A passenger flow guidance planning method and device based on multiple areas of a business district
By dividing heat map area and grid analysis in the business district, the passenger flow is predicted and the guidance correlation intensity is calculated, the problem of difficulty in passenger flow guidance between hot spots in the business district is solved, and more efficient passenger flow distribution and tourists are achieved Recommended paths to improve experience and security.
Patent Information
- Application Number
- CN202411016943.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-29
AI Technical Summary
The traffic orientation between hot spots in the existing business district is difficult, which affects tourists' path planning and play experience, and is prone to safety hazards and road congestion.
By dividing the historical passenger flow data in the business district and grid processing the data, analyzing the changes in time and space distribution, predicting future passenger flow, calculating the intensity of passenger flow orientation between hot spots, and based on the business district grid map, Search for the best guided paths and recommend them to tourists in real time.
It improves the efficiency of passenger flow distribution between various hot spots in the business district, provides convenient and efficient play routes, improves tourist experience and service quality, and reduces safety hazards.
Smart Images

Figure CN118551910B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of path prediction and planning, and in particular to a method and device for passenger flow guidance planning based on multiple areas in a business district. Background Art
[0002] A business district refers to the area of a store that attracts customers by expanding in a certain direction and distance from its location. Simply put, it is the area where customers who come to the store live. Whether it is a large shopping mall or a small store, their sales always have a certain geographical range. This geographical range is centered on the mall and radiates to the places where consumers who may come to the store to buy live.
[0003] The commercial district occupies a larger area, including more shopping malls, restaurants, entertainment, leisure and scenic spots. Since business is determined by consumers' purchasing behavior and the business ability of commercial enterprises, the commercial district is the spatial scope of commercial properties to attract customers, that is, the time distance or spatial distance for consumers to go to commercial places for consumption activities.
[0004] Existing business districts often contain multiple hotspots with dense traffic, and as time goes by, the passenger flow in each hotspot will also change accordingly. During holidays, the crowd density in these hotspots will generally increase significantly, which is easy to cause traffic congestion. When you want to go to other hotspots, the roads are congested and route planning is difficult, which is not conducive to tourists playing and spending in the business district, and reduces the personal experience of tourists. At the same time, hotspots with dense passenger flow are also prone to bring safety hazards, which is not conducive to the personal safety of tourists. Moreover, the denser the hotspots, the more difficult it is to transfer and guide passenger flow, which is not conducive to improving the convenience of passenger flow between various hotspots in the business district. Summary of the invention
[0005] The embodiment of the present application provides a method and device for passenger flow guidance planning based on multiple areas in a business district, which is used to solve the following technical problems: passenger flow guidance between hot spots in existing business districts is relatively difficult, which is not conducive to route planning for tourists during their travels and can easily reduce the tourists' own travel experience.
[0006] The present application embodiment adopts the following technical solutions:
[0007] On the one hand, the embodiment of the present application provides a passenger flow guidance planning method based on multiple areas of a business district, including: according to the geographical boundary range of the target business district, the historical passenger flow set area in the boundary of the business district is divided into regions related to the heat map to obtain a number of hot spots; through a preset spatial topological relationship, the hot spots are gridded, and the passenger flow data collected from the hot spots after data gridding are subjected to hot spot analysis related to the spatiotemporal distribution changes to determine the current passenger flow range data in the hot spots; based on the time scale characteristics and tourist behavior route characteristics in the current passenger flow range data, the current passenger flow range data is analyzed for the trend of tourist transfer changes between hot spots, and the predicted passenger flow range data of the hot spots in the future time period is predicted and obtained; according to the current passenger flow range data According to the predicted passenger flow range data, the correlation calculation of the passenger flow guidance association strength is performed on the route directivities between the several hot spots, and the passenger flow guidance change relationship between the various hot spots in the current time period is obtained; wherein, the passenger flow guidance change relationship is the passenger flow increase and decrease change relationship between the various hot spots; based on the business district grid map corresponding to the target business district, the first position information of the current hot spot area and the second position information of the target hot spot area are searched for the best guidance path under the passenger flow guidance change relationship, and the best guidance recommended route is determined; the best guidance recommended route is visualized on a large screen, and through the visualization large screen, the best guidance recommended route is connected to the terminal device of the tourist in real time to complete the travel guidance recommendation for the tourist in the target business district.
[0008] The embodiment of the present application divides the historical passenger flow collection area in the business district boundary into heat map areas, which helps to more accurately understand the passenger flow distribution in different areas of the business district. It is also possible to use the hot spot analysis of the temporal and spatial distribution changes to determine the current passenger flow range data, which can deeply understand the changing rules of passenger flow in time and space. And based on the characteristics of the current passenger flow range data, the trend of tourist transfer changes between hot spots is analyzed, and the predicted passenger flow range data in the future time period is predicted to provide forward-looking guidance for the operation and management of the business district. At the same time, it can also greatly help optimize the passenger flow distribution between various hot spots in the business district, improve the overall operational efficiency of the business district, and provide tourists with more convenient and efficient tour routes to enhance the experience of tourists. It further facilitates tourists to obtain information, enhances the interactivity between tourists and business districts, and improves service quality.
[0009] In a feasible implementation, according to the geographical boundary range of the target business district, the historical passenger flow set area in the boundary of the business district is divided into regions related to the heat map to obtain several hot spots, specifically including: based on the business data in the smart management platform, the geographical location of each merchant is divided into adjacent areas to determine the merchant adjacent area set; wherein the business data at least includes: merchant operation data, building passenger flow data, popular scenic spot passenger flow data, merchant location table data and merchant rental management data; through the preset remote sensing image geoscience analysis technology, the spatial scale data of the merchant adjacent area set is divided into the latitude and longitude of the scale pattern to obtain the target business district containing any merchant area in the merchant adjacent area set; and based on the regional boundary of the target business district, the geographical boundary range is determined; wherein the spatial scale data includes: the lateral scale of the area occupied by the merchant Boundary shape data and longitudinal structure data of the area occupied by merchants; based on several thermal analysis cameras pre-installed in the target business district, historical passenger flow data are collected in multiple historical time periods to obtain a historical passenger flow data set of each thermal analysis camera; based on the merchant adjacent area set, each historical passenger flow data set is subjected to scale deduction processing of the passenger flow set area to obtain the historical passenger flow set area; the historical passenger flow set area is subjected to characterization conversion of relevant thermal characteristics to obtain a historical thermal characteristic map of the target business district; wherein the historical thermal characteristic map contains historical passenger flow set areas in different colors; based on the heterogeneity between the various thermal characteristics in the historical thermal characteristic map, the historical thermal characteristic map is subjected to hierarchical patch division processing at the macro-regional scale to obtain the hot spot area used to reflect the concentrated change area of passenger flow.
[0010] After collecting historical passenger flow data using a thermal analysis camera, the embodiment of the present application can use the scale deduction processing of the passenger flow aggregation area to achieve better aggregation and clustering. At the same time, based on the hierarchical patch division technology, the historical thermal characteristic map is accurately divided, thereby dividing the hot spots that can reflect the concentrated changes in passenger flow areas.
[0011] In a feasible implementation, the hotspot area is gridded by a preset spatial topological relationship, specifically including: by installing a thermal analysis camera in the hotspot area and based on a preset collection period, collecting a number of point-like periodic passenger flow data in the hotspot area; wherein the collection period at least includes: 5 minutes, 10 minutes and 15 minutes; through the spatial topological relationship, each of the point-like periodic passenger flow data is summarized and converted according to the number of points, so as to obtain the surface-like periodic passenger flow data of each hotspot area; wherein the spatial topological relationship includes: intersection relationship, inclusion relationship and separation relationship; merging a number of the surface-like periodic passenger flow data in the collection period on the same layer to obtain the gridded spatial distribution data of passenger flow in each hotspot area; wherein the gridded spatial distribution data of passenger flow is the initial passenger flow change data within the collection period; based on the end time node of the collection period, the gridded spatial distribution data of passenger flow is time series data marked to obtain the current spatial distribution data of passenger flow at each end time node.
[0012] After gridding the hot spot areas, the embodiment of the present application can convert the point-like periodic passenger flow data collected by each thermal analysis camera into surface passenger flow data, and then use the layer merging processing of the regional gridded image to obtain the gridded spatial distribution data of the passenger flow in each hot spot area. Then, the time series data can be used to obtain the current passenger flow spatial distribution data at the end time node of the collection period, which greatly improves the distribution of passenger flow data on the plane map, thereby grasping a more comprehensive passenger flow distribution situation.
