CN116451963B - Multi-factor pension service facility optimal configuration method based on ensemble learning - Google Patents
Multi-factor pension service facility optimal configuration method based on ensemble learning Download PDFInfo
- Publication number
- CN116451963B CN116451963B CN202310435147.2A CN202310435147A CN116451963B CN 116451963 B CN116451963 B CN 116451963B CN 202310435147 A CN202310435147 A CN 202310435147A CN 116451963 B CN116451963 B CN 116451963B
- Authority
- CN
- China
- Prior art keywords
- pension
- road
- service
- service facility
- pension service
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 238000013145 classification model Methods 0.000 claims abstract description 51
- 238000012216 screening Methods 0.000 claims abstract description 19
- 230000032683 aging Effects 0.000 claims description 27
- 238000012549 training Methods 0.000 claims description 17
- 238000004458 analytical method Methods 0.000 claims description 16
- 238000009826 distribution Methods 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 8
- 230000002776 aggregation Effects 0.000 claims description 6
- 238000004220 aggregation Methods 0.000 claims description 6
- 230000004927 fusion Effects 0.000 claims description 6
- 238000012876 topography Methods 0.000 claims description 6
- 238000003012 network analysis Methods 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 4
- 239000000758 substrate Substances 0.000 claims description 4
- 239000006185 dispersion Substances 0.000 claims description 3
- 230000008901 benefit Effects 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 7
- 238000011160 research Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 5
- 239000002131 composite material Substances 0.000 description 5
- 238000005457 optimization Methods 0.000 description 5
- 238000002360 preparation method Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000012732 spatial analysis Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 241001166076 Diapheromera femorata Species 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Abstract
The invention discloses a multi-factor pension service facility optimal configuration method based on ensemble learning, which comprises the following steps of S10, constructing a layered road network model; s20, calculating and judging the service supply capacity of the pension service facility and the service supply capacity of the region based on the layered road network model; s30, screening effective influence factors from a plurality of possible influence factors of the pension service facility site selection by using a Catboost classification model based on the judging result of the intensity of the pension service supply capability of the region, and constructing an pension service facility site selection index system based on the effective influence factors; s40, predicting a primary site selection result of the pension service facility aiming at the region with weak pension service supply capability based on the trained Catboost classification model; s50, screening out the priority addressing areas by an addressing priority assessment method. The advantages are that: the method can effectively predict the area suitable for site selection and provide references for scientific and effective planning, reasonable site selection and layout of the pension service facilities.
Description
Technical Field
The invention relates to the technical field of homeland space planning, in particular to a multi-factor pension service facility optimal configuration method based on ensemble learning.
Background
Population aging is an objective trend of economic and social development, and China has entered the aging society according to the seventh national population census data. In the social background that the current aging degree is continuously increased, the problem of unbalanced development between the supply and the demand of the pension service facilities becomes an important problem facing the current society, and the society is required to provide more pension service resources to relieve the current pension service pressure. The key for solving the problems is to objectively and accurately evaluate the service supply capacity of the existing pension service facilities, determine facilities and areas with insufficient service supply capacity, analyze factors influencing the site selection of pension service facilities, provide a scientific optimal configuration method, guide the reasonable layout of pension service facilities, improve the utilization rate of pension service facilities and optimize the space layout of facilities. Therefore, a pension service supply capability assessment method based on supply and demand relations needs to be established through a scientific means, a pension service facility site selection index system is formed, a pension service facility site selection model is built, and scientific basis is provided for pension service facility optimal configuration.
At present, in the technical field of homeland space planning, the existing optimal configuration method at home and abroad mainly focuses on the research of site selection by utilizing theories and technologies in the aspects of statistics, operation research, geographic information systems and the like, and along with the development of big data technology and machine learning theory, the construction of site selection models by combining interest point data with machine learning is developed to a certain extent. The first is to apply an operation research method, which regards the site selection process as a location decision process, establishes a hierarchical structure of influencing factors by a hierarchical analysis method, sorts the influencing factors according to importance, determines respective relative contributions of the factors, weights and combines scores to select a target; the second category is to apply a geographic information system method, which combines spatial analysis tools such as superposition analysis, network analysis, buffer area analysis, nuclear density analysis and the like in spatial analysis technology to conduct site selection research and visualize a suitability region of site selection; the third category is to use a statistical method, establish index combination weights of a decision model through a gray correlation analysis method, a multiple linear regression analysis method and the like, fit relationships among factors, and analyze the influence importance of each influence factor on site selection; the fourth class is to apply a machine learning method to generate a model algorithm from data, such as a BP neural network, a support vector machine, a principal component analysis method, a genetic algorithm, an immune algorithm, a heuristic algorithm, etc., and obtain an output model according to a large amount of input data to obtain a suitable site selection area. In the existing method for researching site selection of the pension service facility, when the supply and demand relationship is evaluated, the service supply capacity of the pension service facility and the pension service supply capacity of the area are not accurately described and quantitatively analyzed, and the real and accurate equality evaluation of pension service resources is not given by combining the actual conditions; meanwhile, the selection of the influence factors is greatly influenced by artificial subjective factors, and the importance description and reasonable explanation of the influence factors on the site selection cannot be objectively given.
Therefore, how to build a method for scientifically evaluating the service supply capability of the pension, forming an index system for site selection of the pension service facilities, and building an intelligent site selection model of the pension service facilities is a problem to be solved by those skilled in the art.
