Old-age service institution site selection planning system based on multi-source data and machine learning

Abstract

The invention relates to the technical field of pension services, and discloses a pension service institution site selection planning system based on multi-source data and machine learning. Comprising a planning period calculation module, a site selection area expansion module, a multi-source data acquisition module, a candidate area screening module and a site selection grade division module, and is used for formulating an expansion standard, expanding site selection points into site selection areas, acquiring multi-source site selection data, predicting site selection attributes through a machine learning model, screening out candidate areas and classifying the site selection grades. The site selection grades of the candidate areas are divided; according to the method, the data can be comprehensively acquired from two dimensions of a past time period and a future time period, the acquisition effect of multi-source data on different time lines is realized, and in combination with intelligent prediction of a machine learning model, the site selection attribute of a site selection area can be accurately and quickly predicted, so that the site selection efficiency is improved. The problem that the accuracy of site selection planning is low when evaluation is carried out through data of a single dimension and a single time period is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of elderly care services, and more specifically, to a site selection planning system for elderly care service institutions based on multi-source data and machine learning. Background Art

[0002] Against the backdrop of an increasingly aging population, senior care real estate projects have become a hot research topic and practice area that has attracted much attention. With the accelerating pace of urbanization, how to accurately select reasonable and accurate addresses in cities to build senior care service institutions to meet the housing needs of the elderly has become an issue that needs to be urgently addressed in urban construction.

[0003] The patent application with reference publication number CN119443618A discloses a method for supplementing the site selection of elderly care service institutions based on multi-source data and machine learning, including real-time acquisition of elderly care service multi-source data of the area to be sited, building an optimal elderly care service institution site selection machine learning model, determining the prediction accuracy probability based on the supplementary initial site selection results of the elderly care service institutions and the multi-source data of elderly care services, and dividing the supplementary initial site selection results of the elderly care service institutions based on the elderly care service demand to be sited to obtain priority site selection areas and ordinary site selection areas. It selects sites for the multi-source data of elderly care services through machine learning, reduces labor costs and dependence on manual analysis, and improves the degree of automation of site selection, can dynamically respond to changing needs, ensure that the coverage of elderly care facilities and services matches actual needs, and adapt to dynamic changes in urban and population structures; When planning the site selection of existing elderly care service institutions, real-time multi-source data of the area to be selected is collected and combined with artificial intelligence technology to realize the intelligent site selection operation of the elderly care service institutions. For example, in the above-mentioned patent application, it obtains multi-source data of elderly care services in the area to be selected in real time and combines machine learning to select the site for the multi-source data of elderly care services. Although it can achieve the site selection effect of the elderly care service institution, it only collects real-time multi-source data, which makes it impossible for the elderly care service institution to comprehensively collect data on different time lines of past time periods and future time periods when predicting the site selection, resulting in the limitation of the source channel of multi-source data on the time line, and the inability to achieve the comprehensive collection effect of multi-source data in different time dimensions, which leads to the low accuracy of subsequent site selection prediction.

[0004] In view of this, the present invention proposes a site selection planning system for elderly care service institutions based on multi-source data and machine learning to solve the above problems. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: a site selection and planning system for elderly care service institutions based on multi-source data and machine learning, comprising: The planning cycle calculation module marks the site selection points in the area to be selected according to the point marking criteria, collects the building development cycle and road renovation frequency of the site selection points, calculates the planning duration of the site selection points, and determines the planning cycle of the area to be selected. The point marking criteria are: the intersection of the roads adjacent to the area to be developed is recorded as the site selection point; The site selection area expansion module extracts the planned coverage area, road penetration rate and one-way length peak of the site selection point within the planning period, formulates the expansion standard for radial expansion of the site selection point, and expands the site selection point into the site selection area; The multi-source data collection module records the period from the first development time of the target building in the site selection area to the current time as the historical period, and collects multi-source site selection data of the site selection area in the historical period and planning cycle. The multi-source site selection data includes the floating value of the elderly density, regional traffic activity, accessibility of medical resources, urban greening rate and functional configuration; The candidate area screening module predicts the site selection attributes corresponding to the multi-source site selection data through a machine learning model, and screens candidate areas from the site selection areas; The site selection grade classification module matches the elderly care service data of the elderly care service institutions with the multi-source site selection data of the candidate areas, identifies the matching data, and classifies the site selection grades of the candidate areas into excellent and good.

[0006] Furthermore, the steps for marking the site selection points are: Mark all blank areas in the area to be selected through an electronic map, and record the blank areas with development attributes as to-be-developed areas as to-be-developed areas; Mark all public roads in the area to be selected one by one, and record the areas to be developed that are adjacent to public roads in the east, south, west and north directions as valid areas; The road attributes of the public roads adjacent to the valid area are identified, the public roads with the road attributes of secondary roads are recorded as valid roads, and the road intersection points of any two valid roads are recorded as site selection points, so as to obtain A site selection points.

[0007] Furthermore, the steps for collecting building development cycle and road renovation frequency are as follows: Taking the A site selection point as the extension starting point, extend the effective road adjacent to the A site selection point in both directions to the intersection of the second traffic light to obtain the extended road, and record the completed buildings located on both sides of the extended road as the target buildings; The development and construction time of all target buildings is queried through the city management system, and the maximum development and construction time and the minimum development and construction time are added and averaged to obtain A building development cycles; The number of renovations of effective roads corresponding to A site points within the construction development cycle is counted to obtain A renovation values, and after comparing the A renovation values ​​with the corresponding construction development cycle, the renovation frequency of A roads is obtained.

