visits, which serves as the basis for the standardized combination of the subsequent pre-object visits;

[0102] Step 3: Collect the photos taken before the object P, the average value μP of the photos taken before the object, and the standard deviation σP of the photos taken before the object to calculate the mean standardization of the photos taken before the object, and obtain the mean standardization P of the photos taken before the object std , the specific algorithm formula is:

[0103]

[0104] Among them, this formula is used to calculate the mean standardization of the photos taken in front of the object, which serves as the basis for the standardized combination of the photos taken in front of the object in the future;

[0105] Step ④: Collect the interaction before the object I, the average interaction before the object μI and the standard deviation of the interaction before the object σI to calculate the mean standardization of the interaction before the object, and obtain the mean standardization of the interaction before the object I std , the specific algorithm formula is:

[0106]

[0107] Among them, this formula is used to calculate the mean standardization of interactions before objects, which serves as the basis for the standardized combination of subsequent interactions before objects;

[0108] The implementation principle of calculating the attraction index of the regional object in the attraction index module 22 is:

[0109] Collect the combined object dwell time, visit, photo, interaction standardization data and dwell time weight coefficient W T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I Calculate the attraction index A of the scenic area. The specific algorithm formula is:

[0110] A n =T ns ·W T +V ns ·W V +P ns ·W P +I ns ·W I ;

[0111] Among them, A nRefers to the attraction index of the nth scenic spot area to tourists. This formula is used to calculate the attraction index of the scenic spot area. Through the attraction index of the scenic spot area, early warning of possible congestion can be given, measures can be taken to avoid safety accidents, emergency forces can be deployed in advance, and emergency incidents can be quickly responded to and handled during peak traffic, thereby improving safety management;

[0112] The peak prediction unit 3 includes a regional peak prediction module 31 and a total peak prediction module 32;

[0113] The regional peak prediction module 31 is used to receive the tourist behavior data, historical data, and numbered scenic spot area data in the prediction segmentation module 11 and the regional object attraction index A and the residence time weight coefficient W in the attraction index module 22. T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I , receiving the characteristics of the points of interest in the scenic area analyzed by the point of interest judgment module 12, extracting the data of the dwelling time T in front of the object, the visit V in front of the object, the photo taking P in front of the object, and the interaction I in front of the object from the tourist behavior data, and then extracting the weather data Tq and the historical flow data Hs of the scenic area from the historical data, and using the regional object attraction index A and the set attraction index of the scenic area to judge whether the scenic area is attractive to tourists. When the regional object attraction index A is greater than the set attraction index of the scenic area, it means that the scenic area is attractive to tourists. The regional object attraction index A, weather data and the historical flow data of the scenic area are input into the linear regression model, and the linear regression model unifies the input data into units, and then outputs the corresponding attraction weight coefficient W A , weather weight coefficient W tq and the weight coefficient of historical flow of people in the scenic area W H , using the regional object attraction index A, the dwell time in front of the object T, the visit in front of the object V, the photo in front of the object P, the interaction in front of the object I, the weather data Tq, the historical flow data of the scenic area Hs and the dwell time weight coefficient W T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I , attractiveness weight coefficient W A , weather weight coefficient W tq and the weight coefficient of historical flow of people in the scenic area W HPredict the peak flow of people in the area at time t, and then use the predicted peak flow of people in the area at time t and the set peak flow of people to determine whether there is a peak flow period in the scenic area at time t. By integrating multi-dimensional data, the peak flow of people at specific times and areas can be predicted more accurately, which helps the scenic area management department to make resource allocation in advance, reasonably allocate human resources, facility resources and service resources, avoid resource waste or shortage, and improve management efficiency. At the same time, by predicting the peak flow of people in advance, measures can be taken to guide tourists to avoid crowded areas and improve the comfort of sightseeing;

[0114] The implementation principle of predicting the peak value of regional passenger flow at time t in the regional peak prediction module 31 is as follows:

