CN112884226A - Multi-agent algorithm-based territorial spatial pattern simulation planning method and system - Google Patents
Multi-agent algorithm-based territorial spatial pattern simulation planning method and system Download PDFInfo
- Publication number
- CN112884226A CN112884226A CN202110201118.0A CN202110201118A CN112884226A CN 112884226 A CN112884226 A CN 112884226A CN 202110201118 A CN202110201118 A CN 202110201118A CN 112884226 A CN112884226 A CN 112884226A
- Authority
- CN
- China
- Prior art keywords
- space
- land
- construction
- decision
- agent
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 104
- 238000013439 planning Methods 0.000 title claims abstract description 98
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 28
- 238000004088 simulation Methods 0.000 title claims description 31
- 238000010276 construction Methods 0.000 claims abstract description 156
- 238000011160 research Methods 0.000 claims abstract description 60
- 238000011161 development Methods 0.000 claims abstract description 56
- 230000001413 cellular effect Effects 0.000 claims abstract description 44
- 238000005457 optimization Methods 0.000 claims abstract description 8
- 230000006399 behavior Effects 0.000 claims description 68
- 230000008569 process Effects 0.000 claims description 64
- 238000011156 evaluation Methods 0.000 claims description 52
- 238000006243 chemical reaction Methods 0.000 claims description 32
- 230000009471 action Effects 0.000 claims description 27
- 238000012549 training Methods 0.000 claims description 25
- 238000012360 testing method Methods 0.000 claims description 21
- 230000006870 function Effects 0.000 claims description 16
- 230000003993 interaction Effects 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 8
- 239000011159 matrix material Substances 0.000 claims description 8
- 238000012804 iterative process Methods 0.000 claims description 6
- 238000000638 solvent extraction Methods 0.000 claims description 4
- 230000009466 transformation Effects 0.000 abstract 1
- 239000003795 chemical substances by application Substances 0.000 description 75
- 238000010586 diagram Methods 0.000 description 12
- 238000004458 analytical method Methods 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 5
- 239000002689 soil Substances 0.000 description 5
- 230000000007 visual effect Effects 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 238000005094 computer simulation Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000009034 developmental inhibition Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000001681 protective effect Effects 0.000 description 2
- 230000011664 signaling Effects 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000013316 zoning Methods 0.000 description 1
Images
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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Development Economics (AREA)
- Biophysics (AREA)
- General Business, Economics & Management (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Marketing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Databases & Information Systems (AREA)
- Primary Health Care (AREA)
- Remote Sensing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method and a system for simulating and planning a territorial spatial pattern based on a multi-agent algorithm, and belongs to the technical field of spatial planning. The method comprises the following steps: developing suitability indexes according to ecological environment resource bearing capacity and construction places, and carrying out space element risk identification on the research area to obtain space element risk identification data; extracting a distribution range of a suitability space built in a research area according to the space element risk identification data, taking the distribution range as a base map for simulating a town development boundary, and acquiring a preliminary delimiting result of the town development boundary in the research area based on a cellular automata model; and inputting the space element risk identification data and the preliminary planning result into a multi-agent model for scene optimization, outputting a space pattern division result of a research area, and simulating the planning of three areas and three lines so as to assist the planning of the territorial space planning and respond to the technical requirements of the planning of the territorial space during the transformation period of the current territorial space planning.
Description
Technical Field
The invention belongs to the technical field of space planning, and particularly relates to a method and a system for simulating and planning a territorial space pattern based on a multi-agent algorithm.
Background
The territorial space planning is an important content for the construction of an ecological civilization system and is a basic basis for various development, protection and construction activities. Particularly, at present, the three-region three-line intelligent division becomes important research content in the field of territorial space planning, about the division of three regions of town space, agricultural space and ecological space, and the division of three boundary lines of ecological protection red line, permanent basic farmland protection red line and town development boundary. Wherein, the three regions are mainly used for dividing the main function, and the three lines are mainly used for rigid control of the boundary. At present, methods such as system dynamics, cellular automata, multi-agent, neural network and the like are mainly used for dynamic monitoring of land utilization, evolution of land utilization and urban expansion simulation at home and abroad. The cellular automata and multi-agent technology becomes a widely used scene analysis simulation technology, and a new means is provided for dynamic analysis and simulation prediction of city expansion.
From the description of city growth morphology and law, Cellular Automata (CA) has been widely used in city space growth simulation as a tool for time-space dynamic simulation of complex systems. The key of the space dynamic simulation of the cellular automata is to determine a cellular conversion rule, and the determination method of the conversion rule mainly comprises an artificial neural network method, a logistic regression method, a multi-factor evaluation method and a multi-agent model method. The cellular automata is only based on the interaction of the land units, so that the defects are gradually highlighted in the specific application process, and the national soil space planning and compiling aiming at three-region three-line cannot be effectively carried out, so that a method for simulating and planning the national soil space pattern is urgently needed.
Disclosure of Invention
Aiming at the defects or improvement requirements in the prior art, the invention provides a method and a system for simulating and planning a territorial spatial pattern based on a multi-agent algorithm, and aims to solve the technical problem that the existing cellular automaton is gradually insufficient in a specific application process only based on the interaction of land units.
To achieve the above object, according to a first aspect of the present invention, there is provided a method for simulating and planning a territorial spatial pattern based on a multi-agent algorithm, comprising the steps of:
developing suitability indexes according to ecological environment resource bearing capacity and construction places, and carrying out space element risk identification on the research area to obtain space element risk identification data;
extracting a distribution range of a suitability space built in a research area according to the space element risk identification data, taking the distribution range as a base map for simulating a town development boundary, and acquiring a preliminary delimiting result of the town development boundary in the research area based on a cellular automata model;
and inputting the space element risk identification data and the preliminary planning result into a multi-agent model for scene optimization, and outputting a space pattern dividing result of the research area.
Preferably, the process of developing suitability indexes according to ecological environment resource bearing capacity and construction places to identify the risk of the spatial elements in the research area and obtain spatial element risk identification data further includes: acquiring various basic data of a research area, and unifying coordinates and formats of the various basic data to obtain a basic database; converting map data in a research area into a grid base map, and identifying data in a basic database to the grid base map; for the identified grid in the grid base map, if the area of the development-forbidden construction elements in the grid is larger than a preset threshold value, the grid is used as a development-forbidden land, and if the area of the development-forbidden construction elements in the grid is not larger than the preset threshold value, the grid is used as a development-forbidden land; and respectively dividing the distribution ranges of the ecological suitability space, the agricultural suitability space and the construction suitability space in each developable land, and using the distribution ranges as the space element risk identification data.