[0008] Furthermore, the planning cycle is determined by: Combine the building development cycle of site A with the road renovation frequency to calculate the planning duration of site A; The calculation formula for planning duration is: ; In the formula, For the The planning time for each site selection point, =1,2,...,A, For the The construction development cycle of each site selection point, For the Frequency of road renovation at each site; With A planning durations as the horizontal coordinate and the number of planning durations as the vertical coordinate, a normal distribution diagram of A planning durations is constructed, and a normal curve is drawn; A lower limit value of the number is preset, a horizontal line where the lower limit value of the number is located is drawn horizontally in the normal distribution graph, and the part of the normal curve located above the horizontal line is recorded as a valid curve; Record the planning duration in the effective curve as the effective duration, and obtain The effective time will The effective durations are accumulated and averaged to obtain the planning period.

[0009] Furthermore, the collection steps of road penetration rate are as follows: The planning and design drawings corresponding to the A site selection points are queried in sequence through the urban management system, and the public roads that are the main roads and branch roads in the planning and design drawings are respectively identified and recorded as planned roads; Divide the area to be selected into grids corresponding to the planned coverage area, obtain C regional grids, count the number of planned roads in the C regional grids one by one, and compare the number of planned roads with the area of ​​the regional grids to obtain C sub-population rates; After eliminating the maximum and minimum sub-population rates, the remaining sub-population rates are accumulated and averaged to obtain A road penetration rates.

[0010] Further, the steps for expanding the site selection area are: The planned coverage area, road penetration rate and one-way length peak value of A site selection points are assigned different weight factors and compared to obtain A extension lengths; The extended length is calculated as: ; In the formula, For the The extended length of the site selection point, For the The planned coverage area of ​​each site selection point is For the The road penetration rate of each site selection point, For the The peak one-way length of the site selection point is , , All are weight factors greater than 0; After arranging the A extension lengths from small to large, the third extension length is recorded as the extension standard; Take A site selection points as the expansion base points and the expansion standard as the expansion radius of the site selection points, radially expand the A site selection points into expansion circles, and record the area within the expansion circle as the site selection area, so as to obtain A site selection areas.

[0011] Furthermore, the steps for collecting regional traffic activity are as follows: Taking the site selection point as the midpoint of the division, the A site selection area is equally divided into E sub-areas with fan-shaped structures; During the historical period, the public transportation management system is used to query the number of active public transportation vehicles on weekdays, weekends and holidays in E sub-areas at a time point before the end of the period, and obtain E first active values, E second active values ​​and E third active values; Add the E first activity values, the E second activity values, and the E third activity values ​​and calculate the average, and compare the E averages with the area of ​​the sub-region to obtain E sub-activity levels; After removing the minimum value of the sub-activity, the minimum value of the remaining sub-activity is recorded as the regional traffic activity, and A regional traffic activities are obtained.

[0012] Furthermore, the steps for collecting the accessibility of medical resources are as follows: In A selection areas, the moving distance values ​​of all types of public transportation within a unit time are queried respectively, and the maximum moving distance value is recorded as the standard length; With two standard lengths as the expansion radius and the site selection point as the expansion origin, expand A expansion circles in the A site selection areas and mark the expansion circles at equal distances. non-adjacent expansion points; Respectively The expansion point is the scanning center, and a standard length is the scanning radius. Scan area; Statistics The number of medical service institutions in the scan area is obtained Medical value, The medical quantity values ​​are accumulated and averaged, and compared with the standard length to obtain A medical resource accessibility rates.

[0013] Furthermore, the site selection attributes include available sites and unavailable sites; The steps for screening candidate regions are: When the output of the machine learning model is 1, the location attribute of the site selection area is unavailable for site selection, and the site selection area is not recorded as a candidate area; When the output of the machine learning model is 0, the location attribute of the site selection area is available site selection, and the site selection area of ​​available site selection is recorded as a candidate area, and S candidate areas are obtained.

[0014] Furthermore, the elderly care service data include the elderly density demand value, regional transportation demand, medical resource demand rate, urban greening demand rate and functional configuration demand; The matching data is identified by: When the elderly density demand value is less than or equal to the elderly density floating value, the elderly density floating value is recorded as matching data; When the regional traffic demand is less than or equal to the regional traffic activity, the regional traffic activity is recorded as matching data; When the medical resource demand rate is less than or equal to the medical resource availability rate, the medical resource availability rate is recorded as matching data; When the urban greening demand rate is less than or equal to the urban greening rate, the urban greening rate is recorded as matching data; When the function configuration requirement degree is less than or equal to the function configuration degree, the function configuration degree is recorded as matching data; The steps for excellent and good division are: The number of matching data in the S candidate areas is counted in sequence. When the number of matching data is 5, the site selection level of the candidate area is recorded as excellent. When the number of matching data is 4 or 3, the site selection level of the candidate area is recorded as good.

