Job path planning method and system fusing environmental perception and occupational health

Abstract

This invention provides a method and system for planning work paths that integrates environmental perception and occupational health. Specifically, it involves acquiring a 3D map of a farm and establishing a hazardous gas prediction model associated with ventilation status. Using the personnel's starting and ending points as root nodes, it constructs forward and backward random search trees by biasing sampling towards low-concentration areas based on the model. The path cost is calculated by combining geometric distance and predicted average concentration. The two trees are then reconnected for optimization, updating the parent nodes of neighboring nodes with lower costs. The optimized two trees are then connected to generate an initial path. The cumulative hazardous gas exposure along the path is calculated, where the gas exposure is the integral of the product of residence time and concentration, thereby determining the target operating parameters for each ventilation device along the path. The prediction model is updated using the new parameters, and the path search and ventilation parameter optimization are repeatedly performed until the total cost change of adjacent iterations is less than a convergence threshold. Finally, a coordinated work path and ventilation scheme is output.

Description

Technical Field

[0001] This application belongs to the field of path planning, and in particular relates to a method and system for planning work paths that integrates environmental perception and occupational health. Background Technology

[0002] Farms maintain air quality by diluting and expelling harmful gases through ventilation equipment. Currently, manual start / stop, timed cycles, or threshold-based control using fixed-position sensors are commonly used. However, this lack of awareness of personnel locations and specific tasks prevents on-demand ventilation and makes it difficult to provide timely safety assurance for sudden high-concentration pollution areas or paths personnel are about to traverse. Personnel path planning often relies on individual experience or focuses solely on minimizing walking distance, rarely considering the invisible but potentially deadly risk factors of harmful gas concentration distribution. While mature path planning techniques such as A* and Rapid Expanding Random Tree (RRT) algorithms exist in robotics, their application focuses on obstacle avoidance. Treating personnel path planning and environmental ventilation control as two separate processes lacks information exchange and collaborative decision-making mechanisms. This can lead to personnel unintentionally traversing high-concentration risk areas; and the ventilation system cannot predict personnel's intentions, hindering directional ventilation. Furthermore, existing forward and backward search trees connect based solely on distance, which is not necessarily optimal. Therefore, it is impossible to achieve the best balance between ensuring personnel safety and improving ventilation efficiency, which limits the improvement of safe production and intelligent management in farms. Summary of the Invention

[0003] This invention proposes a work path planning method that integrates environmental perception and occupational health to address the problem that existing technologies cannot achieve the optimal balance between ensuring personnel safety and improving ventilation efficiency, including:

[0004] Obtain a 3D environmental map of the farm and establish a predictive model for the distribution of harmful gases associated with the operating status of the ventilation equipment; using the starting point and target point of the workers as root nodes, construct a forward search tree and a backward search tree respectively by performing bias sampling in the sampling space;

[0005] Reconnection optimization is performed on the forward and backward search trees. When a lower-cost path to a neighboring node via a newly sampled node exists in the search tree, the parent node of the neighboring node is updated to the newly sampled node. The optimized forward and backward search trees are then connected to generate the initial path.

[0006] The sum of the products of the estimated dwell time at each location point along the path and the corresponding concentration of harmful gas is taken as the cumulative exposure to harmful gas along the initial path, and the target operating parameters of each ventilation device along the path are determined based on the exposure amount.

[0007] The hazardous gas distribution prediction model is updated according to the target operating parameters, and the path search and exhaust parameter determination process is repeated based on the updated prediction model until the change in the total cost of the path obtained in two adjacent iterations is less than the preset convergence threshold. The operation path and exhaust coordination scheme are then output.

[0008] In another aspect, the present invention proposes a work path planning system that integrates environmental perception and occupational health, comprising the following modules:

[0009] The module is used to acquire a 3D environmental map of the farm and establish a predictive model for the distribution of harmful gases associated with the operating status of the ventilation equipment. Using the starting point and target point of the workers as root nodes, a forward search tree and a backward search tree are constructed respectively by bias sampling in the sampling space.

[0010] The generation module is used to optimize the reconnection of the forward and backward search trees. When there is a lower-cost path in the search tree that leads to a neighboring node via a newly sampled node, the parent node of the neighboring node is updated to the newly sampled node. The optimized forward and backward search trees are connected to generate the initial path.

[0011] The determination module is used to sum the products of the estimated dwell time at each location point on the path and the corresponding harmful gas concentration as the cumulative harmful gas exposure of the initial path, and to determine the target operating parameters of each exhaust ventilation device along the path based on the exposure amount.

[0012] The output module is used to update the hazardous gas distribution prediction model according to the target operating parameters, and repeat the path search and exhaust parameter determination process based on the updated prediction model until the change in the total cost of the path obtained in two adjacent iterations is less than the preset convergence threshold, and output the operation path and exhaust coordination scheme.

[0013] This invention establishes a hazardous gas distribution prediction model and uses the predicted hazardous gas concentration as a basis in path search. By optimizing the connection method of the forward and backward search trees, it can proactively avoid areas with high concentrations of hazardous gases, obtaining a path that balances safety and efficiency from the initial stage. Path planning and ventilation parameter adjustment are jointly optimized in a closed-loop iterative process. Based on the cumulative hazardous gas exposure along the planned path, the target operating parameters of the ventilation equipment along the route are determined in reverse, achieving coordinated optimization of the work path and ventilation scheme. This not only reduces the cumulative hazardous gas exposure of workers in the farm, effectively protecting occupational health and safety, but also enables on-demand control of ventilation equipment based on the work path, avoiding global or indiscriminate ventilation operations, reducing energy consumption in the farm, and improving environmental management. Attached Figure Description

[0014] Figure 1A flowchart of the first embodiment;

[0015] Figure 2 A diagram illustrating reconnection optimization for the search tree;

[0016] Figure 3 This is a schematic diagram of path assessment and coordinated adjustment of exhaust ventilation. Detailed Implementation

[0017] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0018] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0019] In the first embodiment, the present invention proposes a work path planning method that integrates environmental perception and occupational health, see [link to relevant documentation]. Figure 1 ,include:

[0020] S1. Obtain a 3D environmental map of the farm and establish a predictive model for the distribution of harmful gases associated with the operating status of the ventilation equipment. Using the starting point and target point of the workers as root nodes, construct a forward search tree and a backward search tree respectively by performing bias sampling in the sampling space.

[0021] A lidar scanner was used to perform a comprehensive scan of the farm's interior, generating point cloud data containing physical entities such as walls, fences, and feeding equipment. This point cloud data was then processed into a 3D raster map. Based on computational fluid dynamics, using the 3D raster map as the simulation boundary, real-time concentration data collected by a network of ammonia and hydrogen sulfide sensors deployed throughout the farm were used as initial conditions. The on / off status and fan speed of each ventilation system were used as variable boundary conditions. By solving the Navier-Stokes equations, the gas diffusion process was simulated, constructing a mathematical model capable of predicting the concentration of harmful gases at any location and time point based on ventilation system parameters. The distribution of harmful gases can also be calculated by interpolating the measurements from the harmful gas sensors.

[0022] A 3D coordinate system for a sampling point is randomly generated in a 3D raster map. The aforementioned prediction model is then used to query the predicted hazardous gas concentration at that coordinate system at the current time. An acceptance probability function is defined, which makes sampling points with lower concentrations more likely to be selected. For each selected sampling point, the forward search tree starts from the initial point, finds the node with the closest Euclidean distance, and extends it by a fixed step towards the sampling point to generate a new node. The path cost connecting the old and new nodes is calculated by multiplying their normalized geometric distance by a weighting coefficient α, and then adding the normalized value of the average predicted hazardous gas concentration at multiple points along the path multiplied by a weighting coefficient β. The backward search tree is constructed in reverse from the target point in the same way. In one embodiment, the path cost of a node in the forward search tree is the path length, and the cost of a node in the backward search tree is the average concentration.

[0023] In an optional embodiment, the step of acquiring a three-dimensional environmental map of the farm and establishing a predictive model for the distribution of harmful gases associated with the operating status of the ventilation equipment includes:

[0024] The three-dimensional environmental map of the farm is meshed into a set of voxels of a preset volume;

[0025] Based on computational fluid dynamics, a gridded map, information on harmful gas sources, and the operating status of each ventilation device are input as boundary conditions to perform simulation calculations and obtain the concentration values ​​of harmful gases in each voxel.

[0026] Specifically, taking a pigsty measuring 100m long, 40m wide, and 5m high as an example, the three-dimensional space is divided into 1m×1m×1m cubic units, i.e., voxels, forming a total of 20,000 discrete voxels. Each voxel is uniquely identified by the three-dimensional coordinates (x, y, z) of its center point. A computational fluid dynamics model based on the Reynolds-averaged Navier-Stokes equations is adopted, combined with the standard k-epsilon turbulence model and component transport equations. Boundary conditions are set as follows: the pigpen section is designated as the ammonia release source, with a release rate of 0.05 mg / m² / s; the air inlet at one end of the pigsty is defined as the velocity inlet boundary, with a wind speed of 2 m / s; the exhaust fan position at the other end is defined as the pressure outlet boundary, with a gauge pressure of -20 Pa, representing the operating state of the ventilation equipment. These conditions are input into simulation software, and through iterative solutions, the stable ammonia concentration value for each voxel is obtained, resulting in a three-dimensional concentration data matrix. This matrix serves as the predictive model for the spatial distribution of harmful gases in the farm. Optionally, when calculating path costs, etc., the voxel containing the mouth and nose is determined based on the person's height, and the concentration of other harmful substances corresponding to the voxel is used as the calculation cost or the concentration value of harm to the person's health.

[0027] In an optional embodiment, the bias sampling specifically includes:

[0028] Randomly generate candidate sampling points and obtain predicted harmful gas concentration values;

[0029] The probability of a candidate sampling point being accepted as a sampling node is inversely proportional to the predicted concentration of harmful gas.

[0030] A candidate sampling point is randomly generated within the three-dimensional space of the farm, for example, with coordinates x=30.5m, y=15.2m, z=1.8m. The predicted ammonia concentration at the voxel of the location of this point is obtained by querying the harmful gas distribution prediction model established in the previous step, assuming it to be 15ppm.

[0031] The acceptance probability of the candidate sampling point is calculated, and this probability is inversely proportional to the concentration value. A maximum reference concentration value is set, for example, 50 ppm. The acceptance probability P can then be defined as 1 minus the ratio of the current concentration to the maximum concentration. For a concentration of 15 ppm, the acceptance probability is 0.7. A random number between 0 and 1 is then generated. If this random number is less than 0.7, the candidate point is accepted as a new node in the path search tree; otherwise, it is discarded. Through this mechanism, the algorithm's exploration will tend to avoid high-concentration areas, thus oriented towards a safe environment from the initial planning stages.

[0032] In an optional embodiment, the path cost calculation formula for the forward search tree and the backward search tree is as follows:

[0033]

[0034] in, For the node To the node Path cost, The geometric distance between the two nodes is... This represents the predicted average concentration of harmful gases along the path between the two nodes.

[0035] Specifically, assume the path search tree starts from the existing node at coordinates (10, 5, 2). To the new node with coordinates (12,6,2) To expand, the Euclidean distance between the two nodes is... The value is 2.24m. This is to obtain the average concentration along the path between the two nodes. ,along arrive Uniform sampling is performed on the straight line segment. For example, 5 sampling points are taken, and the concentration values ​​of the points are obtained from the gas distribution model as 8ppm, 9ppm, 11ppm, 10ppm, and 9ppm, with an average value of 9.8ppm. Then the path cost is 21.952. The path cost is used to compare the advantages and disadvantages of different path branches.

[0036] S2, perform reconnection optimization on the forward and backward search trees. When there is a lower-cost path in the search tree that leads to a neighboring node via a newly sampled node, update the parent node of the neighboring node to the newly sampled node; connect the optimized forward and backward search trees to generate the initial path.

[0037] Once a new node is successfully added to the search tree, all neighboring nodes within a specific radius of the new node are searched. For each neighboring node, the total cost of a new path starting from the root node of the tree, passing through the new node, and then reaching that neighboring node is recalculated. If the total cost of the new path is less than the total cost of the original path of the neighboring node, the connection between the neighboring node and its original parent node is broken, and the new node is set as its parent node. Figure 2 .

[0038] In each iteration, after a new node is expanded from the forward search tree or backward search tree, an attempt is made to connect the new node to the nearest node in another tree. It is then checked whether the straight path between the two nodes passes through walls or obstacles on the map. Once an unobstructed connection path is found, the path segment from the starting point to the connection point in the forward search tree and the path segment from the target point to the connection point in the backward search tree are combined to form a complete initial path.

[0039] In an optional embodiment, the reconnection optimization of the forward and backward search trees, where a lower-cost path exists in the search tree leading to a neighboring node via the newly sampled node, and the parent node of the neighboring node is updated to the newly sampled node, includes:

[0040] Using the new sampling node as the center, search for neighboring nodes within a preset radius.

[0041] If the path cost to a neighboring node via the new sampling node is lower, then the parent node of the neighboring node is updated to the new sampling node.

[0042] Assuming a new node is successfully added to the search tree, the total path cost from the starting point is 50. Using the new node as the center, set a search radius, for example, 5m, and search for all existing tree nodes within that neighborhood.

[0043] Suppose a neighboring node is found in the neighborhood, and the total cost of the current path to this node via its old parent node is 60. We attempt to calculate the cost of a new path to the neighboring node via the new node. This new cost equals the path cost of the new node (50) plus the edge cost from the new node to the neighboring node, calculated using the aforementioned weighted sum formula, assumed to be 8. Therefore, the total cost of the new path is 58. Since 58 is less than the original path cost of 60, the new path is superior. We perform a reconnection operation, updating the parent node of the neighboring node from the old parent node to the new node, and recursively updating the path costs of all child nodes.

[0044] In a preferred embodiment, the connection of the optimized forward and backward search trees generates an initial path, including:

[0045] Obtain the complete forward search tree from the starting point to the target point, and the complete backward search tree from the target point to the starting point;

[0046] Obtain at least one region where the distance between the complete forward search tree and the complete backward search tree is less than a preset value. In each region, select a candidate node pair from the set of end nodes of the complete forward search tree and the complete backward search tree located within the region.

[0047] The initial path is obtained using the selected candidate node pairs.

[0048] In a three-dimensional environment, regions where the spatial distance between the two search trees is less than a preset threshold are identified, meaning the two trees are very close to each other. Within each region, the most suitable pair of nodes is selected as candidate node pairs from the end node sets of both the forward and backward trees, considering factors such as path cost, exposure risk, or minimum geometric distance. When multiple regions with distances less than the preset value exist between the complete forward and backward search trees, each region will yield at least one set of candidate node pairs consisting of a forward end node and a backward end node. The connection paths corresponding to these candidate node pairs are evaluated for cost, which can include factors such as path length, operation time, and hazardous gas exposure. The path with the lowest total cost is selected as the initial path.

[0049] In existing technologies, forward and backward search tree connections involve the forward tree growing one step, and the backward tree immediately attempting to connect to the new node. Once connected, the process ends. This often results in highly folded paths for the connection points, and the connection locations are random, potentially falling within areas with high concentrations of harmful gases. In another preferred embodiment, the optimized forward and backward search trees generate an initial path, specifically as follows:

[0050] The growth processes of the forward search tree and the backward search tree are executed independently until the number of nodes in both trees reaches the preset saturation threshold or the overlapping area of ​​the bounding boxes of the two trees reaches the preset ratio, at which point sampling stops.

[0051] Extract all nodes of the forward search tree and the backward search tree located within the overlapping region, and construct the forward front node set and the backward front node set respectively;

[0052] Traverse the node pairs in the forward front node set and the backward front node set, and select node pairs whose Euclidean distance is less than the preset connection threshold as candidate bridge pairs.

[0053] For each candidate bridging pair, calculate the angle between its connection vector and the forward node's in-vector and the backward node's out-vector, and construct a comprehensive bridging evaluation function that includes path length cost, gas concentration cost at the connection point, and connection angle smoothness cost.

[0054] The candidate bridging pair with the smallest comprehensive bridging evaluation function value is selected. This node pair is used as the hub to connect the forward search tree and the backward search tree. Redundant branches that are not involved in the connection are pruned, and the initial path is generated by backtracking.

[0055] This embodiment allows the forward and backward search trees to be searched independently. Assume a rectangular pig pen where workers need to walk from the east gate to the west feed tower, with a high-concentration ammonia accumulation zone in between. First, the forward tree starting from the east gate and the backward tree starting from the west are continuously extended. Even if the two trees have a chance to come into contact near the ammonia zone, sampling continues until the branches of both trees extend to the better-ventilated southern corridor area, forming an overlapping buffer zone containing multiple potential connection points. After obtaining a large number of potential connection points, the cost is calculated. The cost is determined by the smoothness of the connection and the gas concentration, ensuring that the selected connection points are neither bends created for connectivity nor shortcuts that cross the toxic gas center. For example, two connection schemes were identified: Scheme A involves two tree ends only 0.5 meters apart at the edge of the ammonia zone, but the connection forms an acute angle, requiring people to turn around and inhale a significant amount of ammonia gas; Scheme B is located in the south corridor, with the two nodes 1.5 meters apart, a slightly greater distance, but the connection forms a smooth straight line, and the area has good ventilation. Scheme A was eliminated due to its excessively high angle and gas costs. The generated initial path not only connects the trees but also conforms to human walking inertia and ensures occupational health.

[0056] S3, the sum of the products of the estimated dwell time at each location point on the path and the corresponding harmful gas concentration is taken as the cumulative harmful gas exposure of the initial path, and the target operating parameters of each exhaust ventilation device along the path are determined based on the exposure amount.

[0057] The generated initial path is decomposed into a series of tiny line segments. The time required to traverse each tiny line segment is calculated based on the average movement speed of the workers. A predictive model is used to obtain the hazardous gas concentration at the midpoint of each tiny line segment. The travel time for each line segment is multiplied by the corresponding gas concentration, and the results for all line segments are summed to obtain a total exposure value. If this value exceeds a preset safe exposure threshold, the areas with the highest concentrations on the path are identified, and the nearest exhaust equipment to these areas is determined. The target operating parameters are then increased by one or more levels, for example, from low speed to high speed.

[0058] In an optional embodiment, the step of using the sum of the products of the estimated dwell time at each location point along the path and the corresponding hazardous gas concentration as the cumulative hazardous gas exposure of the initial path includes:

[0059] Discretize the initial path into N path segments;

[0060] Based on the preset worker movement speed v, calculate the estimated dwell time for each path segment. ;

[0061] The dwell time of each path segment is correlated with the corresponding predicted concentration of harmful gases. Summing the products of , we obtain the cumulative harmful gas exposure E, calculated using the following formula:

[0062]

[0063] in, Let be the length of the i-th path segment.

[0064] Specifically, after obtaining a complete task path, the path is composed of a series of connected nodes, forming N consecutive path segments. Taking the i-th path segment as an example, its length is... Let v be the straight-line distance between the two nodes, assumed to be 5m. Set the average moving speed of the workers to 1.2 meters per second.

[0065] Calculate the time required for the operator to traverse this path segment. That is, approximately 4.17 seconds. Obtain the representative concentration of harmful gases for this path segment. The predicted concentration value at the midpoint of this segment can be taken, assumed to be 12 ppm. The exposure amount for this segment is 50.04 ppm·second. By calculating and summing the exposure amounts for all N path segments, the cumulative hazardous gas exposure E for the entire path can be obtained, representing the health risk to workers walking along this path.

[0066] In order to proactively optimize the working environment by adjusting the ventilation system, in one optional embodiment, determining the target operating parameters of each exhaust ventilation device along the line based on the exposure level includes:

[0067] In the process of calculating the cumulative exposure to harmful gases, path segments that contribute more than a preset threshold to the total exposure are identified, and the target operating parameters of the exhaust equipment corresponding to the path segments are improved.

[0068] Assume the total exposure for the entire path calculated in the previous step is 800 ppm·s. The contribution threshold is, for example, 10% of the total exposure, or 80 ppm·s. The exposure contribution value of each path segment is checked one by one. If the exposure of the fifth path segment is 95 ppm·s, exceeding the 80 ppm·s threshold, this segment is identified as a high-risk area. Based on the geographical coordinates of this path segment, it is located in the northeast area of ​​the farm. Consulting the equipment layout diagram reveals that exhaust fan number three is responsible for ventilation in this area. If the current operating parameter of exhaust fan number three is 60% power, an adjustment command is generated to increase the target operating parameter to 80%. This new combination of operating parameters will be used in the next round of simulation and path planning iterations.

[0069] S4. Update the hazardous gas distribution prediction model according to the target operating parameters, and repeat the path search and exhaust parameter determination process based on the updated prediction model until the change in the total cost of the path obtained in two adjacent iterations is less than the preset convergence threshold, and output the operation path and exhaust coordination scheme.

[0070] The newly determined operating parameters of the exhaust equipment, such as high-speed operation of fan 1 and medium-speed operation of fan 3, are input as new boundary conditions into the hazardous gas distribution prediction model to recalculate and generate an updated overall hazardous gas concentration distribution. Based on the updated concentration distribution, the random sampling, search tree construction, optimization, and connection steps of the path are repeated from scratch to obtain a new path and a new total cost. The difference between the total cost of the paths obtained in the two iterations is calculated. If the difference is less than a preset minimum value, such as 0.1, the loop terminates. The output scheme includes a path instruction consisting of a series of three-dimensional coordinate points and a sequence of linked exhaust equipment control instructions, specifying the specific operating state that each fan should be adjusted to at the start of the operation, such as... Figure 3 .

[0071] The entire optimization process is a closed-loop iterative process, requiring a termination condition. The relative rate of change of the total path cost obtained from two consecutive iterations is less than a preset convergence threshold. In the (k-1)th iteration, based on the existing ventilation settings, an optimal path is planned, with a total cost equal to the sum of the costs of all segments along the path, calculated to be 150.7. A high-risk area is identified based on this path, ventilation parameters are adjusted, and the process proceeds to the kth iteration. Under the new ventilation conditions, path planning is performed again, resulting in a new optimal path and a total cost of 149.2. The relative rate of change between the two iterations is calculated, which is approximately 0.995%. A convergence threshold is set, for example, 1%. Since the calculated rate of change is less than this threshold, the optimization process is considered to have converged, the iteration terminates, and the ventilation parameters and path obtained in the kth iteration are output as the optimized solution.

[0072] In a second embodiment, the present invention also proposes a work path planning system that integrates environmental perception and occupational health, comprising the following modules:

[0073] The module is used to acquire a 3D environmental map of the farm and establish a predictive model for the distribution of harmful gases associated with the operating status of the ventilation equipment. Using the starting point and target point of the workers as root nodes, a forward search tree and a backward search tree are constructed respectively by bias sampling in the sampling space.

[0074] The generation module is used to optimize the reconnection of the forward and backward search trees. When there is a lower-cost path in the search tree that leads to a neighboring node via a newly sampled node, the parent node of the neighboring node is updated to the newly sampled node. The optimized forward and backward search trees are connected to generate the initial path.

[0075] The determination module is used to sum the products of the estimated dwell time at each location point on the path and the corresponding harmful gas concentration as the cumulative harmful gas exposure of the initial path, and to determine the target operating parameters of each exhaust ventilation device along the path based on the exposure amount.

[0076] The output module is used to update the hazardous gas distribution prediction model according to the target operating parameters, and repeat the path search and exhaust parameter determination process based on the updated prediction model until the change in the total cost of the path obtained in two adjacent iterations is less than the preset convergence threshold, and output the operation path and exhaust coordination scheme.

[0077] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0078] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A work path planning method integrating environmental perception and occupational health, characterized in that, Includes the following steps: Obtain a 3D environmental map of the farm and establish a predictive model for the distribution of harmful gases associated with the operating status of the ventilation equipment; using the starting point and target point of the workers as root nodes, construct a forward search tree and a backward search tree respectively by performing bias sampling in the sampling space; Reconnection optimization is performed on the forward and backward search trees. When there is a lower-cost path in the search tree that leads to a neighboring node via a newly sampled node, the parent node of the neighboring node is updated to the newly sampled node. Connect the optimized forward and backward search trees to generate an initial path; The sum of the products of the estimated dwell time at each location point along the path and the corresponding concentration of harmful gas is taken as the cumulative exposure to harmful gas along the initial path, and the target operating parameters of each ventilation device along the path are determined based on the exposure amount. The hazardous gas distribution prediction model is updated according to the target operating parameters, and the path search and exhaust parameter determination process is repeated based on the updated prediction model until the change in the total cost of the path obtained in two adjacent iterations is less than the preset convergence threshold. The operation path and exhaust coordination scheme are then output.

2. The method according to claim 1, characterized in that, The process of acquiring a three-dimensional environmental map of the farm and establishing a predictive model for the distribution of harmful gases in relation to the operating status of the ventilation equipment includes: The three-dimensional environmental map of the farm is meshed into a set of voxels of a preset volume; Based on computational fluid dynamics, a gridded map, information on harmful gas sources, and the operating status of each ventilation device are input as boundary conditions to perform simulation calculations and obtain the concentration values ​​of harmful gases in each voxel.

3. The method according to claim 1, characterized in that, The bias sampling is specifically as follows: Randomly generate candidate sampling points and obtain predicted harmful gas concentration values; The probability of a candidate sampling point being accepted as a sampling node is inversely proportional to the predicted concentration of harmful gas.

4. The method according to claim 1, characterized in that, The optimized forward and backward search trees are used to generate an initial path, including: Obtain the complete forward search tree from the starting point to the target point, and the complete backward search tree from the target point to the starting point; Obtain at least one region where the distance between the complete forward search tree and the complete backward search tree is less than a preset value. In each region, select a candidate node pair from the set of end nodes of the complete forward search tree and the complete backward search tree located within the region. The initial path is obtained using the selected candidate node pairs.

5. The method according to claim 1 or 4, characterized in that, The reconnection optimization of the forward and backward search trees, where a lower-cost path exists in the search tree leading to a neighboring node via the newly sampled node, involves updating the parent node of the neighboring node to the newly sampled node. Using the new sampling node as the center, search for neighboring nodes within a preset radius. If the path cost to a neighboring node via the new sampling node is lower, then the parent node of the neighboring node is updated to the new sampling node.

6. The method according to any one of claims 1-3, characterized in that, The step of using the sum of the products of the estimated dwell time at each location point along the path and the corresponding hazardous gas concentration as the cumulative hazardous gas exposure of the initial path includes: Discretize the initial path into N path segments; Based on the preset worker movement speed v, calculate the estimated dwell time for each path segment; The cumulative exposure to harmful gases is obtained by summing the products of the dwell time on each path segment and the corresponding predicted concentration of harmful gases.

7. The method according to claim 1, characterized in that, The step of determining the target operating parameters of each ventilation device along the route based on the exposure amount includes: In the process of calculating the cumulative exposure to harmful gases, path segments that contribute more than a preset threshold to the total exposure are identified, and the target operating parameters of the exhaust equipment corresponding to the path segments are improved.

8. A work path planning system integrating environmental perception and occupational health, characterized in that, Includes the following modules: The module is used to acquire a 3D environmental map of the farm and establish a predictive model for the distribution of harmful gases associated with the operating status of the ventilation equipment. Using the starting point and target point of the workers as root nodes, a forward search tree and a backward search tree are constructed respectively by bias sampling in the sampling space. The generation module is used to optimize the reconnection of the forward and backward search trees. When there is a lower-cost path in the search tree that leads to a neighboring node via the newly sampled node, the parent node of the neighboring node is updated to the newly sampled node. Connect the optimized forward and backward search trees to generate an initial path; The determination module is used to sum the products of the estimated dwell time at each location point on the path and the corresponding harmful gas concentration as the cumulative harmful gas exposure of the initial path, and to determine the target operating parameters of each exhaust ventilation device along the path based on the exposure amount. The output module is used to update the hazardous gas distribution prediction model according to the target operating parameters, and repeat the path search and exhaust parameter determination process based on the updated prediction model until the change in the total cost of the path obtained in two adjacent iterations is less than the preset convergence threshold, and output the operation path and exhaust coordination scheme.

9. The system according to claim 8, characterized in that, The process of acquiring a three-dimensional environmental map of the farm and establishing a predictive model for the distribution of harmful gases in relation to the operating status of the ventilation equipment includes: The three-dimensional environmental map of the farm is meshed into a set of voxels of a preset volume; Based on computational fluid dynamics, a gridded map, information on harmful gas sources, and the operating status of each ventilation device are input as boundary conditions to perform simulation calculations and obtain the concentration values ​​of harmful gases in each voxel.

10. The system according to claim 8, characterized in that, The bias sampling is specifically as follows: Randomly generate candidate sampling points and obtain predicted harmful gas concentration values; The probability of a candidate sampling point being accepted as a sampling node is inversely proportional to the predicted concentration of harmful gas.

Images