[0115] Collect the area object attraction index A, the dwell time T, the visit V, the photo P, the interaction I, the weather data Tq, the historical flow data Hs and the dwell time weight coefficient W of the scenic spot area T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I , attractiveness weight coefficient W A , weather weight coefficient W tq and the weight coefficient of historical flow of people in the scenic area W H Predict the peak flow rate of the area at time t, and predict the peak flow rate Fz of the area at time t. The specific algorithm formula is:

[0116] Fz=A n ·W A +T j ·W T +V j ·W V +P j ·W P +I j ·W I +Tq α ·W tq +Hs·W H ;

[0117] Among them, Tq α Refers to the αth weather data. This formula is used to predict the peak flow of people in the region at time t. By integrating multi-dimensional data, the peak flow of people at a specific time and region can be predicted more accurately, which helps the scenic area management department to make resource allocation in advance.

[0118] The specific judgment situations in the regional peak prediction module 31 include:

[0119] Case ①: When the predicted peak flow rate of the area at time t is greater than the set peak flow rate, it means that there is a peak flow rate in the scenic area at time t. The dwell time T, visit V, photo P, and interaction I are input into the linear regression model to analyze the preference characteristics of tourists. The analyzed preference characteristics of tourists include ancient buildings, museums, historical sites, water landscapes, zoos, nature reserves, amusement projects, performances, and entertainment activities. The analyzed preference characteristics of tourists are matched with the characteristics of points of interest in the scenic area to match the points of interest in the scenic area with low peak flow rate. Then, the route is planned according to the points of interest in the scenic area with low peak flow rate. By choosing the scenic area with low flow rate, tourists can avoid the crowded main attractions and enjoy a more comfortable and quiet visit to the relevant points of interest in other scenic areas, avoiding stampedes during the peak flow period, and improving the experience of tourists. At the same time, by guiding tourists to the scenic area with low flow rate, the distribution of people in the scenic area can be effectively balanced, the pressure on a certain scenic area can be reduced, and the utilization of resources in other scenic areas can be improved.

[0120] Case ②: When the predicted peak flow rate of the area at time t is less than the set peak flow rate, it means that there is a low flow rate period in the scenic spot area at time t;

[0121] The total peak prediction module 32 is used to receive the initial passenger flow predicted in the prediction segmentation module 11, the historical data and the regional passenger flow peak predicted at time t in the regional peak prediction module 31, extract the weather data and the historical passenger flow data of the cultural and tourism attractions from the historical data, and use the machine learning model to predict the total passenger flow peak of the cultural and tourism attractions based on the predicted initial passenger flow, the predicted regional passenger flow peak at time t, the weather data and the historical passenger flow data of the cultural and tourism attractions. By accurately predicting the total passenger flow peak of the scenic area, the tourism management department can formulate corresponding management strategies in advance, reasonably allocate human and material resources, and avoid waste or shortage of resources.

[0122] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A peak traffic prediction system for cultural and tourist attractions based on blockchain, characterized by: It includes an interest point prediction unit (1), an attraction index calculation unit (2) and a peak prediction unit (3); The interest point prediction unit (1) obtains tourist behavior data and historical data, divides the cultural and tourist attractions into multiple scenic area areas, analyzes the interest point values ​​based on the historical data, and determines whether the object is an interest point in the scenic area; The attraction index calculation unit (2) is used to receive a command from the interest point prediction unit (1) that the object is an interest point in the scenic area. The attraction index calculation unit (2) obtains the tourist behavior data and historical data from the interest point prediction unit (1), calculates the maximum value of the residence time, visits, photos, and interactions in front of the object, and performs a standardized combination of the residence time, visits, photos, and interactions in front of the object, and calculates the attraction index of the scenic area according to the combined standardized data of the residence time, visits, photos, and interactions in front of the object; The peak prediction unit (3) is used to receive the tourist behavior data and historical data in the prediction point of interest unit (1) and the regional object attraction index in the calculation attraction index unit (2) to determine whether the scenic spot area is attractive to tourists. When attraction is generated, the regional object attraction index, the tourist behavior data and the historical data are used to predict the peak flow of people in the scenic spot area and determine whether there is a peak flow period in the scenic spot area. When there is a peak flow period in the scenic spot area, the tourist preference characteristics are analyzed according to the tourist behavior data and the points of interest in the scenic spot area with low peak flow of people are matched, and then the route is planned according to the points of interest in the scenic spot area with low peak flow of people.