[0013] In a feasible implementation manner, the passenger flow data collected in the hot spot area after data gridding is subjected to hot spot analysis related to the temporal and spatial distribution changes to determine the current passenger flow range data in the hot spot area, specifically including: performing clustering score processing on the random factors in the current passenger flow spatial distribution data to obtain the clustering density of each random factor; wherein the clustering density is proportional to the clustering score value; the random factors are the high and low values of the passenger flow numerical attribute in the current passenger flow spatial distribution data, and based on the clustering density of each random factor, the current passenger flow spatial distribution data is subjected to significant clustering balance processing to obtain the current passenger flow estimation data of each hot spot area; wherein the significant clustering includes: the hot spot significance of the high value and the cold spot significance of the low value; based on the cycle span value in the collection cycle, the current passenger flow estimation data is subjected to curve fitting processing related to the passenger flow peak change to obtain the passenger flow time variation curve; wherein the cycle span value is the dependent variable of the passenger flow time variation curve; according to the current passenger flow estimation data and the corresponding clustering density, the hot spot area is subjected to the passenger flow estimation data based on the passenger flow. The core hot spot area of the calorimetric value is identified and processed, and according to the identified core hot spot area, the hot spot area is diffusely identified on the isoaxial line to divide the hot spot level area; wherein, the hot spot level area includes: core hot spot area, secondary hot spot area and edge hot spot area; based on the hot spot level area, the current passenger flow distribution pattern of the hot spot area is determined; through the current time dimension in the current passenger flow distribution pattern, the time dimension corresponding to the period span value in the passenger flow time change curve is matched with each other; according to the core hot spot area in the current passenger flow distribution pattern, and through the spatial difference algorithm, the passenger flow time change curve and the current passenger flow distribution pattern are subjected to data fusion processing between the benchmark data points to obtain the current spatiotemporal passenger flow estimation data; wherein, the algorithm used for the data fusion is the Geopandas library technology in Python; the benchmark data point is the time dimension data point shared by the current passenger flow distribution pattern and the passenger flow time change curve; the current spatiotemporal passenger flow estimation data is calibrated within a fixed error range to obtain the current passenger flow range data in each of the hot spot areas.
[0014] The embodiment of the present application uses the calculated passenger flow time change curve and the current passenger flow distribution pattern for data fusion, that is, after the spatiotemporal change analysis, it can more accurately calculate the passenger flow range at the current time node. Since the passenger flow changes in real time, it is completely inaccurate to use only the passenger flow data collected at a fixed time node. Therefore, after the fusion processing of the cycle span and the passenger flow distribution pattern, it is possible to more accurately obtain the range data of the passenger flow at the current time node, that is, to keep the passenger flow data within a more accurate numerical range, so as to better reflect the current spatiotemporal passenger flow estimation data in each hot spot area.
[0015] In a feasible implementation manner, based on the time scale characteristics and tourist behavior route characteristics in the current passenger flow range data, the current passenger flow range data is analyzed for the trend of tourist transfer changes between hot spots, and the predicted passenger flow range data of the hot spots in the future time period is predicted and obtained, specifically including: obtaining the historical passenger flow range data in each of the hot spots; wherein the historical passenger flow range data is the data collected in the previous collection cycle; performing a hot spot area correlation analysis on the passenger flow change trend between the historical passenger flow range data and the current passenger flow range data, and determining the associated hot spot areas; wherein the associated hot spot areas are hot spot areas with passenger flow orientation correlation; based on the passenger flow change trend of the associated hot spot areas and the current passenger flow range data, determining the tourist behavior route characteristics in the current passenger flow range data; wherein the tourist behavior route characteristics in the current passenger flow range data are determined. The tourist behavior route feature is the tourist guidance feature between the associated hot spots; based on the tourist behavior route feature of the current hot spot, the remaining hot spots in the associated hot spots are numbered to obtain numbered hot spots; according to the time scale feature of the current hot spot and through the tourist behavior route feature, a two-dimensional matrix is generated for the passenger flow change feature in the numbered hot spots to obtain a tourist quantity feature matrix for each numbered hot spot; wherein the passenger flow change feature includes: passenger flow increase change feature and passenger flow decrease change feature; the passenger flow change data corresponding to the tourist quantity feature matrix in each numbered hot spot is determined; through a preset passenger flow prediction model, the passenger flow change data of the numbered hot spots and the current passenger flow range data of the current hot spot are subjected to heterogeneous data fusion prediction to obtain the predicted passenger flow range data for the current hot spot.
[0016] The embodiment of the present application utilizes the characteristics of tourist behavior routes and tourist quantity to perform heterogeneous data fusion prediction on the passenger flow change data of the numbered hotspot areas and the current passenger flow range data of the current hotspot areas. By utilizing the passenger flow prediction model, it is possible to simultaneously take into account the prediction of tourist behavior trends and the prediction of tourist quantity, thereby greatly improving the accuracy of the predicted passenger flow range data for the current hotspot areas. At the same time, it also weakens the complexity of tourist guidance between related hotspot areas, further improving the reliability of the data input into the passenger flow prediction model.
[0017] In a feasible implementation, a preset passenger flow prediction model is used to perform heterogeneous data fusion prediction on the passenger flow change data of the numbered hotspot area and the current passenger flow range data of the current hotspot area to obtain the predicted passenger flow range data of the current hotspot area, specifically including: generating a tourist behavior route feature matrix based on the tourist behavior route features; extracting the passenger flow change features of the passenger flow change data in the numbered hotspot area through the passenger flow prediction model; and extracting the current passenger flow range features of the current passenger flow range data in the current hotspot area; wherein the passenger flow prediction model is capable of extracting data features; and performing relevant tourist transfer analysis on the passenger flow change features in the numbered hotspot area through the tourist quantity feature matrix and based on the time scale features. Multi-layer dilated causal convolution processing of the changing trend is performed to obtain the first heterogeneous data; wherein, the number of dilated causal convolution kernels used in the first heterogeneous data is half of the number of layers; according to the tourist behavior route feature matrix and based on the time scale feature, the current passenger flow range feature in the current hot spot area is processed with data points under the relevant feature dependency relationship to obtain the second heterogeneous data; wherein, the dilated causal convolution is applied in the second heterogeneous data to capture the dependency relationship between the data points with a long distance; through a preset single hidden layer fully connected network, the dependency relationship between the first heterogeneous data and the second heterogeneous data is processed by data fusion, and based on multiple iterations of the weight parameters in the passenger flow prediction model, the fused heterogeneous data is iteratively predicted to obtain the predicted passenger flow range data of the current hot spot area.
[0018] The embodiment of the present application uses the dilated causal convolution technology on the basis of the same time scale features to perform causal convolution processing on the extracted passenger flow change features, and performs data point capture processing on another extracted current passenger flow range feature under a dependency relationship, and then completes the fusion of heterogeneous data, thereby accurately outputting predicted passenger flow range data. The use of the dilated causal convolution technology can greatly improve the causal fusion accuracy of feature data, and at the same time, can also perform prediction processing based on the current time scale features for the changing passenger flow change features and the current passenger flow range features, which is beneficial to the accurate prediction function of the passenger flow prediction model.
[0019] In a feasible implementation manner, based on the current passenger flow range data and the predicted passenger flow range data, the correlation calculation of the passenger flow guidance association strength is performed on the route directivities between the several hot spots, and the passenger flow guidance change relationship between the various hot spots in the current time period is obtained, which specifically includes: obtaining the current passenger flow range data and the predicted passenger flow range data of each hot spot area in the current time period; based on the tourist behavior route characteristics corresponding to the current hot spot area, the current passenger flow range data and the predicted passenger flow range data of the remaining hot spots are used to calculate the passenger flow change range, and the passenger flow guidance data of each of the remaining hot spots is obtained; wherein the passenger flow guidance data includes increase guidance data and decrease guidance data; the passenger flow guidance data of each of the remaining hot spots is used to calculate the passenger flow change range of the remaining hot spots ... According to the numerical sorting processing, the guidance trend vector of each of the remaining hot spots is obtained; wherein, the guidance trend vector includes: a positive vector corresponding to the increase guidance data and a negative vector corresponding to the decrease guidance data; according to the absolute value of the guidance trend vector, the correlation calculation of the passenger flow guidance association strength is performed on the remaining hot spots, and the correlation level between each of the remaining hot spots and the current hot spot area is determined; wherein, the larger the absolute value of the guidance trend vector is, the higher the corresponding correlation level is; wherein, the correlation level includes: a positive correlation level corresponding to the positive vector and a negative correlation level corresponding to the negative vector; through the correlation level between the remaining hot spots and the current hot spot area, the passenger flow guidance change relationship between each of the hot spots in the current time period is generated.
