Search

CN-122021983-A - Intelligent partition sequencing and line planning method for building distributing machine

CN122021983ACN 122021983 ACN122021983 ACN 122021983ACN-122021983-A

Abstract

The invention discloses an intelligent partition sequencing and line planning method of a building material distributor, which comprises the steps of dividing a construction floor plane into a plurality of regular partitions according to grid data of a building total graph, constructing a graph model based on a communication relation among the partitions, searching a longest path with the most access to the most vertexes in the graph by utilizing a depth-first search algorithm, judging and constructing Euler paths aiming at the vertexes which are not covered by the path, intelligently repartitioning the corresponding partitions to realize a global path capable of repeatedly traversing all the partitions at one time, planning an arch-shaped operation path meeting the requirements of full coverage, high efficiency and obstacle avoidance for the interior of each partition by adopting a genetic algorithm, and finally carrying out construction simulation in a Unity3D digital twin system based on all path results, and dynamically calculating the travelling speed and concrete flow of the material distributor. The invention solves the problem that the distributing machine cannot be used for high-efficiency continuous pouring under the complex partition layout, and remarkably improves the construction efficiency and the building quality.

Inventors

  • ZHANG KAI
  • ZHANG HAIYANG
  • HU ZHENGHUAN
  • LIU SENQI
  • LI YINGSONG
  • YANG YUHENG
  • ZHANG FAN
  • MA QIANG

Assignees

  • 武汉科技大学

Dates

Publication Date
20260512
Application Date
20251027

Claims (8)

  1. 1. The intelligent partition sequencing and line planning method for the building material distributor is characterized by comprising the following steps of: Step 1, dividing a construction floor plane into a plurality of rectangular subareas according to grid data of a building total diagram, judging a communication relation between the subareas according to coordinates of the subareas and barrier information, taking a geometric center of each subarea as one vertex of the building total diagram, and establishing edges between vertexes corresponding to the communicated subareas so as to construct a diagram model; Step 2, based on the graph model constructed in the step 1, searching a longest path which can access the most vertexes and is passed through each vertex only once by adopting a depth-first search algorithm; Step 3, dividing the adjacent partition which is communicated with the residual vertex into a first sub-partition and a second sub-partition aiming at the residual vertex which is not covered by the longest path obtained in the step 2, wherein the first sub-partition is close to the adjacent edges of the residual vertex corresponding partition and the adjacent partition, the second sub-partition is far away from the adjacent edges of the residual vertex corresponding partition and the adjacent partition, the first sub-partition is divided into a third sub-partition and a fourth sub-partition, the third sub-partition and the fourth sub-partition are adjacent to the residual vertex corresponding partition, and the longest path is transformed into a global path which can be repeatedly traversed all the partitions at one time by adding virtual edges of the second sub-partition, the third sub-partition, the residual vertex and the fourth sub-partition to obtain a global path sequence which is formed by the partitions and the sub-partitions through which the global path sequentially passes; step 4, optimizing an arch-shaped operation path in each partition and sub-partition in the sequence by adopting a genetic algorithm according to the global path sequence obtained in the step 3, wherein the path needs to meet the requirements of full coverage, obstacle avoidance and inter-area connection fluency; and 5, performing visual simulation of the construction process in a Unity3D digital twin system based on all the working paths obtained in the previous steps, and dynamically calculating the advancing speed and the concrete flow of the distributing machine.
  2. 2. The method according to claim 1, wherein in step 1, the judging the communication relationship between the partitions is specifically judged by: the minimum gap distance in the horizontal direction of the two partitions satisfies: wherein Representing grid coordinates; representing the wide lengths of the two partitions, respectively; is a preset gap tolerance parameter; the alignment deviation of the two partitions in the vertical direction satisfies: wherein Representing grid coordinates; Is a preset alignment tolerance parameter; The ratio of the obstacle on the joint edges of the two subareas is lower than a preset threshold value of the ratio of the obstacle 。
  3. 3. The method according to claim 1, wherein in step 2, the depth-first search algorithm adopts a main stack and an auxiliary stack to manage a search state and a backtracking process, and specifically comprises: the initialization stage, namely pressing the initial node into a main stack, and emptying an auxiliary stack; The deep search stage comprises the steps of taking out a trestle top node of a main stack as a current node when the main stack is not empty, searching an adjacent node set of the trestle top node, selecting one node to press the main stack if the adjacent node which is not accessed exists, and pressing a residual adjacent node list which is not accessed into an auxiliary stack; And in the path recording stage, when the top node of the main stack is the same as the target node, recording the node sequence in the current main stack as a feasible path.
  4. 4. The method according to claim 1, wherein in step 3, the corresponding physical partition is subdivided, and the partition is divided into three blocks by a T-shaped cutting method or two blocks by a horizontal-vertical cutting method.
  5. 5. The method of claim 1, wherein in step 4, the partition internal paths are optimized using genetic algorithm with the objective of maximizing a comprehensive evaluation function : In the formula, Is a step weight, Is an efficiency weight; the degree of coverage integrity is indicated as such, , , Representing an effective coverage area; representing the total area of the partition; representing the total length of the path; representing a job width of the apparatus; the average working speed is indicated as such, , Indicating the total amount of material required for the partition.
  6. 6. The method according to claim 5, wherein in step 4, the specific iterative steps of the genetic algorithm comprise: ① Initializing a population, namely randomly sampling from a predefined step size parameter set to generate an initial population containing a plurality of individuals, wherein each individual represents a step size scheme; ② Calculate fitness-calculate the comprehensive evaluation function value of its path scheme for each individual in the population As a fitness; ③ Selecting and retaining elite, namely screening individuals with the highest fitness in the current population, and storing the individuals in the elite population; ④ For non elite individuals, randomly selecting a parent from elite population to perform crossing and mutation operation to generate offspring, wherein the crossing operation is to exchange part of step length values, and the mutation operation is to apply random disturbance to the step length values; ⑤ Selecting the winner and the bad, namely comparing the newly generated offspring individual with the original father, and if the offspring has higher adaptability, replacing the original father; ⑥ Combining elite individuals and updated individuals into a next generation population, and repeating the steps ② to ⑤ until the maximum evolution algebra is reached.
  7. 7. The method according to claim 1, wherein in step 4, the planning of the arcuate job path requires determining a start edge and a stop edge of the partition, the start edge selecting a side away from the obstacle, the stop edge selecting a side closer to the next partition to be accessed.
  8. 8. The method of claim 1, wherein in step 5, the dynamic calculation of the traveling speed of the spreader and the concrete flow rate is performed in combination with the path length, the casting area, and the partition size parameters.

Description

Intelligent partition sequencing and line planning method for building distributing machine Technical Field The invention relates to the technical field of building construction, in particular to an intelligent partition sequencing and line planning method of a building spreader, which is particularly suitable for intelligent partition, path optimization and spreader operation control of building floor areas in the concrete pouring process. Background In modern building construction, the partition pouring of concrete by adopting a building spreader is a key link for completing the construction of a main structure of a building. In order to improve the construction efficiency and the structural quality, the floor plane is reasonably divided into a plurality of construction partitions according to a design drawing, and a working path which can sequentially cover all the partitions is planned for the distributing machine. Ideally, the spreader can be used to continuously traverse all the zones at one time without repetition to avoid structural quality problems caused by uneven concrete setting. However, due to the complex spatial topology between the partitions in the building floor plan, there are a large number of irregularities and obstructions that make it extremely difficult to find an optimal path that can traverse all the partitions without repetition. This path planning problem can be essentially abstracted as a "one stroke" problem or euler path problem in graph theory. As shown in the classical Goniosburgh seven-bridge problem, not all connected graphs have paths which do not repeatedly traverse all sides, and when more than two odd vertexes exist in the graph, one stroke cannot be realized. In the actual cloth machine path planning, the path meeting the conditions cannot be found due to the inherent characteristics of the partition layout by directly applying the theory, so that the problem that the planning result is not covered fully or needs to repeatedly pass through certain areas exists, and the continuity and the efficiency of construction are seriously affected. The existing path planning method is difficult to simultaneously meet the engineering requirements of full coverage, high efficiency and no repeated traversal. The method can not process the optimal sorting problem under the complex partition layout, so that the paths are missed or traced in a large amount, or the joint with the global paths, obstacle avoidance and cooperative control of pouring parameters (such as speed and flow) are not fully considered in the process of planning the internal paths of the partitions. Thus, there is an urgent need for an integrated solution that can intelligently handle partition ordering and can construct efficient paths by dynamically adjusting partition topology. Disclosure of Invention Aiming at the defects of the prior art, the intelligent partition sequencing and line planning method for the building distributing machine is used for solving the problem that the distributing machine cannot realize one-time non-repeated full-coverage pouring path planning under complex building partition layout. The method finally generates efficient, continuous and complete global and local operation paths by abstracting physical partitions into graph models and performing intelligent reconstruction. The method comprises the following steps: Step 1, dividing a construction floor plane into a plurality of regular subareas according to grid data of a building total graph, judging a communication relation between the subareas according to coordinates of the subareas and barrier information, and establishing edges between the communicated vertexes by taking a geometric center of each subarea as one vertex of the graph so as to construct a graph model; Step 2, based on the graph model constructed in the step 1, searching a longest path which can access the most vertexes and is passed through each vertex only once by adopting a depth-first search algorithm; Step 3, dividing an adjacent partition which is communicated with the residual vertex into a first sub-partition and a second sub-partition aiming at the vertex which is not covered by the longest path obtained in the step 2, wherein the first sub-partition is close to the adjacent edges of the residual vertex corresponding partition and the adjacent partition, the second sub-partition is far away from the adjacent edges of the residual vertex corresponding partition and the adjacent partition, the first sub-partition is divided into a third sub-partition and a fourth sub-partition, the third sub-partition and the fourth sub-partition are adjacent to the residual vertex corresponding partition, and the longest path is transformed into a global path which can be repeatedly traversed all the partitions at one time by adding virtual edges of the second sub-partition, the third sub-partition, the residual vertex and the fourth sub-partition by utilizing a Euler path pr