Preferably, the process of obtaining preliminary delineation results of the town development boundary within the study area based on the cellular automata model further comprises: based on the map patches in the national survey database, extracting the urban construction land in the research area as an initial land use state of the cellular automata for simulating urban development boundary expansion; determining the total land scale amount of the urban planning according to the existing planning scale, population scale trend, economic development trend and land use growth trend, and determining the scale of the construction land of the central urban area and each village and town within the planning period based on the total land scale amount and the construction land configuration function conducted in multiple levels; and determining the grids covered by the construction land of the town under the current situation based on the scale of the construction land and the initial land state, taking the grids as the initial state, inputting the initial state into a recursive function of a cellular automaton model, calculating based on a conversion rule to obtain a new evaluation value of each grid of the non-construction land, replacing the last iteration evaluation value in the initial state, repeating the cell change iteration process until the total construction land amount reaches the planned construction scale, and obtaining a preliminary planning result.
Preferably, the process of calculating the new evaluation value of each non-construction land grid based on the conversion rule is calculated by any one of the following two formulas:
valuenew＝∑ridvaluerid×typerid；
valuenew＝0.4×∑ridvaluerid×typerid+0.6×valuebefore；
wherein, for any non-construction land grid, valuenewIndicating a new evaluation value obtained by calculating the non-construction land grid, rid indicating the grid number of the neighboring grid around the non-construction land grid, valueridAn evaluation value, type, representing a grid adjacent to the non-construction land gridridIndicating whether the neighboring grid around the non-construction land grid is a construction land, valuebeforeAnd represents the evaluation value of the last iteration of the non-construction land grid.
Preferably, the above iterative process of cell change is repeated until the total construction area reaches the planned construction scale, and the process of obtaining the preliminary planning result further includes:
adjusting internal parameters of the cellular automaton model so that the internal parameters increase at different rates; wherein each adjustment of the internal parameters corresponds to an iterative process;
for any current iteration process, sorting the new evaluation values of each non-construction land grid obtained by calculation in the current iteration process from large to small, selecting the non-construction land grids with the preset percentage to be converted into construction land, judging whether the total construction land quantity formed after conversion reaches the planned construction scale or not, ending the iteration process if the total construction land quantity formed after conversion reaches the planned construction scale, repeating the processes of internal parameter adjustment, evaluation value calculation, evaluation value sorting and non-construction land grid conversion until the total construction land quantity formed after conversion reaches the planned construction scale if the total construction land quantity formed after conversion does not reach the planned construction scale, and outputting a preliminary planning result.
Preferably, the multi-Agent model comprises an Agent decision structure, an Agent decision library and an Agent decision behavior expression;
the Agent decision structure comprises a behavior subject set participating in space development decision and ecological protection decision, an interaction subject set among behavior subjects in the space development or ecological process, a state of the behavior subjects, a weight set of the behavior subjects, a set of space decision actions implemented by the behavior subjects in the space development process and probability estimation results of the behavior subjects on action strategies of other behavior subjects in the model;
the Agent decision library comprises decision rules, and the decision rules are obtained based on limited Boltzmann machine training; the Agent decision behavior expression is used for carrying out game among different behavior bodies based on the space element risk identification data and the preliminary demarcation result, and the space pattern demarcation result of the research area is determined according to the game result.
Preferably, the decision rule is obtained by training based on training data, wherein the training data comprises a land utilization type, a comprehensive factor variable and a decision rule set of a corresponding land type under the action of the comprehensive factor variable; accordingly, the training process of the decision rule further comprises:
initializing the number of layers, the number of nodes of each layer, the learning rate, the iteration period, the connection weight matrix and the bias matrix in the restricted Boltzmann machine, and initializing the structural parameters in the behavior body;
importing a comprehensive factor variable for forward propagation to update the structural parameters in the behavior body;
importing a comprehensive factor variable for back propagation and a decision rule set of corresponding land types under the action of the comprehensive factor variable for back propagation, and adjusting the acquired limited Boltzmann machine based on an error back propagation algorithm;
and importing a comprehensive factor variable for testing and a decision rule set of the corresponding land type under the action of the comprehensive factor variable for testing to test the accuracy of the trained Agent decision library, changing the network internal structure in the limited Boltzmann machine if the test result does not meet the preset condition, retraining, repeating the training process until the test result meets the preset condition, and outputting the final Agent decision library.
Preferably, the comprehensive factor variables include ecological suitability evaluation level, agricultural suitability evaluation level, town suitability evaluation level, subject functional compartmentalization factor, town development strategy factor, major project construction factor, space protection corridor factor, farmland occupation balance factor, index area balance and space elastic reserve.
Preferably, the spatial pattern partitioning process of Agent decision behavior expression further comprises:
judging whether various behavior main bodies can obtain corresponding decision rules from an Agent decision library in the current state or not based on the space element risk identification data and the preliminary planning result;
if various behavior bodies can obtain corresponding decision rules from the Agent decision library in the current state, obtaining the corresponding decision rules and determining the type of the decision behavior, and if various behavior bodies cannot extract the corresponding decision rules from the Agent decision library in the current state, obtaining the decision rules from the newly-set decision rules;
and triggering the space decision actions of various behavior main bodies based on the obtained decision rule so as to carry out game interaction among the various behavior main bodies, repeating the game interaction process until the game is balanced, and outputting the space pattern division result of the research area according to the game result.
According to a second aspect of the present invention, there is provided a multi-agent algorithm based territorial spatial pattern simulation planning system, comprising:
the first module is used for developing suitability indexes according to ecological environment resource bearing capacity and construction places, and carrying out space element risk identification on a research area to obtain space element risk identification data;
the second module is used for extracting a distribution range of a suitability space built in the research area according to the space element risk identification data, using the distribution range as a base map for simulating the town development boundary, and acquiring a preliminary planning result of the town development boundary in the research area based on a cellular automata model;
and the third module is used for inputting the space element risk identification data and the preliminary planning result into the multi-agent model for scene optimization and outputting the space pattern planning result of the research area.
Compared with the prior art, the territorial spatial pattern simulation planning method and the territorial spatial pattern simulation planning system based on the multi-agent algorithm, which are provided by the embodiment of the invention, can obtain the following beneficial effects:
(1) when the risk of the space elements is identified, the idea of developing double evaluations of suitability indexes by using the bearing capacity of ecological environment resources and a construction place is adopted, so that more accurate space element risk identification data can be obtained compared with the existing manual evaluation division mode or single-dimensional evaluation division mode.
(2) When the cellular automaton is used for cellular conversion, the three-region three-line rule can be integrated, so that the territorial space planning compilation can be effectively carried out aiming at the three-region three-line rule, and the technical requirements of the current territorial space planning conversion period on the space pattern planning can be responded in time.
(3) When the spatial pattern is divided, the spatial suitability identification, the cellular automata model and the multi-agent decision model are combined, so that the coordination during the division is good, the three-region three-wire division model with the spatial toughness is integrally constructed, and the division result is stable.
(4) Because a new thought and implementation means for three-region three-line demarcation can be provided for the territorial space planning of different levels, the territorial space planning and planning practice can be effectively promoted.