Disclosure of Invention
The invention aims to provide a multi-factor pension service facility optimal configuration method based on ensemble learning, so as to solve the problems in the prior art.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
a multi-factor pension service facility optimizing configuration method based on integrated learning, which comprises the following steps,
s10, constructing a layered road network model:
constructing a layered road network model considering the height and the gradient by relying on road entity data and terrain data;
s20, calculating and judging the service supply capacity of the pension service facility and the service supply capacity of the region based on the layered road network model:
taking the aging degree and the quantity of the resources of the pension service facilities as the basis of the supply-demand relationship, acquiring service supply capacity indexes of each pension service facility and pension service supply capacity indexes of each region based on a hierarchical road network model, and judging the service supply capacity of the pension service facility and the pension service supply capacity of the region according to the magnitude relationship between each index and the corresponding preset index value;
S30, screening effective influence factors from a plurality of possible influence factors of the pension service facility site selection by using a Catboost classification model based on the judging result of the intensity of the pension service supply capability of the region, and constructing an pension service facility site selection index system based on the effective influence factors:
for areas with weak service supply capability and missing service facilities for the aged, calculating the characteristics of influence factors by using a GIS method, screening effective influence factors by using a Catboost classification model based on the corresponding characteristics, and constructing an address selection index system of the service facilities for the aged based on the effective influence factors;
s40, predicting a primary site selection result of the pension service facility aiming at the region with weak pension service supply capability based on the trained Catboost classification model:
inputting the effective influence factors into a Catboost classification model to train the Catboost classification model, calculating the effective influence factors of areas with weak service supply capacity of the aged, and inputting the effective influence factors into the trained Catboost classification model to judge whether the areas are suitable for site selection;
s50, screening out a priority addressing area by an addressing priority assessment method:
based on the relation between the prediction result of S40 and the land type, the predicted preliminary site selection result is screened multiple times under the double constraint condition of considering the number of the aged population and the endowment service supply capability of the region, and finally whether the region can be used as a site selection candidate region is determined.
Preferably, step S10 includes,
s11, depending on road entity data including an urban trunk road, a secondary trunk road, a branch road and an internal road, extracting a road center line according to a road surface, and connecting the road center line to ensure that a through road network is formed; breaking the road at the road intersection, ensuring connectivity at the road intersection, storing the road network as a road segment set, and constructing an urban road network data set;
s12, depending on road entity data including overpasses, sidewalks and slow roads, extracting central lines of the road entity data according to the plane data of the overpasses, and communicating the central lines with the central lines of urban roads to ensure that the road entity data form a through road network; breaking at the connection position, ensuring connectivity at the intersection point, storing the road network as a road segment set, and fusing the road segment set with an urban road network data set to construct a road network data set of an urban fused walking road;
s13, calculating elevation information from a starting point to an end point of each road segment according to the terrain data, calculating gradient information of each road segment according to the elevation information, and taking the elevation information of each end point of the road segment and the gradient information of each road segment as auxiliary attribute information of a road network data set of the urban fusion walking road;
S14, layering road network data sets of the urban fusion walking roads according to the importance of the roads, constructing a graph model by utilizing graph characteristics of the road network on each layer, optimizing the graph model, and creating a shortcut for important nodes and road sections for improving the shortest path searching efficiency; and constructing a layered road network model considering the height and the gradient by using a coverage method based on graph segmentation and a shortest path algorithm.
Preferably, the step S20 includes,
s21, calculating the actual walking reachable coverage range of the pension service facility in the preset time by depending on the pension service facility, the resident population and the layered road network model, counting the service supply capability difference of the aging population and the pension service facility in the coverage range, and calculating to obtain the service supply capability index PS of the pension service facility i i ；
PS i ＝P-1000×N i /C
Wherein N is i The number of beds available for the pension service i; c is the number of beds which each thousand old people should possess; p is the number of aging population in the coverage range which can be reached by the practical walking of the pension service facility in the preset time;
s22, taking the grid as a substrate, taking the pension service facility i as a center, adopting a layered road network model, and taking into considerationThe walking speed, the terrain and the gradient of the aging population are searched for the sum of the aging population in the coverage range which can be reached by walking in a preset time, and the supply and demand ratio S of the aged service facility i and the aging population is calculated i ；
v k ＝cosθ k ×cotθ k ×α×v
Wherein P is j An aging population number for grid cell j; t is t liTn Presetting a time limit for walking of an aged population; t is t ij A decay time variable, which is a road network based pension service facility i to grid cell j, that is associated with elevation or slope; v k The walking speed of the kth road section considering the gradient and the traveling mode is represented; θ k Is the gradient of the kth road section; alpha is a travel mode; v is walking speed when the gradient is zero; mu (mu) k Is a gradient coefficient; beta is an age-related coefficient; d, d k Representing the plane distance when the gradient of the kth road section is 0; g (t) ij ，t lim ) Is a gaussian time cost decay function; to avoid the situation that the result appears 0, epsilon is set to be a positive number close to 0;
s23, calculating the equality of the resource of the pension service facility acquired by the aging population of the grid unit j by taking the grid as a substrate, and taking the equality as an pension service supply capacity index of the grid unit j; searching for a preset time t with the grid cell j as the center lim Internal walking can reachAll pension services i within the coverage area reached, and the supply-demand ratio S for each pension service i searched i Summing to obtain the pension service supply capability index PG of the grid cell j j ；
Wherein Res is the set of pension service facility resources;
S24 for PS i The pension service facility with the service supply capacity being weak is judged to be the pension service facility with the service supply capacity being weak, when the pension service facility is optimized, the pension service facility is used as a key facility for improving the pension bed number, the service capacity requirement and the index of the related public service facility are referred to, the corresponding bed number is properly increased, and the service supply capacity of the pension service facility is enhanced; for PG j And the grid cells smaller than the second preset index value are judged to be areas with weak endowment service supply capability, and the areas are used as key areas for endowment service facility site selection when the subsequent endowment service facilities are configured.
Preferably, the step S30 includes,
s31, selecting influence factors covering all aspects of clothing, food, living and lines of an aging population in the POI; for the influence factors which do not have the space aggregation characteristic in the data distribution dispersion, space reachability analysis is adopted, and space reachability score values are used as characteristics; for the influence factors which are densely distributed in data and have the space aggregation characteristic, road network analysis is adopted, and the quantity in the walking reachable range in the preset time is used as a characteristic; for the influence factors with balanced data distribution and large data quantity, adopting nuclear density analysis and taking a nuclear density value as a characteristic;
S32, selecting positive and negative samples from the samples, inputting the positive and negative samples into a Catboost classification model for training, obtaining the relative contribution degree of each feature to a prediction result, and selecting an influence factor corresponding to the feature with the relative contribution degree larger than a preset threshold value as an effective influence factor of the pension service facility;
s33, constructing an address selection index system of the pension service facility based on the effective influence factors, and taking the global importance of the effective influence factors as weight values of the effective influence factors in the address selection index system of the pension service facility.