2. The peak flow prediction system for cultural and tourist attractions based on blockchain according to claim 1 is characterized by: The interest point prediction unit (1) comprises a prediction segmentation module (11) and an interest point judgment module (12); The prediction and segmentation module (11) obtains tourist faces and tourist behavior data through monitoring equipment at the entrance of the cultural and tourist attractions, and then obtains historical data, uses blockchain technology to verify the acquired facial data, and then performs a deduplication operation on the successfully verified facial data, and uses time series analysis to predict the initial flow of people based on the deduplication facial data, and then divides the cultural and tourist attractions into multiple scenic area areas, and randomly labels the multiple scenic area areas; The point of interest judgment module (12) is used to receive the scenic spot area data and historical data marked in the prediction segmentation module (11), obtain the historical behavior data of the scenic spot area from the historical data, analyze the point of interest value according to the historical behavior data of the scenic spot area, and then use the analyzed point of interest value and the set point of interest value to judge whether the object is a point of interest in the scenic spot area. When the analyzed point of interest value is greater than the set point of interest value, it means that the object is a point of interest in the scenic spot area, and then record the number of points of interest in the scenic spot area, and perform feature analysis on the points of interest in the scenic spot area.

3. The peak flow prediction system for cultural and tourist attractions based on blockchain according to claim 2 is characterized by: The implementation principle of analyzing the interest point value in the interest point judgment module (12) is: The historical tourists’ stay time Ta in front of the objects in the scenic area, the number of historical tourists’ viewing times Na, and the historical tourists’ interactions with the objects in the scenic area Ne are collected and input into the linear regression model. The linear regression model normalizes the input data and then outputs the corresponding weight coefficients w1, w2, and w3. The interest point values ​​are calculated using the historical tourists’ stay time Ta in front of the objects in the scenic area, the number of historical tourists’ viewing times Na, and the historical tourists’ interactions with the objects in the scenic area Ne and the corresponding weight coefficients w1, w2, and w3. The calculated interest point values ​​are the analyzed interest point values ​​Bs. The specific algorithm formula is: Bs=Ta×w1+Na×w2+Ne×w3.

4. The peak flow prediction system for cultural and tourist attractions based on blockchain according to claim 2 is characterized by: The attraction index calculation unit (2) comprises a data calculation module (21) and an attraction index module (22); The data calculation module (21) is used to receive a command from the point of interest judgment module (12) that the object is a point of interest in the scenic area, and the data calculation module (21) obtains tourist behavior data and historical data from the prediction segmentation module (11), calculates the maximum value of the time spent in front of the object, visits, photos, and interactions based on the tourist behavior data and historical data, and then calculates the average value of the time spent in front of the object, visits, photos, and interactions based on the tourist behavior data; The attraction index module (22) is used to receive the tourist behavior data in the prediction and segmentation module (11) and the average values ​​of the object dwell time, visits, photos, and interactions in the data calculation module (21) and the maximum values ​​of the object dwell time, visits, photos, and interactions, calculate the standard deviation of the object dwell time, visits, photos, and interactions according to the tourist behavior data and the average values ​​of the object dwell time, visits, photos, and interactions, and then perform mean standardization according to the average values ​​of the object dwell time, visits, photos, and interactions and the standard deviation of the tourist behavior data and the object dwell time, visits, photos, and interactions, combine the standardization of the object dwell time, visits, photos, and interactions by using the maximum value standardization of the object dwell time, visits, photos, and interactions and the mean standardization of the object dwell time, visits, photos, and interactions, input the combined object dwell time, visits, photos, and interactions standardized data into a linear regression model, the linear regression model outputs a weight coefficient, and calculates the attraction index of the scenic spot area according to the combined object dwell time, visits, photos, and interactions standardized data and the weight coefficient.

