An intelligent path planning method for concrete placing machine based on evolutionary constrained optimization

Abstract

This invention discloses an intelligent path planning method for a concrete placing boom based on evolutionary constraint optimization. This method plans the global path of the placing boom operating within a target work area map. The method includes obtaining several sub-grid regions and obstacles divided by the target work area map; obtaining the optimal access order for each sub-grid region using a genetic algorithm; obtaining sub-paths within each sub-grid region using a genetic annealing algorithm; and connecting all sub-paths to obtain the global path. This invention, leveraging the principles of evolutionary algorithms, significantly improves the path planning efficiency and execution flexibility of the placing boom in complex working environments, enabling it to achieve optimal performance in both global and local path optimization.

Description

Technical Field

[0001] This invention relates to the field of concrete placing boom path planning, and more specifically, to a path planning method for planning the global path of a concrete placing boom operating in a target work area map. Background Technology

[0002] In highly complex operational scenarios such as building construction and automatic control of concrete placing booms, path planning is crucial for ensuring efficient equipment operation. However, in these scenarios, placing booms often need to perform precise operations in obstacle-filled and irregular working environments, posing a significant challenge to path planning. Traditional path planning algorithms struggle to adapt to such complex environments and cannot guarantee accurate coverage and safe obstacle avoidance across the entire work area.

[0003] Existing full-coverage path planning algorithms used in fields such as concrete placing booms and building construction have many shortcomings in practical applications. While the "ox-plowing" method is simple and efficient, it has poor adaptability to irregular areas and struggles to handle complex obstacles; the spiral method offers compact paths, but suffers from low coverage and flexibility in environments with dense boundaries and obstacles; the random walk method, despite its exploratory nature, has low path planning efficiency and uneven coverage; graph-based path planning methods can adapt to complex scenarios and achieve global optimization, but their computational complexity is high, especially in large-scale or dynamic environments where they are extremely time-consuming. These shortcomings make it difficult for existing technologies to fully meet the requirements in terms of accurate coverage, path optimization, and adaptability to complex environments. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this application provides an intelligent path planning method for a concrete placing boom based on evolutionary constraint optimization, used to plan the global path of the concrete placing boom operating in a target work area map, including the following steps:

[0005] Obtain several sub-grid regions and obstacles divided from the target area map;

[0006] Multiple random sorting sequences composed of all sorted subgrid regions are obtained. The sorting of subgrid regions in the random sorting sequences is transformed using a transformation operation to generate a mutated sorting sequence. The preferred mutated sorting sequence is selected from the mutated sorting sequences to replace the random sorting sequence, and the transformation operation is repeated to obtain the access sequence. The preferred mutated sorting sequence, compared with the replaced random sorting sequence, has fewer intersections in the path formed by the sequence nodes according to the sorting order, and the span between two consecutive subgrid regions in the sequence is smaller. The sequence nodes are subgrid regions.

[0007] Obtain the sub-path that covers the entire sub-grid region within each sub-grid region;

[0008] The sub-path is reconstructed to obtain an obstacle avoidance path, which avoids obstacles and maintains a safe distance from them.

[0009] According to the access sequence, the sub-grid regions that are sequentially adjacent to each other are traversed in turn. The connection path between the end point of the sub-path or obstacle avoidance path in the previous sub-grid region and the start point of the sub-path or obstacle avoidance path in the next sub-grid region is obtained. The connection path has no intersection with the obstacle. The adjacent sub-paths or obstacle avoidance paths are connected sequentially using the connection path to obtain the candidate global path.

[0010] A global path is selected from the candidate global paths, and the global path whose connecting path has the fewest intersections with other connecting paths, obstacle avoidance paths and sub-paths.

[0011] Furthermore, the step of transforming the sorting of subgrid regions in the random sorting sequence using transformation operations to generate a mutated sorting sequence includes:

[0012] Region swapping: Randomly select two subgrid regions from a random sorting sequence and swap their positions in the random sorting sequence;

[0013] Interval Reversal: Select an interval consisting of multiple consecutive subgrid regions from a random sorting sequence, and reverse the sorting of the subgrid regions within that interval;

[0014] Neighborhood fine-tuning: swapping the order of sequentially adjacent subgrid regions in a random sorting sequence;

[0015] The transformation operation randomly selects one of three transformation operations to generate a mutated sorting sequence. The number of mutated sorting sequences before and after the transformation is equal to the number of random sorting sequences.

[0016] Furthermore, the step of selecting a preferred mutated sorting sequence from the mutated sorting sequence to replace the random sorting sequence includes:

[0017] Calculate the objective function for random sorting sequences and mutated sorting sequences; where a smaller objective function indicates fewer intersections in the paths formed by the sorting order of the sequence nodes and / or a smaller span between two consecutive subgrid regions in the sequence.

[0018] If the objective function of the mutated sorting sequence is smaller than that of the random sorting sequence, then the mutated sorting sequence is preferred over the random sorting sequence.

[0019] The repetition ends when the set number of repetitions is reached or the number of selected preferred mutation sorting sequences converges.