Preferably, step S40 includes,
s41, inputting the selected effective influence factors into the Catboost classification model again for training, and continuously adjusting parameters of the Catboost classification model in the training process until the AUC value of the Catboost classification model is highest, so as to obtain a trained Catboost classification model;
s42, calculating effective influence factors corresponding to areas with weak service supply capability of the pension, inputting the effective influence factors into a trained Catboost classification model, judging whether the current area is a suitable address selection area, and preliminarily obtaining a prediction result of the area to be addressed.
Preferably, step S50 includes,
s51, judging the association relation between the prediction result and the land type, and further eliminating the area where the land type is not suitable for building the pension service facility, so as to realize the preliminary screening of the prediction result;
S52, screening the prediction result again under the double constraint conditions of the number of the aged population and the endowment service supply capacity of the area; grid cells j suitable for site selection in a prediction result are calculated one by one, breakpoint layering is carried out on the endowment service supply capacity of the area through a natural breakpoint method, and the endowment service supply capacity interruption value PG of each layer is determined i And by judging the pension service provision capability index PG of the grid cell j j And interrupt value PG i And determining whether the current area can be used as an addressing candidate area.
The beneficial effects of the invention are as follows: the method can accurately measure the service supply capacity of the pension service facility and the region, scientifically and reasonably give the address selection result of the pension service facility, and provide references for scientific and effective planning, reasonable address selection and layout of the pension service facility.
Drawings
Fig. 1 is a schematic flow chart of an optimization configuration method of a pension service facility according to an embodiment of the present invention;
fig. 2 is a detailed flow chart of an optimization configuration method for a pension service facility according to an embodiment of the present invention;
fig. 3 is a schematic diagram of preprocessing road network data according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a layered compression method of a road network according to the embodiment of the present invention;
FIG. 5 is a schematic diagram illustrating the calculation of the reach provided by the present embodiment of the invention;
fig. 6 is a schematic diagram of a spatial distribution of service supply capacity of a service facility for a pension service in a portion of an urban area according to an embodiment of the present invention;
fig. 7 is a schematic diagram of preliminary prediction site selection space distribution of a pension service facility in a part of an experimental area in a certain city according to an embodiment of the present invention;
FIG. 8 is a schematic diagram of a process for evaluating an addressing priority according to an embodiment of the present invention;
fig. 9 is a schematic diagram of spatial distribution of the results of site selection priority classification of a regional care service facility in a city according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the detailed description is presented by way of example only and is not intended to limit the invention.
As shown in fig. 1 and fig. 2, in this embodiment, a method for optimizing configuration of multi-factor pension service facilities based on ensemble learning is provided, which fuses POI data, road network data, pension service facilities, resident population data, performs data preprocessing, builds a layered road network model taking topography and gradient into account, and builds an pension service facility service supply capability assessment model in combination with pension service facilities (number of beds) and resident population data (resident population data), so as to determine pension service facilities needing optimization and areas with weak pension service supply capability; calculating the spatial distribution index of each factor by means of spatial analysis means such as reachability analysis, road network analysis, nuclear density analysis and the like, selecting positive and negative examples by using an integrated learning method, training an integrated learning classification model, screening site selection related factors, determining the importance degree of each related factor and constructing an index system of the site selection related factors of the pension service facility; based on the index system of the site selection factors, re-selecting positive and negative samples and the area to be predicted, and inputting the positive and negative samples and the area to be predicted into an integrated learning classification model to obtain a preliminary judgment result of site selection; and (3) carrying out site selection priority assessment on the preliminary judgment result according to the results of the land sea unit and regional care service supply capability assessment, and finally giving the site selection result of the grading assessment. The method comprises the steps of,
S10, constructing a layered road network model;
s20, calculating and judging the service supply capacity of the pension service facility and the service supply capacity of the region based on the layered road network model;
s30, screening effective influence factors from a plurality of possible influence factors of the pension service facility site selection by using a Catboost classification model based on the judging result of the intensity of the pension service supply capability of the region, and constructing an pension service facility site selection index system based on the effective influence factors;
s40, predicting a primary site selection result of the pension service facility aiming at the region with weak pension service supply capability based on the trained Catboost classification model;
s50, screening out the priority addressing areas by an addressing priority assessment method.
The method comprises five parts, namely, constructing a layered road network model, judging the service supply capacity, constructing an address selection index system of the pension service facility, predicting the primary address selection result of the pension service facility, and determining the address selection candidate area. The following description is made for each part:
1. constructing a layered road network model:
this part corresponds to step S10: and constructing a layered road network model considering the height and the gradient by relying on the road entity data and the terrain data. In particular to the preparation method of the composite material,
1. Road entity in basic geographic entity of certain city is relied on, road entity data (generally planar data) such as city trunk road, sub trunk road, branch road, internal road and the like are extracted according to entity classification codes, road center line is extracted according to road surface, and the road center line is connected to ensure that a through road network is formed; breaking the road at the road crossing point, ensuring connectivity at the road crossing point, removing duplicate road segments, deleting shorter road segments (generally less than 10 meters), storing the road network as a road segment set, and preliminarily constructing the urban road network data set. As shown in fig. 3.
2. Road entity data such as overpasses, sidewalks, and slow roads in basic geographic entities of a certain city are relied on, the central line of the road entity data is extracted according to the plane data of the road entity data, and the central line is communicated with the central line data of the urban roads to ensure that a through road network is formed; breaking at the connection position, ensuring connectivity at the intersection point, storing the road network as a road segment set, fusing the road segment set with the urban road network data set obtained in the step 1, and constructing a road network data set of the urban fused walking road.
3. And calculating elevation information from the starting point to the end point of each road segment according to the terrain data, calculating gradient information of each road segment according to the elevation information, and taking the elevation information of each end point of the road segment and the gradient information of each road segment as auxiliary attribute information of a road network data set of the urban fusion walking road.
4. Layering the road network data set of the urban fusion walking road according to the importance of the road, wherein the higher the importance of the road is, the higher the level of the road is, the higher the general trunk road is, the higher the branch road is, and the like; each layer utilizes the graph characteristics of the road network to construct a graph model, optimizes the graph model, creates a shortcut for important nodes and road sections, and is used for improving the shortest path searching efficiency; and finally, constructing a layered compression road network model considering the elevation and the gradient by using a coverage method based on graph segmentation and a shortest path algorithm.