[0015] The technical effects and advantages of the elderly care service institution site selection planning system based on multi-source data and machine learning of the present invention are as follows: The present invention can calculate the planning cycle of the area to be sited by collecting the building development cycle and road renovation frequency of the site selection point, provide a limiting function of the data collection time for the subsequent expansion operation of the site selection area, and combine the combined calculation of the planning coverage area, the road penetration rate and the one-way length peak value to provide a basis for the calculation of the expansion standard of the site selection point, so that the site selection point of a single point can be expanded into a site selection area with a specific area, avoiding the problem of too small area in the site selection planning operation of a single point. At the same time, by collecting multi-source site selection data of historical time periods and planning cycles, the data affecting the site selection area as the site selection planning object of the elderly care service agency can be comprehensively collected from the two dimensions of the past time period and the future time period, realizing the collection effect of multi-source data on different time lines, and combined with the intelligent prediction of the machine learning model, the site selection attributes of the site selection area can be accurately and quickly predicted, thereby providing real and accurate result support for the subsequent site selection planning of the elderly care service agency, avoiding the problem of low accuracy of site selection planning when evaluating through data of a single dimension and a single time period, and effectively improving the accuracy of site selection planning of the elderly care service agency. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a site selection and planning system for elderly care service institutions based on multi-source data and machine learning provided in Embodiment 1 of the present invention; Figure 2 A flowchart of a method for site selection and planning of elderly care service institutions based on multi-source data and machine learning provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Example 1: Please refer to Figure 1 As shown, the elderly care service institution site selection and planning system based on multi-source data and machine learning described in this embodiment is applied to the site selection and planning platform, including: The planning cycle calculation module marks the site selection points in the area to be selected according to the point marking criteria, collects the planning parameters of the site selection points, calculates the planning duration of the site selection points, and determines the planning cycle of the area to be selected; The area to be sited refers to a larger geographical area within the urban administrative area where the site selection planning of elderly care service institutions is required, and serves as the source location basis for all data in the subsequent site selection planning of elderly care service institutions. The site selection point refers to the initial central location in the area to be sited that can be used as the site for elderly care service institutions, ensuring that the specific location of the final site can be expanded and optimized based on the site selection point.

[0019] Since the site selection point is the initial central position of the site selection of the elderly care service institution, the number of site selection points is usually more than one. When marking the locations of multiple site selection points, they need to be marked according to specific criteria. Therefore, when marking the site selection points in the area to be sited, it is necessary to do so under the constraints of the point marking criteria. The point marking principle is: the intersection of roads adjacent to the area to be developed is recorded as the site selection point; this ensures that the adjacent locations of all site selection points have areas that can meet the site selection, planning and development of elderly care service institutions, thereby meeting the basic needs of subsequent site selection; The steps for marking the site selection points are: Mark all blank areas in the area to be selected through electronic maps, and record the blank areas with development attributes of "to be developed" as areas to be developed; blank areas refer to areas without buildings, and development attributes are used to represent the specific attributes of blank areas, including "to be developed" and "developed". "to be developed" means that there are no complete and delivered buildings on the blank areas, and "developed" means that there are complete and delivered buildings on the blank areas; Mark all public roads in the area to be selected one by one, and record the areas to be developed that are adjacent to public roads in the east, south, west and north directions as valid areas; adjacent locations refer to locations where there are no other areas between public roads and the areas to be developed, so that the two areas in adjacent locations are closely connected in terms of spatial relationship; Identify the road attributes of the public roads adjacent to the valid area, record the public roads with the road attribute of secondary roads as valid roads, and record the intersection of any two valid roads as the site selection point to obtain A site selection points. Road attributes are used to represent the level of public roads, and road attributes include primary roads and secondary roads.

[0020] After marking the site selection points in the area to be selected, it is necessary to collect the planning parameters of the site selection points. The planning parameters at this time can represent the municipal planning strategy and other information of the site selection points in a diversified manner, and evaluate and calculate the planning change time of the site selection points in the municipal planning through the planning parameters, and represent it through the planning duration; Planning parameters include building development cycle and road renovation frequency; building development cycle refers to the duration of the development and construction of buildings at the location corresponding to the site selection, and road renovation frequency refers to the number of renovations per unit time of the public road at the location corresponding to the site selection within the building development cycle; The steps for collecting building development cycle and road renovation frequency are as follows: Taking the A site selection point as the extension starting point, extend the effective road adjacent to the A site selection point in both directions to the intersection of the second traffic light to obtain the extended road, and record the completed buildings located on both sides of the extended road as the target buildings; The development and construction time of all target buildings is queried through the city management system, and the maximum development and construction time and the minimum development and construction time are added and averaged to obtain A building development cycles; The formula for calculating the building development cycle is: ; In the formula, For the The construction development cycle of each site selection point, =1,2,...,A, For the The maximum development and construction time of each site selection point is For the The minimum development and construction time of each site selection point; Count the number of renovations of effective roads corresponding to A site points within the building development cycle to obtain A renovation values, and compare the A renovation values ​​with the corresponding building development cycle to obtain the renovation frequency of A roads; The expression for road renovation frequency is: ; In the formula, For the The frequency of road renovation at each site, For the The renovation value of each site selection point.