Drawings
FIG. 1 is a schematic flow chart of a method for simulating and planning a territorial spatial pattern based on a multi-agent algorithm according to an embodiment of the present invention;
FIG. 2 is a conceptual diagram of a cell automation simulation process provided in an embodiment of the invention;
FIG. 3 is a schematic diagram of the actual operation of the simulation process of the cell robot provided in the embodiment of the present invention;
FIG. 4 is a schematic structural diagram of a network learning framework of a constrained Boltzmann machine according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of three-zone three-wire Agent game coordination provided by an embodiment of the invention;
fig. 6 is a structural diagram of a territorial space planning method for the territorial space pattern provided by the embodiment of the invention;
FIG. 7 is a schematic structural diagram of a simulated planning system for a territorial spatial pattern based on a multi-agent algorithm according to an embodiment of the present invention;
fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
The territorial space planning is an important content for the construction of an ecological civilization system and is a basic basis for various development, protection and construction activities. Particularly, at present, the three-region three-line intelligent division becomes important research content in the field of territorial space planning, about the division of three regions of town space, agricultural space and ecological space, and the division of three boundary lines of ecological protection red line, permanent basic farmland protection red line and town development boundary. Wherein, the three regions are mainly used for dividing the main function, and the three lines are mainly used for rigid control of the boundary. At present, methods such as system dynamics, cellular automata, multi-agent, neural network and the like are mainly used for dynamic monitoring of land utilization, evolution of land utilization and urban expansion simulation at home and abroad. The cellular automata and multi-agent technology becomes a widely used scene analysis simulation technology, and a new means is provided for dynamic analysis and simulation prediction of city expansion.
From the description of city growth morphology and law, Cellular Automata (CA) has been widely used in city space growth simulation as a tool for time-space dynamic simulation of complex systems. The key of the space dynamic simulation of the cellular automata is to determine a cellular conversion rule, and the determination method of the conversion rule mainly comprises an artificial neural network method, a logistic regression method, a multi-factor evaluation method and a multi-agent model method.
Cellular automata are only based on the interaction of land units, and are gradually insufficient in the specific application process, so that a large number of improved models are proposed, such as a SLUTH model, a CLUE-S model, a CA-Markov model, a constrained cellular automata model and the like. However, even though a great deal of improved models are proposed, a perfect intelligent demarcation technical route of the three-region three-wire cellular automaton is not formed yet. A new technical method is proposed for combining the professional knowledge of the territory and the urban planning to provide new intelligent kinetic energy for the territory space planning and solve new problems, scientifically guide the work of the territory space planning and effectively guide the work of the territory space planning and compiling, and therefore a territory space planning simulation planning method is urgently needed.
Based on the above requirements, as shown in fig. 1, an embodiment of the present invention provides a method for simulating and planning a territorial spatial pattern based on a multi-agent algorithm, including:
102, extracting a distribution range of a suitability space built in a research area according to the space element risk identification data, taking the distribution range as a base map for simulating a town development boundary, and acquiring a preliminary demarcation result of the town development boundary in the research area based on a cellular automata model;
and 103, inputting the space element risk identification data and the preliminary planning result into a multi-agent model for scene optimization, and outputting a space pattern planning result of the research area.
In the step 101, basic data of a simulation spatial pattern corresponding to a research area may be obtained, and then two evaluation ideas of suitability indexes are developed according to the bearing capacity of ecological environment resources and a construction place, so as to develop risk identification of spatial elements in the research area. The results of the risk identification of the space elements comprise three types, namely ecological suitability, agricultural suitability and construction suitability, namely division of three regions. The basic data refers to data for describing land use conditions, such as distribution of current land use, use information at the time of historical construction, and the like.
In step 101, the distribution ranges of the spaces corresponding to the ecological suitability, agricultural suitability, and construction suitability of each exploitable land can be obtained. The grid base map data corresponding to the research area may be further divided into a development-prohibited area and a development-allowable area in accordance with the ratio of the development-prohibited construction elements.
In the step 102, after the distribution range of the construction suitability space is extracted, the distribution range can be used as a basis for simulating the town development boundary, that is, as a base map, so that the preliminary delimiting of the town development boundary is completed by combining the cellular automaton model to simulate the town development boundary. Wherein the town development boundary is a control line defined by three lines. The other two types of the three lines are ecological protection red lines and permanent basic farmland protection red lines respectively. In addition, the cellular automaton can further plan the construction land in the base map. Through the step 102, some non-construction land grids can be converted into construction land grids, so that the construction land amount in the base map reaches the upper limit of the built land in the planning time period.
The spatial element risk identification data obtained in the step 101 and the preliminary planning result obtained in the step 102 may be used as a basis for a spatial pattern partitioning decision. In the spatial pattern division, different types of behavior bodies may have division differences with respect to the same grid, for example, for a certain grid, a natural resource department may divide the grid into ecological protection according to its own division standard, an agricultural rural department may divide the grid into farmland protection, and other behavior bodies may have different divisions, so that the cross-overlapped grids have division differences. Therefore, in the step 103, by inputting the two into the multiple intelligent agent model, the different types of behavior subjects can be played, so that the problem of three-region three-line demarcation pattern spot difference is solved.
The method provided by the embodiment of the invention can bring the following beneficial effects:
(1) when the risk of the space elements is identified, the idea of developing double evaluations of suitability indexes by using the bearing capacity of ecological environment resources and a construction place is adopted, so that more accurate space element risk identification data can be obtained compared with the existing manual evaluation division mode or single-dimensional evaluation division mode.
(2) When the cellular automaton is used for cellular conversion, the three-region three-line rule can be integrated, so that the territorial space planning compilation can be effectively carried out aiming at the three-region three-line rule, and the technical requirements of the current territorial space planning conversion period on the space pattern planning can be responded in time.
(3) When the spatial pattern is divided, the spatial suitability identification, the cellular automata model and the multi-agent decision model are combined, so that the coordination during the division is good, the three-region three-wire division model with the spatial toughness is integrally constructed, and the division result is stable.
(4) Because a new thought and implementation means for three-region three-line demarcation can be provided for the territorial space planning of different levels, the territorial space planning and planning practice can be effectively promoted.
Based on the content of the foregoing embodiment, as an optional embodiment, the process of developing a suitability index according to the ecological environment resource bearing capacity and the construction area, and performing spatial element risk identification on the research area to obtain spatial element risk identification data further includes: acquiring various basic data of a research area, and unifying coordinates and formats of the various basic data to obtain a basic database; converting map data in a research area into a grid base map, and identifying data in a basic database to the grid base map; for any grid in the identified grid base map, if the area of the development-forbidden construction elements in the grid is larger than a preset threshold value, the grid is used as a development-forbidden land, and if the area of the development-forbidden construction elements in the grid is not larger than the preset threshold value, the grid is used as a development-forbidden land; and respectively dividing the distribution ranges of the ecological suitability space, the agricultural suitability space and the construction suitability space in each developable land, and using the distribution ranges as the space element risk identification data.