The method mainly comprises dividing and covering the map, wherein the dividing stage mainly reduces the number of boundary arcs (boundary arc) in the covering map, and utilizes natural geographic entities such as river and mountain in the topography and artificial building entities (expressway)Overpass, etc.) as heuristic factors, and representing the segmentation result as a multi-level overlay; the overlay processing stage mainly carries out topology processing on the overlay, ensures that the topology structure is stored only once and can be used for calculation of different measures. Any node (or cut arc) exists at the Level of the overlay G 0 In the layer, therefore, only a set of nodes need to be stored, each node having a flag to identify whether it is Level or not i Boundary nodes (boundary vertexes) on the layer; for each graph-partitioned cell, a bipartite graph can be represented as a border node, each cell has p ingress nodes and q egress nodes, and a p-q matrix W is used to store cost values of all shortcuts (shortcuts); a 2-layer overlay is shown in fig. 4, where boundary arc represents an edge connecting different cells with a level, boundary vertex represents a point on a cell boundary, shortcut represents a shortest path between two points, and clique represents shortcut between any two boundary vertices in a cell.
2. Judging the service supply capability:
this part corresponds to step S20: the method comprises the steps of taking the aging degree and the quantity of the resources of the pension service facilities as the supply and demand relation basis, acquiring the supply capacity index of each pension service facility and the pension service supply capacity index of each area based on a hierarchical road network model, and judging the service supply capacity of the pension service facility and the pension service supply capacity of the area according to the size relation between each index and the corresponding preset index value. In particular to the preparation method of the composite material,
1. Depending on the pension service facility, resident population and layered road network model, calculating the coverage range of the pension service facility which can be actually walked (considering age, topography, gradient and the like) within a preset time (which can be selected according to actual conditions so as to better meet actual demands, and is set to 15 minutes in the embodiment), counting the difference between the aging population in the coverage range and the service supply capacity (number of beds/thousand people) of the pension service facility, and calculating and obtaining the service supply capacity index PS of the pension service facility i i ；
PS i ＝P-1000×N i /C
Wherein N is i The number of beds available for the pension service i; c is the number of beds which each thousand old people should possess; p is the number of aging population in the coverage area that the pension service facility can actually walk within the preset time.
The actual walking reachable coverage calculation flow within 15 minutes of the pension service facility is shown in reference to fig. 5, and is divided into three phases: the first stage is to construct a path space relation from a building (such as a building and the like) to a road on the basis of the layered road network model constructed according to the step S10 by combining basic geographic entity data (such as entities of courtyard, courtyard gate, intra-district road and the like) of a certain city analysis area, so that capturing of a starting point position during path planning is facilitated, and accurate calculation of a walking process is ensured; the second stage is the constraint of population age on walking speed, especially the walking speed of the elderly is different from that of the general adults, as shown in table 1; meanwhile, different impedances are set in consideration of travel modes (such as wheelchairs, carts, walking, bicycles and the like) of the old people, and the different impedances are used for initializing weights of a calculation cost matrix in a layered road network model, as shown in a table 2; and the third stage is to calculate the reachable range of the road network in practice under the constraint conditions of 15 minutes time constraint, age, travel mode and the like.
TABLE 1 age constraint and speed impedance relationship table
TABLE 2 travel mode and speed impedance relationship table
| Travel mode | Speed of speedImpedance of |
| Walk around | 1.0 |
| Bicycle with wheel | 1.2 |
| Electric vehicle | 1.5 |
| Barrows | 0.8 |
| Walking stick | 0.7 |
| Wheelchair | 0.6 |
2. Taking a grid as a base, assuming that the aged population in the grid unit is uniformly distributed, taking the aged service facility i as a center, adopting a layered road network model, simultaneously taking the walking speed, the topography and the gradient of the aged population into consideration, searching the sum of the aged population in a coverage range which can be reached by walking in a preset time, and calculating the supply and demand ratio S of the aged service facility i and the aged population i ；
v k ＝cosθ k ×cotθ k ×α×v
Wherein P is j An aging population number for grid cell j; t is t lim Presetting a time limit for walking of an aged population; t is t ij A decay time variable, which is a road network based pension service facility i to grid cell j, that is associated with elevation or slope; v k The walking speed of the kth road section considering the gradient and the traveling mode is represented; θ k Is the gradient of the kth road section; alpha is a travel mode; v is walking speed when the gradient is zero; mu (mu) k Is a gradient coefficient; beta is an age-related coefficient; d, d k Representing the plane distance when the gradient of the kth road section is 0; g (t) ij ，t lim ) Is a gaussian time cost decay function; to avoid the situation that the result appears 0, epsilon is set to be a positive number close to 0, and the value of epsilon is 0.000001 in the example; referring to the service capability requirement of the pension facility issued by the selected city, the pension bed possessed by each thousand old people needs to be more than 40, and the pension bed is taken as the standard of the service capability of the pension facility, and the value C is 40.
3. Taking the grid as a substrate, calculating the equality of the aged population of the grid unit j for acquiring the resource of the pension service facility, and taking the equality as an pension service supply capacity index of the grid unit j; searching for a preset time t with the grid cell j as the center lim All pension services i within coverage reachable by walking in the house, and the supply-demand ratio S for each pension service i searched i Summing to obtain the service supply capacity index PG of the pension service facility of the grid unit j j ；
Where Res is the set of pension service facility resources.
4. For PS i A pension service facility having a higher than the first preset index value (in this embodiment, the first preset index value is set to 100), which is determined to be a pension service facility having a weak service supply capability, and when the pension service facility is optimized, as a key facility for increasing the number of pension beds, the number of corresponding beds is increased (in this embodiment, the number of increased beds is set to 4) with reference to the service capability requirement of a public service facility in a certain city; for PG j Grid cells smaller than the second preset index value (in this embodiment, the second preset index value is set to 0.496), which are determined as areas where the endowment service supply capability is weak, will be key areas for endowment service facility site selection at the time of subsequent endowment service facility configuration. Taking a study area of a city as an example, the distribution of areas with weak service supply capability of the old people is shown in fig. 6.