[0021] After the planning parameters of the site selection point are collected, the planning time of the site selection point needs to be calculated based on the planning parameters of the site selection point, and the overall planning cycle of the site selection area is determined according to the planning time of the final site selection point, so that the planning cycle can be a comprehensive representation of the length of time for municipal construction in the site selection area; The defining steps of the planning cycle are: Combine the building development cycle of site A with the road renovation frequency to calculate the planning duration of site A; The calculation formula for planning duration is: ; In the formula, For the The planning time for each site selection point; With A planning durations as the horizontal coordinates and the number as the vertical coordinates, a normal distribution diagram of A planning durations is constructed, and a normal curve is drawn; by drawing a normal distribution diagram, the distribution of different numbers and planning durations can be concisely represented, which facilitates the rapid identification of effective durations; A lower limit value is preset, a horizontal line is drawn horizontally in the normal distribution graph, and the part of the normal curve above the horizontal line is recorded as the effective curve; the lower limit value is used to represent the minimum value of the number of effective durations corresponding to the effective curve, which can ensure the accuracy of the subsequent effective duration calculation; Record the planning duration in the effective curve as the effective duration, and obtain The effective time will The effective durations are accumulated and averaged to obtain the planning period; The calculation formula for the planning period is: ; In the formula, For the planning cycle, For the Valid duration.

[0022] The site selection area expansion module extracts the planning characteristics of the site selection point within the planning period, formulates the expansion standard for radially expanding the site selection point based on the planning characteristics, and expands the site selection point into the site selection area; Planning features are specific information about the municipal construction at the location of the site selection point in the future, which can accurately represent the impact range of the municipal construction at the site selection point in the future and serve as the data basis for subsequent expansion operations at the location of the site selection point. Planning characteristics include planning coverage area, road penetration rate and one-way length peak; planning coverage area refers to the coverage area of ​​municipal construction at the corresponding location of the planning point within the planning period, which can represent the impact range of municipal construction at the site selection point within the planning period; When collecting the planned coverage area, it is necessary to take the current time as the query start time, query the planning plan for the area to be selected in the next planning cycle through the urban management system, record the planning plan of the target building containing A site selection points as the target plan, and count the total value of the planned area in all target plans to obtain A planned coverage areas.

[0023] The road penetration rate refers to the degree of penetration of public road construction at the location corresponding to the planning point within the planning period, which can be used to indicate the effect of public road construction in municipal construction at the site selection point within the planning period; The steps for collecting road penetration rate are as follows: The planning and design drawings corresponding to the A site selection points are queried in sequence through the urban management system, and the public roads that are the main roads and branch roads in the planning and design drawings are respectively identified and recorded as planned roads; The area to be selected for site selection corresponding to the planned coverage area is divided into grids to obtain C regional grids, and the number of planned roads in the C regional grids is counted one by one. After comparing the number of planned roads with the area of ​​the regional grids, C sub-population rates are obtained. The grid division method can ensure that the area to be selected for site selection corresponding to the planned coverage area can be divided evenly and orderly, and at the same time, it can ensure that the calculation of the sub-population rate of each regional grid remains relatively consistent. The formula for calculating the sub-population rate is: ; In the formula, For the The first site selection point The sub-population rate of the regional grid, =1,2,...,C, For the The first site selection point The number of planned roads in a regional grid, For the The first site selection point The area of ​​the regional grid; After eliminating the maximum and minimum sub-population rates, the remaining sub-population rates are accumulated and averaged to obtain A road penetration rates; The formula for calculating road penetration rate is: ; In the formula, For the The road penetration rate of each site selection point, For the The first site selection point The sub-population rate of the grid in each region.

[0024] The one-way length peak refers to the maximum length of the main road and the secondary road in a certain direction at the corresponding position of the planning point within the planning period. When collecting the one-way length peak, the length values ​​of the main roads and secondary roads corresponding to A site selection points in the four directions of east, south, west and north are measured one by one through a scale, and the maximum value is taken to obtain A one-way length peaks.

[0025] After obtaining the planning features of A site selection points, the planning features can be extended and analyzed. The relevant features that can be used as extension standards can be analyzed from the planning features to formulate extension standards. The extension standards can be used as the length limit for radially extending the site selection points from points to surfaces, that is, the planning points can be expanded into site selection areas with specific areas. When the site selection point is expanded to a site selection area with a larger area, the site selection area at this time needs to have a relatively closed boundary and a specific range. At this time, each site selection area can be used as a representation of the subsequent construction and development location of the elderly care service institution; The steps for expanding the site selection area are: The planned coverage area, road penetration rate and one-way length peak value of A site selection points are assigned different weight factors and compared to obtain A extension lengths; The extended length is calculated as: ; In the formula, For the The extended length of the site selection point, For the The planned coverage area of ​​each site selection point is For the The peak one-way length of the site selection point is , , are all weight factors greater than 0, and ; After arranging the A extension lengths in order from small to large, the third extension length is recorded as the extension standard; Take A site selection points as the expansion base points and the expansion standard as the expansion radius of the site selection points, radially expand the A site selection points into expansion circles, and record the area within the expansion circle as the site selection area, so as to obtain A site selection areas.

[0026] The multi-source data collection module determines the historical period of the site selection area and collects multi-source site selection data of the site selection area within the historical period and planning period; The historical period refers to the period of time for collecting historical data that can affect whether the site selection area can be used as the site selection and construction of elderly care service institutions, which can provide a time limit for the diversified data of the site selection area in the past time period; Since the historical period is used to limit the time for collecting multi-source and diversified data in the site selection area, when determining the historical period, it needs to start from the moment when multi-source data that meets the requirements appears and end at the current moment; In summary, the historical period is: the first development time of the target building in the A site selection area is taken as the starting point of the period, the current time is taken as the end point of the period, and the period between the starting point and the end point of the period is recorded as the historical period.