[0020] Furthermore, the objective function includes a path intersection penalty term and a region selection constraint penalty term.

[0021] F(τ)=u·Σ 1≤p<q≤k (intersect(τ0))+λ·Σ 1≤p<q≤k constraint(τ p ,τ q );

[0022] Where: u is the weight of the path intersection penalty, λ is the weight of the region selection constraint, and Σ 1≤p<q≤k (intersect(τ0)) represents the number of intersections in the path formed by the sorted order of the sequence nodes, and constraint(τ) represents the number of intersections in the path formed by the sorted order of the sequence nodes. p ,τ q ) represents the penalty coefficient for region selection constraints, τ p ,τ q It is the p-th and q-th subgrid region in a random sorting sequence or a mutated sorting sequence, where q = p + 1;

[0023]

[0024]

[0025] τ p ,τ p Adjacent indicates that the boundary coordinates of two sub-grid regions are adjacent; Δ max For the maximum allowable span, d(τ) p ,τ p ) represents two sub-grid regions τ p and τ p The distance from the center point is δ, which is greater than 1.

[0026] Furthermore, obtaining the sub-paths covering the entire sub-grid region within each sub-grid region includes:

[0027] The subgrid area is divided into a grid map composed of grid cells using mutually perpendicular meridians and parallels; each grid cell without obstacles is marked as passable, and each grid cell with obstacles is marked as impassable.

[0028] Detect each sub-grid region and its next adjacent sub-grid region in the access sequence, find the starting edge and ending edge of each sub-grid region in turn, select the edge in each sub-grid region that is far away from the obstacle as the starting edge, and select the edge in each sub-grid region that is close to the next sub-grid region as the ending edge.

[0029] Select a grid cell from a column or row of grid cells near the starting edge of the sub-grid region as the starting point of the sub-path. Extend the sub-path within the sub-grid region from a direction parallel to the starting edge to form a meandering sub-path composed of interconnected path segments. Adjacent path segments at bends are perpendicular to each other. The end point of the sub-path is located at the terminating edge of the sub-grid region. The sub-path includes several rows of path segments parallel to the starting edge. The sub-path covers the entire sub-grid region and does not intersect within the sub-grid region.

[0030] Furthermore, the process of reconstructing the sub-path to obtain the obstacle avoidance path includes:

[0031] If an intersection portion in the sub-path that coincides with or contacts a grid cell containing an obstacle is detected, the intersection portion is offset to obtain an offset path, and the offset path maintains a safe distance from the obstacle.

[0032] Replace the path segment with an offset path.

[0033] Furthermore, the step of sequentially traversing the adjacent sub-grid regions according to the access sequence to obtain the connection path between the end point of the sub-path or obstacle avoidance path in the previous sub-grid region and the start point of the sub-path or obstacle avoidance path in the next sub-grid region also includes:

[0034] Expand the raster map of the sub-grid area to the entire target area map, mark grid cells without obstacles as passable, and mark grid cells with obstacles as impassable;

[0035] If the connection path is detected to pass through an impassable grid cell, the connection path is offset to bypass the impassable grid cell, and a safe distance is maintained between the connection path and the impassable grid cell.

[0036] Furthermore, the step of obtaining the sub-path covering the entire sub-grid area within each sub-grid area also includes: preferably the step size between several row path segments parallel to the starting edge, so as to maximize the adaptability of the sub-path set according to the step size;

[0037] Fitness is: F(l) = αF coverage (l)+γv avg (R i )

[0038] Where α and γ are the weighting coefficients for coverage and speed, respectively;

[0039] F coverage (l) represents the coverage rate.

[0040] A covered (R i )=L(Ri )·W,L(R i ) represents the subpath length, and W represents the coverage width;

[0041] The total area of ​​the subgrid region is A(R) i The area already covered is A. covered (R i );

[0042] v avg (R i () represents the regional path speed.

[0043] The amount of cement required to be poured in the sub-grid area is M(R) i The total length of the sub-paths is L(R). i ).

[0044] Furthermore, the preferred step includes optimizing the step size using a genetic annealing algorithm:

[0045] Population initialization: Starting from the step size set {l1, l2, ..., l n A set of step sizes is randomly selected as the initial population, and each step size l i As an individual, subpaths are created based on the subpath acquisition steps, and the fitness F(l) of each individual is calculated. i );

[0046] Selection operation: In each generation of the genetic algorithm, individuals in the current population are sorted from largest to smallest according to their fitness value, and the individuals with the top 10% fitness values ​​are selected as superior individuals and placed into the elite population Q.

[0047] A superior individual is randomly selected from the elite population Q as the parent and crossover and mutation operations are performed with the current individual in the current population to generate a new individual. The crossover operation includes: exchanging the step size between some path segments parallel to the starting edge in the superior individual and the current individual; the mutation operation is: applying a random bias to the step size between some path segments parallel to the starting edge in the superior individual and the current individual.