3. Construction of address selection index system of pension service facility
This part corresponds to step S30: for areas with weak service supply capability and missing service facilities for the aged, calculating the characteristics of influence factors by using a GIS method, screening effective influence factors by using a Catboost classification model based on the corresponding characteristics, and constructing an address selection index system of the service facilities for the aged based on the effective influence factors. In particular to the preparation method of the composite material,
1. in view of the fact that the influence factors and the importance of the aged service facility site selection cannot be clearly determined, the influence factors (such as 11 major categories and population structures of living houses, living services, administrative offices, education and scientific researches, medical sanitation, transportation facilities, cultural relics, business services, leisure entertainment, public facilities, greenbelts, squares and the like) covering all aspects of the aged population are selected in POIs as initial indexes to be used as subsequent influence factor screening and factor importance analysis); for the influence factors which do not have the space aggregation characteristic in the data distribution dispersion, space reachability analysis is adopted, and space reachability score values are used as characteristics; for the influence factors which are densely distributed in data and have the space aggregation characteristic, road network analysis is adopted, and the quantity in the walking reachable range in the preset time is used as a characteristic; for the influence factors of balanced data distribution and large data volume, nuclear density analysis is adopted, and nuclear density values are used as characteristics. Combining the PO I data space distribution condition of a research area of a certain city part, and selecting and quantifying indexes of the address of the pension facility, as shown in a table 3;
TABLE 3 selection and quantification of index for pension facility site selection
2. Selecting positive and negative samples from the samples, inputting the positive and negative samples into a Catboost classification model for training, obtaining the relative contribution degree of each feature to a prediction result, and selecting an influence factor corresponding to the feature with the relative contribution degree larger than a preset threshold value as an effective influence factor of the pension service facility;
the Catboost not only has higher prediction precision, but also can obtain feature importance degree sequencing, and can calmly different relative contribution degrees of possible influence factors (namely initial indexes) of the endowment service facilities to a prediction result, wherein the relative contribution degree of a certain feature in a single balance tree is measured by the following formula:
where M is the number of iterations (the number of balanced trees); t (T) m For the mth balanced tree,for feature j in balance tree T m Importance of (3); />For the global importance of the feature j,
wherein t is a node; l is the number of leaf nodes of the tree; l-1 is the number of non-leaf nodes of the tree; v t Is a feature associated with node t; i is a feature indication function; i.e t Loss after splitting for node t;the more the reduction indicates the greater the benefit of this split, meaning that the feature is of higher importance to the feature of the belonging node; final preference- >Features above a certain threshold are taken as endowment service facility impact factors; according to the regional sample data of the part of the study of a certain city, < + >>The index and weight of the selected site selection model are shown in table 4, and the threshold selection of (1) is 0.03, and the method comprises the following steps: 8 major classes and 15 minor classes of medical sanitation, traffic facilities, population structures, business services, life services, education and scientific research, greenbelts, squares, cultural relics and the like are taken as effective influencing factors.
3. Based on the effective influence factors, constructing an address index system of the pension service facility, and taking the global importance of the effective influence factors as weight values of the effective influence factors in the address index system of the pension service facility. And constructing an address selection model for the pension service facility through the Catboost classification model, wherein the output accuracy is 88.97%, and the weight result is shown in Table 4. The index weight can find that the index weight of population factors is the largest among indexes of the population factors, so that the distribution of the population numbers of the aged is the main factor influencing the site selection of the aged facility, and the site selection of the aged service facility takes priority in the region with more aged population under the condition that the distribution of the infrastructure is relatively perfect.
TABLE 4 index and weight for site selection model
4. Predicting primary site selection results of the pension service facilities:
this part corresponds to step S40: the effective influence factors are input into a Catboost classification model to train the Catboost classification model, the effective influence factors of the areas with weak service supply capacity of the aged are calculated, and the effective influence factors are input into the trained Catboost classification model to judge whether the areas are suitable for site selection. In particular to the preparation method of the composite material,
1. inputting the selected effective influence factors into the Catboost classification model again for training, and continuously adjusting parameters of the Catboost classification model in the training process until the AUC value of the Catboost classification model is the highest, so as to obtain a trained Catboost classification model;
and continuously adjusting parameters of the classification model in the training process, and comparing the AUC values of the Catboost classification model to determine the optimal values of the parameters of the classification model so as to determine the optimal parameter combination of the classification model, and inputting the optimal parameter combination into the classification model to obtain the trained classification model.
Determining positive and negative samples according to a sample selection method; the tag value of the positive sample is defined as "1" and the tag value of the negative sample is defined as "0" according to the structure type of the input data. Dividing the sample into two parts without intersection, wherein one part of sample set is a training sample, and the other part of sample set is a test sample, wherein the training sample accounts for 70% of the total number of samples, and the number of the test samples accounts for 30% of the number of samples; the training samples, test samples and the number of the finally determined training samples are shown in table 5.
TABLE 5 number of model samples
| Regional object | Number (personal) |
| Sample of the positive example | 3398 |
| Counterexample sample | 6789 |
| Training sample | 10187 |
| Test sample | 4365 |
| Prediction samples | 390 |
| Total sample | 14552 |
The Catboost classification model uses a balanced tree as a base predictor, and adopts the same segmentation criteria at the same layer of the tree; when using the balanced tree for classification, the corresponding parameters need to be set for controlling the whole decision process, such as the number of trees, the maximum depth of the trees, etc., and the setting of the main parameters in the model and the definition thereof are shown in table 6.
TABLE 6 main parameters of Catboost model
| Parameter name | Parameter value |
| Data slicing | 0.7 |
| Data shuffling | Whether or not |
| Cross validation | Whether or not |
| Number of iterations | 100 |
| Learning rate | 0.03 |
| L2 regularization term | 3 |
| Maximum depth of tree | 10 |
| Overfitting detection threshold | 0 |
| Number of continuing iterations after optimization is achieved | 20 |
| Random subspace | 1 |
| Bayesian bagging control intensity | 1 |
| Fractional standard deviation multiplier | 1 |
| Method for calculating leaf value | Gradient |
| Length coefficient | 2 |
2. And calculating an effective influence factor corresponding to the area with weak service supply capability of the pension, inputting the effective influence factor into a trained Catboost classification model, judging whether the current area is a suitable site selection area, and preliminarily obtaining a prediction result of the area to be site selected. And according to the trained Catboost classification model, the result of preliminary prediction site selection of the regional care service facility of the certain city in the prediction analysis is shown in fig. 7.