[0027] Multi-source site selection data refers to diversified data that can be used to select and build elderly care service institutions in the site selection area in the historical period and planning period. It can be used as data support to determine whether the site selection area meets the site selection and construction needs of elderly care service institutions. It can also provide data representation of the actual multi-source situation of the site selection area in the past and future time periods; Multi-source site selection data include floating values ​​of elderly density, regional traffic activity, accessibility of medical resources, urban greening rate and functional configuration; The floating value of elderly density refers to the growth rate of the number of elderly people per unit area in the site selection area during the historical period, thus providing data support on the density of elderly population for whether the site selection area can be used as the site selection for elderly care service institutions. The larger the floating value of elderly density, the greater the probability that the site selection area will be used as the site selection for elderly care service institutions. When obtaining the floating value of the elderly density, the elderly density values ​​of A selected areas at the start and end of the historical period are queried through the demographic database, and the difference is calculated to obtain the value.

[0028] Regional traffic activity refers to the degree of development and construction of public transportation in the site selection area during the historical period. The greater the regional traffic activity, the higher the degree of development and construction of public transportation in the site selection area, and the greater the probability that the site selection area will be selected as the site for elderly care service institutions; The steps for collecting regional traffic activity are as follows: Taking the site selection point as the midpoint of the segmentation, the A site selection area is equally divided into E sub-areas with a fan-shaped structure. Through the fan-shaped segmentation method, the site selection area can be accurately and evenly segmented, and the consistency of the regional morphology of each sub-area can be maintained, avoiding the random error phenomenon caused by the inconsistent regional morphology when calculating the sub-activity. During the historical period, the public transportation management system is used to query the number of active public transportation vehicles on weekdays, weekends, and holidays in E sub-areas at a time point before the end of the period, and obtain E first active values, E second active values, and E third active values; Add the E first activity values, the E second activity values, and the E third activity values ​​and calculate the average, and compare the E averages with the area of ​​the sub-region to obtain E sub-activity levels; The calculation formula of child activity is: ; In the formula, For the The first The activity of each sub-region, =1,2,...,E, For the The first The first activity value of the sub-region, For the The first The second activity value of the sub-region, For the The first The third activity value of the sub-region, For the The area of ​​the sub-region of the site selection area; After removing the minimum value of the sub-activity, the minimum value of the remaining sub-activity is recorded as the regional traffic activity, and A regional traffic activities are obtained.

[0029] The accessibility rate of medical resources refers to the feasibility of the site selection area to reach the medical service institution within a unit time in the historical period. The greater the accessibility rate of medical resources, the better the performance of the site selection area in medical services, and the greater the probability that the site selection area will be selected as the site of the elderly care service institution; The steps for collecting the accessibility of medical resources are as follows: In A site selection areas, the moving distance values ​​of all types of public transportation within a unit time are queried, and the maximum moving distance value is recorded as the standard length; public transportation means include but are not limited to buses, subways, SkyRail, taxis, ferries, etc.; With two standard lengths as the expansion radius and the site selection point as the expansion origin, expand A expansion circles in the A site selection areas and mark the expansion circles at equal distances. non-adjacent expansion points; by drawing an expansion circle, a distribution path of the scanning starting points covering a certain area can be drawn in the selected area, providing a circular uniform distribution basis for the marking of the subsequent scanning center; Respectively The expansion point is the scanning center, and a standard length is the scanning radius. Scanning area; through the circular scanning method, the accessibility of medical resources in the selected area can be identified and calculated in multiple points and circumferential directions, avoiding the limitations of single-direction and single-point identification and calculation; Statistics The number of medical service institutions in the scan area is obtained Medical value, The medical quantity values ​​are accumulated and averaged, and compared with the standard length to obtain A medical resource accessibility rates; The expression of medical resource accessibility is: ; In the formula, For the The accessibility of medical resources in the selected area, For the The first Medical value, For the The standard length of a site selection area.

[0030] The urban greening rate refers to the maximum coverage rate that can be achieved by municipal greening construction in the site selection area during the planning period, which can be used to represent the greening environment of the site selection area. The higher the urban greening rate, the greater the probability that the site selection area will be selected as the site for a senior care service institution.

[0031] Functional configuration degree refers to the completeness rate of supporting living and entertainment facilities in the site selection area during the planning period, which can be used to represent the functional perfection performance of the site selection area. The greater the functional configuration degree, the greater the probability that the site selection area will be selected as a location for elderly care service institutions. The urban greening rate and functional configuration degree are obtained by querying the greening construction rate and supporting perfection rate of site selection area A at the last time point in the planning period through the urban management system.

