A data processing system for chicken feed finished product warehousing and delivery path planning

Abstract

The present application relates to warehousing logistics and intelligent path planning technical field, specifically to a kind of chicken feed finished product warehousing and delivery path planning data processing system, including data acquisition module, dispatching control module and data processing hub;The grid topological map of warehousing area, idle location set, current inventory distribution data, target material material feature vector, each grid environment state slice and automatic guided vehicle current position are obtained;Based on material mutual exclusion relationship database generation space repulsion matrix, combined with environment state slice and historical weight coefficient generation area suitability heat map, and according to idle location set and automatic guided vehicle current position calculate physical distance cost, combined with grid topological map, the target location and driving track of minimum comprehensive cost are calculated using pathfinding algorithm, generate control instruction and issue to dispatching control module, realize the warehousing and delivery path planning of quality control and turnover efficiency.

Description

Technical Field

[0001] This invention relates to the field of warehousing logistics and intelligent route planning technology, specifically a data processing system for planning the entry and exit routes of finished chicken feed products. Background Technology

[0002] In the existing chicken feed finished product warehousing management process, it is usually necessary to plan the finished product inbound storage location and outbound driving route, and to coordinate and control the automated guided vehicles in conjunction with the warehousing equipment. In related technologies, in order to complete the warehousing and outbound of finished chicken feed, one approach is to allocate storage space mainly based on the availability of storage spaces and the distance from the vehicle to the target location, prioritizing storage spaces with a distance less than a preset distance threshold and generating a travel path with a length less than a preset length threshold; or, to perform regular storage space zoning management based on inventory information in the warehouse management system, so as to realize the basic storage and turnover of different batches of materials. However, given the differences in moisture content, fat ratio, and drug additives in finished chicken feed, as well as the variations in temperature, humidity, and ventilation conditions at different locations within the storage area, the former method struggles to simultaneously address formula isolation, microenvironmental adaptation, and path efficiency, while the latter method is difficult to dynamically adjust based on real-time environmental changes and actual quality feedback. Therefore, the chicken feed finished product warehousing and outbound path planning schemes in related technologies cannot simultaneously achieve quality control, cross-contamination prevention, and storage scheduling efficiency. Summary of the Invention

[0003] To solve the above-mentioned technical problems, the present invention provides a data processing system for planning the warehousing and outbound routes of finished chicken feed products. Specifically, the technical solution of the present invention includes: Furthermore, the data processing system includes: a data acquisition module, which is communicatively connected to the warehouse management system, for acquiring from the warehouse management system a gridded topology map of the warehouse area, a set of available storage locations, current inventory distribution data, and material feature vectors of the target materials of the current batch, and collecting environmental state slices of each grid in the warehouse area, as well as acquiring the current position of the automated guided vehicles in the warehouse area; The scheduling and control module is communicatively connected to the automated guided vehicle (AGV) and is used to control its driving trajectory. The data processing center is communicatively connected to both the data acquisition module and the scheduling and control module. It is used to: query a pre-stored material mutual exclusion relationship database based on the material feature vector, and generate a spatial repulsion matrix of the target material on the current inventory based on the current inventory distribution data. A regional suitability heatmap is generated based on the material feature vector, environmental state slices, and historical weight coefficients of each grid pre-stored in the data processing center; the physical distance cost is calculated based on the set of available storage locations and the current position of the automated guided vehicle. The spatial repulsion force matrix is ​​composed of the repulsion force characteristic values ​​corresponding to each grid, and the regional fitness heatmap is composed of the fitness characteristic values ​​corresponding to each grid. The repulsion force eigenvalues ​​of the spatial repulsion force matrix and the suitability eigenvalues ​​of the regional suitability heatmap are mapped to node cost values. Combined with the physical distance cost and the gridded topology map, a pathfinding algorithm is used to calculate the target storage location and driving trajectory with the minimum comprehensive cost. Control commands that include at least the driving trajectory are generated and sent to the scheduling control module to comprehensively manage the automated guided vehicle.

[0004] Furthermore, the environmental state slice includes various environmental parameters; the material feature vector includes at least an environmental sensitive feature vector containing the sensitivity weights corresponding to the various environmental parameters; the data processing center is also used to perform weighted summation calculations on the various environmental parameters in the environmental state slice, the corresponding sensitivity weights in the environmental sensitive feature vector, and the historical weight coefficients, to generate the regional suitability heatmap.

[0005] Furthermore, the pathfinding algorithm employs a heuristic search algorithm; When the data processing center uses the pathfinding algorithm to calculate the target storage location and driving trajectory with the minimum overall cost, it specifically uses the physical distance cost as the heuristic function of the heuristic search algorithm, the node cost as the actual cost function of the heuristic search algorithm, calculates the overall cost between nodes, and performs iterative calculations to search for the target storage location and driving trajectory with the minimum overall cost.

[0006] Furthermore, the data acquisition module is also used to acquire the actual quality inspection value of the target material when it leaves the warehouse; the data processing center pre-stores the expected quality inspection value; the data processing center is also used to calculate the absolute value of the difference between the actual quality inspection value and the expected quality inspection value to obtain the quality deviation value. The data processing center is also used to fine-tune the historical weight coefficient of the grid where the target storage location is located by using an error feedback-based heuristic weight update algorithm when the quality deviation value is greater than a preset deviation threshold, so as to use the historical weight coefficient to correct the next path planning; the data processing center is also used to keep the historical weight coefficient unchanged when the quality deviation value is less than or equal to the preset deviation threshold.

[0007] Furthermore, the data acquisition module is also used to monitor the data loss rate of the environmental state slice in real time; The data processing hub is also used to trigger a degradation execution logic when the data loss rate is higher than a preset loss rate threshold; the data processing hub is also used to maintain the normal planning mode when the data loss rate is lower than or equal to the preset loss rate threshold. The downgrade execution logic includes suspending the calculation process in the data processing center used to generate the regional suitability heatmap, and calculating the driving trajectory only based on the physical distance cost and the spatial repulsion matrix.

[0008] Furthermore, the target material is finished chicken feed, and the material feature vector is a formula feature vector; The formulation feature vector includes at least moisture content, fat ratio, and drug additive codes; the data processing center is also used to query the material mutual exclusion relationship database based on the drug additive codes to generate the spatial repulsion force matrix.

[0009] Furthermore, the environmental status slices are acquired by an array of wireless temperature and humidity sensors deployed on the shelves in the warehouse area; The data acquisition module is used to periodically retrieve environmental state slices of each grid in the storage area at preset time intervals.

[0010] Furthermore, the data processing center is also used to parse the drug additive code and determine whether the finished chicken feed contains a preset specific drug additive, wherein the preset specific drug additive is a drug additive pre-configured in a high-risk list of cross-contamination. The data processing center is also used to set the repulsion force characteristic value of the storage location within a preset physical distance range around the finished chicken feed product to a preset maximum repulsion force constant when the finished chicken feed product contains a preset specific drug additive, so as to generate the spatial repulsion force matrix. The data processing center is also used to generate the spatial repulsion matrix according to a preset standard repulsion rule when the finished chicken feed does not contain a preset specific drug additive. The standard repulsion rule is that the magnitude of the repulsion force is inversely proportional to the spatial distance between storage locations.

[0011] Furthermore, the data processing center is also used to encapsulate the calculated coordinate sequence of the driving trajectory into control instructions in a target format; The target format is JSON or XML; the data processing center sends the control commands to the scheduling control module through a message queue telemetry transmission protocol.

[0012] Compared with the prior art, the present invention has the following beneficial effects: 1. This system maps the spatial repulsion matrix, which represents the mutual exclusion risk of material storage, and the regional suitability heatmap, which represents the degree of matching between the storage microenvironment and the current batch of feed, into node cost values. Combined with physical distance costs, the system uses a pathfinding algorithm to calculate the target storage location and travel trajectory with the minimum comprehensive cost. This mechanism effectively overcomes the limitations of existing technologies that rely solely on spatial distance or static partitioning. It integrates the feed formulation isolation requirements, storage microenvironment differences, and logistics route scheduling into a unified calculation closed loop, achieving a high balance between quality control, cross-contamination prevention, and turnover efficiency. 2. This system introduces an actual outbound quality comparison mechanism. By calculating the deviation between the actual quality detection value at the time of outbound storage and the expected quality detection value pre-stored in the data processing center, when the deviation value exceeds a preset threshold, a heuristic weight update algorithm based on error feedback is used to fine-tune the historical weight coefficients of the grid. This mechanism can accurately identify and record hidden environmental problems that are difficult to reflect solely through real-time sensors, such as poorly ventilated areas and delayed heat storage. This allows the assessment of regional suitability to no longer rely solely on instantaneous environmental characteristics, but to achieve continuous evolution and self-correction of the planning model through feedback from actual storage quality. 3. This system deeply analyzes the formula feature vector of the target material. In particular, for chicken feed products containing pre-set specific drug additives, it directly sets the storage location repulsion force feature value within a preset range around the product to a preset maximum repulsion force constant. For ordinary materials, the standard rule that the repulsion force is inversely proportional to the distance is adopted. This differentiated treatment directly transforms strict compliance requirements and cross-contamination prevention constraints into underlying pathfinding barriers, establishing an absolute safety isolation boundary for sensitive drug-containing formulas and effectively avoiding the contamination risk when complex formulas are stored together. 4. This system monitors the data loss rate of environmental status slices in real time. When encountering network congestion or local sensor failure leading to excessively high data loss rates, the system automatically triggers degraded execution logic. By suspending the calculation process of the regional suitability heatmap that depends on environmental data, the system continues to calculate the driving trajectory based solely on physical distance costs and spatial repulsion matrix. This mechanism effectively prevents automated guided vehicles from mistakenly entering congested or unsuitable areas due to the adoption of large-area distorted environmental information, ensuring the continuous availability of the system under extreme conditions, and ensuring the safety constraint standards of formula isolation while maintaining basic logistics efficiency. Attached Figure Description

[0013] Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0015] like Figure 1As shown, a data processing system for planning the warehousing and outbound routes of finished chicken feed products is provided. The data processing system includes: The data acquisition module communicates with the warehouse management system and is used to obtain the gridded topology map of the warehouse area, the set of available storage locations, the current inventory distribution data, and the material feature vector of the target material of the current batch from the warehouse management system. It also collects environmental status slices of each grid in the warehouse area and obtains the current position of the automated guided vehicles in the warehouse area. The scheduling and control module communicates with the automated guided vehicle and is used to control its driving trajectory; The data processing hub is connected to the data acquisition module and the scheduling control module respectively; it is used to: query the pre-stored material mutual exclusion relationship database based on the material feature vector, and generate the spatial repulsion matrix of the target material on the current inventory based on the current inventory distribution data; A regional suitability heatmap is generated based on material feature vectors, environmental state slices, and historical weight coefficients of each grid pre-stored in the data processing center; the physical distance cost is calculated based on the set of available storage locations and the current location of the automated guided vehicles. Among them, the spatial repulsion force matrix is ​​composed of the repulsion force characteristic values ​​of each grid, and the regional fitness heatmap is composed of the fitness characteristic values ​​of each grid. The repulsion force eigenvalues ​​of the spatial repulsion force matrix and the suitability eigenvalues ​​of the regional suitability heatmap are mapped to node cost values. Combined with physical distance cost and gridded topology map, a pathfinding algorithm is used to calculate the target storage location and driving trajectory with the minimum comprehensive cost. Control commands, including at least the driving trajectory, are generated and sent to the scheduling control module for comprehensive management of automated guided vehicles.

