Map generation control methods, systems and servers

Abstract

This invention provides a map generation control method, system, and server, relating to the field of map generation control. The method fully utilizes the type and rotation parameters of tiles to dynamically obtain the symmetry results of the tiles, enabling the generation of only the necessary rotation states of the tiles, reducing the resource occupation of the state space, and lowering the computational overhead. In addition, the method can achieve global state caching and precise rollback through global tile state data, reducing the number of restarts and significantly improving the success rate of map generation.

Description

Technical Field

[0001] This invention relates to the field of map generation control, and in particular to a map generation control method, system and server. Background Technology

[0002] Wave Function Collapse (WFC) is a procedural content generation technique based on constraint satisfaction, widely used in the generation of 3D maps, levels, or textures. Its core idea is to simulate quantum state collapse: each grid cell (tile) is initially in a superposition of all possible states (high entropy value). By iteratively selecting the cell with the lowest entropy value for collapse (randomly selecting a certain state), and updating the possible states of other cells according to the constraints of adjacent cells (reducing the entropy value), until all cells have collapsed.

[0003] However, WFC is prone to getting stuck in contradictory states during tile generation, which can severely impact map generation. Specifically, when no valid state is available for a tile after constraints are propagated, traditional backtracking methods only backtrack one step, failing to resolve deep-seated contradictions caused by early random selection. This often requires multiple restarts, resulting in low generation efficiency and unstable success rates; the failure rate increases significantly, especially in complex maps or with strict connection rules. Secondly, for symmetrical tiles (such as those with six identical faces or opposite faces), traditional methods generate independent states for each rotation direction, wasting storage and computational resources. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a map generation control method, system, and server. This method fully utilizes the type parameters and rotation parameters of the tiles to dynamically obtain the symmetry results of the tiles, and can generate only the necessary rotation states of the tiles, reducing the resource occupation of the state space and reducing the computational overhead. In addition, this method can realize global state caching and precise rollback through global tile state data, reducing the number of restarts and significantly improving the success rate of map generation.

[0005] In a first aspect, embodiments of the present invention provide a map generation control method, the method comprising:

[0006] Initialization steps: Initialize the map grid based on the received map generation instructions, determine the symmetry result corresponding to the tile according to the type parameters and rotation parameters of the tile corresponding to the map grid, and generate the first state set corresponding to the tile using the symmetry result;

[0007] Collapse execution steps: Calculate the first entropy value corresponding to the tile through the first state set, determine the target tile in the tile based on the wave function collapse result corresponding to the first entropy value, and obtain the first map element state data corresponding to the target tile after random collapse processing of the target tile;

[0008] Entropy calculation steps: Traverse the adjacent tiles corresponding to the target tile, determine the second state set corresponding to the adjacent tiles based on the edge connection rules corresponding to the adjacent tiles, and use the second state set to calculate the second entropy value corresponding to the target tile;

[0009] Constraint propagation steps: Based on the wave function collapse result corresponding to the second entropy value, the adjacent tiles are randomly collapsed in sequence to obtain the state data of the second map element corresponding to the adjacent tiles;

[0010] Map generation steps: Construct a map grid using the first map element state data and the second map element state data, and generate the map corresponding to the map generation command based on the map grid.

[0011] Optional initialization steps include:

[0012] After receiving the map generation command from the requesting end, the server receives the map generation command based on the preset parameters and initializes the map grid using the map generation command.

[0013] Determine the type parameters of tiles in the map grid based on the surface area type parameters, road structure type parameters, and building type parameters in the map grid.

[0014] The reference surface and rotation axis corresponding to the tile are determined by using the type parameters, and the rotation parameters corresponding to the tile are determined by using the reference surface and rotation axis.

[0015] The symmetry result corresponding to the tile is determined based on the rotation parameters. If all six reference planes of the tile are the same, the symmetry result corresponds to 1 state data. If the horizontal reference groups of the tile are the same but the vertical reference groups are different, the symmetry result corresponds to 2 state data. If the three sets of reference planes of the tile are the same but the mode parameters between each set are different, the symmetry result corresponds to 6 state data. If each reference plane of the tile is different, the symmetry result corresponds to 24 state data.

[0016] The first state set corresponding to the tile is generated based on the number of types of state data in the symmetry result.

[0017] Optionally, the collapse execution steps include:

[0018] Obtain the number of types corresponding to the state type parameters of the tiles in the first state set, and use the number of types to calculate the first entropy value corresponding to the tile;

[0019] The tile corresponding to the minimum entropy value corresponding to the first entropy value in the first state set is determined as the target tile;

[0020] After randomly collapsing the target tile, the state data of the first map element corresponding to the target tile is determined using the road data, surface data, and building data corresponding to the target tile.

[0021] Optionally, before performing random collapse on the target tile, the collapse execution steps may further include:

[0022] The global state data of the tile corresponding to the map grid is determined based on the coordinate parameters, collapse flag, current state type parameters, and first state set of the tile.

[0023] Global state data of tiles is cached in a preset history stack.

[0024] Optional entropy calculation steps include:

[0025] Once the target tile has been detected to have completed the collapse process, the adjacent tiles corresponding to the target tile are obtained based on the position parameters of the target tile, and the edge connection rules between the adjacent tiles are determined.

[0026] Traverse adjacent tiles, use edge connection rules to determine the symmetry result of adjacent tiles, and use the symmetry result to determine the second state set corresponding to adjacent tiles;

[0027] Obtain the number of types corresponding to the state type parameters of adjacent tiles in the second state set, and use the number of types to calculate the second entropy value corresponding to the tile.

[0028] Optional, constraint propagation steps include:

[0029] Based on the wave function collapse result corresponding to the second entropy value, the adjacent tiles are randomly collapsed in sequence, and it is determined whether the second state set is empty.