[0032] The candidate area screening module predicts the site selection attributes of the site selection area corresponding to the multi-source site selection data through a machine learning model, and screens out candidate areas from the site selection area; After collecting multi-source site selection data of the site selection area, the pre-trained machine learning model can be used to predict the prediction results corresponding to the multi-source site selection data, and the prediction results can be represented by site selection attributes, so that the site selection attributes can be used as the output of the machine learning model and as the result of whether the site selection area can be used as the site selection of the elderly care service institution; Site selection attributes include available site selection and unavailable site selection; available site selection means that the site selection area can be used as the site selection for the development and construction of an elderly care service institution, and unavailable site selection means that the site selection area cannot be used as the site selection for the development and construction of an elderly care service institution; site selection attributes are obtained by collecting a large number of historical site selection attributes corresponding to different elderly density floating values, regional traffic activity, medical resource accessibility, urban greening rate and functional configuration.

[0033] The steps for training a machine learning model are: Pre-collecting multiple sets of multi-source site selection data with site selection attributes of available sites and unavailable sites; Mark each group of multi-source site selection data as a training feature, annotate the site selection attributes of each group of training features, the annotations include available sites and unavailable sites, convert the available sites and unavailable sites into digital annotations, for example, convert the available sites into 0, and convert the unavailable sites into 1, divide the annotated training features into a training set and a test set, use 70% of the training features as the training set, and use 30% of the training features as the test set; Use the training set to train the machine learning model, use the test set to test the machine learning model, preset the error threshold, and output the machine learning model when the mean of the prediction errors of all training features in the test set is less than the error threshold.

[0034] Exemplarily, the machine learning model adopts any one of a support vector machine model or a random forest model, and the preset error threshold is pre-set according to the accuracy actually required by the machine learning model.

[0035] Candidate areas refer to areas that can be used as locations for the development and construction of elderly care service institutions. Since candidate areas are not specific areas, the number of candidate areas is not unique, so candidate areas can be used as objects for subsequent optimization analysis of the location of elderly care service institutions. The steps for screening candidate regions are: The collected multi-source site selection data of A site selection areas are input into the machine learning model. When the output of the machine learning model is 1, the site selection attribute of the site selection area is unavailable site selection, and the site selection area is not recorded as a candidate area. When the output of the machine learning model is 0, the location attribute of the site selection area is available site selection, and the site selection area with the location attribute of available site selection is recorded as a candidate area, and S candidate areas are obtained.

[0036] The site selection grading module receives the elderly care service data from the elderly care service institutions, matches the elderly care service data with the multi-source site selection data of the candidate areas, identifies the matching data, and grading the candidate areas; Elderly care service data refers to the target values ​​of various multi-source data formulated and required when developing and building elderly care service institutions, and serves as the reference data for further sorting and screening of candidate areas, thereby providing a data basis for the sorting of candidate areas; Elderly care service data include elderly density demand value, regional traffic demand degree, medical resource demand rate, urban greening demand rate and functional configuration demand degree; specifically, elderly density demand value, regional traffic demand degree, medical resource demand rate, urban greening demand rate and functional configuration demand degree are used to represent the minimum values ​​of elderly population growth, traffic activity, medical resources, greening construction and functional facilities in the site selection, development and construction of elderly care service institutions; After obtaining the elderly care service data, the elderly care service data can be compared with the multi-source site selection data one by one, and the multi-source site selection data that meets the needs can be recorded as matching data, and the candidate areas can be arranged and divided according to the amount of matching data in the candidate areas; The matching data is identified by: The elderly density demand value, regional traffic demand, medical resource demand rate, urban greening demand rate and functional configuration demand are compared with the elderly density floating value, regional traffic activity, medical resource accessibility rate, urban greening rate and functional configuration degree in turn; When the elderly density demand value is less than or equal to the elderly density floating value, it means that the elderly population growth in the candidate area meets the site selection requirements, and the elderly density floating value is recorded as matching data; When the regional traffic demand is less than or equal to the regional traffic activity, it means that the traffic activity of the candidate area meets the site selection requirements, and the regional traffic activity is recorded as matching data; When the medical resource demand rate is less than or equal to the medical resource accessibility rate, it means that the medical resources in the candidate area meet the site selection requirements, and the medical resource accessibility rate is recorded as matching data; When the urban greening demand rate is less than or equal to the urban greening rate, it means that the greening construction of the candidate area meets the site selection requirements, and the urban greening rate is recorded as matching data; When the functional configuration requirement degree is less than or equal to the functional configuration degree, it means that the functional facilities of the candidate area meet the site selection requirements, and the functional configuration degree is recorded as matching data.

[0037] When arranging and ranking candidate areas, it is necessary to count the number of matching data in each candidate area, so as to achieve accurate classification of site selection levels; The site selection grade is used to indicate the suitability of the candidate area as the location of the elderly care service institution. The site selection grades include excellent and good. Among them, the suitability corresponding to excellent and good is from high to low. The steps for excellent and good division are: The number of matching data in the S candidate areas is counted in sequence. When the number of matching data is 5, it means that the candidate area is highly suitable for the location of the elderly care service institution, and the location level of the candidate area is recorded as excellent; When the number of matching data is 4 or 3, it indicates that the candidate area has a medium suitability as a location for a senior care service institution, and the location grade of the candidate area is recorded as good.

[0038] It should be noted that since the candidate area is the area that can be used as the site selection for elderly care service institutions after being predicted by the machine learning model, the overall number of multi-source site selection data in the candidate area can exceed half of the elderly care service data, that is: the number of matching data is 3, 4 or 5, ensuring that the matching data in the candidate area contains at least 3 to meet the basic requirements of the model prediction of the candidate area.