[0048] Replace the current individual according to the simulated annealing acceptance criteria;

[0049] Merge superior individuals with new individuals to generate the next generation of the population;

[0050] If the number of generations reaches the set maximum value, the iteration ends; otherwise, the genetic algorithm operation is repeated for iteration.

[0051] The simulated annealing acceptance criteria are as follows:

[0052]

[0053] Where T represents the current temperature, which gradually decreases as the number of iterations increases, and T = T0·α n T0 is the initial temperature, α∈(0,1) is the temperature decay coefficient, and n is the current iteration number; F(l new F(l) represents the fitness of the new individual. current ) represents the current fitness of the individual, l new It is a new individual, l current is the current individual, and p is a number generated randomly in each iteration that is greater than 0 and less than 1.

[0054] The beneficial effects of this invention are as follows: By proposing a path planning method for placing booms based on a hierarchical genetic annealing algorithm, an innovative solution for efficient path planning of placing booms in complex industrial environments is proposed. This method can not only effectively sort sub-regions at the global level to reduce path intersections, but also achieve efficient coverage and obstacle avoidance functions in local path planning through genetic and simulated annealing algorithms. Finally, through global and local collaborative optimization, this method achieves high precision and high adaptability in placing boom path planning, and can give full play to its advantages in obstacle-dense and irregular operating scenarios, providing reliable technical support for the efficient operation of placing booms. Attached Figure Description

[0055] Figure 1 This is a path pattern diagram for the sub-grid region;

[0056] Figure 2 This is a path obstacle avoidance map for a sub-grid region. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the described embodiments of this disclosure without creative effort are within the scope of protection of this disclosure.

[0058] Example 1

[0059] An intelligent path planning method for a concrete placing boom based on evolutionary constraint optimization is used to plan the global path of the concrete placing boom operating in the target work area map, including the following steps:

[0060] Sub-region division: Obtain the boundary information of the target area through map data, and combine it with the location of obstacles to divide the entire target area into multiple sub-grid areas, ensuring that each sub-grid area has no intersection with the obstacles;

[0061] Subgrid region access sequence sorting: Obtain multiple random sorting sequences composed of all subgrid region sortings. Use transformation operations to transform the sorting of subgrid regions in the random sorting sequences to generate mutated sorting sequences. Repeatedly select the preferred mutated sorting sequence from the mutated sorting sequences to replace the random sorting sequences, and repeat the transformation operation to obtain the access sequence. The preferred mutated sorting sequence, compared with the replaced random sorting sequence, has a smaller objective function value. A smaller objective function value means fewer intersections in the path formed by the sequence nodes according to the sorting order, and a smaller span between two consecutive subgrid regions in the sequence. The sequence node is a subgrid region. The span refers to the center distance between non-adjacent subgrid regions in the map data.

[0062] Sub-path optimization within sub-grid regions: Obtain sub-paths covering the entire sub-grid region within each sub-grid region, and reconstruct the sub-paths to obtain obstacle avoidance paths. The obstacle avoidance paths do not intersect with the obstacles and are separated from the obstacles by a safe distance.

[0063] Optimization of connection paths between sub-grid regions: According to the access sequence, traverse the sub-grid regions that are sequentially adjacent to each other, obtain the connection path between the end point of the sub-path or obstacle avoidance path in the previous sub-grid region and the start point of the sub-path or obstacle avoidance path in the next sub-grid region. The connection path has no intersection with the obstacle. Use the connection path to sequentially connect the adjacent sub-paths or obstacle avoidance paths to obtain candidate global paths.

[0064] A global path is selected from the candidate global paths, and the global path whose connecting path has the fewest intersections with other connecting paths, obstacle avoidance paths and sub-paths.

[0065] Specifically, the subgrid region access sequence sorting aims to optimize the sorting of subgrid regions, connecting them in a certain order to minimize the intersections formed by the paths created by the sequence nodes in the sorted order. To achieve this goal, this invention proposes a region sorting optimization method based on the neighborhood large-scale search (LNS) algorithm. By introducing region selection constraints, the order of subgrid regions is gradually adjusted to ensure the rationality of the sorted paths and minimize intersections.

[0066] The specific optimization sorting process includes the following key steps:

[0067] 1. Initialize the sorting scheme: First, abstract each sub-grid region as a point and generate a random sorting scheme τ0 = (τ1, τ2, ..., τ0). k The initial sequence is used as the parent individual. This initial sequence serves as the starting point for the optimization process, and the objective function value F(τ0) of the randomized sequence is calculated. This value will be used to evaluate the merits of subsequent optimization schemes.

[0068] The objective function design includes path intersection penalty terms and region selection constraint penalty terms. The objective function is defined as follows:

[0069]

[0070] Where: τ p ,τ q λ is the p-th and q-th consecutive sub-grid regions in the sorting scheme, where q = p + 1; u is the weight of the path intersection penalty, used to control the impact of intersections on paths; λ is the weight of the region selection constraint, ensuring that the combination of sub-grid regions meets the requirements of the actual application scenario.

[0071] The components of the objective function are explained as follows: ∑ 1≤p<q≤k (intersect(τ0)): Represents the number of intersections in the path formed by the sequence nodes of the random sorting scheme according to the sorting order. It is used to measure the intersection situation of the paths in the sorting scheme. After each sub-grid region is abstracted into points, it is connected into a complete path by sorting. The number of intersections in the path is counted as a measure of path complexity. The more intersections, the higher the path complexity. Therefore, the optimization process will increase the penalty for this scheme to guide the path to be optimized in a smoother direction with fewer intersections.

[0072] constraint(τ p ,τ q ): This represents the penalty coefficient for region selection constraints, used to evaluate whether the random sorting scheme violates constraints such as span. If the sorting scheme exceeds the maximum allowed span limit, this term will impose a larger penalty to guide the optimization process towards convergence in the direction that satisfies the constraints.

[0073] Region selection constraint penalty settings:

[0074] To improve the rationality of sorting and meet the requirements of practical applications, the following region selection constraint penalty terms are defined:

[0075] Adjacent region priority constraints:

[0076] To ensure that physically close subgrid regions are arranged adjacently during sorting, the adjacency function is defined as follows:

[0077]

[0078] Calculate the two consecutive subregions τ in the sequence in sequence. p and τ q The adjacency situation is determined based on the two sub-grid regions τ. p and τ q Determine the boundary coordinates of the two sub-grid regions τ p and τ qAre they completely adjacent? If they are adjacent, then adjacency(τ) p ,τ q If adjacency(τ) = 0; if they are not adjacent, then adjacency(τ) = 0. p ,τ q ) = 1;

[0079] For cases where the sequence consists of consecutive but non-adjacent subgrid regions, a maximum allowable span constraint is set:

[0080] To avoid excessive span in the path, a maximum span constraint Δ is set. max Define d(τ) p ,τ q ) represents two consecutive but non-adjacent subgrid regions τ in the sequence. p and τ q The distance from the center point;

[0081]

[0082] For two consecutive subregions τ in sequence sorting p and τ p If the elements are adjacent, the penalty coefficient is 0; if they are not adjacent, d(τ) is further calculated. p ,τ p )

[0083] Where δ is the penalty coefficient for exceeding the allowed span. δ is greater than 1 to ensure that the regions are not too scattered in the sorting. The span refers to the center distance of the sub-grid regions that are consecutive but not adjacent in the sequence sorting of the map data.

[0084] 2. Local Search and Adjustment: Based on the current parent individuals, the random sorting scheme is refined to obtain a mutated sorting sequence. This includes the following operations:

[0085] Region swap: Randomly select two sub-grid regions within the parent individual. and They then swap their positions within the parent individual. This sub-grid region swapping reduces intersections while maintaining overall structural stability, balancing path continuity and rationality.

[0086] Interval Reversal: Selecting an interval τ consisting of multiple consecutive sub-grid regions from the parent individual. i:j This operation reverses the sorting of subgrid regions within the interval. The aim is to break the inherent path intersections within the intervals, reduce path complexity through reverse sorting, and improve the overall rationality of the sorting.

[0087] Neighborhood Fine-tuning: This function fine-tunes the order of sequentially adjacent sub-grid regions within a path by incorporating neighborhood information. This operation is particularly useful when there are many local path intersections. By making small adjustments to sequentially adjacent sub-grid regions, it further reduces the number of path intersections and satisfies region selection constraints.

[0088] During local search and adjustment, one of the three transformation operations is randomly selected for mutation. The number of mutated sorted sequences before and after the transformation is equal to the number of randomly sorted sequences.

[0089] 3. Objective Function Recalculation and Constraint Check: After each local adjustment, the objective function value F(τ) of the new mutated sorting sequence is recalculated, and constraint conditions are checked. Specifically, it is ensured that the order adjustment of all sub-grid regions meets the requirements of adjacent region priority constraint and maximum span constraint. If the mutated sorting sequence does not meet the constraints as well as the parent individual, a penalty weight is applied to ensure that the mutated sorting sequence is reasonably restricted by the constraint mechanism during evaluation.

[0090] 4. Adaptive Optimization Strategy: Based on the comparison between the mutated sorting sequence and the objective function value of the parent individual, a decision is made on whether to update the sorting scheme. If the objective function value of the mutated sorting sequence is lower than that of the parent individual, the parent individual is replaced with the new parent individual using the mutated sorting sequence. Otherwise, the original parent individual is kept unchanged to ensure the convergence of the optimization process.

[0091] 5. Termination and Output of Optimal Solution: When the set number of iterations is reached or the optimization effect of the solution tends to stabilize, the algorithm stops and outputs the optimal sorting scheme τ that satisfies all constraints. * As an access sequence, this scheme maximizes the rationality of region sorting while minimizing path intersections, meeting the needs of the application scenario.

[0092] This method strikes a good balance between local search and global optimization, making region sorting not only optimized in terms of path intersection, but also more practical under the constraints of real-world applications.

[0093] Specifically, sub-path optimization within a sub-grid region includes:

[0094] The subgrid region is divided into a raster map composed of grid cells by mutually perpendicular meridians and parallels; each grid cell is marked as passable or impassable, that is, F(x,y)=0 indicates that the grid cell is passable, while F(x,y)=1 indicates that the grid cell is impassable.

[0095] According to the access sequence, find the starting edge and ending edge of each sub-grid region in turn. Select the edge in each sub-grid region that is far away from the obstacle as the starting edge, and select the edge in each sub-grid region that is close to the next sub-grid region as the ending edge. The next sub-grid region is the sub-grid region that is sequentially adjacent to the sub-grid region according to the access sequence.

[0096] Select a grid cell from a column or row of grid cells near the starting edge of the sub-grid area as the starting point of the sub-path. The sub-path is typically planned in a "bow-shaped" pattern. When planning a bow-shaped path, it starts from the starting edge of the sub-grid area, proceeds parallel to the starting edge, and upon reaching the boundary of that row, the concrete placing boom turns and enters the adjacent row, continuing in the opposite direction, row by row, until it covers the entire sub-grid area and reaches the endpoint at the termination edge, forming a meandering sub-path composed of interconnected path segments. The spacing between adjacent rows is the step size. This path pattern is suitable for the working characteristics of the concrete placing boom, namely, simultaneously applying cement during its movement. Since cement naturally flows to both sides after spraying, choosing a bow-shaped path ensures that each movement is parallel and adjacent to the previous row, reducing omissions and overlaps, and helping to cover the entire sub-area.

[0097] like Figure 1 As shown, by adjusting different step sizes and different directions, there are different combinations of sub-paths. The purpose of this step is to select a suitable sub-grid area planning mode.

[0098] Adjacent path segments at bends in a subpath are perpendicular to each other, and the subpath covers the entire subgrid area without intersecting within the subgrid area.

[0099] Obtaining the sub-paths covering the entire sub-grid region within each sub-grid region includes: selecting the step size between several row path segments parallel to the starting edge based on the genetic annealing algorithm.

[0100] In the current subgrid region R i Within the sub-path planning of the fabric placing machine, the selection of the step size *l* is crucial, as it directly impacts the covering efficiency and smoothness of the machine within the sub-grid area. Therefore, a genetic annealing algorithm is introduced. By leveraging the population search capability of the genetic algorithm and the acceptance criteria of simulated annealing, the optimal step size *l* is found. * .

[0101] In the step size optimization process, firstly, within the step size range l∈[l... min ,l max [Take multiple discrete step size values ​​l1, l2, ..., l] n Each step generates a sub-path. The cloth placing machine moves along these sub-paths from the sub-mesh region R. i The starting point S(R) iIt begins by moving in a reciprocating motion until it covers the entire subgrid area.

[0102] To find the optimal step size, the steps of the genetic annealing algorithm are as follows:

[0103] 1. Population initialization: Starting from the step size set {l1, l2, ..., l...} n A set of step sizes is randomly selected as the initial population, and each step size represents a candidate sub-path scheme. For each step size l i Subpaths are created based on the subpath generation rules of the fabric placing machine, and their fitness function F(l) is calculated. i ).

[0104] 2. Fitness function evaluation: To ensure that subpaths have good coverage and smooth inter-regional connections, the fitness function F(l) comprehensively considers the following factors: coverage and regional path speed;

[0105] The overall formula for the fitness function is:

[0106] F(l)=αF coverage (l)+γv avg (R i (5)

[0107] Where α and γ are the weighting coefficients for coverage and speed, respectively.

[0108] F coverage (l) Coverage: The fabric laying machine at R i The higher the coverage within a sub-path, the larger its fitness function value. Let the total area of ​​the sub-grid region be A(R). i The area already covered is A. covered (R i If ), then the coverage rate can be expressed as:

[0109]

[0110] Let the pouring width of the concrete placing machine be W, and A. covered (R i )=L(R i )·W;L(R i ) represents the subpath length, and W represents the coverage width.

[0111] v avg (R i ) represents the path speed within a region: assuming each sub-region has a path speed of R. i The required amount of cement is M(R) i The speed of the concrete placing boom should be adjusted according to the cement demand. This can be achieved by adjusting the cement quantity M(R). i Divide by the total length of the subpaths L(R) iTo calculate the average speed v of the placing boom in that area. avg (R i ), the formula is as follows:

[0112]

[0113] in:

[0114] M(R i ) is the fabric placing machine in sub-region R i The amount of cement required to be laid.

[0115] L(R i ) is the total path length of the placing boom within this sub-region.

[0116] 3. Selection operation: In each generation of the genetic algorithm, individuals in the current population are sorted from largest to smallest fitness value, and the top 10% of individuals in the fitness value ranking are selected as superior individuals and placed into the elite population Q; while retaining individuals with better step size, individuals with low fitness value are moderately retained to maintain population diversity.

[0117] 4. Crossover and Mutation Operations: Crossover: Randomly select a superior individual from the elite population Q as the parent and perform crossover and mutation operations with the current individual in the current population to generate a new individual. Exchange the step lengths between some path segments parallel to the starting edge in the superior individual and the current individual. The crossover operation utilizes existing gene combinations to generate new step lengths, potentially leading to better solutions. Mutation: Perform small mutations on the step lengths between some path segments parallel to the starting edge in the superior individual and the current individual, i.e., apply a random bias to the step length values. This introduces randomness into the search space, preventing the algorithm from getting trapped in local optima.

[0118] 5. Simulated Annealing Acceptance Criterion: In each generation update of the genetic algorithm, the simulated annealing acceptance criterion is used to determine whether to replace the current individual with a new individual. The acceptance criterion is as follows:

[0119]

[0120] Here, T represents the current temperature, which gradually decreases as the number of iterations increases. p is a randomly generated number greater than 0 and less than 1 during each iteration. This acceptance criterion allows the algorithm to accept moderately degraded solutions in the early stages to escape local optima, while in later stages it gradually tends to accept only optimal solutions.

[0121] 6. Temperature Decrease Strategy: Temperature T decreases algebraically, controlling the randomness of the acceptance criterion. The temperature decrease formula is:

[0122] T = T0·α n (6)

[0123] Where T0 is the initial temperature, α∈(0,1) is the temperature decay coefficient, and n is the current iteration number. The decreasing temperature causes the system to gradually accept better solutions, reducing the probability of worse solutions, and eventually converging to the global optimum.

[0124] 7. Termination Condition and Step Size Determination: After multiple generations of genetic evolution and simulated annealing optimization, the algorithm terminates when the population fitness value tends to converge or reaches the maximum number of iterations. At this point, the step size l with the highest fitness is selected. * This step size is considered the optimal step size. It performs excellently in terms of coverage and area connectivity, meeting the requirements of the fabric placement machine's path planning.

[0125] Specifically, reconstructing sub-paths to obtain obstacle avoidance paths includes:

[0126] If a path segment in a sub-path is detected to intersect with an obstacle in a sub-grid region, the path segment is offset to obtain an offset path, which maintains a safe distance from the obstacle.

[0127] Specifically, if a path segment coincides with or contacts a grid cell containing an obstacle, it is considered that the path segment intersects with the obstacle in the sub-grid region.

[0128] Locate the intersection point on the path segment where it coincides with or contacts a grid cell containing an obstacle. This intersection point is the section of the path where the placing boom needs to be adjusted to ensure it avoids the obstacle.

[0129] Replace the intersection portion with an offset path.

[0130] The goal of reconstructing the path segment is to minimize the deviation from the shape of the atomic path and ensure that the new path segment seamlessly connects to the unobstructed portion of the atomic path. Generating the offset path segment: When bypassing an obstacle location, each point (x′) of the offset path segment... p ,y′ p All offset path segments should maintain a safe distance from mesh cells containing obstacles to avoid the path being too close to the obstacle. The shape and direction of the offset path segments should be adjusted according to the direction of the placing boom's movement and the relative position of the mesh cells containing obstacles to ensure that the reconstructed offset path connects naturally with the atomic path.

[0131] Path connection: After the offset path segment that bypasses the obstacle is generated, the offset path segment is reconnected with the obstacle-free area of ​​the atomic path to form a continuous obstacle avoidance path, ensuring full coverage of the material placing machine within the sub-grid area.

[0132] Through the above obstacle avoidance processing, the concrete placing machine can effectively bypass obstacles next to sub-grid areas, ensuring that the obstacle avoidance path is not disturbed by obstacles, and maintaining the smoothness of the obstacle avoidance path and the overall coverage effect during the obstacle avoidance process. Figure 2 As shown.

[0133] Specifically, the optimization of connection paths between sub-grid regions aims to: Given a determined access order σ = (σ1, σ2, ..., σ...) for the sub-grid regions... k After that, different combinations of complete sub-paths are generated based on this order, and the optimal candidate global path is selected. total .

[0134] Each subgrid region Its internal subpaths have been generated Therefore, it is necessary to connect the sub-paths of each sub-grid area to ensure that the concrete placing machine can move smoothly and cover the entire working area.

[0135] For two adjacent subgrid regions in order of sorting and The fabric placing machine needs to start from the end point of the i-th sub-grid region. Move to the starting point of the next subgrid region The coordinates of the endpoint and the starting point are respectively and By generating connection paths between sub-mesh regions A complete candidate global path can be gradually formed.

[0136] Since different subgrid region orders correspond to different combinations of subgrid region connection paths, each sorting result will generate a unique candidate global path. To select the optimal solution, a fitness function F(path) is defined. total The fitness function measures the rationality of a path by the number of intersections between the candidate global path's connecting paths and other connecting paths, obstacle avoidance paths, and sub-paths. The fitness function is expressed as follows:

[0137] F(path total ) = cross(path) total )

[0138] Where, cross(path) total ) represents the candidate global path. total The fewer the number of intersections in the candidate global path, the smoother and more efficient the candidate global path. By minimizing the number of intersections in the candidate global path, unnecessary path interference can be avoided, thereby improving the working efficiency of the concrete placing machine.

[0139] By evaluating and filtering the number of intersections of sub-paths and connection paths in different sub-grid areas, the candidate global path with the fewest intersections is finally selected to ensure smooth connection and unobstructed coverage of the concrete placing machine throughout the entire working area.

[0140] Optimization of connection paths between sub-mesh regions also includes:

[0141] The grid map of the sub-grid area is expanded to the entire target area map. Grid cells without obstacles are marked as passable, and grid cells with obstacles are marked as impassable; that is, F(x,y)=0 indicates that the cell is passable, while F(x,y)=1 indicates that there is an impassable obstacle.

[0142] Intersection Detection: First, each path segment in the connecting path is checked to determine whether it intersects with a certain obstacle location in the obstacle set. If a path segment coincides with or touches an obstacle location, it is considered that the connecting path and the obstacle have an intersection.

[0143] Intersection localization: Once an intersection between a connecting path and an obstacle is detected, the path segment l is located. intersect The intersection section. The intersection path segment is the part of the path that the placing boom needs to adjust in this area to ensure that it avoids obstacles.

[0144] Path Reconstruction: The offset intersection path forms an offset intersection path segment that bypasses impassable grid cells, maintaining a safe distance between the offset intersection path segment and the impassable grid cells. The shape and orientation of the offset intersection path segment are adjusted according to the direction of travel of the concrete placing machine and the relative position of the obstacles to ensure that the reconstructed offset intersection path segment connects naturally with the original connecting path.

[0145] Path connection: After the offset connection path segment that bypasses the obstacle is generated, it is reconnected with the unobstructed area of ​​the original connection path to form a continuous connection path.

[0146] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

[0147] In summary, the method of this invention not only effectively sorts sub-regions at the global level to reduce path intersections, but also achieves efficient coverage and obstacle avoidance functions in local path planning through genetic and simulated annealing algorithms. Ultimately, through global and local collaborative optimization, this method achieves high precision and high adaptability in the path planning of the concrete placing boom, fully leveraging its advantages in obstacle-dense and irregular operating scenarios, and providing reliable technical support for the efficient operation of the concrete placing boom.

Claims

1. A method for intelligent path planning of a concrete placing boom based on evolutionary constraint optimization, used to plan the global path of the concrete placing boom operating in a target work area map, characterized in that, Includes the following steps: Obtain several sub-grid regions and obstacles divided from the target area map; Multiple random sorting sequences composed of all sorted subgrid regions are obtained. The sorting of subgrid regions in the random sorting sequences is transformed using a transformation operation to generate a mutated sorting sequence. The preferred mutated sorting sequence is selected from the mutated sorting sequences to replace the random sorting sequence, and the transformation operation is repeated to obtain the access sequence. The preferred mutated sorting sequence, compared with the replaced random sorting sequence, has fewer intersections in the path formed by the sequence nodes according to the sorting order, and the span between two consecutive subgrid regions in the sequence is smaller. The sequence nodes are subgrid regions. Obtain the sub-path that covers the entire sub-grid region within each sub-grid region; The sub-path is reconstructed to obtain an obstacle avoidance path, which avoids obstacles and maintains a safe distance from them. According to the access sequence, the sub-grid regions that are sequentially adjacent to each other are traversed in turn. The connection path between the end point of the sub-path or obstacle avoidance path in the previous sub-grid region and the start point of the sub-path or obstacle avoidance path in the next sub-grid region is obtained. The connection path has no intersection with the obstacle. The adjacent sub-paths or obstacle avoidance paths are connected sequentially using the connection path to obtain the candidate global path. A global path is selected from candidate global paths, wherein the global path has the fewest intersections with other connecting paths, obstacle avoidance paths, and sub-paths. The step of selecting a preferred mutated sorting sequence to replace the random sorting sequence from the mutated sorting sequence includes: Calculate the objective function for random sorting sequences and mutated sorting sequences; where a smaller objective function indicates fewer intersections in the paths formed by the sorting order of the sequence nodes and / or a smaller span between two consecutive subgrid regions in the sequence. If the objective function of the mutated sorting sequence is smaller than that of the random sorting sequence, then the mutated sorting sequence is preferred over the random sorting sequence. The repetition ends when the number of repetitions reaches a set number or when the number of selected preferred mutation sorting sequences converges. The objective function includes a path intersection penalty term and a region selection constraint penalty term. ； Where: u is the weight of the path intersection penalty, and λ is the weight of the region selection constraint. The number of intersections in the path formed by the sorted order of the sequence nodes. The penalty coefficient for region selection constraints, τ p , τ q It is the p-th and q-th subgrid region in a random sorting sequence or a mutated sorting sequence, where q = p + 1; ； ； τ p , τ q This indicates that the boundary coordinates of two sub-grid regions are adjacent; Δ max For the maximum allowable span, d(τ) p , τ q ) represents two sub-grid regions τ p and τ q The distance from the center point is δ, which is greater than 1.