The acquired basic data of the research area can be data sets such as a research range, a permanent basic farmland, an ecological red line, a digital elevation model, a third national soil survey database, an interest point, population heating power, current farmland distribution, a flood storage area, geological disaster susceptibility, soil thickness, soil texture, gradient, microscopic landform and current construction land of a historical construction area. By performing data processing such as coordinate unification and format unification on the data, a basic database for analysis can be formed.
After the basic database is obtained, the data in the research area can be processed into appropriate grid base map data according to the analysis precision requirement, and all the data in the basic database are marked to the grid base map to form data gridding. Based on the identified grid base map data, the grid can be partitioned into exploitable places and forbidden exploitable places. Specifically, if the area of development-prohibited construction elements such as permanent basic farmlands, ecological redlines, water protection systems, terrain slopes, and the like in any of the divided grids is greater than 30% of the grid area, the grid can be used as a development-prohibited land. Wherein, 30% is the preset threshold.
In the development-prohibited elements, the protective water system may be a water source protective land and a first-and second-level river, and the terrain slope may be a high land area having a slope area of more than 25%. In addition, in actual practice, for the sake of easy distinction, the attribute value of the mesh assignment corresponding to the development inhibition point (that is, the type value) may be 0, and the attribute value of the mesh assignment corresponding to the development inhibition point may be 1.
After the above process is completed, the grids with the grid attribute value of 1 can be further subdivided, and evaluation can be specifically performed according to the resource environment bearing capacity and the technical guideline (trial) for evaluating the suitability for developing the homeland space issued by the department of natural resources, so that the distribution ranges of three types of suitability spaces, namely ecological suitability, agricultural suitability and construction suitability, are partitioned.
Based on the above description of the embodiments, as an alternative embodiment, the process of obtaining the preliminary delimiting result of the town development boundary in the research area based on the cellular automaton model further includes: based on the map patches in the national survey database, extracting the urban construction land in the research area as an initial land use state of the cellular automata for simulating urban development boundary expansion; determining the total land scale amount of the urban planning according to the existing planning scale, population scale trend, economic development trend and land use growth trend, and determining the scale of the construction land of the central urban area and each village and town within the planning period based on the total land scale amount and the construction land configuration function conducted in multiple levels; and determining the grids covered by the construction land of the town under the current situation based on the scale of the construction land and the initial land state, taking the grids as the initial state, inputting the initial state into a recursive function of a cellular automaton model, calculating based on a conversion rule to obtain a new evaluation value of each grid of the non-construction land, replacing the last iteration evaluation value in the initial state, repeating the cell change iteration process until the total construction land amount reaches the planned construction scale, and obtaining a preliminary planning result.
The above process of determining the scale of the construction land for planning the time limit of the central urban area and each town may refer to the following formulas:
in the above formula, SkRepresents the newly added construction land in the central urban area of the kth network quality, and k belongs to [1, num]. Where num denotes the total number of grids in the investigation region, XkLand use distribution coefficient, F, representing the kth gridkShowing the proportion of the construction land in the central city region of the kth grid-lattice to the county region, S showing the newly added construction land in the central city region of all grid-lattices, XnkRepresenting the distribution coefficient of the land for the nth cell after the kth grid is divided into cells, n representing the total number of the cells after the kth grid is divided, fkThe current land area, a, is established for the central city of the kth gridkThe total amount of construction site for the current situation of the kth grid.
In the above process, after the scale of the construction site and the initial site status are determined, the grid covered with the construction site of the present town may be determined and referred to as T0 as the initial status. The land for the T0 state is input into a recursive function of the cellular automata model, and a new evaluation value (namely a new value) of each non-construction land grid is calculated based on the conversion rule to replace the original evaluation value in the T0 state. And (3) carrying out cell state change iteration of a plurality of rounds, namely a state change process from the non-construction land to the construction land of the plurality of rounds until the total construction land amount reaches the planned construction scale. It should be noted that, in order to follow the principle of flexible planning, a coefficient of 1.2 may be multiplied by the planned construction scale to serve as the final planned construction scale. Fig. 2 is a conceptual diagram of a cell automatic simulation process.
Based on the contents of the above-described embodiment, as an alternative embodiment, the process of calculating the new evaluation value of each non-construction land grid based on the conversion rule is calculated by any one of the following two formulas:
valuenew＝∑ridvaluerid×typerid；
valuenew＝0.4×∑ridvaluerid×typerid+0.6×valuebefore；
wherein, for any non-construction land grid, valuenewIndicating a new evaluation value obtained by calculating the non-construction land grid, rid indicating the grid number of the neighboring grid around the non-construction land grid, valueridAn evaluation value, type, representing a grid adjacent to the non-construction land gridridIndicating whether the neighboring grid around the non-construction land grid is a construction land, valuebeforeAnd represents the evaluation value of the last iteration of the non-construction land grid.
The two formulas are two conversion rules, the new suitability evaluation value of the central grid in rule 1 is calculated by the sum of products of values and types of surrounding adjacent grids, and the new suitability evaluation value of the central grid in rule 2 is calculated by the sum of products of values and types of surrounding adjacent grids and the original suitability evaluation value of the central grid, namely the evaluation value of the last iteration.
Based on the content of the foregoing embodiment, as an optional embodiment, the above iterative process of cell change is repeated until the total construction area reaches the planned construction scale, and the process of obtaining the preliminary planning result further includes: adjusting internal parameters of the cellular automaton model so that the internal parameters increase at different rates; wherein each adjustment of the internal parameters corresponds to an iterative process; for any current iteration process, sorting the new evaluation values of each non-construction land grid obtained by calculation in the current iteration process from large to small, selecting the non-construction land grids with the preset percentage to be converted into construction land, judging whether the total construction land quantity formed after conversion reaches the planned construction scale or not, ending the iteration process if the total construction land quantity formed after conversion reaches the planned construction scale, repeating the processes of internal parameter adjustment, evaluation value calculation, evaluation value sorting and non-construction land grid conversion until the total construction land quantity formed after conversion reaches the planned construction scale if the total construction land quantity formed after conversion does not reach the planned construction scale, and outputting a preliminary planning result.
When the non-construction land grids with the preset percentage are converted into construction lands before selection, the type value of the non-construction land grids can be changed from 0 to 1. In addition, the planned construction scale refers to an upper limit of the construction land within a preset time period, such as the upper limit of the construction land of 2035 years. The output preliminary delimiting result is the result Tn of using the ground boundary to increase, and n is the recursion times. The calculation process of the two conversion rules can refer to fig. 3, and fig. 3 is a schematic diagram of the actual operation of the automatic simulation process of the cell.