[0039] In this embodiment, by collecting the construction development cycle and road renovation frequency of the site selection point, the planning cycle of the area to be sited can be calculated, and the data collection time limitation function is provided for the subsequent expansion operation of the site selection area. In combination with the combined calculation of the planned coverage area, the road penetration rate and the one-way length peak value, a basis can be provided for the calculation of the expansion standard of the site selection point, so that the single-point site selection point can be expanded into a site selection area with a specific area, avoiding the problem of too small area in the site selection planning operation of the single point. At the same time, by collecting multi-source site selection data of historical time periods and planning cycles, the data affecting the site selection area as the site selection planning object of the elderly care service institution can be comprehensively collected from the two dimensions of the past time period and the future time period, realizing the collection effect of multi-source data on different time lines, and combined with the intelligent prediction of the machine learning model, the site selection attributes of the site selection area can be accurately and quickly predicted, thereby providing real and accurate result support for the subsequent site selection planning of the elderly care service institution, avoiding the problem of low accuracy of site selection planning when evaluating through data of a single dimension and a single time period, and effectively improving the accuracy of site selection planning of the elderly care service institution.

[0040] Example 2: Please refer to Figure 2 As shown, the part not described in detail in this embodiment is described in the first embodiment, and a method for site selection and planning of a nursing service institution based on multi-source data and machine learning is provided, which is applied to a site selection and planning platform and is implemented by a site selection and planning system for a nursing service institution based on multi-source data and machine learning, including: S1: According to the point marking criteria, the site selection points are marked in the area to be selected, the construction development cycle and road renovation frequency of the site selection points are collected, the planning duration of the site selection points is calculated, and the planning cycle of the area to be selected is determined. The point marking criteria are: the intersection of the roads adjacent to the area to be developed is recorded as the site selection point; S2: Extract the planned coverage area, road penetration rate and one-way length peak of the site selection point within the planning period, formulate the expansion standard for radial expansion of the site selection point, and expand the site selection point into the site selection area; S3: The period from the first development time of the target building in the site selection area to the current time is recorded as the historical period, and multi-source site selection data of the site selection area in the historical period and planning cycle are collected. The multi-source site selection data includes the floating value of the elderly density, regional traffic activity, accessibility of medical resources, urban greening rate and functional configuration; S4: Predict the site selection attributes corresponding to the multi-source site selection data through the machine learning model, and screen candidate areas from the site selection area; S5: Match the elderly care service data of the elderly care service institutions with the multi-source site selection data of the candidate areas, identify the matching data, and classify the site selection levels of the candidate areas into excellent and good.

[0041] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A site selection and planning system for elderly care service institutions based on multi-source data and machine learning, applied to the site selection and planning platform, characterized by: include: The planning cycle calculation module marks the site selection points in the area to be selected according to the point marking criteria, collects the building development cycle and road renovation frequency of the site selection points, calculates the planning duration of the site selection points, and determines the planning cycle of the area to be selected. The point marking criteria are: the intersection of the roads adjacent to the area to be developed is recorded as the site selection point; The site selection area expansion module extracts the planned coverage area, road penetration rate and one-way length peak of the site selection point within the planning period, formulates the expansion standard for radial expansion of the site selection point, and expands the site selection point into the site selection area; The multi-source data collection module records the period from the first development time of the target building in the site selection area to the current time as the historical period, and collects multi-source site selection data of the site selection area in the historical period and planning cycle. The multi-source site selection data includes the floating value of the elderly density, regional traffic activity, accessibility of medical resources, urban greening rate and functional configuration; The candidate area screening module predicts the site selection attributes corresponding to the multi-source site selection data through a machine learning model, and screens candidate areas from the site selection areas; The site selection grade classification module matches the elderly care service data of the elderly care service institutions with the multi-source site selection data of the candidate areas, identifies the matching data, and classifies the site selection grades of the candidate areas into excellent and good.

2. The elderly care service institution site selection planning system based on multi-source data and machine learning according to claim 1 is characterized in that: The steps for marking the site selection points are: Mark all blank areas in the area to be selected through an electronic map, and record the blank areas with development attributes as to-be-developed areas as to-be-developed areas; Mark all public roads in the area to be selected one by one, and record the areas to be developed that are adjacent to public roads in the east, south, west and north directions as valid areas; The road attributes of the public roads adjacent to the valid area are identified, the public roads with the road attributes of secondary roads are recorded as valid roads, and the road intersection points of any two valid roads are recorded as site selection points, so as to obtain A site selection points.

3. The elderly care service institution site selection planning system based on multi-source data and machine learning according to claim 2 is characterized in that: The steps for collecting building development cycle and road renovation frequency are as follows: Taking the A site selection point as the extension starting point, extend the effective road adjacent to the A site selection point in both directions to the intersection of the second traffic light to obtain the extended road, and record the completed buildings located on both sides of the extended road as the target buildings; The development and construction time of all target buildings is queried through the city management system, and the maximum development and construction time and the minimum development and construction time are added and averaged to obtain A building development cycles; The number of renovations of effective roads corresponding to A site points within the construction development cycle is counted to obtain A renovation values, and after comparing the A renovation values ​​with the corresponding construction development cycle, the renovation frequency of A roads is obtained.

4. The elderly care service institution site selection planning system based on multi-source data and machine learning according to claim 3 is characterized in that: The defining steps of the planning cycle are: Combine the building development cycle of site A with the road renovation frequency to calculate the planning duration of site A; The calculation formula for planning duration is: ； In the formula, For the The planning time for each site selection point, =1,2,...,A, For the The construction development cycle of each site selection point, For the Frequency of road renovation at each site; With A planning durations as the horizontal coordinate and the number of planning durations as the vertical coordinate, a normal distribution diagram of A planning durations is constructed, and a normal curve is drawn; A lower limit value of the number is preset, a horizontal line where the lower limit value of the number is located is drawn horizontally in the normal distribution graph, and the part of the normal curve located above the horizontal line is recorded as a valid curve; Record the planning duration in the effective curve as the effective duration, and obtain The effective time will The effective durations are accumulated and averaged to obtain the planning period.

5. The elderly care service institution site selection planning system based on multi-source data and machine learning according to claim 4 is characterized in that: The steps for collecting road penetration rate are as follows: The planning and design drawings corresponding to the A site selection points are queried in sequence through the urban management system, and the public roads that are the main roads and branch roads in the planning and design drawings are respectively identified and recorded as planned roads; Divide the area to be selected into grids corresponding to the planned coverage area, obtain C regional grids, count the number of planned roads in the C regional grids one by one, and compare the number of planned roads with the area of ​​the regional grids to obtain C sub-population rates; After eliminating the maximum and minimum sub-population rates, the remaining sub-population rates are accumulated and averaged to obtain A road penetration rates.

6. The elderly care service institution site selection planning system based on multi-source data and machine learning according to claim 5 is characterized in that: The steps for expanding the site selection area are: The planned coverage area, road penetration rate and one-way length peak value of A site selection points are assigned different weight factors and compared to obtain A extension lengths; The extended length is calculated as: ； In the formula, For the The extended length of the site selection point, For the The planned coverage area of ​​each site selection point is For the The road penetration rate of each site selection point, For the The peak one-way length of the site selection point is , , All are weight factors greater than 0; After arranging the A extension lengths from small to large, the third extension length is recorded as the extension standard; Take A site selection points as the expansion base points and the expansion standard as the expansion radius of the site selection points, radially expand the A site selection points into expansion circles, and record the area within the expansion circle as the site selection area, so as to obtain A site selection areas.

7. The elderly care service institution site selection planning system based on multi-source data and machine learning according to claim 6 is characterized in that: The steps for collecting regional traffic activity are as follows: Taking the site selection point as the midpoint of the division, the A site selection area is equally divided into E sub-areas with fan-shaped structures; During the historical period, the public transportation management system is used to query the number of active public transportation vehicles on weekdays, weekends and holidays in E sub-areas at a time point before the end of the period, and obtain E first active values, E second active values ​​and E third active values; Add the E first activity values, the E second activity values, and the E third activity values ​​and calculate the average, and compare the E averages with the area of ​​the sub-region to obtain E sub-activity levels; After removing the minimum value of the sub-activity, the minimum value of the remaining sub-activity is recorded as the regional traffic activity, and A regional traffic activities are obtained.

8. The elderly care service institution site selection planning system based on multi-source data and machine learning according to claim 7 is characterized in that: The steps for collecting the accessibility of medical resources are as follows: In A selection areas, the moving distance values ​​of all types of public transportation within a unit time are queried respectively, and the maximum moving distance value is recorded as the standard length; With two standard lengths as the expansion radius and the site selection point as the expansion origin, expand A expansion circles in the A site selection areas and mark the expansion circles at equal distances. non-adjacent expansion points; Respectively The expansion point is the scanning center, and a standard length is the scanning radius. Scan area; Statistics The number of medical service institutions in the scan area is obtained Medical value, The medical quantity values ​​are accumulated and averaged, and compared with the standard length to obtain A medical resource accessibility rates.

9. The elderly care service institution site selection planning system based on multi-source data and machine learning according to claim 8, characterized in that: Site selection attributes include available sites and unavailable sites; The steps for screening candidate regions are: When the output of the machine learning model is 1, the location attribute of the site selection area is unavailable for site selection, and the site selection area is not recorded as a candidate area; When the output of the machine learning model is 0, the location attribute of the site selection area is available site selection, and the site selection area of ​​available site selection is recorded as a candidate area, and S candidate areas are obtained.

10. The elderly care service institution site selection planning system based on multi-source data and machine learning according to claim 9, characterized in that: Elderly care service data include elderly density demand value, regional transportation demand, medical resource demand rate, urban greening demand rate and functional configuration demand; The matching data is identified by: When the elderly density demand value is less than or equal to the elderly density floating value, the elderly density floating value is recorded as matching data; When the regional traffic demand is less than or equal to the regional traffic activity, the regional traffic activity is recorded as matching data; When the medical resource demand rate is less than or equal to the medical resource availability rate, the medical resource availability rate is recorded as matching data; When the urban greening demand rate is less than or equal to the urban greening rate, the urban greening rate is recorded as matching data; When the function configuration requirement degree is less than or equal to the function configuration degree, the function configuration degree is recorded as matching data; The steps for excellent and good division are: The number of matching data in the S candidate areas is counted in sequence. When the number of matching data is 5, the site selection level of the candidate area is recorded as excellent. When the number of matching data is 4 or 3, the site selection level of the candidate area is recorded as good.

Images