A digital simulation method for village and town settlements based on cellular automata

Abstract

The present invention provides a digital simulation method for village and town settlements based on cellular automata. First, basic data on village and town settlement development is collected, cleaned, and refined. The decisions and behaviors of various stakeholders are dynamically captured through a multi-agent system and incorporated into a preliminary analysis. These basic data and analysis results are input into a cellular automaton based on machine learning, and ultimately the transformation rules of the cellular automaton are learned. Cell states are iteratively updated according to these rules to simulate and record land behavior changes over time. The land use conversion probability of each cell is calculated, and land use change trends are analyzed. Land development projects are randomly selected using a parallel Monte Carlo method, and the total land supply is determined. Finally, a determination is made as to whether the total supply meets a preset supply threshold. If not, iterative adjustments are made. If so, the results are summarized and visualized. This method provides a systematic, intelligent, and efficient solution for complex land management needs.

Description

Technical Field

[0001] The present invention relates to the technical field of settlement simulation, and in particular to a digital simulation method for village and town settlements based on cellular automata. Background Art

[0002] Basic concepts of cellular automata: Cellular automata (CA) is a discrete model that divides space into grids (or cells). The state of a cell is affected by other surrounding cells and the state is updated according to specific rules, thereby simulating the dynamic behavior of complex systems.

[0003] Cellular automata are widely used in many fields, including ecology, meteorology, urban planning, and traffic flow simulation. Through rule-based dynamic evolution, cellular automata can effectively simulate processes such as human activities and land use changes in villages and towns.

[0004] The formation and development of village and town settlements are influenced by many factors, such as geographical environment, economic development, and social structure. Simulations based on cellular automata can integrate these complex factors and conduct dynamic analysis in space and time. Building a cellular automaton simulation system usually includes the following steps: defining the state of cells and their neighborhood relationships, setting transition rules (such as population growth, building expansion, etc.), initializing the system state, and performing time-stepping simulation. The advantage of this method lies in its simplicity and flexibility, which can simulate a variety of scenarios under different conditions. However, the challenge lies in how to accurately set the model parameters and transition rules to ensure the realism and effectiveness of the simulation results.

[0005] The prior art discloses a digital simulation method for the spatial development of village and town settlements, which includes the following steps: step a: collecting basic data on village and town settlement spatial planning, data on adjustment of villagers' willingness, and basic data on the total scale of land supply, incentive and restriction policies; step b: simulating the main behaviors of planners, villagers, governments and developers based on the collected data sets to update planning constraints, villagers' willingness constraints, government willingness constraints and developers' willingness constraints; step c: calculating the land use conversion probability of cells based on the economic and environmental factors of the cells based on the multi-agent model; step d: calculating the land use conversion probability of cells based on the land use during the simulation period; and The paper presents a novel method for simulating the land supply in year t based on the land use conversion trend calculated by the planner. The paper also proposes a method for randomly selecting land transfer project locations using the Monte Carlo method based on the land use conversion probability calculated by the planner. The government selects the scope of the land transfer project. The paper also proposes a method for randomly selecting land transfer project locations using the Monte Carlo method based on the land use conversion probability calculated by the planner. The paper also proposes a method for randomly selecting land transfer project locations using the Monte Carlo method based on the land use conversion probability calculated by the planner. The paper also proposes a method for randomly selecting land transfer project locations using the Monte Carlo method based on the land use conversion probability calculated by the planner. The paper also proposes a method for randomly selecting land transfer project locations using the Monte Carlo method based on the land use conversion probability calculated by the government ... However, the traditional cellular automata in this technology rely on static, pre-set rule sets and cannot flexibly respond to dynamically changing socio-economic environments and policy changes. Summary of the Invention

[0006] In order to overcome the shortcomings of the existing technology, the purpose of the present invention is to provide a digital simulation method for village and town settlements based on cellular automata, which can provide a systematic, intelligent and efficient solution for complex land management needs.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A digital simulation method for village and town settlements based on cellular automata, comprising:

[0009] Collect basic data on village and town settlement development, and perform cleaning, filtering and pre-processing;

[0010] Use a multi-agent system to dynamically capture the decision-making and behavior data of various stakeholders and incorporate them into the preliminary simulation stage to obtain simulation output;

[0011] Analyzing the basic data and the simulation output using a machine learning algorithm to automatically learn and generate conversion rules of a cellular automaton;

[0012] The cell states in the cellular automaton model are updated according to the transformation rules, and simulation iterations are performed to record the land use state changes at each time step;

[0013] Based on the updated cellular automaton model and the preset economic and environmental factors, the land use conversion probability of each cell is calculated;

[0014] Analyze the changes in land use status over multiple time steps to obtain the land use change trend;

[0015] Based on a parallel Monte Carlo method, randomly selecting land development projects according to the land use change trend and the land use conversion probability, and determining the total land supply according to the land use change trend and the land use conversion probability;

[0016] Determine whether the total land supply meets the preset supply threshold. If not, return to the step "based on the parallel execution of the Monte Carlo method, randomly select land development projects according to the land use change trend and the land use conversion probability, and determine the total land supply according to the land use change trend and the land use conversion probability". If so, summarize the simulation results of the multi-agent system and the cellular automaton model and perform visual display.

[0017] Preferably, the basic data include: village and town spatial geographic information data, planning data, villagers' willingness data, total land supply, incentive and restriction policies.

[0018] Preferably, the stakeholders include planners, villagers, government and developers.

[0019] Preferably, a multi-agent system is used to dynamically capture the decision-making and behavior data of various stakeholders and incorporate them into the preliminary simulation phase to obtain simulation outputs, including:

[0020] Create a corresponding agent for each stakeholder and define the attributes, behavior rules and decision-making mechanism of each agent:

[0021] Defining a simulation environment and setting interaction rules between the agents; the simulation environment includes land geographic information, land use types, resource allocation, and policy constraints;

[0022] Collect historical data and current environment data to initialize the state and environment parameters of the agent;

[0023] Start the simulation of the agent so that the agent makes decisions based on the behavior rules and the state of the environment:

[0024] At each time step, record the agent's decision and behavior data;

[0025] Generate preliminary simulation output based on the decision and behavior data of the agent.

[0026] Preferably, the attributes of the intelligent agent include: basic information, state variables and environmental information; the behavioral rules include: decision rules, interaction rules and adaptive rules; the decision mechanism includes: decision sub-model, information processing sub-model and risk assessment sub-model.

[0027] Preferably, the formula of the decision sub-model is: U=f(x1, x2, ..., xn); wherein U is the utility value, and xi is the i-th variable affecting the decision; the formula of the information processing sub-model is: Where P(H|E) is the posterior probability of hypothesis H when evidence E is observed; the calculation formula of the risk assessment sub-model is: Where ER is the expected return, P(i) is the probability of occurrence of each outcome i, and U(i) is the utility value of the corresponding outcome.

[0028] Preferably, the basic data and the simulation output are analyzed using a machine learning algorithm to generate a conversion rule of a cellular automaton, including:

[0029] Collect basic data sets or simulation output data sets related to land use, environmental characteristics, socioeconomic factors, and demographic factors;

[0030] Build a convolutional neural network;

[0031] The convolutional neural network is trained according to the basic data set and the simulated output data set to obtain a trained first classifier and a trained second classifier;

[0032] Cascade the trained first classifier and the second classifier to obtain a classification network;

[0033] Connecting a trained LSTM neural network after the classification network to obtain a conversion rule extraction model;

[0034] The conversion rules are extracted according to the conversion rule extraction model; the conversion rules include state transition conditions, probability transition rules and time dependency.

[0035] Preferably, the calculation formula for the land use conversion probability of each cell is:

[0036]

[0037] Among them, PY i =j|X i Indicates that cell i has a given feature X i The probability of transitioning to state j under the condition of i is the target state of cell i, X i is the characteristic vector that affects the cell state transformation, β jk : with feature X ik The relevant regression coefficient, representing the feature X ik The degree of influence on the probability of the cell transitioning to state j.

[0038] Preferably, the land use change trend is obtained by analyzing the land use state changes at multiple time steps, including:

[0039] Encoding the data of land use state changes at the multiple time steps into numerical form, and performing missing value processing on the encoded numerical values;

[0040] By comparing the coded values ​​of adjacent time steps, the transition of each land use type and the rate of change of each time step are calculated;

[0041] The land use change trend is determined based on the transition status of each land use type and the change rate of each time step using a time series analysis method.

[0042] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0043] The present invention provides a digital simulation method for village and town settlements based on cellular automata, comprising: collecting basic data on village and town settlement development; using a multi-agent system to dynamically capture the decision-making and behavior data of various stakeholders, and incorporating the data into a preliminary simulation phase to obtain simulation output; using a machine learning algorithm to analyze the basic data and the simulation output to generate a conversion rule for the cellular automata; updating the cell state in the cellular automata model according to the conversion rule, performing simulation iterations to record the change in land use state at each time step; calculating the land use conversion probability of each cell based on the updated cellular automata model and preset economic and environmental factors; analyzing the change in land use state at multiple time steps to obtain Land use change trend; based on the parallel execution of the Monte Carlo method, randomly select land development projects according to the land use change trend and the land use conversion probability, and determine the total land supply based on the land use change trend and the land use conversion probability; determine whether the total land supply meets the preset supply threshold, if not, return to the step "based on the parallel execution of the Monte Carlo method, randomly select land development projects according to the land use change trend and the land use conversion probability, and determine the total land supply based on the land use change trend and the land use conversion probability", if so, summarize the simulation results of the multi-agent system and the cellular automaton model and perform a visual display. By comprehensively utilizing multiple technologies, the present invention provides a systematic, intelligent, efficient, and flexible solution that can better meet complex land management needs, support scientific decision-making, and help promote sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 A flow chart of a method provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] Figure 1 A flow chart of the method provided in the embodiment of the present invention is shown in FIG. Figure 1 As shown, the present invention provides a digital simulation method for village and town settlements based on cellular automata, comprising:

[0049] Step 100: Collect basic data on village and town settlement development, and perform cleaning, filtering, and preprocessing;

[0050] Step 200: Utilize the multi-agent system to dynamically capture the decision-making and behavior data of each stakeholder and incorporate them into the preliminary simulation phase to obtain simulation output;

[0051] Step 300: Analyze the basic data and the simulation output using a machine learning algorithm to learn the conversion rules of a cellular automaton;

[0052] Step 400: updating the cell states in the cellular automaton model according to the conversion rule, performing simulation iterations to record the land use state changes at each time step;

[0053] Step 500: Calculate the land use conversion probability of each cell based on the updated cellular automaton model, preset economic factors and environmental factors;

[0054] Step 600: Analyze the land use status changes at multiple time steps to obtain the land use change trend;

[0055] Step 700: Based on the parallel execution of the Monte Carlo method, randomly select land development projects according to the land use change trend and the land use conversion probability, and determine the total land supply according to the land use change trend and the land use conversion probability;

[0056] Step 800: Determine whether the total land supply meets the preset supply threshold. If not, return to step "Based on the parallel execution of the Monte Carlo method, randomly select land development projects according to the land use change trend and the land use conversion probability, and determine the total land supply according to the land use change trend and the land use conversion probability". If so, summarize the simulation results of the multi-agent system and the cellular automaton model and perform visual display.

[0057] Preferably, the basic data include: village and town spatial geographic information data, planning data, villagers' willingness data, total land supply, incentive and restriction policies.

[0058] Specifically, step 100 of this embodiment includes:

[0059] 1. Determine data requirements, including:

[0060] ① Land use data: including the area, distribution and changes of different land use types (such as residential, commercial, agricultural, public facilities, etc.).

[0061] ② Population data: including the total population of villages and towns, population density, age structure, migration patterns, etc.

[0062] ③ Economic data: including local economic activities, income levels, employment rates, industrial structure, etc.

[0063] ④ Environmental data: including topography, climate, soil type, water resources, ecologically sensitive areas, etc.

[0064] ⑤Infrastructure data: including transportation networks, public service facilities (such as schools and hospitals), water supply and drainage systems, etc.

[0065] 2. Obtain data sources, including:

[0066] ① Government statistics: Access statistical yearbooks, census data, and other relevant reports published by local governments or the National Bureau of Statistics to obtain official land use, population, and economic data.

[0067] ② Remote sensing data: Use satellite imagery and aerial photography to obtain data on land use changes, topography, and environmental characteristics. Open-source remote sensing data (such as Landsat and Sentinel) can be used for analysis.

[0068] ③ Geographic Information System (GIS) data: Use GIS platforms (such as ArcGIS and QGIS) to collect and analyze spatial data, including land use, infrastructure, and environmental characteristics.

[0069] ④ Questionnaires and interviews: Design questionnaires or conduct interviews to collect opinions and data on land use, economic activities and social behavior from village and town residents, businesses and government officials.

[0070] ⑤Academic research and literature: Consult relevant academic papers, research reports and case studies to obtain existing research data and analysis results.

[0071] 3. Data collation and processing

[0072] ① Data cleaning: clean the collected data, process missing values, outliers and duplicate data, and ensure the accuracy and consistency of the data.

[0073] ②Data integration: integrate data from different sources, establish a unified data framework, and ensure the comparability and correlation between data.

[0074] ③Data formatting: Organize the data into a format suitable for analysis, such as CSV, Excel, or database format, for subsequent analysis and modeling.

[0075] 4. Data verification and update

[0076] ① Data Verification: Verify the collected data to ensure its source is reliable and the data is accurate. The credibility of the data can be improved by cross-validating data from different sources.

[0077] ② Regular updates: Establish a data update mechanism to regularly collect and update basic data to reflect the latest developments in village and town settlement development.

[0078] 5. Data storage and management

[0079] ① Database management: Use database management systems (such as MySQL, PostgreSQL) to store and manage collected data to ensure data security and accessibility.

[0080] ②Data backup: Back up data regularly to prevent data loss or damage.

[0081] Through the above steps, this example systematically collects basic data on the development of villages and towns, providing a solid data foundation for subsequent simulations and analysis. This data will provide important support for understanding the dynamic changes in villages and towns, the decision-making behavior of stakeholders, and the transformation of land use.

[0082] Preferably, the stakeholders include planners, villagers, government and developers.

[0083] Preferably, a multi-agent system is used to dynamically capture the decision-making and behavior data of various stakeholders and incorporate them into the preliminary simulation phase to obtain simulation outputs, including:

[0084] Create a corresponding agent for each stakeholder and define the attributes, behavior rules and decision-making mechanism of each agent:

[0085] Defining a simulation environment and setting interaction rules between the agents; the simulation environment includes land geographic information, land use types, resource allocation, and policy constraints;

[0086] Collect historical data and current environment data to initialize the state and environment parameters of the agent;

[0087] Start the simulation of the agent so that the agent makes decisions based on the behavior rules and the state of the environment:

[0088] At each time step, record the agent's decision and behavior data;

[0089] Generate preliminary simulation output based on the decision and behavior data of the agent.

[0090] Preferably, the attributes of the intelligent agent include: basic information, state variables and environmental information; the behavioral rules include: decision rules, interaction rules and adaptive rules; the decision mechanism includes: decision sub-model, information processing sub-model and risk assessment sub-model.

[0091] Preferably, the formula of the decision sub-model is: U=f(x1, x2, ..., xn); wherein U is the utility value, and xi is the i-th variable affecting the decision; the formula of the information processing sub-model is: Where P(H|E) is the posterior probability of hypothesis H when evidence E is observed; the calculation formula of the risk assessment sub-model is: Where ER is the expected return, P(i) is the probability of occurrence of each outcome i, and U(i) is the utility value of the corresponding outcome.

[0092] Specifically, step 200 of this embodiment includes:

[0093] 1. Define stakeholders and identify the main stakeholders related to the spatial evolution of village and town settlements, including:

[0094] planners (government departments);

[0095] villagers (residents);

[0096] Government (policy makers);

[0097] Developer (investor).

[0098] 2. Build a multi-agent model, create corresponding agents for each stakeholder, and define their attributes, behavioral rules, and decision-making mechanisms: among them, planners focus on land use planning and policy formulation and have long-term goals; villagers focus on quality of life and land use rights and have short-term needs; the government focuses on public interests and policy implementation and has regulatory functions; developers focus on return on investment and market demand and have profit goals.

[0099] 3. Set the environment and interaction rules

[0100] Define the simulation environment, including land use types, resource allocation, policy constraints, etc., and set the interaction rules between agents, for example:

[0101] Feedback mechanisms between villagers and planners (e.g., solicitation of opinions).

[0102] Policy implementation and market response between government and developers.

[0103] 4. Data collection and initialization

[0104] Collect historical data and current environmental data, initialize the state and environmental parameters of the intelligent agent, and set initial conditions such as land use status, population distribution, economic activities, etc.

[0105] 5. Simulation run: Start the simulation, and the agent makes decisions based on its behavioral rules and the state of the environment. Specifically:

[0106] Planners, aligning land use planning with policy objectives.

[0107] Villagers choose land use methods (such as agriculture or residence) according to their living needs.

[0108] The government will supervise and adjust policies based on their implementation.

[0109] Developers select development projects based on market demand and policy environment.

[0110] 6. Dynamically capture decisions and behaviors. At each time step, record the agent’s decision and behavior data, including:

[0111] The state changes of each agent (such as changes in land use type).

[0112] The results of interactions between agents (such as policy feedback and market response).

[0113] These data will be used for subsequent analysis and model optimization.

[0114] 7. Generate preliminary simulation output. Based on the agent’s decision and behavior data, generate preliminary simulation output, including:

[0115] Land use changes (e.g. changes in the area of ​​different types of land).

[0116] Satisfaction of various stakeholders (such as villagers' satisfaction with their living environment).

[0117] Policy effectiveness evaluation (such as the impact of policies on land development).

[0118] 8. Result analysis and feedback: Analyze the preliminary simulation output, evaluate the decision-making effects of each stakeholder, and adjust the behavior rules and interaction mechanisms of the intelligent agent based on the analysis results to improve the accuracy and adaptability of the model.

[0119] 9. Iterative optimization: Repeat steps 5 to 8 for multiple iterations to gradually optimize the model and ensure the reliability and effectiveness of the simulation results.

[0120] Through the above steps, this embodiment uses a multi-agent system to dynamically capture the decision-making and behavior data of various stakeholders and incorporates them into the preliminary simulation stage. It can effectively simulate the spatial evolution process of village and town settlements and provide a scientific basis for subsequent planning and decision-making.

[0121] Preferably, the basic data and the simulation output are analyzed using a machine learning algorithm to generate a conversion rule of a cellular automaton, including:

[0122] Collect basic data sets or simulation output data sets related to land use, environmental characteristics, socioeconomic factors, and demographic factors;

[0123] Build a convolutional neural network;

[0124] The convolutional neural network is trained according to the basic data set and the simulated output data set to obtain a trained first classifier and a trained second classifier;

[0125] Cascade the trained first classifier and the second classifier to obtain a classification network;

[0126] Connecting a trained LSTM neural network after the classification network to obtain a conversion rule extraction model;

[0127] The conversion rules are extracted according to the conversion rule extraction model; the conversion rules include state transition conditions, probability transition rules and time dependency.

[0128] Optionally, the attributes of this embodiment are the basic characteristics and status of the agent, which generally include the following aspects:

[0129] Basic Information:

[0130] Name: An identifier for the agent, such as "Planner", "Villager", etc.

[0131] Type: The category of the agent, indicating its role.

[0132] State variables:

[0133] Resources: The amount of resources the agent has, such as money, land area, population, etc.

[0134] Satisfaction: The agent’s satisfaction with the current state, which may affect its subsequent decisions.

[0135] Goal: The agent's primary goal. For example, a planner's goal might be to optimize land use, while a villager's goal might be to improve their quality of life.

[0136] Environmental Information:

[0137] External environment: The environmental status of the intelligent agent, such as land use status, policies and regulations, etc.

[0138] Other agent status: relationships and status with other agents, such as the developer's project progress, changes in villagers' needs, etc.

[0139] Furthermore, the behavioral rules of this embodiment are the actions or reactions taken by the agent under specific conditions, and generally include:

[0140] Decision rules:

[0141] Condition-action rules: For example, if the villagers’ satisfaction falls below a certain threshold, a request to improve their living environment is made.

[0142] Priority rules: Among multiple options, the agent chooses actions based on priority, such as developers giving priority to developing land with high returns.

[0143] Interaction rules:

[0144] Communication mechanism: How agents exchange information, such as villagers feeding back their needs to planners or developers communicating policies with the government.

[0145] Cooperation and competition: cooperative or competitive relationships between intelligent agents, such as market competition between developers and cooperation between villagers and the government.

[0146] Adaptive rules:

[0147] Learning mechanism: How the agent adjusts its behavior based on environmental changes and historical experience, such as adjusting development strategies based on market feedback.

[0148] Furthermore, the decision-making mechanism of this embodiment is the process of how an agent makes choices when faced with uncertainty and complex environments, which generally includes:

[0149] Decision sub-model:

[0150] Rule-based decision making: The agent makes decisions based on preset rules and conditions.

[0151] Optimization-based decision-making: The agent uses mathematical models or algorithms (such as linear programming, genetic algorithms, etc.) to find the optimal solution.

[0152] Information processing sub-model:

[0153] Data collection: How the agent collects and processes information about the environment, such as through sensors, historical data, etc.

[0154] Information analysis: How the agent analyzes collected data to support decision-making, such as using statistical analysis, machine learning, and other methods.

[0155] Risk assessment sub-model:

[0156] Uncertainty handling: How agents assess and respond to uncertainty in decision-making, for example through simulations, scenario analysis, etc.

[0157] By defining the attributes, behavioral rules, and decision-making mechanisms of each stakeholder, this embodiment can construct a realistic and dynamic intelligent agent model. These agents can interact, make decisions, and adapt in complex environments, effectively simulating the spatial evolution of villages and towns.

[0158] Preferably, the calculation formula for the land use conversion probability of each cell is:

[0159]

[0160] Among them, PY i =j|X i Indicates that cell i has a given feature X i The probability of transitioning to state j under the condition of i is the target state of cell i, X i is the characteristic vector that affects the cell state transformation, β jk : with feature X ik The relevant regression coefficient, representing the feature X ik The degree of influence on the probability of the cell transitioning to state j.

[0161] Steps for building a transformation rule extraction model

[0162] Collection of basic data sets and simulation output data sets

[0163] Specifically, the model building method in step 300 of this embodiment is as follows:

[0164] 1. Collect data related to land use, environmental characteristics, and socioeconomic factors to build a basic dataset. Specifically, collect simulation output data as labeled data for training. This output data should include changes in land status.

[0165] 2. Model construction: Convolutional neural network (CNN) is constructed. CNN is used to process image data, extract spatial features through convolution layers and pooling layers, and output feature vectors. The CNN of this embodiment is designed as multiple convolution layers and pooling layers, and finally outputs the classification results through a fully connected layer. Then, the first CNN classifier is trained using the basic data set to obtain a trained first classifier. The second CNN classifier is trained using the simulated output data set to obtain a trained second classifier. The trained first classifier and the second classifier are cascaded to form a comprehensive classification network. This network integrates the features learned from the basic data and the simulated data.

[0166] 3. Connecting to an LSTM network: Add an LSTM neural network after the cascaded classification network. LSTM is used to process time series data and capture the temporal dependencies of land use change. The LSTM network can input the feature vectors output by the cascaded classification network to generate dynamic change predictions over long periods of time.

[0167] 4. Training of the transformation rule extraction model: Use the fused classification network and LSTM network for training to extract effective transformation rules. Set the input as historical land use data and the expected output as future land use status changes to build a transformation rule extraction model.

[0168] Furthermore, this embodiment can extract the following conversion rules based on the trained conversion rule extraction model:

[0169] State transition conditions define the conditions for the current cell state change, usually including:

[0170] Neighboring cell state: If the neighboring cells of a cell have a certain land use type in a specific proportion, the state transfer occurs.

[0171] Environmental factors: conditions that affect land use changes, such as climate change and geographical characteristics.

[0172] Probabilistic transition rules define the probability of a cell state transition under specific conditions. For example, there might be a 70% probability of agricultural land transitioning to residential land, while the probability of commercial land transitioning is 30%. This rule can be derived through model predictions, such as using LSTM to generate a probability distribution for future state predictions.

[0173] Temporal dependence, identifying and representing temporal dependence rules in time series, for example, the probability of certain land use changes increases significantly in certain months or seasons, such as the probability of land conversion to residential housing increases after the agricultural harvest season.

[0174] Example of a conversion rule

[0175] Rule 1: If three of the five cells surrounding a cell are residential land, the current cell will be converted into residential land with a probability of 70% in the next time step.

[0176] Rule 2: The current land use of the cell is agriculture. If the climate change index is greater than 0.5, the probability of conversion to commercial land in the next 12 months is 60%.

[0177] Rule 3: In the spring of each year (such as March to May), the probability of agricultural land being converted to residential land is 50% higher than at other times.

[0178] By building a transformation rule extraction model combining a convolutional neural network and LSTM, this embodiment can effectively analyze land use change, defining these rules as state transitions and probabilistic transition rules for cellular automata, thereby enabling dynamic simulation and prediction of land use change. This approach, combined with the powerful data processing capabilities of deep learning, can extract potential influencing factors and patterns from complex data sets.

[0179] According to the transformation rules, the cell states in the cellular automaton (CA) model are updated and simulation iterations are performed to record the changes in land use status at each time step. The specific steps are as follows:

[0180] Optionally, this embodiment defines a cellular automaton model, first defining the state of the cell, which usually represents different land use types. For example:

[0181] 0: Undeveloped land

[0182] 1: Agricultural land

[0183] 2: Residential land

[0184] 3: Commercial land

[0185] Then, the environment is set up to define the spatial structure of the model (such as a two-dimensional grid) and boundary conditions (such as periodic boundaries or fixed boundaries).

[0186] Subsequently, this embodiment initializes the cell state based on the basic data set or actual data. For example, different land use types can be assigned to each cell randomly or based on actual conditions.

[0187] Furthermore, this embodiment defines transition rules. Based on the previously extracted transition rules, a function is written that determines how a cell updates its state under given conditions. This function takes into account the states of neighboring cells and any external factors. The code example is as follows:

[0188]

[0189]

[0190] Each land use model is automatically learned and calculated using a machine learning algorithm. This example then iterates the simulation, creating a main loop to iteratively update the cell state at each time step. For each cell, the state of its neighbors is checked, a transition rule is applied, and the state is updated. A function is then created to record the state. This function can save the state at each time step to a file, list, or array for subsequent analysis and visualization.

[0191] After the simulation is completed, a visualization library (such as Matplotlib) can be used to display the changes in land use status at each time step.

[0192] Through the above steps, this embodiment can update the cell states in the cellular automaton model according to the transition rules and perform simulation iterations. The changes in land use state at each time step are recorded for subsequent analysis and display. This simulation process can effectively reflect the dynamic changes in land use and the self-organizing behavior of the system.

[0193] Preferably, the land use change trend is obtained by analyzing the land use state changes at multiple time steps, including:

[0194] Encoding the data of land use state changes at the multiple time steps into numerical form, and performing missing value processing on the encoded numerical values;

[0195] By comparing the coded values ​​of adjacent time steps, the transition of each land use type and the rate of change of each time step are calculated;

[0196] The land use change trend is determined based on the transition status of each land use type and the change rate of each time step using a time series analysis method.

[0197] Furthermore, this embodiment analyzes the land use status changes at multiple time steps to obtain land use change trends, including:

[0198] 1. Data preparation: Collect time series data to ensure that you have land use status data at multiple time steps. This data can be used as input for cellular automaton simulations, recording the state changes of each cell at different time steps. The data is then organized into a format suitable for analysis, typically a two-dimensional array or data frame, where rows represent time steps and columns represent the states of different cells.

[0199] 2. Data preprocessing: Encode land use status into numerical form (e.g., 0 for undeveloped, 1 for agriculture, 2 for residential, and 3 for commercial) for analysis. Check the data for missing values ​​and perform appropriate processing (e.g., filling, interpolation, or deletion).

[0200] 3. Calculate change indicators

[0201] Calculate the state transition matrix: By comparing the states of adjacent time steps, the transition of each land use type is calculated. A transition matrix can be constructed to represent the frequency of transitions from one state to another.

[0202]

[0203] Among them, Pij represents the probability of transitioning from state i to state j.

[0204] Calculate the rate of change: For each land use type, calculate its rate of change at each time step, for example:

[0205]

[0206] 4. Trend analysis: Use time series analysis methods (such as moving averages and exponential smoothing) to identify changing trends. This analysis can be performed using statistical software or programming languages ​​(such as the pandas and statsmodels libraries in Python). Plot a change curve for each land use type, showing changes in area or proportion at different time steps.

[0207] 5. Statistical analysis

[0208] Regression analysis: Linear regression or other regression models can be used to analyze the factors affecting land use change and determine the significance of trends.

[0209] Analysis of variance: If there are multiple factors that affect the change in land use status, analysis of variance (ANOVA) can be conducted to determine the degree of influence of each factor.

[0210] 6. Interpret and report the results. Based on the change curves and statistical analysis results, summarize the changing trends of land use. For example, agricultural land has gradually decreased over the past few years, while residential land has gradually increased. Organize the analysis results into a report that includes charts, data interpretation, and policy recommendations.

[0211] 7. Policy recommendations and decision support: Based on the analysis results, recommendations are made for land use management and planning, such as protecting agricultural land and optimizing urban expansion; data support is provided to relevant decision makers to help them make more scientific decisions in land use planning and management.

[0212] Through the above steps, this example can systematically analyze land use status changes over multiple time steps, identify trends, and provide data support for land use management and policy formulation. This approach combines data analysis, visualization, and statistical methods to effectively reveal the dynamic changes in land use and the factors that influence them.

[0213] Optionally, the Monte Carlo method based on parallel execution of this embodiment randomly selects land development projects according to the land use change trend and the land use conversion probability, and determines the total land supply according to the land use change trend and the land use conversion probability, including:

[0214] 1. Data preparation

[0215] Land use change trend, the land use change trend data obtained from the previous analysis, including the changes in area or proportion of different land use types at each time step.

[0216] The land use conversion probability is the land use conversion probability of each cell calculated based on the aforementioned logistic regression model or other probability models.

[0217] Land development project data: prepare a database containing potential land development projects and record the characteristics of each project (such as location, intended use, development cost, etc.).

[0218] 2. Define Monte Carlo simulation parameters

[0219] Number of simulations: Determine the number of Monte Carlo simulations (e.g., 1000) to ensure statistical significance of the results.

[0220] The random selection strategy defines how to randomly select land development projects based on the land use conversion probability. For example, this embodiment uses a random number generator to perform weighted selection based on the conversion probability.

[0221] 3. Implement Monte Carlo simulation

[0222] Initialize the simulation: Create a loop that executes each iteration of the Monte Carlo simulation.

[0223] 4. Calculate the total land supply

[0224] Summarize selected results: In each simulation, record the land development projects selected and summarize the results after all simulations are completed.

[0225] 5. Analyze the results and calculate the total supply of each land use type based on the summarized selection results.

[0226] The beneficial effects of the present invention are as follows:

[0227] (1) This invention integrates multiple technologies, including cellular automata, machine learning, and multi-agent systems, to comprehensively consider the complexity of village settlements at different levels, taking into account multiple influencing factors such as the environment, economy, and society. By dynamically capturing the decisions and behaviors of stakeholders, it provides a more realistic simulation scenario, enabling the model to reflect actual social behavior and economic interactions.

[0228] (2) The present invention uses machine learning algorithms to analyze basic data and simulation outputs, and can automatically generate and adjust the transformation rules of cellular automata, making the simulation more accurate and adaptable. For example, as data is updated, the rules can self-adjust to reflect new trends and patterns. In addition, the multi-agent system can simulate the behavior and decision-making of different stakeholders (such as government, enterprises, and communities), which helps to understand the role of different roles in the development of villages and towns, and how to optimize the interaction between these roles.

[0229] (3) The present invention uses parallel processing technology to improve simulation efficiency. By randomly selecting land development projects, the effectiveness of different development plans can be quickly evaluated to adapt to different policy objectives and development needs. Furthermore, the present invention supports the rapid adjustment of land development strategies based on real-time updated land use conversion probabilities and change trends, enabling flexible response to changes in the external environment or adjustments in policy guidance.

[0230] (4) The simulation output and calculation results of this invention can provide quantitative data support for policymakers, helping them assess whether land supply meets actual demand and make more scientific decisions based on the data. Visualizing the simulation results can intuitively demonstrate the development trends of villages and towns, making it easier for stakeholders to understand complex data and better communicate and coordinate.

[0231] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0232] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A digital simulation method for village and town settlements based on cellular automata, characterized in that: include: Collect basic data on village and town settlement development, and perform cleaning, filtering and pre-processing; Use a multi-agent system to dynamically capture the decision-making and behavior data of various stakeholders and incorporate them into the preliminary simulation stage to obtain simulation output; Analyzing the basic data and the simulation output using a machine learning algorithm to learn a conversion rule of a cellular automaton; The cell states in the cellular automaton model are updated according to the transformation rules, and simulation iterations are performed to record the land use state changes at each time step; Based on the updated cellular automaton model and the preset economic and environmental factors, the land use conversion probability of each cell is calculated; Analyze the changes in land use status over multiple time steps to obtain the land use change trend; Based on a parallel Monte Carlo method, randomly selecting land development projects according to the land use change trend and the land use conversion probability, and determining the total land supply according to the land use change trend and the land use conversion probability; Determine whether the total land supply meets a preset supply threshold. If not, return to step "randomly select land development projects based on the land use change trend and the land use conversion probability based on the parallel execution Monte Carlo method, and determine the total land supply based on the land use change trend and the land use conversion probability." If so, summarize the simulation results of the multi-agent system and the cellular automation model and visualize them. Analyzing the basic data and the simulation output using a machine learning algorithm to generate a cellular automaton transformation rule includes: Collect basic data sets or simulation output data sets related to land use, environmental characteristics, socioeconomic factors, and demographic factors; Build a convolutional neural network; The convolutional neural network is trained according to the basic data set and the simulated output data set to obtain a trained first classifier and a trained second classifier; Cascade the trained first classifier and the second classifier to obtain a classification network; Connecting a trained LSTM neural network after the classification network to obtain a conversion rule extraction model; The conversion rules are extracted according to the conversion rule extraction model; the conversion rules include state transition conditions, probability transition rules and time dependency.

2. The digital simulation method for village and town settlements based on cellular automata according to claim 1 is characterized in that: The basic data include: village and town spatial geographic information data, planning data, villagers' willingness data, total land supply, incentive and restriction policies.

3. The digital simulation method for village and town settlements based on cellular automata according to claim 1 is characterized in that: The stakeholders include planners, villagers, government and developers.

4. The method for digital simulation of villages and towns based on cellular automata according to claim 1, characterized in that: The decision-making and behavior data of various stakeholders are dynamically captured using a multi-agent system and incorporated into the preliminary simulation phase to obtain simulation outputs, including: Create a corresponding agent for each stakeholder and define the attributes, behavior rules and decision-making mechanism of each agent: Defining a simulation environment and setting interaction rules between the agents; the simulation environment includes land geographic information, land use types, resource allocation, and policy constraints; Collect historical data and current environment data to initialize the state and environment parameters of the agent; Start the simulation of the agent so that the agent makes decisions based on the behavior rules and the state of the environment: At each time step, record the agent's decision and behavior data; Generate preliminary simulation output based on the decision and behavior data of the agent.

5. The method for digital simulation of villages and towns based on cellular automata according to claim 4 is characterized in that: The attributes of the intelligent agent include: basic information, state variables and environmental information; the behavioral rules include: decision rules, interaction rules and adaptive rules; the decision mechanism includes: decision sub-model, information processing sub-model and risk assessment sub-model.

6. The digital simulation method for village and town settlements based on cellular automata according to claim 5 is characterized in that: The formula of the decision sub-model is: U=f(x1,x2,...,xn); where U is the utility value and xi is the i-th variable affecting the decision; the formula of the information processing sub-model is: ;in To observe evidence Assume that The posterior probability of the risk assessment sub-model is calculated as follows: ;in, is the expected return, Each result The probability of occurrence, is the utility value of the corresponding outcome.

7. The method for digital simulation of villages and towns based on cellular automata according to claim 1 is characterized in that: The calculation formula for the land use conversion probability of each cell is: ； in, Represents a cell In a given feature In the case of transition to state probability; For cells The target state, is the characteristic vector that affects the cell state transformation, : With features The relevant regression coefficient, indicating the characteristics Transform the cell into a state The degree of influence of probability.

8. The method for digital simulation of villages and towns based on cellular automata according to claim 1 is characterized in that: Analyze the changes in land use status over multiple time steps to obtain land use change trends, including: Encoding the data of land use state changes at the multiple time steps into numerical form, and performing missing value processing on the encoded numerical values; By comparing the coded values ​​of adjacent time steps, the transition of each land use type and the rate of change of each time step are calculated; The land use change trend is determined based on the transition status of each land use type and the change rate of each time step using a time series analysis method.

Images