Based on the content of the embodiment, as an optional embodiment, the multi-Agent model comprises an Agent decision structure, an Agent decision library and Agent decision behavior expression; the Agent decision structure comprises a behavior subject set participating in space development decision and ecological protection decision, an interaction subject set among behavior subjects in the space development or ecological process, a state of the behavior subjects, a weight set of the behavior subjects, a set of space decision actions implemented by the behavior subjects in the space development process and probability estimation results of the behavior subjects on action strategies of other behavior subjects in the model; the Agent decision library comprises decision rules, and the decision rules are obtained based on limited Boltzmann machine training; the Agent decision behavior expression is used for carrying out game among different behavior bodies based on the space element risk identification data and the preliminary demarcation result, and the space pattern demarcation result of the research area is determined according to the game result.
Specifically, the definition Agent decision structure can be expressed as follows, Agent ═ Z, G, s, W, a, B >. Wherein, the Agent represents a behavior main body participating in space development decision and ecological protection decision. Z represents an Agent set, Z ═ Z1, Z2, Z3, Z4, … …, zn }. G is an interaction theme set among the agents in the space development or ecological process, such as traffic zones, ecological protection, development difficulty, farmland protection and the like, and G is { G1, G2, G3, … …, gn }.
s is the state of Agent, and s is < T, tmax, U, W >. Wherein, T is the type of Agent and can be divided into a natural resource department, an ecological environment department, an agricultural rural department, a development enterprise, the public, an environmental protection organization and the like. tmax is Agent interaction time limit, U is decision utility, U is expressed by the product of interaction theme weight and theme value in a combined sum mode, and W is an Agent weight set.
a is a set of space decision actions implemented by various types of agents in the space development process, and can be specifically divided into four categories, namely a space development action, an ecological protection action, a farmland protection action and a land utilization conflict coordination action. The four types of behaviors can specifically comprise various space decision actions such as main body function division, town development strategy, major project construction, space protection corridor, land occupation and compensation balance, index area balance, space elastic reserve and the like. And B is the belief of the Agent and is a probability estimation of other Agent action strategies in the model.
Based on the defined Agent decision structure, the Agent decision rule of the multi-Agent and the deep belief network can be extracted and coupled. The embodiment of the invention couples a multi-Agent system theory and an Agent decision rule intelligent algorithm under a deep signaling network learning framework. In addition, the depth signaling network adopted by the embodiment of the invention is specifically a limited boltzmann machine which is used as a basic modeling unit.
The network learning framework of the restricted boltzmann machine has 2 layers in common, as shown in fig. 4. The first layer is a visual layer (V) also called an input layer, and consists of m visual nodes, and the emphasis point of the embodiment of the invention mainly comprises each factor in DX concentration. The second layer is a hidden layer (H), namely a feature extraction layer, mainly comprises a set of decision rules of various types and consists of n hidden nodes.
For ease of understanding, the limiting boltzmann machine will now be described as follows. Specifically, the constraint boltzmann machine is a deep learning model with energy, and the joint configuration energy of the constraint boltzmann machine in a specific state (v, h) can be represented by the following formula:
where E represents the joint configuration energy in a particular state (v, h), and v and h represent given parameters. And θ is a parameter set, a represents the bias of the visible layer node, b represents the bias of the hidden layer node, and W represents a weight set. The underlined v, a and b indicate the mean values under the set.
With respect to limiting the boltzmann machine to a particular state, a joint probability distribution for a given parametric model (v, h) can be obtained:
p(v，h|θ)＝e-E＝(v，h|θ)/Z(θ)；
wherein Z (theta) is Σv，he-E＝(v，h|θ)Is a normalization factor. In addition, under the premise that the visible layer (hidden layer) is known, each node of the hidden layer (visible layer) is independent from each other and does not affect each other, and the conditional distribution between the visible layer and the hidden layer can refer to the following formula:
where i and j represent different nodes, viGiven parameter, h, representing the ith node of the visible layerjA given parameter representing the jth node of the hidden layer.
In addition, given the state of the nodes of the visual layer, the conditional probability distribution of the jth node of the hidden layer can refer to the following formula:
P(hj＝1|v)＝σ(bj+∑iviwij)；
where l represents the state of a given visual level node, bjIndicating the j section of the hidden layerOffset of the point, wijAnd the ith node of the visible layer and the jth node of the hidden layer correspond to the weight.
Also, given the state of a hidden layer node, the conditional probability distribution of the ith node of the visible layer can be referenced by the following formula:
P(vj＝1|h)＝σ(aj+∑iviwij)；
wherein, ajAnd representing the bias of the jth node of the visual layer, wherein the sigma (x) function is an activation function of sigmoid, and when x is less than 0, the function value is 0. When x is greater than 0, the function value is 1.
Based on the content of the embodiment, as an optional embodiment, the decision rule is obtained by training based on training data, wherein the training data comprises a land utilization type, a comprehensive factor variable and a decision rule set of a corresponding land type under the action of the comprehensive factor variable; accordingly, the training process of the decision rule further comprises: initializing the number of layers, the number of nodes of each layer, the learning rate, the iteration period, the connection weight matrix and the bias matrix in the restricted Boltzmann machine, and initializing the structural parameters in the behavior body; importing a comprehensive factor variable for forward propagation to update the structural parameters in the behavior body; importing a comprehensive factor variable for back propagation and a decision rule set of corresponding land types under the action of the comprehensive factor variable for back propagation, and adjusting the acquired limited Boltzmann machine based on an error back propagation algorithm; and importing a comprehensive factor variable for testing and a decision rule set of the corresponding land type under the action of the comprehensive factor variable for testing to test the accuracy of the trained Agent decision library, changing the network internal structure in the limited Boltzmann machine if the test result does not meet the preset condition, retraining, repeating the training process until the test result meets the preset condition, and outputting the final Agent decision library.
In the above process, a data structure of the training data may be first constructed, and specifically, Dt ═ Dc, Dx, Dy }, D may be specifically constructedcE, a, U. Wherein Dc represents the land utilization type, and Dx is a comprehensive factorAnd the sub-variable Dy is a decision rule set of corresponding land types under the action of the Dx. E, A and U respectively represent ecological land, agricultural land and urban land, and specifically respectively correspond to an ecological suitability space, an agricultural suitability space and a construction suitability space.
Based on the constructed data structure, the restricted Boltzmann machine can be trained to obtain the decision rule of each type of Agent. Before training, the historical current construction land data processed in the previous step, the collected Dx data and the relevant decision rule (consisting of several Dy) of the corresponding Agent can be used as a training set sample _ set (Dr _ x, Dr _ y, Dn _ x, Dn _ y, test _ x and test _ y) and input into a model for training. In combination with the above training process, further, the following steps may be referred to for a specific training process:
(1) firstly, initializing the number L of layers, the number N of nodes of each layer, a learning rate mu, an iteration period k, a connection weight matrix W and a bias matrix b in a limited Boltzmann machine;
(2) next, the behavior Agent is initializeduThe structural parameter of (1);
(3) introducing Dr _ x and Dr _ training restricted Boltzmann machine and updating AgentuStructural parameters of (a);
(4) introducing Dn _ x and Dn _ y, and fine-tuning the acquired restricted Boltzmann machine by adopting an error back propagation algorithm;
(5) importing test _ x and test _ y, and testing the accuracy of the trained multi-agent decision rule base based on the limited Boltzmann machine;
(6) if the tested accuracy does not meet the requirement, returning to the initialization state of the limited Boltzmann machine, changing the internal structure of the limited Boltzmann machine, and restarting the training process;
(7) and finally, outputting a decision rule base after the accuracy of the deep neural network output by the test meets the requirement.