5. And determining an address candidate area:
this part corresponds to step S50: based on the relation between the prediction result of the last step and the land type, the predicted preliminary site selection result is screened for multiple times under the double constraint condition of considering the number of the old population and the endowment service supply capability of the area, and finally whether the area can be used as the site selection candidate area is judged. In particular to the preparation method of the composite material,
1. judging the association relation between the prediction result and the land type, and further eliminating the area where the land type is not suitable for building the pension service facility, so as to realize the preliminary screening of the prediction result; as shown in fig. 8.
2. Screening the prediction results again under the double constraint condition of the number of the aged population and the endowment service supply capacity of the region, as shown in fig. 8; grid cells j suitable for site selection in a prediction result are calculated one by one, breakpoint layering is carried out on the endowment service supply capacity of the area through a natural breakpoint method, and the endowment service supply capacity interruption value PG of each layer is determined i And by judging the pension service provision capability index PG of the grid cell j j And interrupt value PG i And determining whether the current area can be used as an addressing candidate area.
As shown in table 7, assuming that the high, middle, and low 3-class candidate areas are classified, the pension service provision capability interruption value is expressed as PG i I is more than or equal to 1 and less than or equal to 3, and the endowment service supply capability index PG of the grid j is judged j And breakpoint PG i And determining whether the current area is a candidate area for addressing.
TABLE 7 priority rating
Finally, according to the addressing priority assessment method, a final addressing result is obtained, and is subjected to space visualization, the addressing priority result is screened as shown in fig. 9, the addressing result is divided into a priority-high addressing area, a priority-medium addressing area and a priority-low addressing area, the divided result considers the aging degree and the endowment service supply capacity of the area, the final addressing result is realized based on the relation between the requirement and the supply, and the result not only solves the problem that whether the area is suitable for building the endowment service facility, but also solves the problem that whether the area needs to be built with the endowment service facility. For the decision maker, the higher priority address selection area can be preferentially selected as the decision result of the address selection according to the actual planning.
In the embodiment, a hierarchical road network model considering the topography and the gradient is constructed based on an integrated learning multi-factor pension service facility optimization configuration method, and an pension service supply capability assessment model is constructed based on the aging degree and the pension service supply capability; for areas with weak service supply capability, constructing an integrated learning classification model based on a Catboost algorithm, determining relevant factors and importance degrees of the service facility site selection of the aged, and further constructing an index system of the service facility site selection of the aged; the training samples are selected and input into an integrated learning classification model, so that a preliminary prediction result of the address selection of the pension service facility is obtained; and giving a final priority grading and site selection result by a site selection priority grading method. The method can predict the area suitable for site selection and provide references for scientific and effective planning, reasonable site selection and layout of the pension service facilities.
By adopting the technical scheme disclosed by the invention, the following beneficial effects are obtained:
the invention provides an integrated learning-based multi-factor pension service facility optimal configuration method, which can accurately measure service supply capacity of pension service facilities and areas, scientifically and reasonably give out site selection results of pension service facilities, and provide references for scientific and effective planning, reasonable site selection and layout of pension service facilities.
The foregoing is merely a preferred embodiment of the present invention and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present invention, which is also intended to be covered by the present invention.
Claims (6)
1. A multi-factor pension service facility optimal configuration method based on ensemble learning is characterized in that: comprises the following steps of the method,
s10, constructing a layered road network model:
constructing a layered road network model considering the height and the gradient by relying on road entity data and terrain data;
s20, calculating and judging the service supply capacity of the pension service facility and the service supply capacity of the region based on the layered road network model:
Taking the aging degree and the quantity of the resources of the pension service facilities as the basis of the supply-demand relationship, acquiring service supply capacity indexes of each pension service facility and pension service supply capacity indexes of each region based on a hierarchical road network model, and judging the service supply capacity of the pension service facility and the pension service supply capacity of the region according to the magnitude relationship between each index and the corresponding preset index value;
s30, screening effective influence factors from a plurality of possible influence factors of the pension service facility site selection by using a Catboost classification model based on the judging result of the intensity of the pension service supply capability of the region, and constructing an pension service facility site selection index system based on the effective influence factors:
for areas with weak service supply capability and missing service facilities for the aged, calculating the characteristics of influence factors by using a GIS method, screening effective influence factors by using a Catboost classification model based on the corresponding characteristics, and constructing an address selection index system of the service facilities for the aged based on the effective influence factors;
s40, predicting a primary site selection result of the pension service facility aiming at the region with weak pension service supply capability based on the trained Catboost classification model:
inputting the effective influence factors into a Catboost classification model to train the Catboost classification model, calculating the effective influence factors of areas with weak service supply capacity of the aged, and inputting the effective influence factors into the trained Catboost classification model to judge whether the areas are suitable for site selection;
S50, screening out a priority addressing area by an addressing priority assessment method:
based on the relation between the prediction result of S40 and the land type, the predicted preliminary site selection result is screened multiple times under the double constraint condition of considering the number of the aged population and the endowment service supply capability of the region, and finally whether the region can be used as a site selection candidate region is determined.