[0020] The embodiment of the present application utilizes the positive correlation level and negative correlation level in the passenger flow direction change relationship, which can accurately display the passenger flow changes and regional correlation strength between various hot spots, reflect the changes in tourist direction between each hot spot area to the greatest extent, and improve the accuracy of basic data in complex passenger flow changes, thereby laying a data foundation for the subsequent optimal guidance recommendation route.
[0021] In a feasible implementation, based on the business district grid map corresponding to the target business district, the first position information of the current hot spot area and the second position information of the target hot spot area are searched for the best guidance path under the passenger flow guidance change relationship to determine the best guidance recommended route, which specifically includes: based on the business district grid map corresponding to the target business district, and through the target area path generation algorithm, the first position information of the current hot spot area and the second position information of several target hot spot areas are tracked and positioned to obtain a linked list array containing all paths; wherein the target hot spot area is any one of the remaining hot spot areas; if the passenger flow guidance change relationship between the target hot spot area and the current hot spot area is a passenger flow increase change relationship, then based on the The positive correlation level between the first location information and the second location information is used to search and sort the related paths in the linked list array, and the guidance path corresponding to the lowest positive correlation level is determined as the optimal guidance recommended route through the search connection of the line segment sequence with the minimum Manhattan distance between the two end points; if the passenger flow guidance change relationship between the target hot spot area and the current hot spot area is a passenger flow reduction change relationship, then based on the negative correlation level between the first location information and the second location information, the related paths in the linked list array are searched and sorted, and the guidance path corresponding to the highest negative correlation level is determined as the optimal guidance recommended route; wherein the positive correlation level and the negative correlation level are both included in the passenger flow guidance change relationship.
[0022] The embodiment of the present application is based on the passenger flow guidance change relationship between different hot spots, that is, the positive correlation level and the negative correlation level, so as to realize the path screening processing of the linked list array and obtain the best guidance recommended route. It only needs to consider the passenger flow change between the related hot spots to reflect the actual passenger flow situation of the target hot spots to the greatest extent. Since the passenger flow changes rapidly in real time and the amount of calculation is huge, if all areas are taken into consideration, the cost is too high, and the passenger flow prediction results and guidance recommended routes obtained will also be difficult to determine and process due to the redundant data. Therefore, the passenger flow guidance change relationship between the target hot spot area and the current hot spot area is used as the recommendation data basis for the route, which can complete the recommendation quickly and at a low cost, and push the best guidance recommended route to tourists as much as possible.
[0023] In a feasible implementation manner, the best guided recommended route is visualized on a large screen, and the best guided recommended route is connected to the terminal device of the tourist in real time through the visualized large screen, specifically comprising: installing the visualized large screen in the core hot spot area of the hot spot area, and visually displaying the best guided recommended route recommended by the current hot spot area to achieve guided recommendation for tourists in the current hot spot area; based on the scannable QR code in the visualized large screen, sending the best guided recommended routes recommended by the remaining hot spots to the terminal device of the tourist to obtain multi-domain best guided recommended routes; and estimating the walking time from the current hot spot area to the best guided recommended routes corresponding to the remaining hot spots to obtain an estimated walking time; generating the global best guided recommended route of the target business district through the estimated walking time and the multi-domain best guided recommended routes, and sending it to the terminal device; based on the push processing of the best guided recommended route, the multi-domain best guided recommended routes and the global best guided recommended routes, completing the travel guided recommendation for the tourist in the target business district.
[0024] In order to make it more convenient for tourists to have a shopping experience, this application implements that they can scan the QR code on the visual large screen, and then implement the best guidance recommended route from their own hot spot area to the target hot spot area, and even the large screen far away can still grasp the implementation information. At the same time, it can also implement the grasp of the passenger flow between the various hot spots in the entire target business district, that is, the congestion situation, and can obtain the best guidance recommended route between the remaining hot spots in advance, that is, the multi-domain best guidance recommended route, and can also grasp the best guidance recommended route for the entire domain, which is more convenient for tourists to travel, save congestion time, and play or go shopping efficiently, etc., which is conducive to the route planning of tourists when playing, and greatly improves the tourists' own play experience.
[0025] On the other hand, an embodiment of the present application also provides a passenger flow guidance planning device based on multiple areas in a business district, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, so that the at least one processor can execute a passenger flow guidance planning method based on multiple areas in a business district as described in any of the above embodiments.
[0026] The present application provides a method and device for passenger flow guidance planning based on multiple areas of a business district. Compared with the prior art, the embodiments of the present application have the following beneficial technical effects:
[0027] 1. Accurate regional division: Based on the geographical boundary of the target business district, the historical passenger flow collection area within the business district boundary is divided into heat map regions to obtain several hot spots, which helps to more accurately understand the passenger flow distribution in different areas within the business district.
[0028] 2. Spatiotemporal analysis: By performing data gridding processing on hot spots and hot spot analysis of spatiotemporal distribution changes, the current passenger flow range data can be determined, which can provide an in-depth understanding of the changing patterns of passenger flow in time and space.
[0029] 3. Forecasting function: Based on the characteristics of the current passenger flow range data, analyze the changing trend of tourist transfer between hot spots, predict the predicted passenger flow range data in the future time period, and provide forward-looking guidance for the operation and management of the business district.
[0030] 4. Customer flow optimization: By calculating the correlation strength of customer flow between hot spots, the change relationship of customer flow can be obtained, which helps to optimize the customer flow distribution between hot spots in the business district and improve the overall operational efficiency of the business district.
[0031] 5. Best route recommendation: Based on the business district grid map, search for the best recommended route, which can provide tourists with more convenient and efficient travel routes and enhance the tourists' experience.
[0032] 6. Visual display and real-time connection: The best recommended route is displayed on a large screen and connected to the tourists’ terminal devices in real time, making it easier for tourists to obtain information, enhancing the interactivity between tourists and the business district and improving service quality. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:
[0034] Figure 1 A flow chart of a passenger flow guidance planning method based on multiple areas of a business district provided in an embodiment of the present application;
[0035] Figure 2 A schematic diagram of dividing hotspot level areas provided in an embodiment of the present application;
[0036] Figure 3 A schematic diagram of association of numbered hotspot areas provided in an embodiment of the present application;
[0037] Figure 4A schematic diagram of a data prediction process based on heterogeneous data fusion provided in an embodiment of the present application;
[0038] Figure 5 A schematic diagram of the structure of a passenger flow guidance planning device based on multiple areas of a business district provided in an embodiment of the present application. DETAILED DESCRIPTION
[0039] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.
[0040] It should be noted that the hotspot area is a representative area of the passenger flow aggregate in the business district, which can indicate that there are a large number of tourists and the number of tourists entering and leaving the area. The passenger flow represents the number of tourists traveling between different areas in the business district. The larger the passenger flow in a hotspot area, the more people there are in the area, which means that it is more crowded, the higher the safety risks are, and the lower the tourists' travel experience is.
[0041] It should also be noted that the embodiment of the present application only considers the correlation between the current hot spot area and the remaining hot spot areas. Since the accuracy of the passenger flow change data is difficult to guarantee in actual activities, only the passenger flow change between the current hot spot area and the target hot spot area is considered, including the current passenger flow range data and the predicted passenger flow range data, etc. In this way, the passenger flow change between the two hot spots can be calculated with relatively accurate quantitative data. From the field experiment, only by using the above-mentioned implementation method, the real passenger flow change can be estimated more accurately. In addition, the application only obtains the recommended route that is convenient for tourists to experience and alleviate the congestion problem of the play route. Therefore, a passenger flow guidance planning method based on multiple areas of a business district based on the present application can make it more convenient for tourists to travel, save congestion time, and efficiently play or go shopping, etc. It is very beneficial to the route planning of tourists when playing, and greatly improves the tourists' own play experience.
[0042] The present application embodiment provides a method for planning passenger flow guidance based on multiple areas of a business district, such as Figure 1 As shown, the passenger flow guidance planning method based on multiple areas of a business district specifically includes steps S101-S106:
[0043] S101. According to the geographical boundary of the target business district, the historical passenger flow collection area in the boundary of the business district is divided into regions related to the heat map to obtain a number of hot spots.
[0044] Specifically, it is necessary to first divide the geographical location of each merchant into adjacent areas based on the commercial data in the smart management platform, and determine the collection of merchant adjacent areas, that is, through the smart management platform, obtain the commercial data in a certain area, such as shopping malls, popular attractions, tourist attractions, pedestrian streets, and large-scale building store groups, etc. Among them, commercial data at least includes: merchant operation data, building passenger flow data, popular attraction passenger flow data, merchant location table data, and merchant rental management data.
[0045] Furthermore, the preset remote sensing image geoscience analysis technology is used to divide the spatial scale data of the merchant adjacent area set into latitude and longitude of the relevant scale pattern, and obtain the target business district that includes any merchant area in the merchant adjacent area set. And based on the regional boundaries of the target business district, the geographical boundary range is determined. Among them, the spatial scale data includes: the horizontal boundary shape data of the merchant area and the vertical structure data of the merchant area.
[0046] In one embodiment, a smart management platform for commercial data management is first constructed using the docked big data platform, and the smart management platform can regularly update data types. Then, using a regionalized data management method, the geographical location of each merchant in the target area is divided into adjacent areas, that is, a set of merchant adjacent areas within a certain range of association is determined, and then remote sensing image geoscientific analysis technology is used to divide the spatial scale data pattern of the above merchant adjacent area set, that is, using the merchant adjacent area set, the merchant area within a certain range is determined, and the merchant set within these ranges is defined as the target business district.
[0047] Furthermore, based on a number of thermal analysis cameras pre-installed in the target business district, historical passenger flow data is collected in multiple historical time periods to obtain a historical passenger flow data set of each thermal analysis camera.
[0048] Furthermore, by using the merchant adjacent area set, each historical passenger flow data set is subjected to a scale deduction process related to the passenger flow set area to obtain the historical passenger flow set area.
[0049] Furthermore, it is also necessary to transform the historical passenger flow collection area into a representation of relevant thermal characteristics to obtain a historical thermal characteristic map of the target business district. The historical thermal characteristic map includes historical passenger flow collection areas in different colors.
[0050] Furthermore, based on the heterogeneity between each thermal feature in the historical thermal characteristic map, the historical thermal characteristic map is processed by level patch division at the macro-regional scale. That is, it is necessary to utilize the heterogeneity between each thermal feature and perform regional scale division processing on the historical thermal characteristic map based on the image level patch technology, and finally obtain the hot spot area reflecting the concentrated change area of passenger flow.
[0051] In one embodiment, in order to facilitate the management of passenger flow in the target business district, the ordinary surveillance cameras in the business district can be replaced with thermal analysis cameras, and then the passenger flow data in multiple historical periods can be collected. Then, the scale deduction of the geographic image of the target business district is used to express the multi-scale nature of the region to ensure that the observation scale of the image acquisition information matches the intrinsic scale of the surface structure information, so as to determine the historical passenger flow collection area on the remote sensing map, so that the data obtained subsequently maintains the consistency of the geographic scale. At the same time, with the help of the historical thermal characteristic map of the target business district, that is, the heterogeneity between the various thermal characteristics in the historical thermal characteristic map, the already formed thermal area is processed by the regional scale division under the image level patch technology, that is, the different thermal areas are divided into patches to accurately reflect the hot spots in the area where the passenger flow is concentrated and changed.
[0052] S102. Data gridding is performed on the hotspot area through a preset spatial topological relationship, and hotspot analysis of the temporal and spatial distribution changes of the passenger flow data collected in the hotspot area after data gridding is performed to determine the current passenger flow range data in the hotspot area.
[0053] Specifically, firstly, a thermal analysis camera installed in the hot spot area is used to collect a number of point-shaped periodic passenger flow data in the hot spot area based on a preset collection period, wherein the collection period includes at least: 5 minutes, 10 minutes and 15 minutes.
[0054] Furthermore, through the spatial topological relationship, each point-shaped periodic passenger flow data is summarized and converted into the number of points, and the surface-shaped periodic passenger flow data of each hot spot area is obtained. Among them, the spatial topological relationship includes: intersection relationship, inclusion relationship and separation relationship.
[0055] Furthermore, it is also necessary to merge several area-shaped periodic passenger flow data in the collection period into the same layer to obtain the passenger flow grid spatial distribution data of each hotspot area. Among them, the passenger flow grid spatial distribution data is the initial passenger flow change data in the collection period.
[0056] As a feasible implementation method, data gridding refers to converting point positioning data into surface data through spatial topological relationship analysis. The spatial topological relationship between points and surfaces mainly includes three types: intersection, inclusion, and separation, among which inclusion is a special case of intersection. The surface grid data counts the positioning points that intersect with it in turn, and summarizes the positioning numbers of all points to obtain the total passenger flow data of the surface. Data gridding can simplify the amount of data, thereby improving the efficiency of data analysis. At the same time, the periodic passenger flow data collected during the collection period can be very closely combined with the hot spot area image to complete the same layer of merging processing, that is, to obtain the gridded spatial distribution data of passenger flow in each hot spot area, thereby ensuring the consistency of passenger flow data in the regional and time domain dimensions.
[0057] Furthermore, the terminal time node of the collection cycle is used to perform time series data labeling processing on the gridded spatial distribution data of passenger flow to obtain the current spatial distribution data of passenger flow at each terminal time node.
[0058] Furthermore, it is also necessary to process the random factors in the current passenger flow spatial distribution data with relevant clustering scores, that is, it is necessary to score the random factors, provide data basis for the high and low values of the passenger flow numerical attributes, and then obtain the clustering density of each random factor. Among them, the clustering density is proportional to the clustering score. The random factors are the high and low values of the passenger flow numerical attributes in the current passenger flow spatial distribution data.
[0059] Furthermore, based on the clustering density of each random factor, the current passenger flow spatial distribution data is processed with significant clustering equilibrium to obtain the current passenger flow estimation data of each hot spot area. Among them, significant clustering includes: high-value hot spot significance and low-value cold spot significance.
[0060] In one embodiment, hot spot analysis of the temporal and spatial distribution changes of passenger flow data can be performed by obtaining the spatial clustering positions of high-value or low-value elements. In order to make the current passenger flow spatial distribution data statistically good, the random factor method is used, that is, the high-value and low-value scoring method is used to judge the clustering tightness of each hot spot area, that is, whether the passenger flow size of each hot spot area is obviously significant, so as to integrate the current passenger flow spatial distribution data of the entire hot spot area by using the high-value hot spot significance and the low-value cold spot significance, that is, finally obtain the current passenger flow estimation data that can better reflect the current terminal time node of each hot spot area.
[0061] Furthermore, based on the cycle span value in the collection cycle, the current passenger flow estimation data is subjected to curve fitting processing under the change of passenger flow peak value to obtain the passenger flow time change curve, wherein the cycle span value is the dependent variable of the passenger flow time change curve.
[0062] Furthermore, using the above-obtained current passenger flow estimation data and the corresponding aggregation density, the hotspot area is processed for core hotspot area identification based on passenger flow thermal value, and according to the identified core hotspot area, the hotspot area is diffused and identified along the same axis to divide the hotspot level area. The hotspot level area includes: core hotspot area, secondary hotspot area and edge hotspot area.
[0063] In one embodiment, Figure 2 A schematic diagram of dividing hotspot level areas provided in an embodiment of the present application is shown in FIG. Figure 2 As shown in the figure, each hotspot area contains at least core hotspot areas, secondary hotspot areas, and edge hotspot areas. In other words, by using the passenger flow time variation curve after curve fitting and combining the density of the above-mentioned entire hotspot area, the core hotspot area identification process based on the passenger flow thermal value can be carried out. First, the core hotspot area with the densest aggregation is identified, and then it is diffused in sequence according to the equal axis lines to divide the hotspot level areas with different passenger flow sizes in sequence.
[0064] Further, based on the hotspot level area, the current passenger flow distribution pattern of the hotspot area is determined. Then, through the current time dimension in the current passenger flow distribution pattern, the time dimension corresponding to the cycle span value in the passenger flow time change curve is subjected to consistency matching processing.
[0065] Furthermore, according to the core hotspot areas in the current passenger flow distribution pattern, the spatial difference algorithm is used to perform data fusion processing between the benchmark data points between the passenger flow time change curve and the current passenger flow distribution pattern to obtain the current spatiotemporal passenger flow estimation data. Among them, the algorithm used for data fusion is the Geopandas library technology in Python. And the benchmark data point is the time dimension data point shared by the current passenger flow distribution pattern and the passenger flow time change curve. Finally, the current spatiotemporal passenger flow estimation data is calibrated within a fixed error range to obtain the current passenger flow range data in each hotspot area.
[0066] As a feasible implementation method, the hotspot level area can be used to determine the current passenger flow distribution pattern of the hotspot area, that is, the current passenger flow distribution of each hotspot area. Then, the spatial difference algorithm is used to insert the passenger flow time change curve into the current passenger flow distribution pattern under the benchmark data point, and then the Geopandas library technology in Python is combined to fuse the above passenger flow time change curve with the current passenger flow distribution pattern, that is, to use the data difference between the two data to optimize and calibrate the current spatiotemporal passenger flow estimation data to the greatest extent, and finally obtain the current passenger flow range data.
[0067] S103. Based on the time scale characteristics and tourist behavior route characteristics in the current passenger flow range data, the current passenger flow range data is analyzed for the trend of tourist transfer changes between hot spots, and the predicted passenger flow range data of the hot spots in the future time period is predicted and obtained.
[0068] Specifically, the historical passenger flow range data in each hotspot area is first obtained through the smart management platform, wherein the historical passenger flow range data is the data collected in the previous collection cycle.
[0069] Furthermore, the hot spot area correlation analysis of the passenger flow change trend is performed between the historical passenger flow range data and the current passenger flow range data to determine the associated hot spot area, wherein the associated hot spot area is a hot spot area with passenger flow orientation correlation.
[0070] Furthermore, the passenger flow change trend of the associated hotspot area and the current passenger flow range data is used to determine the tourist behavior route characteristics in the current passenger flow range data. Among them, the tourist behavior route characteristics are the tourist guidance characteristics between the associated hotspot areas. And based on the tourist behavior route characteristics of the current hotspot area, the remaining hotspot areas in the associated hotspot area are numbered as sub-areas to obtain numbered hotspot areas.
[0071] In one embodiment, Figure 3 A schematic diagram of the association of numbered hot spots provided in an embodiment of the present application is shown in FIG. Figure 3As shown, the No. 1 hotspot area is set as the current hotspot area where tourists are located, and the remaining hotspot areas are No. 2, No. 3, No. 4, No. 5 and No. 6. Among them, the remaining hotspot areas are the associated hotspot areas between the historical passenger flow range data and the current passenger flow range data, where there is a correlation between the hotspot areas related to the passenger flow change trend. Then, by using the associated hotspot areas and the passenger flow change trend of the current passenger flow range data, the characteristics of the tourist behavior route in the current passenger flow range data can be determined, that is, the determination of the tourist behavior route with a tourist-oriented trend. Finally, these associated hotspot areas with tourist behavior route characteristics are numbered and located in sub-areas, thereby obtaining numbered hotspot areas.
[0072] Furthermore, it is also necessary to generate a two-dimensional matrix of the passenger flow change characteristics in the numbered hotspot area according to the time scale characteristics of the current hotspot area and the characteristics of the tourist behavior route, so as to obtain the tourist quantity characteristic matrix of each numbered hotspot area. Among them, the passenger flow change characteristics include: passenger flow increase change characteristics and passenger flow decrease change characteristics. Then determine the passenger flow change data corresponding to the tourist quantity characteristic matrix in each numbered hotspot area.
[0073] Furthermore, through the preset passenger flow prediction model, the passenger flow change data of the numbered hot spot area and the current passenger flow range data of the current hot spot area are fused and predicted as heterogeneous data to obtain the predicted passenger flow range data of the current hot spot area.
[0074] As a feasible implementation method, based on the same time scale characteristics in the numbered hotspot area, combined with the characteristics of tourist behavior routes and the current passenger flow change range data of the same numbered hotspot area, a tourist quantity feature matrix based on the above two characteristics can be generated, that is, the passenger flow change characteristics in the numbered hotspot area can be obtained, that is, the increase change characteristics of the passenger flow inflow of tourists and the decrease change characteristics of the passenger flow outflow of tourists. Then, using the trained passenger flow prediction model, the passenger flow change data of the numbered hotspot area and the current passenger flow range data of the current hotspot area are used as the input of the model, and the two heterogeneous data are fused by using the dilated causal convolution method, and the predicted passenger flow range data can be obtained as the model at the output.
[0075] Among them, the predicted passenger flow range data of the current hot spot area obtained by the passenger flow prediction model can be:
[0076] 1) First, based on the characteristics of tourist behavior routes, generate a tourist behavior route feature matrix.
[0077] 2) Then, the passenger flow prediction model is used to extract the passenger flow change characteristics of the passenger flow change data in the numbered hotspot area. Also, the current passenger flow range characteristics of the current passenger flow range data in the current hotspot area are extracted. The passenger flow prediction model can perform data feature extraction processing.
[0078] 3) Then, through the tourist quantity feature matrix and based on the time scale feature, the multi-layer dilated causal convolution processing on the tourist transfer change trend is performed on the passenger flow change characteristics in the numbered hotspot area to obtain the first heterogeneous data. Among them, the number of dilated causal convolution kernels used in the first heterogeneous data is half of the number of layers.
[0079] 4) Then, according to the tourist behavior route feature matrix and based on the time scale feature, the current passenger flow range feature in the current hot spot area is processed by capturing data points under the feature dependency relationship to obtain the second heterogeneous data. In the second heterogeneous data, the dilated causal convolution is applied to capture the dependency relationship between data points that are far away.
[0080] 5) Finally, the dependency relationship between the first heterogeneous data and the second heterogeneous data is fused through a preset single hidden layer fully connected network, and the fused heterogeneous data is iteratively predicted based on multiple iterations of the weight parameters in the passenger flow prediction model to obtain the predicted passenger flow range data of the current hot spot area.
[0081] In one embodiment, Figure 4 A schematic diagram of a data prediction process based on heterogeneous data fusion provided in an embodiment of the present application is shown in FIG. Figure 4 As shown in the figure, using the tourist quantity feature matrix and based on the time scale feature, the multi-layer dilated causal convolution processing of the tourist transfer change trend is performed on the passenger flow change characteristics in the numbered hotspot area. That is, in order to obtain the first heterogeneous data, the input of the four-layer dilated causal convolution is a time series of length 16, corresponding to 16 nodes, and the dilation factor is selected as 2, then the convolution kernel size is 1×2, and the multi-layer dilated causal convolution can be completed after four layers of dilated convolution, that is, the feature quantity output by each layer of dilated convolution is processed by L layers of dilated convolution, and then a 1×1 convolution is used to obtain the final predicted first heterogeneous data. For the second heterogeneous data, it is necessary to apply one convolution kernel to every n elements of the input vector in the dilated convolution based on the expansion factor and the number of channel layers to capture the dependency between data points that are far away, where the dependency may include periodic, long-term and short-term relationships, and finally predict the above second heterogeneous data.
[0082] In a real-time example, Figure 4As shown in the figure, a single hidden layer fully connected neural network is first used. The weight parameter in the passenger flow prediction model is the weight parameter in the single hidden layer fully connected neural network. The hidden layer is obtained by the tourist quantity feature matrix, the tourist behavior route feature matrix and the weight parameter matrix. Then the hidden layer matrix is multiplied by the weight parameter matrix to obtain the final output matrix. The weight parameter obtains the optimal value in multiple iterations. Each element in the output matrix represents the probability of correlation between the current passenger flow range data in the current hotspot area and the passenger flow change data in the numbered hotspot area. According to the correlation probability, the final output after fusion is obtained by combining the tourist quantity feature matrix and the tourist behavior route feature matrix, which is the predicted passenger flow range data of the current hotspot area.
[0083] S104: Based on the current passenger flow range data and the predicted passenger flow range data, the correlation calculation of the passenger flow orientation correlation strength is performed on the route directivity between the several hot spots, and the passenger flow orientation change relationship between the hot spots in the current time period is obtained. The passenger flow orientation change relationship is the passenger flow increase and decrease change relationship between the hot spots.
[0084] Specifically, it is necessary to first obtain the current passenger flow range data and the predicted passenger flow range data of each hot spot area in the current time period.
[0085] Furthermore, based on the tourist behavior route characteristics corresponding to the current hot spot area, the current passenger flow range data and the predicted passenger flow range data of the other hot spots are used to calculate the passenger flow change range, and the passenger flow guidance data of each other hot spot area is obtained. The passenger flow guidance data includes increase guidance data and decrease guidance data.
[0086] Furthermore, it is also necessary to sort the passenger flow guidance data of each remaining hot spot area under numerical sorting to obtain the guidance trend vector of each remaining hot spot area. The guidance trend vector includes: a positive vector corresponding to the increase guidance data and a negative vector corresponding to the decrease guidance data.
[0087] Furthermore, the absolute value of the guiding trend vector is used to calculate the correlation of the passenger flow guiding association strength for the remaining hot spots, and the correlation level between each remaining hot spot area and the current hot spot area is determined. The larger the absolute value of the guiding trend vector, the higher the corresponding correlation level. The correlation level includes: a positive correlation level corresponding to a positive vector and a negative correlation level corresponding to a negative vector.
[0088] Furthermore, taking the correlation levels between the remaining hot spots and the current hot spots as reference, a passenger flow orientation change relationship between the hot spots in the current time period is generated.
[0089] As a feasible implementation method, Figure 3 As shown in the figure, based on the current passenger flow range data and predicted passenger flow range data of each hot spot area at the end time node of the current time period, referring to the tourist behavior route characteristics corresponding to the current hot spot area, the passenger flow guidance result of the current hot spot area is judged, that is, which area the passenger flow increases or which area the passenger flow decreases to, so as to judge the numerical sorting result of the passenger flow guidance data between the current hot spot area No. 1 and No. 2, No. 3, No. 4, and No. 5, that is, the guidance trend vector, and then use the absolute value of the guidance trend vector to complete the correlation calculation of the remaining hot spots with positive and negative guidance trends. For example, No. 2 and No. 3 are positive vectors, and the absolute value of the guidance trend vector of No. 2 is greater than that of No. 3, which means that the tourists flowing out of No. 1 are the largest outflow from No. 2, indicating that the trend of tourists guiding No. 2 is obvious, and No. 2 is determined as the maximum positive correlation level; similarly, the larger the negative correlation level corresponding to the negative vector, the more obvious the trend between the corresponding numbered hot spot area and No. 1 is, and the passenger flow flowing into No. 1 is large.
[0090] S105. Based on the grid map of the business district corresponding to the target business district, the first position information of the current hot spot area and the second position information of the target hot spot area are searched for the best guidance path under the relationship of passenger flow guidance change, and the best guidance recommended route is determined.
[0091] Specifically, firstly, based on the grid map of the business district corresponding to the target business district, and through the target area path generation algorithm, the first position information of the current hot spot area and the second position information of several target hot spot areas are tracked and positioned, and a linked list array containing all paths is obtained, that is, based on the use of navigation technology, the path line between the first position information and the second position information is generated. The first position information and the second position information can be the position information of any hot spot area, that is, many path lines. Therefore, after tracking and positioning, a linked list array of route links is generated, which includes all path lines. Among them, the target hot spot area is any one of the remaining hot spots.
[0092] If the passenger flow guidance change relationship between the target hot spot area and the current hot spot area is an increase in passenger flow, then based on the positive correlation level between the first position information and the second position information, the relevant paths in the linked list array are searched and sorted, and the guidance path corresponding to the lowest positive correlation level is determined as the best guidance recommended route through the search connection of the line segment sequence with the smallest Manhattan distance between the two end points.
[0093] If the passenger flow guidance change relationship between the target hot spot area and the current hot spot area is a passenger flow decrease change relationship, then based on the negative correlation level between the first position information and the second position information, the relevant paths in the linked list array are searched and sorted, and the guidance path corresponding to the highest negative correlation level is determined as the best guidance recommended route.
[0094] Among them, both the positive correlation level and the negative correlation level are included in the passenger flow orientation change relationship, that is, the data of the positive correlation level and the negative correlation level are derived from the above passenger flow orientation change relationship.
[0095] As a feasible implementation method, a coordinate system can be first established based on the hot spot area, and then a path search of the target hot spot area coordinates (second position information) is performed from the current hot spot area as the origin coordinates (first position information), and the terminating boundary is the boundary range of the target business district. Then, all uncovered points are obtained within a fixed range. At the same time, if the horizontal and vertical coordinate errors of the above points are lower than the endpoints of the grid, the target node selection is expanded based on this, and finally a linked list array containing all paths is generated. Then, the change relationship of passenger flow increase and the change relationship of passenger flow decrease are used as reference data, and the improved A* algorithm is used to complete the connectivity of global and local waypoints and line connection, so as to form a fully covered line. That is, it is necessary to search and connect the line sequence with the minimum Manhattan distance between the two end points, and finally determine the guidance path corresponding to the lowest level of positive correlation as the best guidance recommended route and / or determine the guidance path corresponding to the highest level of negative correlation as the best guidance recommended route. In this way, it is possible to better avoid congested hot spots in the future and have an efficient travel experience. At the same time, based on the passenger flow guidance change relationship, it is possible to prevent trampling or crowding accidents, do a good job in crowd diversion in advance, prevent excessive crowding, and enable the smart management platform to obtain accurate current passenger flow range data, predicted passenger flow range data and passenger flow guidance change relationship to the greatest extent.
[0096] S106. The best recommended route is displayed on a large-screen visualization screen, and the best recommended route is connected to the terminal device of the tourist in real time through the large-screen visualization screen to complete the recommendation of tourist travel guidance in the target business district.
[0097] Specifically, in order to better feedback the best recommended route to tourists, direct visual communication or mobile applets can be used: install a large visual screen in the core hot spot area of the hot spot area, and visualize the best recommended route recommended by the current hot spot area to achieve guidance recommendations for tourists in the current hot spot area.
[0098] Furthermore, based on the scannable QR code in the visualization screen, the best guide recommended routes for other hot spots can be sent to the terminal device of the tourists to obtain the best guide recommended routes in multiple domains. The walking time from the current hot spot to the best guide recommended routes corresponding to other hot spots can be estimated and calculated to obtain the estimated walking time.
[0099] Furthermore, the estimated walking time and the multi-domain optimal navigation recommended route are used to generate the global optimal navigation recommended route of the target business district and send it to the terminal device.
[0100] Furthermore, based on the push processing of the best guided route, the multi-domain best guided route and the global best guided route, that is, the best guided route for different tourists' needs is comprehensively generated and pushed to complete the travel guide recommendation for tourists in the target business district.
[0101] In one embodiment, the use of a large visual screen can alleviate the crowd guidance planning in the core hot spots among the hot spots, and help tourists intelligently recommend uncrowded travel routes or shopping areas. At the same time, a mini-program scanning function is also launched, allowing people to view the best guidance recommended routes for the current hot spots on their mobile phones, as well as the guidance recommended routes for the remaining hot spots or all hot spots, which can further enable tourists to obtain the most convenient way to travel, greatly save congestion time, improve tourist management in the business district, and optimize the distribution of passenger flow between various hot spots in the business district.
[0102] In one embodiment, by using the global best navigation recommended route, the smart management platform can also plan the route of the smart garbage collection vehicle. That is, by using the uncongested navigation route, the smart garbage collection vehicle can quickly shuttle between different hot spots. It can also avoid too many tourists, making it difficult for the vehicle to move and even causing secondary congestion. Based on the global best navigation recommended route, the tracking planning of the smart garbage collection vehicle can be realized, thereby improving the smart sanitation operation efficiency of the business district.
[0103] In addition, the embodiment of the present application also provides a passenger flow guidance planning device based on multiple areas of a business district, such as Figure 5 As shown, the passenger flow guidance planning device 500 based on multiple areas of a business district specifically includes:
[0104] At least one processor 501. And a memory 502 in communication with the at least one processor 501. The memory 502 stores instructions that can be executed by the at least one processor 501, so that the at least one processor 501 can execute:
[0105] According to the geographical boundary of the target business district, the historical passenger flow collection area in the business district boundary is divided into regions related to the heat map to obtain several hot spots;
[0106] Through the preset spatial topological relationship, the hotspot area is processed into data grids, and the passenger flow data collected in the hotspot area after data gridding is subjected to hotspot analysis on the changes in spatiotemporal distribution to determine the current passenger flow range data in the hotspot area;
[0107] Based on the time scale characteristics and tourist behavior route characteristics in the current passenger flow range data, the current passenger flow range data is analyzed for the trend of tourist transfer changes between hot spots, and the predicted passenger flow range data of the hot spots in the future time period is predicted and obtained;
[0108] According to the current passenger flow range data and the predicted passenger flow range data, the correlation calculation of the passenger flow orientation correlation strength is performed on the route directivity between several hot spots, and the passenger flow orientation change relationship between the hot spots in the current time period is obtained; wherein the passenger flow orientation change relationship is the passenger flow increase and decrease change relationship between the hot spots;
[0109] Based on the grid map of the business district corresponding to the target business district, the first position information of the current hot spot area and the second position information of the target hot spot area are searched for the best guidance path under the relationship of passenger flow guidance change, and the best guidance recommended route is determined;
[0110] The best recommended route is displayed on a large-screen visualization screen, and the best recommended route is connected to the tourists' terminal devices in real time through the visualization screen to complete the recommendation of tourist guides in the target business district.
[0111] The embodiment of the present application divides the historical passenger flow collection area in the business district boundary into heat map areas, which helps to more accurately understand the passenger flow distribution in different areas of the business district. It is also possible to use the hot spot analysis of the temporal and spatial distribution changes to determine the current passenger flow range data, which can deeply understand the changing rules of passenger flow in time and space. And based on the characteristics of the current passenger flow range data, the trend of tourist transfer changes between hot spots is analyzed, and the predicted passenger flow range data in the future time period is predicted to provide forward-looking guidance for the operation and management of the business district. At the same time, it can also greatly help optimize the passenger flow distribution between various hot spots in the business district, improve the overall operational efficiency of the business district, and provide tourists with more convenient and efficient tour routes to enhance the experience of tourists. It further facilitates tourists to obtain information, enhances the interactivity between tourists and business districts, and improves service quality.
[0112] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0113] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0114] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiments of the present application should be included in the scope of the specification of the present application.
Claims
1. A passenger flow guidance planning method based on multiple areas of a business district, characterized in that: The method comprises: According to the geographical boundary of the target business district, the historical passenger flow collection area in the business district boundary is divided into regions related to the heat map to obtain several hot spots; Through the preset spatial topological relationship, the hotspot area is subjected to data gridding processing, and the passenger flow data collected in the hotspot area after data gridding is subjected to hotspot analysis of the temporal and spatial distribution changes to determine the current passenger flow range data in the hotspot area, specifically including: The random factors in the current passenger flow spatial distribution data are processed with scores related to aggregation to obtain the aggregation density of each random factor; wherein the aggregation density is proportional to the aggregation score value; the random factors are the high value and the low value of the passenger flow numerical attribute in the current passenger flow spatial distribution data; Based on the clustering density of each of the random factors, the current passenger flow spatial distribution data is subjected to significant clustering equilibrium processing to obtain the current passenger flow estimation data of each of the hot spot areas; wherein the significant clustering includes: the hot spot significance of the high value and the cold spot significance of the low value; Based on the cycle span value in the collection cycle, the current passenger flow estimation data is subjected to curve fitting processing under the change of the passenger flow peak value to obtain a passenger flow time change curve; wherein the cycle span value is the dependent variable of the passenger flow time change curve; According to the current passenger flow estimation data and the corresponding aggregation density, the hot spot area is subjected to core hot spot area identification processing based on the passenger flow thermal value, and according to the identified core hot spot area, the hot spot area is subjected to isoaxial diffusion identification to divide the hot spot level area; wherein the hot spot level area includes: core hot spot area, secondary hot spot area and edge hot spot area; Based on the hotspot level area, determining a current passenger flow distribution pattern of the hotspot area; Through the current time dimension in the current passenger flow distribution pattern, the time dimension corresponding to the cycle span value in the passenger flow time variation curve is subjected to consistency matching processing; According to the core hotspot area in the current passenger flow distribution pattern, the passenger flow time variation curve and the current passenger flow distribution pattern are subjected to data fusion processing between the benchmark data points through a spatial difference algorithm to obtain the current spatiotemporal passenger flow estimation data; wherein the algorithm used for the data fusion is the Geopandas library technology in Python; the benchmark data point is the time dimension data point shared by the current passenger flow distribution pattern and the passenger flow time variation curve; Calibrate the current spatiotemporal passenger flow estimation data within a fixed error range to obtain the current passenger flow range data in each hotspot area; Based on the time scale characteristics and tourist behavior route characteristics in the current passenger flow range data, the current passenger flow range data is analyzed for the tourist transfer change trend between hot spots, and the predicted passenger flow range data of the hot spots in the future time period is predicted and obtained; According to the current passenger flow range data and the predicted passenger flow range data, the correlation calculation of the passenger flow orientation correlation strength is performed on the route directivities between the plurality of hot spots, so as to obtain the passenger flow orientation change relationship between the various hot spots in the current time period; wherein the passenger flow orientation change relationship is the passenger flow increase and decrease change relationship between the various hot spots; Based on the grid map of the business district corresponding to the target business district, searching and processing the best guidance path under the passenger flow guidance change relationship between the first position information of the current hot spot area and the second position information of the target hot spot area, and determining the best guidance recommended route; The best recommended route is visualized on a large screen, and the best recommended route is connected to the terminal device of the tourist in real time through the large screen to complete the travel guide recommendation for the tourist in the target business district.
2. According to claim 1, a passenger flow guidance planning method based on multiple areas of a business district is characterized in that: According to the geographical boundary of the target business district, the historical passenger flow collection area in the business district boundary is divided into regions related to the heat map, and several hot spots are obtained, including: Based on the commercial data in the smart management platform, the geographical location of each merchant is divided into adjacent areas to determine the merchant adjacent area set; wherein the commercial data at least includes: merchant operation data, building passenger flow data, popular scenic spot passenger flow data, merchant location table data and merchant rental management data; By using a preset remote sensing image geoscience analysis technology, the spatial scale data of the merchant adjacent area set is divided into longitude and latitude of the relevant scale pattern to obtain the target business district containing any merchant area in the merchant adjacent area set; and based on the regional boundary of the target business district, the geographical boundary range is determined; wherein the spatial scale data includes: the horizontal boundary shape data of the merchant area and the vertical structure data of the merchant area; According to a plurality of thermal analysis cameras pre-installed in the target business district, data collection is performed on historical passenger flow data in multiple historical time periods to obtain a set of historical passenger flow data of each thermal analysis camera; According to the merchant adjacent area set, each of the historical passenger flow data sets is subjected to scale deduction processing on the passenger flow set area to obtain the historical passenger flow set area; The historical passenger flow collection area is converted into a representation of relevant thermal characteristics to obtain a historical thermal characteristic map of the target business district; wherein the historical thermal characteristic map includes historical passenger flow collection areas in different colors; Based on the heterogeneity between the various thermal characteristics in the historical thermal characteristic map, the historical thermal characteristic map is divided into hierarchical patches at the macro-regional scale to obtain the hot spot area reflecting the concentrated change area of passenger flow.
3. The method for passenger flow guidance planning based on multiple areas of a business district according to claim 1 is characterized in that: The hotspot area is processed into data grids through a preset spatial topological relationship, specifically including: By installing a thermal analysis camera in the hot spot area, based on a preset collection period, a plurality of point-shaped periodic passenger flow data in the hot spot area are collected; wherein the collection period includes at least: 5 minutes, 10 minutes and 15 minutes; Through the spatial topological relationship, each of the point-shaped periodic passenger flow data is summarized and converted according to the number of points, so as to obtain the surface-shaped periodic passenger flow data of each hot spot area; wherein the spatial topological relationship includes: intersection relationship, inclusion relationship and separation relationship; Merge the passenger flow data of several planar cycles in the collection period on the same layer to obtain the passenger flow grid spatial distribution data of each hot spot area; wherein the passenger flow grid spatial distribution data is the initial passenger flow change data in the collection period; Based on the terminal time node of the collection period, the gridded spatial distribution data of passenger flow is subjected to time series data labeling processing to obtain the current spatial distribution data of passenger flow at each terminal time node.
4. The method for passenger flow guidance planning based on multiple areas of a business district according to claim 1 is characterized in that: Based on the time scale characteristics and tourist behavior route characteristics in the current passenger flow range data, the current passenger flow range data is analyzed for the tourist transfer change trend between hot spots, and the predicted passenger flow range data of the hot spots in the future time period is predicted and obtained, specifically including: Obtaining historical passenger flow range data in each of the hot spots; wherein the historical passenger flow range data is data collected in the last collection cycle; Performing a hot spot area correlation analysis on the passenger flow change trend between the historical passenger flow range data and the current passenger flow range data to determine a related hot spot area; wherein the related hot spot area is a hot spot area with passenger flow orientation correlation; Based on the passenger flow change trend of the associated hotspot area and the current passenger flow range data, determining the tourist behavior route characteristics in the current passenger flow range data; wherein the tourist behavior route characteristics are tourist guidance characteristics between the associated hotspot areas; Based on the tourist behavior route characteristics of the current hot spot area, the remaining hot spots in the associated hot spot area are numbered as sub-areas to obtain numbered hot spots; According to the time scale characteristics of the current hot spot area and the characteristics of the tourist behavior route, a two-dimensional matrix is generated for the passenger flow change characteristics in the numbered hot spot area to obtain a tourist quantity characteristic matrix for each numbered hot spot area; wherein the passenger flow change characteristics include: passenger flow increase change characteristics and passenger flow decrease change characteristics; Determine the passenger flow change data corresponding to the tourist quantity characteristic matrix in each numbered hotspot area; Through a preset passenger flow prediction model, the passenger flow change data of the numbered hot spot area and the current passenger flow range data of the current hot spot area are subjected to heterogeneous data fusion prediction to obtain the predicted passenger flow range data of the current hot spot area.
5. The method for passenger flow guidance planning based on multiple areas of a business district according to claim 4 is characterized in that: Through the preset passenger flow prediction model, the passenger flow change data of the numbered hot spot area and the current passenger flow range data of the current hot spot area are subjected to heterogeneous data fusion prediction to obtain the predicted passenger flow range data of the current hot spot area, specifically including: Based on the tourist behavior route characteristics, generating a tourist behavior route characteristic matrix; By means of the passenger flow prediction model, the passenger flow change characteristics of the passenger flow change data in the numbered hot spot area are extracted; and the current passenger flow range characteristics of the current passenger flow range data in the current hot spot area are extracted; wherein the passenger flow prediction model can perform data feature extraction processing; Through the tourist quantity feature matrix and based on the time scale feature, multi-layer dilated causal convolution processing is performed on the passenger flow change feature in the numbered hotspot area regarding the tourist transfer change trend to obtain first heterogeneous data; wherein the number of dilated causal convolution kernels used in the first heterogeneous data is half the number of layers; According to the tourist behavior route feature matrix and based on the time scale feature, the current passenger flow range feature in the current hot spot area is subjected to data point capture processing under the feature dependency relationship to obtain second heterogeneous data; wherein, the second heterogeneous data is subjected to dilated causal convolution to capture the dependency relationship between the data points with a long distance; The dependency relationship between the first heterogeneous data and the second heterogeneous data is fused through a preset single hidden layer fully connected network, and based on multiple iterations of the weight parameters in the passenger flow prediction model, the fused heterogeneous data are iteratively predicted to obtain the predicted passenger flow range data for the current hot spot area.
6. The method for passenger flow guidance planning based on multiple areas of a business district according to claim 1 is characterized in that: According to the current passenger flow range data and the predicted passenger flow range data, the correlation calculation of the passenger flow orientation correlation strength is performed on the route directivities between the hot spots, and the passenger flow orientation change relationship between the hot spots in the current time period is obtained, specifically including: Obtain the current passenger flow range data and the predicted passenger flow range data of each hot spot area in the current time period; Based on the tourist behavior route characteristics corresponding to the current hot spot area, the current passenger flow range data and the predicted passenger flow range data of the remaining hot spots are used to calculate the passenger flow change range, and the passenger flow guidance data of each of the remaining hot spots is obtained; wherein the passenger flow guidance data includes increase guidance data and decrease guidance data; The passenger flow guidance data of each of the remaining hot spots are sorted numerically to obtain a guidance trend vector of each of the remaining hot spots; wherein the guidance trend vector includes: a positive vector corresponding to the increase guidance data and a negative vector corresponding to the decrease guidance data; According to the absolute value of the guiding trend vector, the correlation calculation of the passenger flow guiding association strength is performed on the remaining hot spots, and the correlation level between each of the remaining hot spots and the current hot spots is determined; wherein, the greater the absolute value of the guiding trend vector, the higher the corresponding correlation level; wherein, the correlation level includes: a positive correlation level corresponding to the positive vector and a negative correlation level corresponding to the negative vector; The passenger flow orientation change relationship between the hot spots in the current time period is generated through the correlation levels between the remaining hot spots and the current hot spots.
7. The method for passenger flow guidance planning based on multiple areas of a business district according to claim 1 is characterized in that: Based on the grid map of the business district corresponding to the target business district, searching and processing the best guidance path under the passenger flow guidance change relationship between the first position information of the current hot spot area and the second position information of the target hot spot area is performed to determine the best guidance recommended route, specifically including: Based on the grid map of the business district corresponding to the target business district, and through the target area path generation algorithm, the first position information of the current hot spot area and the second position information of several target hot spot areas are tracked and positioned to obtain a linked list array containing all paths; wherein the target hot spot area is any one of the remaining hot spot areas; If the passenger flow guidance change relationship between the target hot spot area and the current hot spot area is a passenger flow increase change relationship, then based on the positive correlation level between the first location information and the second location information, the relevant paths in the linked list array are searched and sorted, and the guidance path corresponding to the lowest positive correlation level is determined as the optimal guidance recommended route through the search connection of the line segment sequence with the minimum Manhattan distance between the two end points; If the passenger flow guidance change relationship between the target hot spot area and the current hot spot area is a passenger flow reduction change relationship, then based on the negative correlation level between the first location information and the second location information, the relevant paths in the linked list array are searched and sorted, and the guidance path corresponding to the highest negative correlation level is determined as the optimal guidance recommended route; Wherein, the positive correlation level and the negative correlation level are both included in the passenger flow guidance change relationship.
8. The method for passenger flow guidance planning based on multiple areas of a business district according to claim 1 is characterized in that: The best guidance recommended route is displayed on a large-screen visualization screen, and the best guidance recommended route is connected to the terminal device of the tourist in real time through the large-screen visualization screen, specifically including: Install the visualization large screen in the core hot spot area of the hot spot area, and visualize the best recommended route for guidance recommended by the current hot spot area to achieve guidance recommendation for tourists in the current hot spot area; Based on the scannable QR code in the visualization large screen, the best guided recommended routes recommended by the other hot spots are sent to the terminal device of the tourist to obtain the multi-domain best guided recommended routes; and the walking time from the current hot spot area to the best guided recommended routes corresponding to the other hot spots is estimated and calculated to obtain the estimated walking time; Generate the global best guidance recommended route of the target business district through the estimated walking time and the multi-domain best guidance recommended route, and send it to the terminal device; Based on the push processing of the best guidance recommended route, the multi-domain best guidance recommended route and the global best guidance recommended route, the travel guidance recommendation for the tourists in the target business district is completed.
9. A passenger flow guidance planning device based on multiple areas of a business district, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, so that the at least one processor can execute the passenger flow guidance planning method based on multiple areas of a business district according to any one of claims 1-8.
Citation Information
Patent Citations
System and method for evacuating people flow in shopping mall
CN106897798A
Method, device and system for predicting degree of passenger flow congestion in scenic spots
CN109034460A
A hub passenger flow space-time distribution prediction modeling method based on multi-source data fusion
CN109543883A
Passenger flow distribution prediction method and device based on multi-view passenger flow change trend modeling
CN114169632A
Rail transit passenger flow volume prediction system and method based on digital twinning
CN117131999A