Based on the content of the above embodiments, as an optional embodiment, the comprehensive factor variables include an ecological suitability evaluation level, an agricultural suitability evaluation level, a town suitability evaluation level, a subject function zoning factor, a town development strategy factor, a major project construction factor, a space protection corridor factor, a cultivated land occupation balance factor, an index area balance and a space elasticity reserve.
It should be noted that the spatial protection corridor factor may include an ecological corridor, the cultivated land occupation balance factor may include the total cultivated land balance, the index area balance may include the urban level protection/development index cooperation, the rural level protection/development index cooperation, and the spatial elastic reserve mainly refers to the reserved space as the subsequent protection or development use.
Based on the content of the foregoing embodiment, as an optional embodiment, the spatial pattern partitioning process for Agent decision behavior expression further includes: judging whether various behavior main bodies can obtain corresponding decision rules from an Agent decision library in the current state or not based on the space element risk identification data and the preliminary planning result; if various behavior bodies can obtain corresponding decision rules from the Agent decision library in the current state, obtaining the corresponding decision rules and determining the type of the decision behavior, and if various behavior bodies cannot extract the corresponding decision rules from the Agent decision library in the current state, obtaining the decision rules from the newly-set decision rules; and triggering the space decision actions of various behavior main bodies based on the obtained decision rule so as to carry out game interaction among the various behavior main bodies, repeating the game interaction process until the game is balanced, and outputting the space pattern division result of the research area according to the game result.
Specifically, in the land utilization process of the preset time period Ti, whether various behavior subjects can obtain corresponding decision rules from the decision rule base in the current state can be judged according to the space element risk identification data and the preliminary planning result. If the decision-making behavior can be obtained, extracting a corresponding rule, and judging which type of decision-making behavior belongs to space development, ecological protection and farmland protection. And if the rule can not be obtained, obtaining the rule from the newly-set decision rule, and further triggering the space decision action of various behavior subjects through the rule. And repeating the steps until the game is balanced. By continuously repeating the interaction process among different behavior bodies, the problem of three-region three-line drawing spot difference can be solved, and the game coordination schematic diagram of the three-region three-line Agent can refer to fig. 5. With reference to the content of the foregoing embodiment, reference may be made to fig. 6 for a structure diagram of a planning method corresponding to a flow of a territorial space planning method oriented to a territorial space pattern.
Based on the content of the above embodiment, the embodiment of the present invention provides a multi-agent algorithm based geospatial layout simulation planning system, which is used for executing the multi-agent algorithm based geospatial layout simulation planning method provided in the above method embodiment. Referring to fig. 7, the system includes:
a first module 701, which is used for developing suitability indexes according to ecological environment resource bearing capacity and construction places, and carrying out space element risk identification on a research area to obtain space element risk identification data;
a second module 702, configured to extract a distribution range of a suitability space built in a research area according to the spatial element risk identification data, use the distribution range as a base map for simulating a town development boundary, and obtain a preliminary planning result of the town development boundary in the research area based on a cellular automata model;
a third module 703, configured to input the spatial element risk identification data and the preliminary delineation result into the multi-agent model for scenario optimization, and output a spatial configuration delineation result of the research area.
Fig. 8 illustrates a physical structure diagram of an electronic device, and as shown in fig. 8, the electronic device may include: a processor (processor)810, a communication Interface 820, a memory 830 and a communication bus 840, wherein the processor 810, the communication Interface 820 and the memory 830 communicate with each other via the communication bus 840. The processor 810 may call logic instructions in the memory 830 to perform the following method: developing suitability indexes according to ecological environment resource bearing capacity and construction places, and carrying out space element risk identification on the research area to obtain space element risk identification data; extracting a distribution range of a suitability space built in a research area according to the space element risk identification data, taking the distribution range as a base map for simulating a town development boundary, and acquiring a preliminary delimiting result of the town development boundary in the research area based on a cellular automata model; and inputting the space element risk identification data and the preliminary planning result into a multi-agent model for scene optimization, and outputting a space pattern dividing result of the research area.
In addition, the logic instructions in the memory 830 may be implemented in software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, an electronic device, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. A territorial spatial pattern simulation planning method based on a multi-agent algorithm is characterized by comprising the following steps:
and developing suitability indexes according to the bearing capacity of ecological environment resources and a construction place, and carrying out space element risk identification on the research area to obtain space element risk identification data.
Extracting a distribution range of a suitability space built in the research area according to the space element risk identification data, taking the distribution range as a base map for simulating town development boundaries, and acquiring a preliminary demarcation result of the town development boundaries in the research area based on a cellular automata model;
and inputting the space element risk identification data and the preliminary planning result into a multi-agent model for scene optimization, and outputting a space pattern planning result of the research area.
2. The multi-agent algorithm-based territorial spatial pattern simulation planning method of claim 1, wherein the process of developing suitability indexes according to ecological environment resource bearing capacity and construction places to identify spatial element risks in a research area and obtain spatial element risk identification data further comprises:
acquiring various basic data of the research area, and unifying coordinates and formats of the various basic data to obtain a basic database;
and converting map data in the research area into a grid base map, and identifying the data in the basic database to the grid base map.
For any grid in the identified grid base map, if the area of development-forbidden construction elements in the grid is larger than a preset threshold value, taking the grid as a development-forbidden land, and if the area of development-forbidden construction elements in the grid is not larger than the preset threshold value, taking the grid as a developable land;
and respectively dividing the distribution ranges of the ecological suitability space, the agricultural suitability space and the construction suitability space in each developable land, and using the distribution ranges as the risk identification data of the space elements.
3. The multi-agent algorithm-based territorial spatial pattern simulation planning method of claim 1 or 2, wherein the process of obtaining preliminary delineation results of the boundaries of the development of the towns in the research area based on the cellular automaton model further comprises:
based on the map patches in the national survey database, extracting the urban construction land in the research area as an initial land use state of the cellular automata for simulating urban development boundary expansion;
determining a total land gauge modulus for urban planning according to the existing planning scale, population scale trend, economic development trend and land use growth trend, and determining the scale of the construction land in the planning period of a central urban area and each village and town based on the total land gauge modulus and a construction land configuration function conducted in multiple levels;
and determining grids covered by the construction land of the town under the current situation based on the construction land scale and the initial land state, taking the grids as an initial state, inputting the initial state into a recursive function of the cellular automaton model, calculating based on a conversion rule to obtain a new evaluation value of each non-construction land grid, replacing the last iteration evaluation value in the initial state, repeating the cell change iteration process until the total construction land amount reaches the planned construction scale, and obtaining the initial planning result.