2. The method for optimizing configuration of an integrated learning-based multi-factor pension services facility of claim 1, wherein: the step S10 includes the steps of,
s11, depending on road entity data including an urban trunk road, a secondary trunk road, a branch road and an internal road, extracting a road center line according to a road surface, and connecting the road center line to ensure that a through road network is formed; breaking the road at the road intersection, ensuring connectivity at the road intersection, storing the road network as a road segment set, and constructing an urban road network data set;
s12, depending on road entity data including overpasses, sidewalks and slow roads, extracting central lines of the road entity data according to the plane data of the overpasses, and communicating the central lines with the central lines of urban roads to ensure that the road entity data form a through road network; breaking at the connection position, ensuring connectivity at the intersection point, storing the road network as a road segment set, and fusing the road segment set with an urban road network data set to construct a road network data set of an urban fused walking road;
S13, calculating elevation information from a starting point to an end point of each road segment according to the terrain data, calculating gradient information of each road segment according to the elevation information, and taking the elevation information of each end point of the road segment and the gradient information of each road segment as auxiliary attribute information of a road network data set of the urban fusion walking road;
s14, layering road network data sets of the urban fusion walking roads according to the importance of the roads, constructing a graph model by utilizing graph characteristics of the road network on each layer, optimizing the graph model, and creating a shortcut for important nodes and road sections for improving the shortest path searching efficiency; and constructing a layered road network model considering the height and the gradient by using a coverage method based on graph segmentation and a shortest path algorithm.
3. The method for optimizing configuration of an integrated learning-based multi-factor pension services facility of claim 1, wherein: the step S20 includes the steps of,
s21, calculating the actual walking reachable coverage range of the pension service facility in the preset time by depending on the pension service facility, the resident population and the layered road network model, counting the service supply capability difference of the aging population and the pension service facility in the coverage range, and calculating to obtain the service supply capability index PS of the pension service facility i i ；
PS i ＝P-1000×N i /C
Wherein N is i The number of beds available for the pension service i; c is the number of beds which each thousand old people should possess; p is the number of aging population in the coverage range which can be reached by the practical walking of the pension service facility in the preset time;
s22, taking the grid as a base, taking the pension service facility i as a center, adopting a layered road network model, simultaneously taking the walking speed, the topography and the gradient of the aging population into consideration, searching the sum of the aging population in a coverage range which can be reached by walking in a preset time, and calculating the supply and demand ratio S of the pension service facility i and the aging population i ；
v k ＝cosθ k ×cotθ k ×α×v
Wherein P is j An aging population number for grid cell j; t is t lim Presetting a time limit for walking of an aged population; t is t ij A decay time variable, which is a road network based pension service facility i to grid cell j, that is associated with elevation or slope; v k The walking speed of the kth road section considering the gradient and the traveling mode is represented; θ is the gradient of the kth road segment; alpha is a travel mode; v is walking speed when the gradient is zero; mu (mu) k Is a gradient coefficient; beta is an age-related coefficient; d, d k Representing the plane distance when the gradient of the kth road section is 0; g (t) ij ,t lim ) Is a gaussian time cost decay function; to avoid the situation that the result appears 0, epsilon is set to be a positive number close to 0;
S23, calculating the equality of the resource of the pension service facility acquired by the aging population of the grid unit j by taking the grid as a substrate, and taking the equality as an pension service supply capacity index of the grid unit j; searching for a preset time t with the grid cell j as the center lim All pension services i within coverage reachable by walking in the house, and the supply-demand ratio S for each pension service i searched i Summing to obtain the pension service supply capability index PG of the grid cell j j ；
Wherein Res is the set of pension service facility resources;
s24 for PS i The pension service facility with the service supply capacity being weak is judged to be the pension service facility with the service supply capacity being weak, when the pension service facility is optimized, the pension service facility is used as a key facility for improving the pension bed number, the service capacity requirement and the index of the related public service facility are referred to, the corresponding bed number is properly increased, and the service supply capacity of the pension service facility is enhanced; for PG j And the grid cells smaller than the second preset index value are judged to be areas with weak endowment service supply capability, and the areas are used as key areas for endowment service facility site selection when the subsequent endowment service facilities are configured.
4. The method for optimizing configuration of an integrated learning-based multi-factor pension services facility of claim 1, wherein: the step S30 includes the steps of,
S31, selecting influence factors covering all aspects of clothing, food, living and lines of an aging population in the POI; for the influence factors which do not have the space aggregation characteristic in the data distribution dispersion, space reachability analysis is adopted, and space reachability score values are used as characteristics; for the influence factors which are densely distributed in data and have the space aggregation characteristic, road network analysis is adopted, and the quantity in the walking reachable range in the preset time is used as a characteristic; for the influence factors with balanced data distribution and large data quantity, adopting nuclear density analysis and taking a nuclear density value as a characteristic;
s32, selecting positive and negative samples from the samples, inputting the positive and negative samples into a Catboost classification model for training, obtaining the relative contribution degree of each feature to a prediction result, and selecting an influence factor corresponding to the feature with the relative contribution degree larger than a preset threshold value as an effective influence factor of the pension service facility;
s33, constructing an address selection index system of the pension service facility based on the effective influence factors, and taking the global importance of the effective influence factors as weight values of the effective influence factors in the address selection index system of the pension service facility.
5. The method for optimizing configuration of an integrated learning-based multi-factor pension services facility of claim 1, wherein: the step S40 includes the steps of,
S41, inputting the selected effective influence factors into the Catboost classification model again for training, and continuously adjusting parameters of the Catboost classification model in the training process until the AUC value of the Catboost classification model is highest, so as to obtain a trained Catboost classification model;
s42, calculating effective influence factors corresponding to areas with weak service supply capability of the pension, inputting the effective influence factors into a trained Catboost classification model, judging whether the current area is a suitable address selection area, and preliminarily obtaining a prediction result of the area to be addressed.