5. The peak flow prediction system for cultural and tourist attractions based on blockchain according to claim 4 is characterized by: The implementation principle of calculating the attraction index of the regional object in the attraction index module (22) is as follows: Collect the combined object dwell time, visit, photo, interaction standardization data and dwell time weight coefficient W T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I Calculate the attraction index A of the scenic area. The specific algorithm formula is: A n =T ns ·W T +V ns ·W V +P ns ·W P +I ns ·W I ; Among them, A n Refers to the attractiveness index of the nth scenic spot area to tourists.

6. The peak flow prediction system for cultural and tourist attractions based on blockchain according to claim 4 is characterized by: The peak prediction unit (3) comprises a regional peak prediction module (31) and a total peak prediction module (32); The regional peak prediction module (31) is used to receive the tourist behavior data, historical data, and numbered scenic spot area data in the prediction segmentation module (11) and the regional object attraction index and weight coefficient in the attraction index module (22), receive the interest point characteristics of the scenic spot area analyzed by the interest point judgment module (12), use the regional object attraction index to judge whether the scenic spot area is attractive to tourists, and when the scenic spot area is attractive to tourists, use the regional object attraction index, tourist behavior data, historical data and weight coefficient to predict the peak value of regional human flow at time t, and then use the predicted regional human flow peak at time t to judge whether there is a human flow peak period in the scenic spot area at time t; The total peak prediction module (32) is used to receive the initial passenger flow predicted by the prediction segmentation module (11), historical data and the regional passenger flow peak value predicted at time t by the regional peak prediction module (31), and use a machine learning model to predict the total passenger flow peak value of the cultural and tourism scenic spot based on the predicted initial passenger flow, the predicted regional passenger flow peak value at time t and historical data.

7. The peak flow prediction system for cultural and tourist attractions based on blockchain according to claim 6 is characterized by: The implementation principle of predicting the peak value of regional passenger flow at time t in the regional peak prediction module (31) is as follows: Collect the area object attraction index A, the dwell time T, the visit V, the photo P, the interaction I, the weather data Tq, the historical flow data Hs and the dwell time weight coefficient W of the scenic spot area T , access weight coefficient W V , Photo weight coefficient W P , interaction weight coefficient W I , attractiveness weight coefficient W A , weather weight coefficient W tq and the weight coefficient of historical flow of people in the scenic area W H Predict the peak flow rate of the area at time t, and predict the peak flow rate Fz of the area at time t. The specific algorithm formula is: Fz=A n ·W A +T j ·W T +V j ·W V +P j ·W P +I j ·W I +Tq α ·W tq +Hs·W H ; Among them, Tq α Refers to the αth weather data.

8. The peak flow prediction system for cultural and tourist attractions based on blockchain according to claim 6 is characterized by: The specific judgment situations in the regional peak prediction module (31) include: Case ①: When the predicted peak flow rate of the area at time t is greater than the set peak flow rate, it means that there is a peak flow rate in the scenic spot area at time t. The tourist behavior data is input into the linear regression model to analyze the tourist preference characteristics. The analyzed tourist preference characteristics are used to perform similarity matching with the analyzed points of interest characteristics of the scenic spot area to match the points of interest in the scenic spot area with low peak flow rate, and then the route is planned according to the points of interest in the scenic spot area with low peak flow rate; Case ②: When the predicted peak flow rate of the area at time t is less than the set peak flow rate, it means that there is a low flow rate period in the scenic spot area at time t.

Images