4. The multi-agent algorithm-based territorial spatial pattern simulation planning method of any one of claims 1 to 3, wherein the process of calculating the new evaluation value of each non-construction land grid based on the conversion rule is calculated by any one of the following two formulas:
valuenew＝∑ridvaluerid×typerid；
valuenew＝0.4×∑ridvaluerid×typerid+0.6×valuebefore；
wherein, for any non-construction land grid, valuenewRepresenting a new evaluation value obtained after the calculation of any one of the grids for non-construction land, rid representing the grid number and value of the adjacent grid around the grid for non-construction landridAn evaluation value, type, representing a grid adjacent to the grid of any one of the non-construction landsridIndicating whether the neighboring grid around the non-construction land grid is a construction land, valuebeforeAnd representing the evaluation value of the last iteration of any one of the non-construction land grids.
5. The multi-agent algorithm-based territorial spatial pattern simulation planning method of claim 3, wherein said repeating said iterative process of cell change until the total construction area reaches the planned construction scale, and obtaining said preliminary planning result further comprises:
adjusting internal parameters of the cellular automaton model so that the internal parameters increase at different rates; wherein each adjustment of the internal parameters corresponds to an iterative process;
for any current iteration process, sorting the new evaluation values of each non-construction land grid obtained by calculation in the current iteration process from large to small, selecting the non-construction land grids with the preset percentage to be converted into construction land, judging whether the total construction land quantity formed after conversion reaches the planned construction scale or not, ending the iteration process if the total construction land quantity formed after conversion reaches the planned construction scale, repeating the processes of internal parameter adjustment, evaluation value calculation, evaluation value sorting and non-construction land grid conversion until the total construction land quantity formed after conversion reaches the planned construction scale if the total construction land quantity formed after conversion does not reach the planned construction scale, and outputting a preliminary planning result.
6. The multi-Agent algorithm-based geospatial spatial pattern simulation planning method of claim 1 wherein the multi-Agent model comprises an Agent decision structure, an Agent decision library and an Agent decision behavior expression.
The Agent decision structure comprises a behavior subject set participating in space development decision and ecological protection decision, an interaction subject set among behavior subjects in the space development or ecological process, a state of the behavior subject, a weight set of the behavior subject, a set of space decision actions implemented by the behavior subject in the space development process and probability estimation results of the behavior subject on action strategies of other behavior subjects in the model;
the Agent decision library comprises decision rules, and the decision rules are obtained based on limited Boltzmann machine training; and the Agent decision behavior expression is used for carrying out game among different behavior bodies based on the space element risk identification data and the preliminary demarcation result, and determining the space pattern demarcation result of the research area according to the game result.
7. The multi-agent algorithm-based territorial spatial pattern simulation planning method of claim 6, wherein the decision rules are trained based on training data, the training data comprising a land utilization type, a comprehensive factor variable and a set of decision rules for a corresponding land type under the action of the comprehensive factor variable; accordingly, the training process of the decision rule further comprises:
initializing the number of layers, the number of nodes of each layer, the learning rate, the iteration period, the connection weight matrix and the bias matrix in the restricted Boltzmann machine, and initializing the structural parameters in the behavior body;
importing a comprehensive factor variable for forward propagation to update the structural parameters in the behavior body;
importing a comprehensive factor variable for back propagation and a decision rule set of corresponding land types under the action of the comprehensive factor variable for back propagation, and adjusting the acquired limited Boltzmann machine based on an error back propagation algorithm;
and importing a comprehensive factor variable for testing and a decision rule set of the corresponding land type under the action of the comprehensive factor variable for testing to test the accuracy of the trained Agent decision library, changing the network internal structure in the limited Boltzmann machine if the test result does not meet the preset condition, retraining, repeating the training process until the test result meets the preset condition, and outputting the final Agent decision library.
8. The multi-agent algorithm-based territorial spatial pattern simulation planning method of claim 7, wherein the comprehensive factor variables comprise ecological suitability evaluation level, agricultural suitability evaluation level, town suitability evaluation level, subject function compartmentalization factor, town development strategy factor, major project construction factor, space protection corridor factor, cultivated land occupation balance factor, index area balance and space elastic reserve.
9. The multi-Agent algorithm-based territorial spatial pattern simulation planning method of claim 6, wherein the spatial pattern partitioning process expressed by the Agent decision behaviors further comprises:
judging whether various behavior bodies can obtain corresponding decision rules from an Agent decision library in the current state or not based on the space element risk identification data and the preliminary planning result;
if various behavior bodies can obtain corresponding decision rules from the Agent decision library in the current state, obtaining the corresponding decision rules and determining the type of the decision behavior, and if various behavior bodies cannot extract the corresponding decision rules from the Agent decision library in the current state, obtaining the decision rules from the newly-set decision rules;
and triggering the space decision actions of various behavior main bodies based on the obtained decision rule so as to carry out game interaction among the various behavior main bodies, repeating the game interaction process until the game is balanced, and outputting the space pattern division result of the research area according to the game result.
10. A territorial spatial pattern simulation planning system based on a multi-agent algorithm is characterized by comprising:
the first module is used for developing suitability indexes according to ecological environment resource bearing capacity and construction places, and carrying out space element risk identification on a research area to obtain space element risk identification data;
the second module is used for extracting a distribution range of a suitability space built in the research area according to the space element risk identification data, using the distribution range as a base map for simulating a town development boundary, and acquiring a preliminary planning result of the town development boundary in the research area based on a cellular automaton model;
and the third module is used for inputting the space element risk identification data and the preliminary planning result into a multi-agent model for scene optimization and outputting a space pattern planning result of the research area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110201118.0A CN112884226A (en) | 2021-02-23 | 2021-02-23 | Multi-agent algorithm-based territorial spatial pattern simulation planning method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110201118.0A CN112884226A (en) | 2021-02-23 | 2021-02-23 | Multi-agent algorithm-based territorial spatial pattern simulation planning method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112884226A true CN112884226A (en) | 2021-06-01 |
Family
ID=76053785
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110201118.0A Withdrawn CN112884226A (en) | 2021-02-23 | 2021-02-23 | Multi-agent algorithm-based territorial spatial pattern simulation planning method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112884226A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113361774A (en) * | 2021-06-07 | 2021-09-07 | 四川农业大学 | Ecological space intensive utilization optimization method and system |
| CN113449936A (en) * | 2021-08-31 | 2021-09-28 | 北京市城市规划设计研究院 | Urban space evolution simulation prediction method, device, electronic equipment and storage medium |
| CN113792942A (en) * | 2021-11-15 | 2021-12-14 | 江西省国土空间调查规划研究院 | Dynamic simulation prediction system for ecological effect development and utilization of homeland space |
| CN113935532A (en) * | 2021-10-18 | 2022-01-14 | 广东工业大学 | Urban land space optimization configuration method based on pareto frontier degradation |
| CN113988574A (en) * | 2021-10-22 | 2022-01-28 | 广州市城市规划勘测设计研究院 | Ecological protection red line marking method and device, terminal equipment and storage medium |
| CN114549223A (en) * | 2022-02-24 | 2022-05-27 | 安徽耘联大数据有限公司 | Grid cultivated land protection method and system based on big data |
| CN114676568A (en) * | 2022-01-17 | 2022-06-28 | 中国地质大学（北京） | Regional geological structure evolution method and device based on cellular automaton |
| CN114925974A (en) * | 2022-04-13 | 2022-08-19 | 清华大学 | Classification gradual double-evaluation method for city and county territory space planning |
| CN115329591A (en) * | 2022-08-29 | 2022-11-11 | 华中科技大学 | City spatial pattern design method, system and computer storage medium |
| CN115099715B (en) * | 2022-08-08 | 2023-04-14 | 北京新兴科遥信息技术有限公司 | Big data-based analysis processing system for three-line planning of territorial space planning |
| CN116563073A (en) * | 2023-04-10 | 2023-08-08 | 四川师范大学 | Self-adapting rural residential point layout method |
| CN118052477A (en) * | 2024-04-16 | 2024-05-17 | 菏泽市国土综合整治服务中心 | Method and system for evaluating ecological restoration effect of comprehensive land space restoration area |
-
2021
- 2021-02-23 CN CN202110201118.0A patent/CN112884226A/en not_active Withdrawn
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113361774A (en) * | 2021-06-07 | 2021-09-07 | 四川农业大学 | Ecological space intensive utilization optimization method and system |
| CN113449936A (en) * | 2021-08-31 | 2021-09-28 | 北京市城市规划设计研究院 | Urban space evolution simulation prediction method, device, electronic equipment and storage medium |
| CN113935532B (en) * | 2021-10-18 | 2024-05-10 | 广东工业大学 | Urban land space optimal configuration method based on pareto front degradation |
| CN113935532A (en) * | 2021-10-18 | 2022-01-14 | 广东工业大学 | Urban land space optimization configuration method based on pareto frontier degradation |
| CN113988574A (en) * | 2021-10-22 | 2022-01-28 | 广州市城市规划勘测设计研究院 | Ecological protection red line marking method and device, terminal equipment and storage medium |
| CN113792942A (en) * | 2021-11-15 | 2021-12-14 | 江西省国土空间调查规划研究院 | Dynamic simulation prediction system for ecological effect development and utilization of homeland space |
| CN113792942B (en) * | 2021-11-15 | 2022-02-08 | 江西省国土空间调查规划研究院 | Dynamic simulation prediction system for ecological effect development and utilization of homeland space |
| CN114676568A (en) * | 2022-01-17 | 2022-06-28 | 中国地质大学（北京） | Regional geological structure evolution method and device based on cellular automaton |
| CN114676568B (en) * | 2022-01-17 | 2022-09-23 | 中国地质大学（北京） | Regional geological structure evolution method and device based on cellular automaton |
| CN114549223A (en) * | 2022-02-24 | 2022-05-27 | 安徽耘联大数据有限公司 | Grid cultivated land protection method and system based on big data |
| CN114925974A (en) * | 2022-04-13 | 2022-08-19 | 清华大学 | Classification gradual double-evaluation method for city and county territory space planning |
| CN115099715B (en) * | 2022-08-08 | 2023-04-14 | 北京新兴科遥信息技术有限公司 | Big data-based analysis processing system for three-line planning of territorial space planning |
| CN115329591B (en) * | 2022-08-29 | 2023-06-20 | 华中科技大学 | Urban space pattern design method, system and computer storage medium |
| CN115329591A (en) * | 2022-08-29 | 2022-11-11 | 华中科技大学 | City spatial pattern design method, system and computer storage medium |
| CN116563073A (en) * | 2023-04-10 | 2023-08-08 | 四川师范大学 | Self-adapting rural residential point layout method |
| CN116563073B (en) * | 2023-04-10 | 2024-04-26 | 四川师范大学 | Self-adapting rural residential point layout method |
| CN118052477A (en) * | 2024-04-16 | 2024-05-17 | 菏泽市国土综合整治服务中心 | Method and system for evaluating ecological restoration effect of comprehensive land space restoration area |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112884226A (en) | Multi-agent algorithm-based territorial spatial pattern simulation planning method and system | |
| Yeh et al. | Simulation of development alternatives using neural networks, cellular automata, and GIS for urban planning | |
| CN106780089B (en) | Permanent basic farmland planning method based on neural network cellular automaton model | |
| CN112163367B (en) | Firefly algorithm and cellular automaton fused city expansion simulation prediction method | |
| CN102750411B (en) | Urban dynamic micro-simulation method based on multi-agent discrete choice model | |
| CN113297174B (en) | Land utilization change simulation method based on deep learning | |
| Yeh et al. | Urban simulation using neural networks and cellular automata for land use planning | |
| CN110619432A (en) | Feature extraction hydrological forecasting method based on deep learning | |
| CN113468803A (en) | Improved WOA-GRU-based flood flow prediction method and system | |
| CN113505510B (en) | Ecological safety pattern recognition method fusing landscape index and random walk model | |
| CN110826244B (en) | Conjugated gradient cellular automaton method for simulating influence of rail transit on urban growth | |
| CN102646164A (en) | Land use change modeling method and system implemented in combination with spatial filtering | |
| Dai et al. | Agent-based model of land system: Theory, application and modelling framework | |
| CN113902580A (en) | Historical farmland distribution reconstruction method based on random forest model | |
| CN112784503A (en) | Gate dam dispatching launching bird habitat suitability calculation method and device | |
| Yang et al. | A multitarget land use change simulation model based on cellular automata and its application | |
| Wagener et al. | Identification and evaluation of watershed models | |
| CN113011455A (en) | Air quality prediction SVM model construction method | |
| CN113449936B (en) | Urban space evolution simulation prediction method, device, electronic equipment and storage medium | |
| CN113240219A (en) | Land utilization simulation and prediction method | |
| CN117610208A (en) | River water quality and quantity prediction method based on urban water system model interaction | |
| CN112749907A (en) | Spatial feature analysis method and device of cultural ecosystem | |
| Subiyanto et al. | Modeling changes in land use using the integration of MLP-NN, CA-Markov models and GIS for settlement development in Tembalang District | |
| CN115759346A (en) | AO algorithm based carbon emission prediction method, device and equipment | |
| Pérez et al. | Model testing and assessment: perspectives from a swarm intelligence, agent-based model of forest insect infestations |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20210601 |