6. The method for optimizing configuration of an integrated learning-based multi-factor pension services facility of claim 1, wherein: the step S50 includes the steps of,
s51, judging the association relation between the prediction result and the land type, and further eliminating the area where the land type is not suitable for building the pension service facility, so as to realize the preliminary screening of the prediction result;
s52, screening the prediction result again under the double constraint conditions of the number of the aged population and the endowment service supply capacity of the area; grid cells j suitable for site selection in a prediction result are calculated one by one, breakpoint layering is carried out on the endowment service supply capacity of the area through a natural breakpoint method, and the endowment service supply capacity interruption value PG of each layer is determined i And by judging the pension service provision capability index PG of the grid cell j j And interrupt value PG i And determining whether the current area can be used as an addressing candidate area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310435147.2A CN116451963B (en) | 2023-04-21 | 2023-04-21 | Multi-factor pension service facility optimal configuration method based on ensemble learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310435147.2A CN116451963B (en) | 2023-04-21 | 2023-04-21 | Multi-factor pension service facility optimal configuration method based on ensemble learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116451963A CN116451963A (en) | 2023-07-18 |
| CN116451963B true CN116451963B (en) | 2023-11-07 |
Family
ID=87125328
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310435147.2A Active CN116451963B (en) | 2023-04-21 | 2023-04-21 | Multi-factor pension service facility optimal configuration method based on ensemble learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116451963B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117114286A (en) * | 2023-07-31 | 2023-11-24 | 青岛正信德宇信息科技有限公司 | Intelligent pension auxiliary planning method and system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021103323A1 (en) * | 2019-11-29 | 2021-06-03 | 青岛理工大学 | Historical city protection and development cooperative control scheme aided design system |
| CN112966900A (en) * | 2021-02-03 | 2021-06-15 | 上海应用技术大学 | Accessibility evaluation method for endowment institution based on improved potential model |
| CN113011671A (en) * | 2021-03-31 | 2021-06-22 | 深圳市规划国土发展研究中心 | GIS (geographic information System) spatial geographic analysis based forest fire prevention helideck layout site selection method |
| CN113935608A (en) * | 2021-10-08 | 2022-01-14 | 南京工业大学 | Evaluation method and device for community endowment service facilities |
| WO2022169874A1 (en) * | 2021-02-03 | 2022-08-11 | Atlas Ai P.B.C. | Computer system with economic development decision support platform and method of use thereof |
| CN114925985A (en) * | 2022-04-25 | 2022-08-19 | 深圳市蕾奥规划设计咨询股份有限公司 | Municipal facility site selection planning method based on GIS decision |
| CN115187311A (en) * | 2022-08-03 | 2022-10-14 | 上海维智卓新信息科技有限公司 | Shop site selection method and device suitable for multiple industries |
-
2023
- 2023-04-21 CN CN202310435147.2A patent/CN116451963B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021103323A1 (en) * | 2019-11-29 | 2021-06-03 | 青岛理工大学 | Historical city protection and development cooperative control scheme aided design system |
| CN112966900A (en) * | 2021-02-03 | 2021-06-15 | 上海应用技术大学 | Accessibility evaluation method for endowment institution based on improved potential model |
| WO2022169874A1 (en) * | 2021-02-03 | 2022-08-11 | Atlas Ai P.B.C. | Computer system with economic development decision support platform and method of use thereof |
| CN113011671A (en) * | 2021-03-31 | 2021-06-22 | 深圳市规划国土发展研究中心 | GIS (geographic information System) spatial geographic analysis based forest fire prevention helideck layout site selection method |
| CN113935608A (en) * | 2021-10-08 | 2022-01-14 | 南京工业大学 | Evaluation method and device for community endowment service facilities |
| CN114925985A (en) * | 2022-04-25 | 2022-08-19 | 深圳市蕾奥规划设计咨询股份有限公司 | Municipal facility site selection planning method based on GIS decision |
| CN115187311A (en) * | 2022-08-03 | 2022-10-14 | 上海维智卓新信息科技有限公司 | Shop site selection method and device suitable for multiple industries |
Non-Patent Citations (3)
| Title |
|---|
| Research on the Configuration of the Medical Facilities for Elderly of Rural Community;Wang Ping 等;《Humanities and Social Sciences》;第5卷(第6期);210-214 * |
| 基于GIS的合肥市包河区养老设施选址适宜性评价;李璐;;《居业》(第02期);26-27，29 * |
| 面向"15分钟社区生活圈"规划的养老设施建设测度;忻静;;《遥感信息》;第34卷(第02期);118-123 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116451963A (en) | 2023-07-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107943880B (en) | Geological disaster susceptibility improvement and evaluation method based on analytic hierarchy process | |
| Xu et al. | Modelling urban expansion guided by land ecological suitability: A case study of Changzhou City, China | |
| CN108446470B (en) | Medical facility accessibility analysis method based on vehicle trajectory data and population distribution | |
| CN110334864B (en) | GIS-based multi-specification-in-one urban and rural space partition method | |
| CN110176141B (en) | Traffic cell division method and system based on POI and traffic characteristics | |
| CN111651545A (en) | Urban marginal area extraction method based on multi-source data fusion | |
| CN116451963B (en) | Multi-factor pension service facility optimal configuration method based on ensemble learning | |
| CN110413855B (en) | Region entrance and exit dynamic extraction method based on taxi boarding point | |
| CN113112068A (en) | Method and system for addressing and layout of public facilities in villages and small towns | |
| Acar et al. | Assessment of road network planning by using GIS-based multi-criteria evaluation for conversion of coppice forest to high forest | |
| CN115063276A (en) | Urban ecological corridor space dividing method based on MSPA and circuit theory | |
| Yang et al. | How to improve urban transportation planning in big data era? A practice in the study of traffic analysis zone delineation | |
| CN114662774A (en) | City block vitality prediction method, storage medium and terminal | |
| CN114154880A (en) | Method for evaluating walking accessibility of public service facility on street unit level | |
| CN113191582B (en) | Road torrential flood vulnerability evaluation method based on GIS and machine learning | |
| CN114219521A (en) | Sports center site selection evaluation method based on multi-source data | |
| CN111008730B (en) | Crowd concentration prediction model construction method and device based on urban space structure | |
| CN116911507A (en) | Urban block vitality evaluation method and storage medium based on space projection pursuit | |
| Huo et al. | Using fuzzy clustering of user perception to determine the number of level-of-service categories for bus rapid transit | |
| Liu et al. | Road density analysis based on skeleton partitioning for road generalization | |
| CN114881309A (en) | Method for measuring characteristic correlation between urban vitality and carbon emptying | |
| CN114862137A (en) | Street walking traffic environment evaluation method based on analytic hierarchy process | |
| Haghparast et al. | Comprehensive Environmental Monitoring based on Stations of Environmental Pollutants (Air, Water and Soil) in Tehran | |
| CN116644809B (en) | Urban development boundary demarcation method integrating geographic big data and machine learning | |
| CN108446784A (en) | A kind of highway coordinates horizontal forecast method with Development of Urbanization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |