Dynamic self-adaptive obstacle avoidance path planning method for hoisting operation area

Abstract

The invention relates to a dynamic self-adaptive obstacle avoidance path planning method for a hoisting operation area, and belongs to the technical field of path planning in hoisting operation. The method comprises the following steps: scanning point cloud data by using three-dimensional laser, and establishing a static three-dimensional working environment model with a known environment; dividing a space by using a grid method, and dispersing a continuous space into a limited three-dimensional point set; setting a safe distance d; redefining a heuristic function eta; defining a volatilization coefficient; constructing a self-adaptive adjustment function considering the path length, the hoisting height and the safety factor; and updating pheromones, carrying out iteration, stopping iteration after an optimal effect is achieved, and obtaining a three-dimensional path. The method is improved on the basis of the ant colony algorithm, and efficient and safe path planning of the hoisting equipment in a complex environment is realized by combining strategies such as environment precise modeling, dynamic visual area definition, three-dimensional space rasterization processing and algorithm core optimization.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of path planning in hoisting operation, and particularly relates to a dynamic self-adaptive obstacle avoidance path planning method for a hoisting operation area. BACKGROUND

[0002] Hoisting equipment such as cranes is an important object of safety management work in power construction sites, and plays an extremely important role in safety production and enterprise economic benefits. Due to the complex power operation construction environment, and the types and sizes of these large machines are different, relying on manual observation and control has serious loopholes and observation blind spots. A slight mistake can easily lead to insufficient early warning threshold, especially due to insufficient perception of hoist arm movement space, lack of intelligent obstacle avoidance path planning capability, and collision with power facilities, thereby causing large-scale power outages, secondary disasters, and social instability.

[0003] Traditional path planning methods are mostly based on static environment for path planning, and do not fully consider the dynamic and uncertain nature of the hoist arm of the hoisting operation. These methods perform well in cases where there are few obstacles and the environment is stable, but their flexibility and adaptability are obviously insufficient in the face of sudden obstacles or environmental changes, which can lead to operation interruption or safety accidents. The grid method can effectively simplify the path search process by dividing the operation area into a number of grids. However, pure grid method will lead to a dramatic increase in calculation amount when there are many obstacles or complex environment, affecting the planning efficiency and leading to unreasonable planning path. In recent years, with the development of artificial intelligence technology, many intelligent algorithms such as ant colony algorithm, particle swarm algorithm, artificial fish swarm algorithm, etc. have been applied to the path planning of cranes. These algorithms improve the intelligent level of path planning to some extent, and can better adapt to complex environments. However, single algorithm often has limitations, such as slow convergence speed of ant colony algorithm and easy falling into local optimum of particle swarm algorithm. In view of the three-dimensional space characteristics of hoisting operation, some researches have begun to explore three-dimensional path planning methods. However, these methods still need to be improved in terms of real-time performance, accuracy and adaptability, especially in handling dynamic obstacles and unknown environments, and their effectiveness is often not satisfactory.

[0004] Therefore, how to overcome the shortcomings of the prior art is a problem that needs to be solved in the field of path planning in hoisting operation at present. SUMMARY

[0005] The purpose of the present application is to solve the shortcomings of the prior art and provide a dynamic self-adaptive obstacle avoidance path planning method for a hoisting operation area.

[0006] To achieve the above purpose, the technical scheme adopted by the present application is as follows: A dynamic self-adaptive obstacle avoidance path planning method for a hoisting operation area, comprising the following steps: Step (1), using three-dimensional laser scanning point cloud data, a static and known environment three-dimensional working environment model is established; then the grid method is used to divide the space, according to the safety accuracy requirements of the hoisting operation (such as the minimum interval between the boom and the obstacle), the grid resolution is set, according to the three-dimensional working environment model, the occupation state of each grid is judged, the continuous space is discretized into a finite three-dimensional point set, each point has position and occupation attribute; Step (2), the sum of the maximum allowable swing range d1 of the load hoisted by the crane and the distance d2 of the safe passing of the obstacle is set as the safety distance d; Step (3), based on the ant colony optimization algorithm, pheromone is introduced as an influencing factor, and the heuristic function η is redefined; Step (4), a volatility coefficient is defined, which represents the degree of gradual reduction of pheromone with time, thereby improving the diversity and global search ability of the search; Step (5), an adaptive adjustment function considering path length, lifting height and safety factor is constructed , which is used to comprehensively evaluate the quality of the path constructed by the ant colony, and according to which the global pheromone volatility coefficient is dynamically adjusted , and the search for better paths is strengthened; Step (6), update the pheromone, iterate, stop iteration when the optimal effect is reached, and get the three-dimensional path.

[0007] Further, the specific method of step (2) is: The safety distance d is set as: (1) (2) (3) Wherein, L is the length of the sling; is the maximum swing angle under safe conditions; is an empirical coefficient; is the current wind speed; P is the distance from the obstacle to the nearest point of the crane; r hook is the radius of the hook; r load is the radius of the load; d1 is the maximum swing of the load; d2 is the minimum distance for safe passing of the obstacle.

[0008] Further, 0.15.

[0009] Further, the specific method of step (3) is: The heuristic function η(i, j, k) is used to guide the ants to select the path in the ant colony algorithm, and the function considers various factors from the current position node i to the next position node j and finally to the target position node k. The pheromone is integrated into the heuristic function. The redefined heuristic function is as follows: (4) wherein, D(i, j) represents the distance from node i to node j; Q(j, k) represents the distance from node j to target k; S(i, j, k) represents the visualization ratio of node i, node j and node k; M(i, j) represents the height calculation result of node i and node j; τ(i, j) represents the pheromone concentration on the path / connection edge (i, j); α is the pheromone weight factor; ω1, ω2, ω3 are weight coefficients corresponding to Q(j, k), S(i, j, k) and M(i, j), respectively; (5) (6) (7) wherein, τ(i, j)(t+1) represents the pheromone concentration on the path / connection edge (i, j) after the completion of the t+1th iteration, which is the updated value; is the pheromone evaporation coefficient; represents the original pheromone concentration on the path / connection edge (i, j) after the completion of the tth iteration, which is the value before updating.

[0010] wherein, num represents the number of visible points, that is, the number of grid points (or test points) on the line from the current path point (or observation point) to the candidate next point (or target area) which is not blocked by any obstacle; unum represents the number of invisible points, that is, the number of grid points blocked by obstacles or in the shadow area, which indicates that there may be a collision risk or the need for concealment in the path segment; H is the height of the next point, that is, the next candidate moving node considered by the ant currently located at node i in the ant colony algorithm, which is to guide the algorithm to preferentially select the path with low height, thereby reducing the operation risk and energy consumption; is the evaporation coefficient.

[0011] Further, ω1=0.3, ω2=0.2, ω3=0.5, and α=0.5.

[0012] Further, the specific method of step (4) is as follows: (8) wherein, is the evaporation coefficient; represents the minimum value of the evaporation coefficient, which is usually close to 0; represents the maximum value of the volatility coefficient; NC represents the optimal iteration number; and NCM represents the highest iteration number.

[0013] Further, the specific method of step (5) is: An adaptive function is defined: (9) wherein the path length represents the path length from the starting point to the ith node and then to the jth node. represents the lifting height of the current path segment. is a safety factor. , , respectively represent the weight coefficients of the path length , the lifting height , and the safety factor .

[0014] Further, (10) wherein (x i , y i , z i ) and (x j , y j , z j ) respectively represent the spatial coordinates of the ith node and the jth node. (11) (12) wherein, is the number of obstacles on the path; is a safety factor; is the total number of obstacles scanned in the three-dimensional space. The calculation formulae of the path length , the lifting height , and the safety factor are brought into equation (9) to obtain the specific expression of the adaptive adjustment function as follows: (13).

[0015] Further, the specific method of step (6) is: In each iteration process, the pheromone concentration is updated based on the current path selection condition. After each iteration, the quality of the current solution is evaluated using an adaptive fitness function; if a new better solution is found, continue adjusting the pheromone concentration and the next iteration; if no better solution is found or the preset maximum number of iterations is reached, stop the iteration process; the optimal three-dimensional path considering the path length, lifting height and safety factor is obtained.

[0016] Further, the iteration loop process is as follows: 1) Path construction: each ant starts from the starting point, selects the next node according to the pheromone concentration of the current node and the redefined heuristic function (the value is negatively related to the estimated cost of the node to the target point, the safety factor), and moves step by step; when selecting, it must be ensured that the distance between the new node and all obstacle grids is not less than the safety distance, so as to construct the collision-free candidate path online; 2) Path evaluation and adaptive adjustment: when all ants complete path construction, calculate the adaptive adjustment function fitness(i), which comprehensively considers the length, maximum lifting height and safety margin of the path; according to the fitness distribution of all ants in the current generation, dynamically adjust the volatilization coefficient: if the quality of the solution tends to be stagnant (falls into local optimum), reduce it to strengthen positive feedback; if the diversity is sufficient, encourage exploration; 3) Global update of pheromone: first, the pheromone on all paths is uniformly volatilized according to the current volatilization coefficient; then, according to the fitness(i) value of each ant, new pheromone is released to the corresponding path node in a certain proportion, and the path with the better fitness(i) value (the smaller the better) obtains more enhanced pheromone; This "construction-evaluation-adjustment-update" process is repeated iteratively until the optimal path of consecutive generations no longer improves or the maximum number of iterations is reached, the algorithm terminates and outputs the historical optimal three-dimensional path.

[0017] The present application comprehensively considers the advantages and disadvantages of various methods, combines the particularity of the lifting and hoisting operation and the actual demand of the obstacle avoidance path planning of the lifting boom movement, and fuses various intelligent algorithms and dynamic perception technology, so that real-time perception and accurate path planning of complex dynamic environment are realized, which is particularly important for lifting and hoisting operation.

[0018] The present application is based on the ant colony algorithm, combines accurate environment modeling, dynamic visible area definition, three-dimensional space gridding processing and algorithm core optimization strategies, realizes efficient and safe path planning of the lifting equipment in complex environment, and is particularly suitable for lifting and hoisting operation in the complex environment of power mechanical construction site.

[0019] The application solves a key safety problem in current hoisting operation, i.e. collision dynamic avoidance of obstacles in a key section during hoisting operation. In view of the lack of flexibility and adaptability of traditional two-dimensional path planning and the lack of dynamic self-adaptive adjustment capability of the original fish swarm algorithm, the application proposes an innovative solution, which combines traditional path planning, three-dimensional grid division of the key section, optimization of the ant colony algorithm, introduction of an adaptive adjustment function of the evaporation coefficient, and construction of a more intelligent calculation model to dynamically avoid obstacles during operation. The core of the method is to comprehensively consider path length, lifting height and safety factors to realize dynamic intelligent planning of the hoisting path.

[0020] The method has the characteristics that real-time perception and accurate path planning are realized for a complex dynamic environment by combining multiple intelligent algorithms and dynamic perception technologies. Advanced sensor technologies and environmental modeling methods are used to provide accurate data support for hoisting operation path planning, optimize the traditional ant colony algorithm, improve the global search capability, and improve the intelligent level and adaptability of path planning.

[0021] The application constructs an improved ant colony algorithm to prevent the ant colony algorithm from falling into a local optimal solution, as shown in the principle. Figure 3 An adaptive adjustment function of the evaporation coefficient is introduced on the basis of the heuristic function η of the ant colony algorithm, as shown in the principle. The multiple key factors such as path length, lifting operation height limit and safety redundancy factor are comprehensively considered to realize dynamic optimization and flexible adjustment of the planned path. This feature ensures that the crane can quickly respond to various unexpected situations, flexibly avoid obstacles, and ensure smooth and efficient hoisting operation, thereby providing strong technical support for power construction safety.

[0022] In step (1), a grid processing method is used to discretize the three-dimensional space during path planning, and the grid size is proportional to the size of the actual environment, so as to better guide the actual operation and facilitate calculation and algorithm implementation. A three-dimensional operation environment model of the crane is planned based on the actual size of the crane. During path planning, the load and hook are regarded as a particle to simplify the calculation model, and the environment model is as shown in the principle. Figure 1 The model not only includes the working range of the crane itself, but also details the position and size information of all fixed obstacles around the crane.

[0023] In step (2), the load may swing when the crane accelerates or decelerates during hoisting, and the load is prone to collide with obstacles during swinging, which may cause accidents. The crane dynamics model is as shown in the principle. Figure 2As shown in the figure, M represents the crane body which can move in horizontal direction, m represents the load object, φ represents the swing state of the load, i.e. the angle relative to the horizontal line, f x is the lateral force caused by wind, etc. l represents the pulling force provided by the crane, d1 is the distance generated when the load swings to the maximum (under the premise of ensuring safety), and d2 is the minimum distance for safely passing through the obstacle. Therefore, the safety distance of the present application is set as: (1) (2) (3) wherein L represents the length of the hoisting rope, which is an important factor determining the swing range of the load; is the maximum swing angle under the safety condition, reflecting the position change of the load under the maximum allowable swing state; is an empirical coefficient, which is pre-set according to the type of the crane and the operating environment, reflecting the degree of influence of wind speed on the safety distance, is the current wind speed, which can be obtained in real time through the sensor installed on the crane, then the safety additional distance calculated based on the real-time dynamic factors can be represented; P is the distance from the obstacle to the nearest point of the crane; r hook is the radius of the hook; r load is the radius of the load.

[0024] In step (3), the heuristic function η(i, j, k) is used to guide the ants in the ant colony algorithm to select the direction of the path. This function considers various factors from the current position i to the next position j and finally to the target position k. In the ant colony algorithm, we can integrate the important factor of pheromone in the heuristic function. The redefined heuristic function can be expressed as follows: (4) wherein D(i, j) represents the distance from node i to node j; Q(j, k) represents the distance from node j to target k; S(i, j, k) represents the visualization ratio of i, j, k; M(i, j) represents the height calculation result of nodes i, j; τ(i, j) represents the pheromone concentration on nodes i to j; α is the pheromone weight factor, used to control the influence degree of pheromone on the heuristic function; ω1, ω2, ω3 are weight coefficients corresponding to Q, S, M respectively, used to balance the influence of each factor on path selection.

[0025] (5) (6) (7) Wherein, num represents the number of visible points, unum represents the number of invisible points; H is the height of the next point; is the evaporation coefficient, which can be regarded as a constant.

[0026] Here, pheromone refers to the virtual secretion released by ants on the path during the iteration process, and its concentration reflects the degree to which a certain path (or node) is preferred in historical search. It carries the "group experience" and is a key positive feedback factor guiding the convergence of the algorithm. In a three-dimensional grid environment, pheromone is usually deposited on nodes (i.e. discretized three-dimensional grid points) or on the connecting edges between nodes.

[0027] Node refers to a three-dimensional point formed by discretization through the grid method, and each node corresponds to a unique three-dimensional coordinate (x, y, z), representing a candidate position that the crane load can occupy in space. Nodes are divided into "free" (walkable) and "occupied" (obstacles), and path planning is carried out in the network composed of free nodes.

[0028] Integrating pheromone into the heuristic function η means that the pheromone concentration of the current node i to the candidate node j (such as τ(i, j) or τ(j)) is taken as an input variable for calculating the heuristic value, which works together with other factors (such as the estimated cost of node j to the target node k, safety margin, etc.) to dynamically evaluate the expected attractiveness of moving to node j. This makes the ant's choice not only based on static environmental heuristic information, but also integrates the dynamically accumulated group search experience.

[0029] In step (4), the evaporation coefficient of the ant colony algorithm determines the evaporation rate of pheromone, i.e. how pheromone decreases over time. The value of the evaporation coefficient is between 0 and 1, and the closer the value is to 0, the slower the pheromone evaporates, while the closer the value is to 1, the faster the pheromone evaporates. In order to dynamically adjust the evaporation coefficient, the value of can be adjusted according to the current path quality and search state; if the quality of the current path (fitness value) is significantly better than the historical record, the evaporation coefficient can be appropriately reduced; if the quality of the current path does not improve significantly, or even deteriorates, the evaporation coefficient can be appropriately increased, so that the pheromone evaporates faster, prompting the algorithm to jump out of the current local optimal solution and find a new path; let represent the minimum value of the evaporation coefficient, which is usually close to 0; represent the maximum value of the evaporation coefficient, which is usually close to 1; let NC represent the optimal number of iterations, which is determined according to environmental parameters in the specific implementation process; NCM represents the highest number of iterations, which is determined according to environmental parameters in the specific implementation process. The specific adjustment formula is: (8) In step (5), the fitness function is a key component to evaluate the path quality, which determines which path is more worthy to be selected. Considering the particularity of the crane lifting operation, the fitness function used can simultaneously consider the path length , lifting height and safety factor , which are three objectives. The fitness function is defined as: (9) wherein, path length represents the path length from the starting point to the i-th node and then to the j-th node, which can be measured by calculating the Euclidean distance between two points, wherein (x i , y i , z i ) and (x j , y j , z j ) represent the spatial coordinates of the i-th node and the j-th node, respectively, and are specifically represented as: (10) Lifting height represents the lifting height of the current path segment, which is represented as the maximum value of the z coordinates of all points on the path. Higher lifting height will bring greater difficulty and risk, so this part will also affect the fitness of the path, and is specifically calculated as: (11) Safety factor is a parameter that comprehensively considers the number, type and other factors that may cause danger of obstacles on the path, wherein is the number of obstacles on the path, is the total number of obstacles scanned in the three-dimensional space, which is a regulating parameter to control the influence degree of the obstacles, and the parameter setting depends on the specific construction environment. The safety factor is defined as: (12) , , respectively represent the weight coefficients of path length , lifting height and safety factor . The calculation formula of path length , lifting height and safety factor is brought into formula (9), and the specific expression of the fitness adjustment function is as follows: (13) Further, in step (6), the pheromone concentration is updated based on the current path selection situation in each iteration. For the selected path segment, the pheromone concentration on it will increase, while the pheromone on the unselected path segment will gradually decrease over time according to the evaporation coefficient. After each iteration, the quality of the current solution is evaluated using an adaptive fitness function. If a new better solution is found, the pheromone concentration is adjusted and the next iteration is continued; if no better solution is found or the preset maximum number of iterations is reached, the iteration process is stopped.

[0030] During the iteration process, the improved heuristic function η(i, j, k) is continuously used to guide the selection behavior of the ants, while the dynamically adjusted pheromone evaporation coefficient ρ is used to balance exploration and development, ensuring that the algorithm can jump out of the local optimal solution and find the global optimal path. When the iteration process converges to the optimal solution, the best three-dimensional path considering multiple objectives such as path length, lifting height, and safety factor is obtained.

[0031] Compared with the prior art, the present application has the following beneficial effects: (1) Significantly improve the accuracy and accuracy of path planning: The present application significantly improves the accuracy of the path planning of the lifting and hoisting operation by integrating environmental precise modeling, dynamic visible area definition, three-dimensional space gridding processing and improved ant colony algorithm. The environmental modeling technology ensures the accurate simulation of the working environment, and the dynamic visible area definition technology ensures the unobstructed view of the crane in complex environment, making the path planning more accurate and reliable.

[0032] (2) Optimize algorithm efficiency and response speed: The dynamic adjustment mechanism of the pheromone evaporation coefficient, combined with factors such as path length, lifting height and number of obstacles, greatly improves the global search ability and path diversity of the algorithm. This method not only speeds up the processing speed from environmental perception to path planning, realizes instant path adjustment, but also enables the operator to quickly respond to sudden obstacles or environmental changes, effectively reducing the risk of operation.

[0033] (3) Improve the practicality and compatibility of the path planning system: The present application overcomes the limitations of traditional path planning methods in complex environments, such as large amount of calculation, poor adaptability, etc. By introducing intelligent algorithms and dynamic perception technology, a smart and practical path planning system is created, which is suitable for various complex working environments. Its simple configuration, maintenance convenience and low use cost make the system more easily applied to power machinery construction lifting and hoisting operations, improving the compatibility and popularity of the system.

[0034] (4) Enhance the safety and reliability of the operation: the present application can monitor the changes in the working environment in real time, such as sudden obstacles or wind speed changes, and adjust the path planning accordingly through the dynamic adaptive obstacle avoidance path planning method. This not only reduces the interruption of work caused by unexpected situations, but also effectively avoids safety accidents, ensuring the safety of the workers and the safe operation of the equipment.

[0035] (5) Improve the efficiency and quality of the operation: the present application can adaptively adjust the operation path of the crane according to the actual situation of the operation site, avoiding obstacles while choosing the optimal path, thereby improving the efficiency of the operation. In addition, precise path planning also helps to reduce the impact on the surrounding environment and improve the overall quality of the operation.

[0036] (6) Compared with the existing grid-based operation path planning method, the present application method is improved based on ant colony algorithm, combining strategies such as accurate environmental modeling, dynamic visible area definition, three-dimensional space gridding processing and algorithm core optimization, to realize efficient and safe path planning of the hoisting equipment in complex environment. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 Modeling of the gridded crane three-dimensional environment; Figure 2 Schematic diagram of the crane dynamics model; Figure 3 Flowchart of the improved ant colony algorithm; Figure 4 Iteration process for calculating fitness value; Figure 5 Flowchart of the present application method; Figure 6 Three-dimensional path planning result. DETAILED DESCRIPTION

[0038] The present application will be further described in detail below in conjunction with the examples.

[0039] Those skilled in the art will understand that the following examples are only for illustration of the present application, and should not be considered as limiting the scope of the present application. If the specific technology or condition is not specified in the examples, it is carried out according to the technology or condition described in the literature in the art or according to the product instruction. If the manufacturer of the material or equipment is not specified, it is a conventional product that can be obtained by purchase.

[0040] Example 1 A dynamic adaptive obstacle avoidance path planning method for a hoisting and lifting operation area, comprising the following steps: Step (1), a three-dimensional laser scanning point cloud data is used to establish a static and known three-dimensional working environment model; then a grid method is used to divide the space, a grid resolution is set according to the safety precision requirement of the hoisting operation (such as the minimum interval between the boom and the obstacle), and the occupancy state of each grid is judged according to the three-dimensional working environment model, so that the continuous space is discretized into a finite three-dimensional point set, and each point has a position and occupancy attribute; Step (2), the sum of the maximum allowed swing range d1 of the load hoisted by the crane and the distance d2 for safely passing through the obstacle is set as a safety distance d; Step (3), based on the ant colony optimization algorithm, pheromone is introduced as an influencing factor, and the heuristic function η is redefined; Step (4), a volatilization coefficient is defined, which represents the degree of gradual reduction of pheromone over time, thereby improving the diversity and global search ability of the search; Step (5), an adaptive adjustment function is constructed considering the path length, lifting height and safety factor, which is used to comprehensively evaluate the quality of the path constructed by the ant colony, and the global pheromone volatilization coefficient is dynamically adjusted according to the quality, so as to strengthen the search for the better path; Step (6), the pheromone is updated, iteration is performed, and when the optimal effect is reached, the iteration is stopped, and a three-dimensional path is obtained.

[0041] Embodiment 2 A dynamic and adaptive obstacle avoidance path planning method for a hoisting operation area, comprising the following steps: Step (1), a three-dimensional laser scanning point cloud data is used to establish a static and known three-dimensional working environment model; then a grid method is used to divide the space, a grid resolution is set according to the safety precision requirement of the hoisting operation (such as the minimum interval between the boom and the obstacle), and the occupancy state of each grid is judged according to the three-dimensional working environment model, so that the continuous space is discretized into a finite three-dimensional point set, and each point has a position and occupancy attribute; Step (2), the sum of the maximum allowed swing range d1 of the load hoisted by the crane and the distance d2 for safely passing through the obstacle is set as a safety distance d; Step (3), based on the ant colony optimization algorithm, pheromone is introduced as an influencing factor, and the heuristic function η is redefined; Step (4), a volatilization coefficient is defined, which represents the degree of gradual reduction of pheromone over time, thereby improving the diversity and global search ability of the search; Step (5), an adaptive adjustment function The quality of the path constructed by the ant colony is comprehensively evaluated, and the global pheromone evaporation coefficient is dynamically adjusted according to the quality , and the search for the better path is strengthened Step (6), updating the pheromone, iterating, stopping iteration when the optimal effect is achieved, and obtaining a three-dimensional path.

[0042] The specific method of step (2) is: The safety distance d is set as: (1) (2) (3) Wherein, L is the length of the sling; is the maximum swing angle in the safety case; is an empirical coefficient; is the current wind speed; P is the distance from the obstacle to the nearest point of the crane; r hook is the radius of the hook; r load is the radius of the load; d1 is the maximum swing of the load; d2 is the minimum distance for safely passing through the obstacle.

[0043] is 0.15.

[0044] The specific method of step (3) is: The heuristic function η(i, j, k) is used to guide the ants in the ant colony algorithm to select the direction of the path, which considers various factors from the current position node i to the next position node j and finally to the target position node k. The pheromone, an important factor, is integrated in the heuristic function. The redefined heuristic function is as follows: (4) Wherein, D(i, j) represents the distance from node i to node j; Q(j, k) represents the distance from node j to target k; S(i, j, k) represents the visualization ratio of node i, node j and node k; M(i, j) represents the height calculation result of node i and node j; τ(i, j) represents the pheromone concentration on node i to node j; α is the pheromone weight factor; ω1, ω2, ω3 are weight coefficients corresponding to Q(j, k), S(i, j, k) and M(i, j) respectively; (5) (6) (7) Wherein, τ(i, j)(t+1) is the pheromone concentration on the path / connection edge (i, j) after the completion of the t+1th iteration, which is the updated value; is the pheromone evaporation coefficient; is the original pheromone concentration on the path / connection edge (i, j) after the completion of the tth iteration, which is the pre-updated value.

[0045] Wherein, num represents the number of visible points, that is, the number of grid points (or test points) on the connecting line from the current path point (or observation point) to the candidate next point (or target area) which are not blocked by any obstacle; unum represents the number of invisible points, that is, the number of grid points blocked by obstacles or in the shadow area, which indicates that there may be a collision risk or the need for concealment in the path segment; H is the height of the next point, that is, the next candidate moving node that the ant is currently considering in the ant colony algorithm, which is to guide the algorithm to preferentially select a low-height path, thereby reducing the risk and energy consumption of the operation; is the evaporation coefficient.

[0046] ω1=0.3, ω2=0.2, ω3=0.5, α=0.5.

[0047] The specific method of step (4) is: (8) Wherein, is the evaporation coefficient; represents the minimum value of the evaporation coefficient, which is usually close to 0; represents the maximum value of the evaporation coefficient; let NC represent the optimal number of iterations; NCM represents the highest number of iterations.

[0048] The specific method of step (5) is: Define the adaptive function: (9) Wherein, the path length represents the path length from the starting point to the i th node and then to the j th node; represents the lifting height of the current path segment; is the safety factor; , , respectively represent the weight coefficients of the path length , the lifting height , and the safety factor .

[0049] (10) Wherein, (x i , y i , zi ) and (x j , y j , z j ) represent the spatial coordinates of the ith node and the jth node, respectively; (11) (12) wherein, is the number of obstacles on the path; is the safety factor; is the total number of obstacles scanned in the three-dimensional space; the path length the lifting height and the safety factor are substituted into equation (9), the specific expression of the adaptive adjustment function is as follows: (13).

[0050] The specific method of step (6) is as follows: In each iteration process, the pheromone concentration is updated based on the current path selection situation; After each iteration, the quality of the current solution is evaluated using the adaptive fitness function; if a new better solution is found, the pheromone concentration is adjusted and the next iteration is continued; if no better solution is found or the preset maximum number of iterations is reached, the iteration process is stopped; thus, the best three-dimensional path considering the path length, lifting height and safety factor is obtained.

[0051] The iteration loop process is as follows: 1) Path construction: each ant starts from the starting point and selects the next node according to the pheromone concentration of the current node and the redefined heuristic function (its value is negatively related to the estimated cost of the node to the target point, safety factor) probability, and moves step by step; when selecting, it must ensure that the distance between the new node and all obstacle grids is not less than the safety distance, so as to construct a collision-free candidate path online; 2) Path evaluation and adaptive adjustment: when all ants complete path construction, calculate the adaptive adjustment function fitness(i), which integrates the length, maximum lifting height and safety margin of the path; according to the fitness distribution of the whole colony of ants, dynamically adjust the evaporation coefficient: if the quality of the solution tends to be stagnant (trapped in local optimum), reduce it to strengthen positive feedback; if the diversity is sufficient, encourage exploration; 3) Global pheromone update: First, pheromones on all paths are uniformly volatilized according to the current volatilization coefficient; then, according to the fitness(i) value of each ant, new pheromones are released to the corresponding path nodes in a certain proportion, and the path with the better (smaller) fitness(i) value receives more pheromone enhancement. This “construction-evaluation-adjustment-update” process iterates repeatedly until the optimal path for multiple consecutive generations no longer improves or reaches the maximum number of iterations. At this point, the algorithm terminates and outputs the historical best 3D path.

[0052] Application Examples The following example illustrates how the method of this invention is used for intelligent path planning during hoisting operations of a crane under an overhead power transmission line in a hemispherical three-dimensional space along a railway section. In this example, the hoisting path space is set to 20*20*120 (m), with the starting point at (1, 10, 40) and the ending point at (21, 8, 50). The algorithm involves the following parameters: population size 10; maximum lateral movement of ants 1; maximum vertical movement of ants 3; maximum number of iterations 100. The specific steps include: Step (1): This invention studies global three-dimensional path planning in a static environment where obstacles are known. The research object is a crane with a lifting capacity not exceeding 50t, a span of 20m, and a maximum hook lifting height of 15m. Based on the crane's dimensions, a three-dimensional operating environment model is constructed as follows: Figure 1 As shown. In path planning, to simplify the calculation model, the load and hook are considered as a single point mass. To accurately describe obstacles in the environment, a height matrix based on obstacle height is used. This height matrix is ​​a point cloud numerical matrix formed by scanning the obstacle's laser point cloud. In this invention, the height matrix is ​​specifically: Step (2), the crane dynamics model is as follows Figure 2 As shown, the crane is performing a lifting operation, where L is the rope length, which is taken as 1.2m in this invention; φ max For the maximum swing angle under safe conditions, this invention uses 4.9°; k d It is an empirical coefficient, preset to 0.15 based on the crane type and operating environment. The wind speed v during operation is obtained in real time through sensors installed on the crane. w The speed is 7.6 m / s. First, the angle 4.9° is converted to radians, which is approximately 0.0855π. According to formula (2), the distance generated when the load swings to its maximum can be calculated: To ensure the crane can safely pass over obstacles, the radius r of the hook needs to be subtracted from the distance P = 2m from the nearest point of the obstacle to the crane. hook= 0.15m and the radius of the load r load = 1.05m to determine the actual safe passing distance. The minimum distance of the crane safely passing the obstacle is calculated according to formula (3): = 2m - 0.15m - 1.05m = 0.8m The distance d1 generated when the load swings to the maximum and the minimum distance d2 of the crane safely passing the obstacle are brought into formula (1) to obtain the safety distance: Step (3), in the path planning process, it is assumed that there are 8 visible points and 2 invisible points on the path, which are brought into formula (5) to obtain the visualization ratio: The height H of the next point is measured to be 35m, which is brought into formula (6) to obtain the height of point i and j: The distance D(i, j) of node i to node j is 10m, the distance Q(j, k) of node j to target k is 20m, the weight coefficients ω1=0.3, ω2=0.2, ω3=0.5, the pheromone weight factor α=0.5, and the initial pheromone τ(i, j)(0)=τ0=0.8. The heuristic function η is obtained by substituting the above values: Step (4), according to the specific construction environment and hoisting operation requirements of the present example, it is set that: The maximum number of iterations is 50 times, the evaporation coefficient is set , According to the iteration process of step (5), the optimal iteration number is 5 times, which is brought into formula (8) to calculate: Step (5), according to the specific environmental conditions and hoisting operation requirements of the present example, the initial hoisting operation path length of the crane is 40m, the hoisting height is 85m, the maximum variation distance of the next path length of the hoisting operation is not more than 3m, and the maximum change of the hoisting height is not more than 8m. The path length , the hoisting height and the weight coefficient of the safety factor are set to 0.2, 0.7 and 0.1 respectively. The total number of obstacles scanned in the three-dimensional space is 9, and the adaptive value after the first iteration is obtained according to formula (13): ​​​ As Figure 4 The iteration process of calculating the fitness value is shown. The optimal value is reached when the iteration is about the fifth time, at this time, the path length of the hoisting operation is 37m, the lifting height is 77m, the number of obstacles is 4, and the fitness value calculated is: Step (6), the calculation application of the algorithm in this example has a fast convergence speed, the pheromone is updated, the iteration is carried out, after the most effective result is reached, the iteration is stopped, and the optimal three-dimensional path based on the fitness function of the path length, the lifting height and the safety factor is obtained, as shown in Figure 6 The specific path coordinates are shown in Table 1. Table 1 x i ]]> [[ y i ]]> z i ]]> 1 10 40 2 9 40 3 7 40 4 5 50 5 7 40 6 5 80 7 4 80 8 5 40 9 6 40 10 8 40 11 7 40 12 6 80 13 7 40 14 8 44 15 9 85 16 5 40 17 11 120 18 9 68 19 9 45 20 9 38 21 8 50 The basic principles, main features and advantages of the present application are shown and described above. It should be understood by those skilled in the art that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for dynamically adaptive obstacle avoidance path planning for a crane lifting operation area, characterized in that, Includes the following steps: Step (1): Use 3D laser scanning point cloud data to establish a static 3D working environment model with known environment; then use the grid method to divide the space, set the grid resolution according to the safety accuracy requirements of hoisting operation, and determine the occupancy status of each grid according to the 3D working environment model. The continuous space is discretized into a finite set of 3D points, each point with position and occupancy attributes. Step (2): Set the sum of the maximum permissible swing range d1 of the load lifted by the crane and the safe distance d2 for passing through the obstacle as the safe distance d; Step (3) Based on the ant colony optimization algorithm, an improvement is made by introducing pheromones as an influencing factor and redefining the heuristic function η; Step (4), defining a volatility coefficient , indicating the degree of gradual reduction of pheromone over time, thus increasing the diversity of search and global search capability; Step (5), constructing an adaptive adjustment function considering path length, lifting height, and safety factor , which is used to comprehensively evaluate the quality of the path constructed by the ant colony and dynamically adjust the evaporation coefficient of global pheromone according to the evaluation result , and strengthen the search for better paths Step (6): Update the pheromones, iterate until the optimal effect is achieved, then stop the iteration and obtain the three-dimensional path.

2. The dynamic self-adaptive obstacle avoidance path planning method for a crane lifting operation area according to claim 1, characterized in that, The specific method for step (2) is as follows: The safe distance d is set as follows: （1） （2） （3） where L is the length of the sling; is the maximum swing angle for safety; is an empirical coefficient; is the current wind speed; P is the distance of the obstacle to the nearest point of the crane; hook is the radius of the hook; r load is the radius of the load; d1 is the maximum swing of the load; d2 is the minimum distance to safely pass the obstacle.

3. The dynamic adaptive obstacle avoidance path planning method for lifting and hoisting operation areas according to claim 1, characterized in that, is 0.

15.

4. The dynamic self-adaptive obstacle avoidance path planning method for a crane lifting operation area according to claim 1, characterized in that, The specific method for step (3) is as follows: The heuristic function η(i, j, k) guides the ants in the ant colony algorithm to choose their path. This function considers various factors from the current node i to the next node j, and finally to the target node k. The important factor of pheromone is integrated into the heuristic function. The redefined heuristic function is expressed as follows: （4） Where D(i,j) represents the distance from node i to node j; Q(j,k) represents the distance from node j to target k; S(i,j,k) represents the visualization ratio of nodes i, j, and k; M(i,j) represents the calculated height of nodes i and j; τ(i,j) represents the pheromone concentration from node i to node j; α is the pheromone weight factor; ω1, ω2, and ω3 are the weight coefficients corresponding to Q(j,k), S(i,j,k), and M(i,j), respectively. （5） （6） （7） wherein τ(i, j)(t + 1) is the pheromone concentration on the path (i, j) after the completion of the t+1th iteration; is the pheromone evaporation coefficient; is the pheromone concentration on the path (i, j) after the completion of the tth iteration; num is the number of visible points; unum is the number of invisible points; H is the height of the next point; is the evaporation coefficient.

5. The dynamic self-adaptive obstacle avoidance path planning method for a crane lifting operation area according to claim 4, characterized in that, ω1=0.3, ω2=0.2, ω3=0.5, α=0.

5.

6. The dynamic self-adaptive obstacle avoidance path planning method for a crane lifting operation area according to claim 1, characterized in that, The specific method for step (4) is as follows: （8） wherein, is the volatility factor; denotes the minimum value of the volatility factor, typically close to 0; denotes the maximum value of the volatility factor; let NCdenote the optimal number of iterations; let NCMdenote the maximum number of iterations.

7. The dynamic self-adaptive obstacle avoidance path planning method for a crane lifting operation area according to claim 1, characterized in that, The specific method for step (5) is as follows: Define the adaptive function: （9） Among them, path length This represents the path length from the starting point to the i-th node and then to the j-th node; Indicates the lifting height of the current path segment; As a safety factor; , , Representing path length Lifting height and safety factors The weighting coefficients.

8. The dynamic adaptive obstacle avoidance path planning method for lifting and hoisting operation areas according to claim 7, characterized in that, （10） where (x i , y i , z i ) and (x j , y j , z j ) represent the spatial coordinates of the i-th node and the j-th node, respectively; （11） （12） in, It is the number of obstacles on the path; As a safety factor; The total number of obstacles scanned in three-dimensional space; path length Lifting height and safety factors Substituting the calculation formula into equation (9), we obtain the adaptive adjustment function. The specific expression is as follows: （13）。 9. The dynamic adaptive obstacle avoidance path planning method for lifting and hoisting operation areas according to claim 1, characterized in that, The specific method for step (6) is as follows: In each iteration, the pheromone concentration is updated based on the current path selection. After each iteration, the quality of the current solution is evaluated using the degree of adaptation function; If a new, better solution is found, the pheromone concentration is adjusted and the next iteration is performed; if no better solution is found or the preset maximum number of iterations has been reached, the iteration process is stopped; thus, the optimal three-dimensional path that takes into account path length, lifting height, and safety factors is obtained.

10. The dynamic adaptive obstacle avoidance path planning method for lifting and hoisting operation areas according to claim 9, characterized in that, The iterative process is as follows: 1) Path construction: Each ant starts from the starting point and selects the next node based on the pheromone concentration of the current node and the probability of the redefined heuristic function, moving step by step; when selecting, it must be ensured that the distance between the new node and all obstacle grids is not less than the safe distance, so as to construct a collision-free candidate path online; 2) Path evaluation and adaptive adjustment: After all ants have completed path construction, the adaptive adjustment function fitness(i) is calculated; the volatile coefficient is dynamically adjusted according to the fitness distribution of all ants in the current era: if the quality of the solution tends to stagnate, it is reduced to strengthen positive feedback; if the diversity is sufficient, exploration is encouraged. 3) Global pheromone update: First, pheromones on all paths are uniformly volatilized according to the current volatilization coefficient; then, according to the fitness(i) value of each ant, new pheromones are released to the corresponding path nodes in a certain proportion, and the path with the better fitness(i) value receives more pheromone enhancement. This process iterates repeatedly until the optimal path for multiple consecutive generations no longer improves or the maximum number of iterations is reached. The algorithm then terminates and outputs the historical best 3D path.

Images