[0030] If not, the status data of the second map element corresponding to the target tile is determined using the road data, surface data, and building data corresponding to the adjacent tiles.

[0031] Optionally, the constraint propagation step may also include:

[0032] If the second state set is empty, then the global state data of the adjacent tiles is obtained using the time parameter corresponding to the historical record stack.

[0033] The second state set is updated based on the global state data of the tiles, and the marking results of adjacent tiles are updated to the historical record stack.

[0034] Optional map generation steps include:

[0035] Determine whether adjacent tiles have completed the collapse treatment;

[0036] If yes, then construct a map grid using the state data of the first and second map elements; otherwise, perform the constraint propagation step.

[0037] Secondly, the present invention provides a map generation and control system, the system comprising:

[0038] Initialization module: Used to initialize the map grid based on the received map generation instructions. After determining the symmetry result of the tile according to the type parameters and rotation parameters of the tile corresponding to the map grid, the first state set corresponding to the tile is generated using the symmetry result.

[0039] Collapse Execution Module: Used to calculate the first entropy value corresponding to the tile through the first state set, determine the target tile in the tile based on the wave function collapse result corresponding to the first entropy value, and obtain the first map element state data corresponding to the target tile after random collapse processing of the target tile;

[0040] Entropy calculation module: used to traverse the adjacent tiles corresponding to the target tile, determine the second state set corresponding to the adjacent tiles based on the edge connection rules corresponding to the adjacent tiles, and use the second state set to calculate the second entropy value corresponding to the target tile;

[0041] Constraint propagation module: used to perform random collapse processing on adjacent tiles sequentially based on the wave function collapse result corresponding to the second entropy value, and then obtain the state data of the second map element corresponding to the adjacent tiles;

[0042] Map generation module: Used to construct a map grid using the state data of the first map element and the state data of the second map element, and generate the map corresponding to the map generation command based on the map grid.

[0043] Thirdly, embodiments of the present invention also provide a server, the server including a processor and a memory, the memory storing computer-executable instructions that can be executed by the processor, the processor executing the computer-executable instructions to implement the steps of the map generation control method provided in the first aspect.

[0044] Fourthly, embodiments of the present invention also provide a storage medium storing computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the steps of the map generation control method provided in the first aspect.

[0045] This invention provides a map generation control method, system, and server. In controlling the map generation process, the method first initializes a map grid based on a received map generation command. After determining the symmetry result of the tile corresponding to the map grid based on the tile type and rotation parameters, it generates a first state set corresponding to the tile using the symmetry result. Then, it calculates the first entropy value corresponding to the tile using the first state set, determines the target tile based on the wave function collapse result corresponding to the first entropy value, and performs random collapse processing on the target tile to obtain the first map element state data corresponding to the target tile. Subsequently, it traverses the adjacent tiles corresponding to the target tile, determines the second state set corresponding to the adjacent tiles based on the edge connection rules corresponding to the adjacent tiles, and calculates the second entropy value corresponding to the target tile using the second state set. Then, based on the wave function collapse result corresponding to the second entropy value, it sequentially performs random collapse processing on the adjacent tiles to obtain the second map element state data corresponding to the adjacent tiles. Finally, it constructs a map grid using the first and second map element state data, and generates the map corresponding to the map generation command based on the map grid. This method fully utilizes the tile type and rotation parameters to dynamically obtain the symmetry results of the tiles, enabling the generation of only the necessary rotation states of the tiles, reducing the resource consumption of the state space and lowering the computational overhead. In addition, this method can achieve global state caching and precise rollback through global tile state data, reducing the number of restarts and significantly improving the success rate of map generation.

[0046] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0048] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0049] Figure 1 A flowchart of a map generation control method provided in an embodiment of the present invention;

[0050] Figure 2This is a flowchart of the initialization step S101 in a map generation control method provided in an embodiment of the present invention;

[0051] Figure 3 This is a flowchart of the collapse execution step S102 in a map generation control method provided by an embodiment of the present invention;

[0052] Figure 4 A flowchart of the collapse execution step S102 in another map generation control method provided in an embodiment of the present invention;

[0053] Figure 5 This is a flowchart of the entropy calculation step S103 in a map generation control method provided by an embodiment of the present invention;

[0054] Figure 6 This is a flowchart of the constraint propagation step S104 in a map generation control method provided in an embodiment of the present invention;

[0055] Figure 7 This is a flowchart of map generation step S105 in a map generation control method provided in an embodiment of the present invention;

[0056] Figure 8 This is a map rendering obtained from a map generation control method provided in an embodiment of the present invention;

[0057] Figure 9 This is a schematic diagram of the structure of a map generation and control system provided in an embodiment of the present invention;

[0058] Figure 10 This is a schematic diagram of the structure of a server provided in an embodiment of the present invention.

[0059] icon:

[0060] 910 - Initialization Module; 920 - Collapse Execution Module; 930 - Entropy Calculation Module; 940 - Constraint Propagation Module; 950 - Map Generation Module;

[0061] 101 - Processor; 102 - Memory; 103 - Bus; 104 - Communication interface. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0063] To facilitate understanding of this embodiment, a map generation control method disclosed in this invention will first be introduced. This method addresses the problems of state-space redundancy and weak contradiction backoff capability inherent in traditional wave function collapse (WFC) map generation techniques. Through tile symmetry optimization and global state management, it achieves efficient and high-success-rate map generation. The method is specifically as follows... Figure 1 As shown, it includes:

[0064] Initialization step S101: Initialize the map grid based on the received map generation command, determine the symmetry result corresponding to the tile according to the type parameter and rotation parameter of the tile corresponding to the map grid, and generate the first state set corresponding to the tile using the symmetry result.

[0065] This step responds to the received map generation command, completes the initial construction of the map grid, and clarifies the position and attributes to be filled for each cell in the grid. For each tile to be assigned within the grid, its inherent type parameters (such as terrain type and functional module category) and rotation parameters (such as the range of rotatable angles and rotational symmetry features) are extracted. Based on these two types of parameters, symmetry analysis is performed to determine the equivalence of the tile's state under different rotation operations (for example, any rotation of a tile with six identical faces is an equivalent state), and redundant rotation states are eliminated. Finally, using the symmetry results after redundancy removal, a first state set for each tile is generated. This set contains only the necessary and valid states of the tile, which can effectively compress the state space and reduce subsequent computational overhead.

[0066] Collapse execution step S102: Calculate the first entropy value corresponding to the tile through the first state set, determine the target tile in the tile based on the wave function collapse result corresponding to the first entropy value, and obtain the first map element state data corresponding to the target tile after random collapse processing of the target tile.

[0067] This step uses the first state set of each tile as a basis to calculate the first entropy value for each tile. The magnitude of the entropy value characterizes the degree of uncertainty of the current state of the tile (the more states there are and the more uniform their distribution, the higher the entropy value). Following the core principle of prioritizing the lowest entropy, the tile with the smallest first entropy value is selected as the target tile (i.e., the unit with the lowest current uncertainty). A collapse process is then performed on the target tile using random sampling to determine its unique deterministic state, generating the first map element state data corresponding to the target tile, laying the foundation for the subsequent filling of map elements.

[0068] Entropy calculation step S103: Traverse the adjacent tiles corresponding to the target tile, determine the second state set corresponding to the adjacent tiles based on the edge connection rules of the adjacent tiles, and use the second state set to calculate the second entropy value corresponding to the target tile.

[0069] This step involves locating all adjacent tile units of the target tile within the grid, retrieving preset edge connection rules (which define the legal splicing conditions for different tile states at adjacent positions, such as road tiles needing to be adjacent to road / intersection tiles, and water area tiles not being directly spliced ​​to land tiles, etc.); based on the edge connection rules, the first state set of each adjacent tile is filtered, eliminating illegal states that conflict with the collapsed state of the target tile, resulting in the second state set of adjacent tiles; based on the second state set, the second entropy value corresponding to each adjacent tile is recalculated. At this point, the entropy value has incorporated the constraint influence of the target tile and can reflect the uncertainty of adjacent tiles under the current constraints.

[0070] Constraint propagation step S104: Based on the wave function collapse result corresponding to the second entropy value, the adjacent tiles are randomly collapsed in sequence to obtain the state data of the second map element corresponding to the adjacent tiles.

[0071] This step, based on the second entropy value of each adjacent tile, continues to execute the lowest entropy priority collapse strategy, and performs random collapse processing on adjacent tiles in order of entropy value from low to high. During the collapse process, the legality of the connection between adjacent tiles is continuously checked to ensure that the collapse result of each step meets the global constraint requirements. Finally, the state data of the second map elements corresponding to all adjacent tiles is generated, realizing the layer-by-layer propagation of constraints from the target tile to the surrounding area.

[0072] Map generation step S105: Construct a map grid using the first map element state data and the second map element state data, and generate the map corresponding to the map generation command based on the map grid.

[0073] The system integrates the first map element state data of the target tile with the second map element state data of all adjacent tiles, maps these deterministic state data one by one to the corresponding grid cells, and completes the full state filling of the map grid. The filled grid is then subjected to global validity verification to confirm that all tiles meet the preset connection rules and symmetry requirements. Finally, a complete map matching the initial map generation command is output.

[0074] Initialization step S101 can be further refined into the following sub-steps: through layered processing of map generation instructions, tile parameters, and symmetry analysis, a precise set of redundant tile states is constructed. Optionally, initialization step S101, such as... Figure 2 As shown, it includes:

[0075] Step S201: After receiving the map generation instruction from the requesting end, the server receives the map generation instruction based on the preset settings and initializes the map grid through the map generation instruction.

[0076] First, the server receives map generation instructions from requesting clients (such as clients or editor interfaces) according to a preset communication protocol and receiving mechanism. These instructions include core parameters such as map size, scene type (such as roads, buildings, and terrain types), and functional requirements. After validating the received instructions (such as parameter completeness and format correctness), the server initializes the map grid based on the map specification parameters in the instructions, specifically defining the number of rows and columns. Finally, it divides the grid into independent tile-filling units and presets the basic attributes of each unit (such as coordinate information and initial empty state), laying the foundation for subsequent tile allocation and state management.

[0077] Step S202: Determine the type parameters of the tiles in the map grid based on the surface area type parameters, road structure type parameters, and building type parameters in the map grid.

[0078] Based on the scene requirements corresponding to the initialized map grid, three core parameters are extracted and integrated to define the tile type parameters: First, the surface area type parameter, used to distinguish the surface attributes corresponding to the tile (such as plains, mountains, water bodies, deserts, etc.); second, the road structure type parameter, clarifying whether the tile contains a road, road level (main road, secondary road), road direction, etc.; and third, the building type parameter, identifying whether the tile belongs to a building area, building function (residential, commercial, public facilities), number of floors, etc. By combining and judging these three types of parameters, the unique type parameter of each tile to be filled in the map grid is finally determined, providing a basis for the subsequent determination of the reference surface and rotation axis.

[0079] Step S203: Use the type parameters to determine the reference surface and rotation axis corresponding to the tile, and use the reference surface and rotation axis to determine the rotation parameters corresponding to the tile.

[0080] Based on the tile type parameters obtained in step S202, the core reference planes for this type of tile are first identified. The reference planes are typically the six outer surfaces of the tile (top, bottom, front, back, left, and right). Different types of tiles have different functional attributes for their reference planes (e.g., the reference planes for road tiles need to distinguish between road connection surfaces and non-connection surfaces, and the reference planes for building tiles need to distinguish between wall surfaces and roof surfaces). Subsequently, based on the functional attributes and spatial distribution of the reference planes, the legal rotation axes of the tiles are determined (e.g., rotation axes based on the vertical direction, horizontal direction, etc.), clarifying the dimensions and range within which the tiles can be rotated. Finally, by defining the attributes of the reference planes and limiting the range of the rotation axes, the rotation parameters of the tiles are integrated. These parameters directly determine the direction of subsequent symmetry analysis of the tiles.

[0081] Step S204: Determine the symmetry result corresponding to the tile based on the rotation parameters; wherein, if all six reference planes of the tile are the same, the symmetry result corresponds to 1 state data; if the horizontal reference groups of the tile are the same but the vertical reference groups are different, the symmetry result corresponds to 2 state data; if the three sets of reference planes of the tile are the same but the mode parameters between each set are different, the symmetry result corresponds to 6 state data; if each reference plane of the tile is different, the symmetry result corresponds to 24 state data.

[0082] Based on the rotation parameters obtained in step S203, a full-dimensional rotation simulation of the tile is performed. The equivalence of the tile's six-sided pattern (top, bottom, left, right, front, back) under 3D rotation operations (cube symmetry group, 24 possible rotations) is analyzed, specifically divided into four core scenarios:

[0083] First, if the six reference faces of a tile have the same properties (such as a cube-shaped decorative block or a homogeneous ground tile), then the tile state is equivalent after any rotation, and the symmetry result corresponds to one state data (all rotations are equivalent, and there is no need for redundant variants).

[0084] Secondly, if the horizontal reference group (upper and lower surfaces) of the tile has the same attributes, but the vertical reference group (front, back, left, and right surfaces) has different attributes (e.g., tiles with only the upper and lower surfaces being the ground and the surrounding surfaces being different walls), then the symmetry result corresponds to two types of state data (rotating around the vertical axis at 0° and 90° can cover all valid directions; 180° and 270° are equivalent to 0° or 90°, because the horizontal surfaces are the same).

[0085] Third, if the three sets of relative reference surfaces of the tile (top-bottom, front-back, left-right) are the same, but the attribute patterns between the three sets are different (e.g., the three sets of reference surfaces are tiles on the ground, wall, and roof respectively), then the symmetry result corresponds to 6 state data (3 selections of the vertical axis direction × 2 selections of horizontal rotation, instead of the traditional 24).

[0086] Fourth, if the six reference faces of a tile have different attributes (such as complex building component tiles with unique texture directions and functional markings), there will be no equivalent state after rotation, and the symmetry result corresponds to 24 state data (standard 3D rotation).

[0087] Step S205: Generate the first state set corresponding to the tile based on the number of types of state data in the symmetry result.

[0088] Based on the symmetry result determined in step S204, the number of state data types is counted, all redundant states with rotational equivalence are removed, and only the unique legal state data corresponding to each unique symmetry state is retained, ultimately generating the first state set corresponding to the tile. This set achieves simplification of the state space through symmetry optimization, ensuring the legality of the subsequent collapse process while significantly reducing storage usage and state traversal overhead in subsequent calculations.

[0089] The collapse execution step S102 can be further refined into the following sub-steps, with the core revolving around the logic of entropy calculation, target selection, and collapse rights confirmation, ensuring the rationality of the collapse process and the validity of the results. Optionally, the collapse execution step S102, such as... Figure 3 As shown, it includes:

[0090] Step S301: Obtain the number of types corresponding to the state type parameters of the tiles in the first state set, and calculate the first entropy value corresponding to the tile using the number of types.

[0091] First, the first state set corresponding to each tile is traversed, and all independent state type parameters within the set are extracted. The number of categories corresponding to these state type parameters is then counted and determined (i.e., the total number of legal states in which the tile is currently in a superposition state). Subsequently, the first entropy value of the tile is calculated based on the counted number of state categories. Entropy is a core indicator characterizing system uncertainty. In the WFC technical framework, an entropy calculation logic positively correlated with the number of states is adopted: the more state categories, the higher the uncertainty of the tile, and the larger the corresponding first entropy value; conversely, the fewer state categories, the lower the uncertainty, and the smaller the first entropy value. Through this calculation method, the degree of uncertainty of the current superposition state of each tile can be accurately quantified, providing data support for the selection of subsequent target tiles.

[0092] Step S302: Determine the tile corresponding to the minimum entropy value corresponding to the first entropy value in the first state set as the target tile.

[0093] The first entropy value data of all tiles are collected, sorted, and compared. The tile with the smallest first entropy value is selected as the target tile for this collapse process. The core reason for choosing the tile with the smallest entropy value as the target is that such tiles have the lowest uncertainty. Collapsing them first can minimize the probability of conflict between the collapse result and subsequent constraint rules, reducing the occurrence of contradictory states. If multiple tiles simultaneously have the smallest first entropy value (i.e., their entropy values ​​are tied for minimum), a unique target tile can be determined from these tiles using a preset random selection rule, ensuring the orderly progress of the collapse process.

[0094] Step S303: After performing random collapse processing on the target tile, determine the state data of the first map element corresponding to the target tile using the road data, surface data and building data corresponding to the target tile.

[0095] A random sampling method is used to perform collapse processing on the target tile. Specifically, a valid state type parameter is randomly selected from the first state set of the tile, causing the target tile to collapse from a multi-state superposition state to a single-state deterministic state. After collapse, the core attribute data corresponding to the target tile is retrieved, including road data (such as road direction, road grade, connection port information, etc.), surface data (such as surface type, surface texture, terrain relief parameters, etc.), and building data (such as building structure type, building function identifier, building appearance parameters, etc.). These attribute data are then correlated and integrated with the collapsed state type parameter. Data verification ensures the consistency of each attribute data with the collapsed state. Finally, the first map element state data corresponding to the target tile is determined and generated. This data clarifies the specific attributes and presentation form of the target tile on the map and serves as the basic rights confirmation data for subsequent map grid filling.

[0096] Optionally, the collapse execution step S102 is as follows: Figure 4 As shown, it includes:

[0097] Step S301: Obtain the number of types corresponding to the state type parameters of the tiles in the first state set, and calculate the first entropy value corresponding to the tile using the number of types.

[0098] Step S302: Determine the tile corresponding to the minimum entropy value corresponding to the first entropy value in the first state set as the target tile.

[0099] Step S304: Determine the global state data of the tile corresponding to the map grid based on the coordinate parameters, collapse flag, current state type parameters, and first state set of the tile.

[0100] Before performing collapse processing on the target tile, it is necessary to construct and record the global state data of the current map grid tiles. This data provides a complete state basis for possible subsequent rollback of contradictions. Specifically, the global state data is determined by integrating four types of core information: First, the coordinate parameters corresponding to the tile, which clearly defines the unique spatial location of each tile in the map grid (such as two-dimensional coordinates composed of row number and column number); second, the collapse flag, which indicates whether each tile has completed collapse (0 for not collapsed, 1 for collapsed), clearly distinguishing between overlay state and deterministic state tiles; third, the current state type parameter, which records the details of the current overlay state of uncollapsed tiles (i.e., the current state corresponding to the first state set) and the deterministic state parameters of collapsed tiles; fourth, the first state set, which retains the baseline of the legal state set of each tile after initial redundancy removal. By integrating the above four types of information, global state data that can completely reflect the state of all tiles in the current map grid is formed.

[0101] Step S305: Cache the global state data of the tile into the preset history stack.

[0102] The tile global state data constructed in step S304 is cached into a preset historical record stack according to the temporal order of the collapse operations. This historical record stack adopts a last-in-first-out storage rule, and the current global state data is pushed onto the stack before each collapse operation. The core purpose of this is that when contradictory states occur during subsequent constraint propagation (the state set of a certain tile is empty), the global state data before the corresponding collapse step can be retrieved from the historical record stack to achieve precise multi-step backtracking, rather than the single-step backtracking of traditional WFC. This effectively solves the deep contradictions caused by early random selection, reduces the number of map generation restarts, and improves the generation success rate and efficiency.

[0103] Step S303: After performing random collapse processing on the target tile, determine the state data of the first map element corresponding to the target tile using the road data, surface data and building data corresponding to the target tile.

[0104] In the above implementation process, before the random collapse of the target tile, the collapse execution step S102 adds a global state data recording and caching step to achieve accurate rollback in case of subsequent conflicts.

[0105] The core purpose of entropy calculation step S103 is to transfer the collapse constraint of the target tile to adjacent tiles. By filtering legal states and updating the entropy value, the influence of the constraint is quantified, providing a basis for the collapse order of subsequent constraint propagation. Optionally, entropy calculation step S103, as follows: Figure 5 As shown, it includes:

[0106] Step S501: After the target tile is detected to have completed the collapse process, the adjacent tiles corresponding to the target tile are obtained according to the position parameters of the target tile, and the edge connection rules between the adjacent tiles are determined.

[0107] The system monitors the collapse state of the target tile in real time. Once the collapse of the target tile is detected and the first map element state data is generated, the adjacent tile retrieval process is immediately triggered. Based on the position parameters of the target tile (such as two-dimensional coordinates and three-dimensional spatial coordinates in the map grid), and according to preset adjacency determination rules (such as four directions in the two-dimensional grid: up, down, left, and right; and six directions in the three-dimensional grid), all tile units directly adjacent to the target tile (i.e., adjacent tiles) are accurately located and obtained. At the same time, preset edge connection rules are retrieved. These rules are the core basis for ensuring the legality of tile splicing and clarify the matching requirements of different tile states on adjacent boundary surfaces (such as edge attribute identification, interface form, functional compatibility, etc., for example: the connecting edge of a road tile must be consistent with the road edge attribute of the adjacent tile and cannot be directly spliced ​​with a non-road edge). Based on the relative orientation of the target tile and the adjacent tile, the corresponding edge connection relationship between them is determined (such as the right edge of the target tile and the left edge of the adjacent tile being a matching edge).

[0108] Step S502: Traverse adjacent tiles, use edge connection rules to determine the symmetry result of adjacent tiles, and use the symmetry result to determine the second state set corresponding to adjacent tiles.

[0109] Each adjacent tile is traversed one by one. Based on the edge connection rules determined in step S501, the symmetry result and corresponding first state set of each adjacent tile are subjected to validity screening. Specifically, the symmetry result (including all equivalent rotation states) determined during the initialization phase of the adjacent tile is retrieved first. Then, for each state type parameter in the first state set, the attributes of its corresponding matching edge are checked to see if they conform to the edge connection rules (i.e., consistent with the attributes of the corresponding edge after the target tile collapses). If the edge attribute corresponding to a certain state type parameter does not meet the connection rules, that state and its rotationally equivalent redundant states are removed. Finally, all valid states that conform to the edge connection rules are retained, forming the second state set corresponding to the adjacent tile. This set is a subset of the first state set, and the constraint screening significantly reduces the uncertainty of the adjacent tile's state.

[0110] Step S503: Obtain the number of types corresponding to the state type parameters of adjacent tiles in the second state set, and calculate the second entropy value corresponding to the tile using the number of types.

[0111] For each adjacent tile's second state set, all independent state type parameters within the set are extracted, and the number of categories corresponding to these legal state type parameters is counted and determined. Following the same calculation logic as the first entropy value (entropy value is positively correlated with the number of legal state categories), the second entropy value for each adjacent tile is calculated based on the statistically obtained number of categories. Unlike the first entropy value, the second entropy value incorporates the collapse constraint of the target tile, accurately reflecting the degree of uncertainty of adjacent tiles under the current constraint conditions. The smaller the second entropy value, the more significant the constraint of the target tile on the adjacent tile, and the fewer legal states can be selected. Subsequent collapse processing should be prioritized to reduce the risk of conflict.

[0112] The core of constraint propagation step S104 is to propagate the collapse constraint of the target tile layer by layer to adjacent areas, ensuring the legality of tile splicing through ordered collapse, while simultaneously verifying the validity of the state in real time. Optionally, constraint propagation step S104, such as... Figure 6 As shown, it includes:

[0113] Step S601: Based on the wave function collapse result corresponding to the second entropy value, perform random collapse processing on adjacent tiles in sequence, and determine whether the second state set is empty.

[0114] Based on the second entropy value obtained in the entropy calculation step (S103) above, following the core collapse criterion of prioritizing the lowest entropy, the collapse order of adjacent tiles is determined, and random collapse processing is performed on adjacent tiles in order of increasing second entropy value. The core reason for this order selection is that adjacent tiles with smaller second entropy values ​​are more significantly constrained by the target tile, have fewer legal states, and lower uncertainty. Prioritizing collapse can minimize the risk of subsequent splicing conflicts. During the random collapse of each adjacent tile (randomly selecting a legal state from its second state set), the second state set of the adjacent tile is simultaneously checked in real time to see if it is empty: if the second state set is empty, it means that there are no legal states to choose from under the current constraints, i.e., a conflicting state has occurred, and the subsequent global state rollback process needs to be triggered; if the second state set is not empty, the collapse operation continues.

[0115] Step S602: If not, determine the second map element status data corresponding to the target tile using the road data, surface data and building data corresponding to the adjacent tiles.

[0116] When the judgment result is "No" (i.e., the second state set of adjacent tiles is not empty, and the collapse operation is legally completed), the construction of the second map element state data corresponding to the adjacent tile begins. Specifically, core attribute data matching the state type parameters after collapse of the adjacent tile is retrieved, including road data (such as road direction, connection port specifications, road grade, etc.), surface data (such as surface texture type, terrain slope, vegetation cover information, etc.), and building data (such as building component type, functional zoning identifier, appearance detail parameters, etc.). These attribute data are associated and integrated with the determined state type parameters after collapse. After data consistency verification (ensuring that the attribute data does not conflict with the collapsed state and conforms to the edge connection rules), the second map element state data corresponding to the adjacent tile is finally determined and generated. This data has natural splicing compatibility with the first map element state data of the target tile, providing a legal basic unit for the subsequent complete construction of the map mesh.

[0117] Optionally, the constraint propagation step S104 further includes:

[0118] Step S603: If the second state set is empty, then use the time parameter corresponding to the historical record stack to obtain the global state data of the adjacent tiles.

[0119] If the judgment result is "yes" (i.e., the second state set of adjacent tiles is empty, resulting in a contradictory state), the current collapse propagation process is immediately terminated, and the contradiction resolution mechanism is initiated. The core is to accurately backtrack to the contradictory-free state baseline using the historical record stack cached in the previous collapse execution steps: by using the time parameters stored in the historical record stack (which correspond one-to-one with the timing of each collapse operation, marking the push order of global state data), the global state data of the tile related to the currently contradictory adjacent tile before the previous valid collapse operation is located. That is, the complete global state data (including the coordinates of each tile, collapse flags, initial valid state set, and other core information) of the adjacent tile and related tiles when they were still in a valid state before the contradiction occurred is retrieved, providing data for subsequent state recovery and updates.

[0120] Step S604: Update the second state set based on the global state data of the tiles, and update the marking results of adjacent tiles to the historical record stack.

[0121] Based on the global tile state data retrieved in step S603, the second state set of the adjacent tiles in the current conflict is updated in reverse: the legal state set of the adjacent tile before the conflict occurred is restored (by removing erroneous constraint filtering results that caused the conflict), and the state set of its associated tiles is updated simultaneously (to avoid new constraint conflicts caused by single tile rollback), ensuring that the updated second state set has legal state type parameters again. After the state set update is completed, the key information of this conflict handling, namely the marking results of adjacent tiles (which may include rollback markers, updated state identifiers, conflict handling sequence, etc.), is pushed into the history stack. This not only ensures the integrity of the historical state chain, but also provides a traceable state node for new conflicts that may occur later, avoiding repeated occurrences of the same type of conflict and improving the stability of constraint propagation.

[0122] The core of map generation step S105 is to verify the collapse progress of adjacent tiles and advance map mesh construction according to different scenarios to ensure that the final generated map is complete and conforms to the constraint rules. Optionally, map generation step S105, such as... Figure 7 As shown, it includes:

[0123] Step S701: Determine whether the adjacent tiles have completed the collapse process.

[0124] After completing the target tile collapse and the first round of constraint propagation, a comprehensive verification of the collapse status of adjacent tiles is initiated. The verification is based on the aforementioned defined collapse flag (0 for no collapse, 1 for collapsed) and the completeness of the second map element status data: each tile unit adjacent to the target tile is checked to determine whether it has completed random collapse processing (i.e., the collapse flag is 1) and whether it has generated valid second map element status data. This verification clarifies the current constraint propagation coverage progress and determines whether the basic conditions for map grid construction have been met.

[0125] Step S702: If yes, construct a map grid using the first map element state data and the second map element state data; if no, perform the constraint propagation step.

[0126] Based on the verification result of step S701, differential processing is performed: if the verification result is "yes" (all adjacent tiles have completed collapse and have complete second map element state data), then the map grid construction stage begins. Specifically, the first map element state data of the target tile and the second map element state data of all adjacent tiles are accurately mapped to the corresponding map grid cells according to their respective coordinate parameters; at the same time, a global stitching compliance check is performed on the mapped grid to ensure that the edge connection attributes and functional attributes of each tile are completely matched (such as smooth road connections and reasonable transition of surface types), and finally the construction of the map grid for the current area is completed.

[0127] If the verification result is "No" (there are adjacent tiles that have not completed collapse, or no valid second map element state data has been generated), the process immediately returns and re-executes the constraint propagation step (S104). Subsequently, the uncollapsed adjacent tiles will be used as the new starting point for constraint propagation, and the logic of entropy sorting-random collapse-state verification will continue until all adjacent tiles have completed valid collapse. Then, the process will enter the map mesh construction stage to ensure that the final generated map mesh has no splicing contradictions and is state-complete.

[0128] The map generation control method described above can be executed in typical scenarios of procedurally generated open-world terrain (such as the Unreal Engine environment), and is applicable to scenarios such as procedurally generated open-world terrain, city street networks, underground city layouts, or puzzle-like levels. For example, the generated city street grid map is as follows: Figure 8 As shown, Figure 8 The map now strictly satisfies all predefined connection rule constraints, has a reasonable structure and is free of contradictions. During map generation, it can automatically and reliably recover to a valid state from any contradictory point through global state caching and precise rollback, without manual intervention or a complete restart, significantly improving the success rate of map generation. In addition, the map generation process dynamically identifies tile symmetry, generates only the necessary rotation states, reduces the size of the state space, reduces memory usage and entropy calculation overhead, and further improves computational efficiency.

[0129] As can be seen from the above map generation control method, this method makes full use of the tile type parameters and rotation parameters to dynamically obtain the symmetry results of the tiles, and can generate only the necessary rotation states of the tiles, reducing the resource occupation of the state space and reducing the computational overhead. In addition, this method can realize global state caching and precise rollback through global tile state data, reducing the number of restarts and significantly improving the success rate of map generation.

[0130] Corresponding to the above-described map generation control method embodiments, this invention also provides a map generation control system, such as... Figure 9 As shown, the system includes:

[0131] Initialization module 910: Used to initialize the map grid based on the received map generation command, determine the symmetry result corresponding to the tile according to the type parameter and rotation parameter of the tile corresponding to the map grid, and generate the first state set corresponding to the tile using the symmetry result;

[0132] Collapse execution module 920: is used to calculate the first entropy value corresponding to the tile through the first state set, determine the target tile in the tile based on the wave function collapse result corresponding to the first entropy value, and obtain the first map element state data corresponding to the target tile after random collapse processing of the target tile;

[0133] Entropy calculation module 930: used to traverse the adjacent tiles corresponding to the target tile, determine the second state set corresponding to the adjacent tiles based on the edge connection rules corresponding to the adjacent tiles, and use the second state set to calculate the second entropy value corresponding to the target tile;

[0134] Constraint propagation module 940: used to perform random collapse processing on adjacent tiles sequentially based on the wave function collapse result corresponding to the second entropy value, and then obtain the second map element state data corresponding to the adjacent tiles;

[0135] Map generation module 950: Used to construct a map grid using the first map element state data and the second map element state data, and generate a map corresponding to the map generation command based on the map grid.

[0136] As can be seen from the above map generation control system, the system makes full use of the tile type parameters and rotation parameters to dynamically obtain the symmetry results of the tiles, and can generate only the necessary rotation states of the tiles, reducing the resource occupation of the state space and reducing the computational overhead. In addition, the system can realize global state caching and accurate rollback through global tile state data, reducing the number of restarts and significantly improving the success rate of map generation.

[0137] The map generation control system provided in this embodiment of the invention has the same implementation principle and technical effects as the aforementioned map generation control method embodiment. For the sake of brevity, any parts not mentioned in the system embodiment can be referred to the corresponding content in the aforementioned map generation control method embodiment.

[0138] This embodiment also provides a server, the structural diagram of which is shown below. Figure 10 As shown, the device includes a processor 101 and a memory 102; wherein, the memory 102 is used to store one or more computer instructions, which are executed by the processor to implement the steps of the map generation control method described above.

[0139] Figure 10 The server shown also includes a bus 103 and a communication interface 104. The processor 101, the communication interface 104, and the memory 102 are connected via the bus 103.

[0140] The memory 102 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. The bus 103 may be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 10 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0141] The communication interface 104 is used to connect to at least one user terminal and other network units through a network interface, and to send encapsulated IPv4 packets or IPv4 packets to the user terminal through the network interface.

[0142] Processor 101 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 101 or by instructions in software form. The processor 101 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this disclosure can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 102. The processor 101 reads the information in memory 102 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments.

[0143] This invention also provides a storage medium storing a computer program, which, when run by a processor, executes the steps of the map generation control method described in the foregoing embodiments.

[0144] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, devices, and methods can be implemented in other ways. The system embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0145] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0146] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0147] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, electronic device, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0148] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A map generation control method, characterized in that, The method includes: Initialization steps: Initialize the map grid based on the received map generation instructions; determine the symmetry result corresponding to the tile according to the type parameters and rotation parameters of the tile corresponding to the map grid; and generate the first state set corresponding to the tile using the symmetry result. Collapse execution steps: Calculate the first entropy value corresponding to the tile through the first state set, determine the target tile in the tile based on the wave function collapse result corresponding to the first entropy value, and obtain the first map element state data corresponding to the target tile after random collapse processing of the target tile; Entropy calculation steps: Traverse the adjacent tiles corresponding to the target tile, determine the second state set corresponding to the adjacent tiles based on the edge connection rules corresponding to the adjacent tiles, and use the second state set to calculate the second entropy value corresponding to the target tile; Constraint propagation steps: Based on the wave function collapse result corresponding to the second entropy value, the adjacent tiles are sequentially subjected to random collapse processing to obtain the second map element state data corresponding to the adjacent tiles; Map generation steps: Construct the map grid using the first map element state data and the second map element state data, and generate the map corresponding to the map generation instruction based on the map grid; The collapse execution step includes: obtaining the number of types corresponding to the state type parameters of the tiles in the first state set; calculating the first entropy value corresponding to the tile using the number of types; determining the tile corresponding to the minimum entropy value corresponding to the first entropy value in the first state set as the target tile; performing random collapse processing on the target tile; and determining the first map element state data corresponding to the target tile using the road data, surface data, and building data corresponding to the target tile. The entropy calculation step includes: after detecting that the target tile has completed the collapse process, obtaining the adjacent tiles corresponding to the target tile according to the position parameters of the target tile, and determining the edge connection rules between the adjacent tiles; traversing the adjacent tiles, using the edge connection rules to determine the symmetry result of the adjacent tiles, and using the symmetry result to determine the second state set corresponding to the adjacent tiles; obtaining the number of types corresponding to the state type parameters of the adjacent tiles in the second state set, and using the number of types to calculate the second entropy value corresponding to the tile; The constraint propagation step includes: performing random collapse processing on the adjacent tiles sequentially based on the wave function collapse result corresponding to the second entropy value, and determining whether the second state set is empty; if not, determining the second map element state data corresponding to the target tile using the road data, surface data and building data corresponding to the adjacent tiles.

2. The map generation control method according to claim 1, characterized in that, The initialization step includes: After receiving the map generation instruction from the requesting end, the server receives the map generation instruction based on a preset configuration and initializes the map grid using the map generation instruction. The type parameters of the tiles in the map grid are determined based on the surface area type parameters, road structure type parameters, and building type parameters in the map grid. The reference surface and rotation axis corresponding to the tile are determined using the type parameters, and the rotation parameters corresponding to the tile are determined using the reference surface and rotation axis. The symmetry result corresponding to the tile is determined based on the rotation parameters; wherein, if all six reference planes of the tile are the same, the symmetry result corresponds to 1 state data; if the horizontal reference groups of the tile are the same but the vertical reference groups are different, the symmetry result corresponds to 2 state data; if the three sets of reference planes of the tile are the same but the mode parameters between each set are different, the symmetry result corresponds to 6 state data; if each reference plane of the tile is different, the symmetry result corresponds to 24 state data. Based on the number of types corresponding to the state data in the symmetry result, a first state set corresponding to the tile is generated.

3. The map generation control method according to claim 1, characterized in that, Before performing random collapse processing on the target tile, the collapse execution step further includes: The global state data of the tile corresponding to the map grid is determined based on the coordinate parameters, collapse flag, current state type parameter and the first state set of the tile. The global state data of the tiles is cached in a preset history stack.

4. The map generation control method according to claim 3, characterized in that, The constraint propagation step also includes: If the second state set is empty, then the global state data of the adjacent tile is obtained using the time parameter corresponding to the historical record stack; The second state set is updated based on the global state data of the tiles, and the marking results of the adjacent tiles are updated to the historical record stack.

5. The map generation control method according to claim 1, characterized in that, The map generation steps include: Determine whether the adjacent tiles have completed the collapse process; If yes, then the map grid is constructed using the first map element state data and the second map element state data; if no, then the constraint propagation step is executed.

6. A map generation and control system, characterized in that, The system includes: Initialization module: used to initialize map grid based on received map generation instructions, determine the symmetry result corresponding to the tile according to the type parameters and rotation parameters of the tile corresponding to the map grid, and generate the first state set corresponding to the tile using the symmetry result; Collapse execution module: used to calculate the first entropy value corresponding to the tile through the first state set, determine the target tile in the tile based on the wave function collapse result corresponding to the first entropy value, and obtain the first map element state data corresponding to the target tile after performing random collapse processing on the target tile; Entropy calculation module: used to traverse the adjacent tiles corresponding to the target tile, determine the second state set corresponding to the adjacent tiles based on the edge connection rules corresponding to the adjacent tiles, and use the second state set to calculate the second entropy value corresponding to the target tile; Constraint propagation module: used to perform random collapse processing on the adjacent tiles sequentially based on the wave function collapse result corresponding to the second entropy value, and then obtain the second map element state data corresponding to the adjacent tiles; Map generation module: used to construct the map grid using the first map element status data and the second map element status data, and generate the map corresponding to the map generation instruction based on the map grid; The collapse execution module is further configured to: obtain the number of types corresponding to the state type parameters of the tiles in the first state set; calculate the first entropy value corresponding to the tile using the number of types; determine the tile corresponding to the minimum entropy value corresponding to the first entropy value in the first state set as the target tile; and after performing random collapse processing on the target tile, determine the first map element state data corresponding to the target tile using the road data, surface data, and building data corresponding to the target tile. The entropy calculation module is further configured to: after detecting that the target tile has completed the collapse process, obtain the adjacent tiles corresponding to the target tile according to the position parameters of the target tile, and determine the edge connection rules between the adjacent tiles; traverse the adjacent tiles, use the edge connection rules to determine the symmetry result of the adjacent tiles, and use the symmetry result to determine the second state set corresponding to the adjacent tiles; obtain the number of types corresponding to the state type parameters of the adjacent tiles in the second state set, and use the number of types to calculate the second entropy value corresponding to the tile; The constraint propagation module is further configured to: perform random collapse processing on the adjacent tiles sequentially based on the wave function collapse result corresponding to the second entropy value, and determine whether the second state set is empty; if not, determine the second map element state data corresponding to the target tile using the road data, surface data and building data corresponding to the adjacent tiles.

7. A server, characterized in that, The server includes a processor and a memory, the memory storing computer-executable instructions that can be executed by the processor, the processor executing the computer-executable instructions to implement the steps of the map generation control method according to any one of claims 1 to 5.

Images