2. The method according to claim 1, characterized in that, The method of transforming the sorting of subgrid regions in a random sorting sequence using transformation operations to generate a mutated sorting sequence includes: Region swapping: Randomly select two subgrid regions from a random sorting sequence and swap their positions in the random sorting sequence; Interval Reversal: Select an interval consisting of multiple consecutive subgrid regions from a random sorting sequence, and reverse the sorting of the subgrid regions within that interval; Neighborhood fine-tuning: swapping the order of sequentially adjacent subgrid regions in a random sorting sequence; The transformation operation randomly selects one of three transformation operations to generate a mutated sorting sequence. The number of mutated sorting sequences before and after the transformation is equal to the number of random sorting sequences.

3. The method according to claim 1, characterized in that, The step of obtaining the sub-path that covers the entire sub-grid region within each sub-grid region includes: The subgrid area is divided into a grid map composed of grid cells using mutually perpendicular meridians and parallels; each grid cell without obstacles is marked as passable, and each grid cell with obstacles is marked as impassable. Detect each sub-grid region and its next adjacent sub-grid region in the access sequence, find the starting edge and ending edge of each sub-grid region in turn, select the edge in each sub-grid region that is far away from the obstacle as the starting edge, and select the edge in each sub-grid region that is close to the next sub-grid region as the ending edge. Select a grid cell from a column or row of grid cells near the starting edge of the sub-grid region as the starting point of the sub-path. Extend the sub-path within the sub-grid region from a direction parallel to the starting edge to form a meandering sub-path composed of interconnected path segments. Adjacent path segments at bends are perpendicular to each other. The end point of the sub-path is located at the terminating edge of the sub-grid region. The sub-path includes several rows of path segments parallel to the starting edge. The sub-path covers the entire sub-grid region and does not intersect within the sub-grid region.

4. The method according to claim 1, characterized in that, The process of reconstructing the sub-path to obtain the obstacle avoidance path includes: If an intersection portion in the sub-path that coincides with or contacts a grid cell containing an obstacle is detected, the intersection portion is offset to obtain an offset path, and the offset path maintains a safe distance from the obstacle. Replace the path segment with an offset path.

5. The method according to claim 1, characterized in that, The step of sequentially traversing adjacent sub-grid regions according to the access sequence to obtain the connection path between the end point of the sub-path or obstacle avoidance path in the previous sub-grid region and the start point of the sub-path or obstacle avoidance path in the next sub-grid region further includes: Expand the raster map of the sub-grid area to the entire target area map, mark grid cells without obstacles as passable, and mark grid cells with obstacles as impassable; If the connection path is detected to pass through an impassable grid cell, the connection path is offset to bypass the impassable grid cell, and a safe distance is maintained between the connection path and the impassable grid cell.

6. The method according to claim 3, characterized in that, The process of obtaining sub-paths covering the entire sub-grid area within each sub-grid area further includes: selecting the step size between several row path segments parallel to the starting edge to maximize the adaptability of the sub-paths set according to the step size. Fitness is: ; Where α and γ are the weighting coefficients for coverage and speed, respectively; For coverage, ; L(R) i ) represents the subpath length, and W represents the coverage width; The total area of ​​the subgrid region is A(R) i The area already covered is A. covered (R i ); v avg (R i () represents the regional path speed. ; The amount of cement required to be poured in the sub-grid area is M(R) i The total length of the sub-paths is L(R). i ).

7. The method according to claim 6, characterized in that, The selection includes the step of selecting the step size using a genetic annealing algorithm: Population initialization: from step size set A set of step sizes is randomly selected as the initial population, and each step size... l i As an individual, subpaths are created based on the subpath acquisition steps, and the fitness F( ) of each individual is calculated. l i ); Selection operation: In each generation of the genetic algorithm, individuals in the current population are sorted from largest to smallest according to their fitness value, and the individuals with the top 10% fitness values ​​are selected as superior individuals and placed into the elite population Q. A superior individual is randomly selected from the elite population Q as the parent and crossover and mutation operations are performed with the current individual in the current population to generate a new individual. The crossover operation includes: exchanging the step size between some path segments parallel to the starting edge in the superior individual and the current individual; the mutation operation is: applying a random bias to the step size between some path segments parallel to the starting edge in the superior individual and the current individual. Replace the current individual according to the simulated annealing acceptance criteria; Merge superior individuals with new individuals to generate the next generation of the population; If the number of generations reaches the set maximum value, the iteration ends; otherwise, the genetic algorithm operation is repeated for iteration. The simulated annealing acceptance criteria are as follows: Where T represents the current temperature, which gradually decreases as the number of iterations increases. T0 is the initial temperature. Here, n is the temperature decay coefficient, and n is the current iteration number. It is the fitness of the new individual. It is the current fitness level of the individual. l new It is a new individual. l current is the current individual, and p is a number generated randomly in each iteration that is greater than 0 and less than 1.

Images