[0016] This embodiment provides a data processing mechanism for planning the entry and exit routes of finished chicken feed. Specifically, this embodiment uses an automated warehouse of a broiler feed production enterprise as a unified scenario: two batches of finished chicken feed are produced from the production line in the morning, one batch being high-fat fattening broiler feed and the other being health feed containing medicated additives; the warehouse already contains chick feed, layer hen feed, and non-medicated formula feed stored the previous day, and in the afternoon of summer, the upper temperature of the warehouse area is higher than or equal to a preset temperature threshold, and the airflow near the ventilation corridor is greater than or equal to a preset flow rate threshold; under this continuous business link, the system simultaneously completes the entry location selection, the automatic guided vehicle route issuance, and subsequent exit tracking. Specifically, the data acquisition module reads a gridded topology map of the storage area from the warehouse management system. This map divides the storage area into multiple regular grids, each corresponding to an actual passage node or storage location node. The gridded spatial division mode is based on the physical characteristics that the quality status of finished chicken feed is constrained by the target storage location environment, the environment of the travel route area, and the repulsive force of adjacent stored materials. For example, a grid near the roof heat accumulation area may be unsuitable for storing high-fat feed even if the geometric distance to the target storage location is less than a preset distance threshold, because the temperature may be higher than a preset temperature threshold for more than a preset time threshold. Similarly, a grid near the storage of medicated feed may be a shortcut near an empty storage location, but it may increase the risk of cross-settling of dust. Based on this, the data acquisition module further reads the set of available storage locations, the current inventory distribution data, and the material feature vector of the target material for the current batch, and collects environmental state slices for each grid, while also obtaining the current position of the automated guided vehicle. The environmental state slice can be understood as a snapshot of the local storage environment of the grid at a certain moment, reflecting the real micro-environmental conditions such as temperature and humidity near the location. The material feature vector reflects the sensitivity of the batch of feed to the external environment and the surrounding formula. Since chicken feed products usually have residual heat and have different types of oil, moisture content, and drug additives, their tendency to clump, moisture absorption, and cross-contamination sensitivity in storage are not the same. Therefore, the system configures them as customized finished products with multi-dimensional environmental sensitivity and formula isolation constraints for data processing. After acquiring the aforementioned data, the data processing center first queries the material mutual exclusion database based on the material feature vector, and then generates a spatial repulsion matrix by combining it with the current inventory distribution data. Spatial repulsion is a data feature used to characterize the degree to which target materials are not suitable for close proximity storage. If the target material has a formula isolation requirement with existing materials in a certain area, the repulsion force of that area and its surrounding grids will increase to a preset high threshold. If there is no compatibility conflict exceeding the preset conflict judgment standard, the repulsion force will remain at or below the preset repulsion baseline value. Then, combining the environmental state slice and the historical weight coefficients of each grid, the data processing center generates a regional suitability heat map. This heat map expresses the degree of matching between a batch of feed and the current microenvironment of the storage area. Areas with high suitability are more suitable for that batch of feed to complete cooling, moisture dissipation, and quality stabilization. The data processing center calculates the physical distance cost based on the set of available storage locations and the current location of the automated guided vehicles (AGVs). This cost retains the focus on transportation efficiency in traditional logistics systems, preventing excessive detours in pursuit of quality protection. Then, the repulsion force eigenvalues ​​in the spatial repulsion force matrix and the suitability eigenvalues ​​in the regional suitability heatmap are mapped to node cost values. The specific mapping logic for node cost values ​​is as follows: using a maximum-minimum normalization algorithm, the repulsion force eigenvalues ​​in the spatial repulsion force matrix and the suitability eigenvalues ​​in the regional suitability heatmap are uniformly mapped and scaled to [the appropriate value]. The interval is used to obtain the dimensionless normalized repulsive force. With normalized goodness Then, the node cost is output through the mapping calculation formula. ,in and The pre-calibrated cost conversion weight constant ensures that the higher the suitability of the grid, the smaller the comprehensive node cost value output. The system combines a gridded topology map with a preset pathfinding algorithm to calculate the target storage location and driving trajectory with the lowest overall cost. Through the above mapping logic, the system simultaneously incorporates logistics efficiency parameters and environmental cross-contamination parameters during the calculation of the target location. For example, the storage area can be simplified into a nine-square grid, where the two middle squares are vacant storage spaces, the upper left square is adjacent to a ventilation opening, and the upper right square is adjacent to a drug-containing storage space. If the automated guided vehicle is located at the lower entrance, and the target batch is high-fat broiler feed, the upper right empty space, although having the fewest path turning nodes, is assigned a node cost greater than the baseline value because its heat accumulation exceeds the preset environmental safety threshold and it is adjacent to a drug-containing storage space. The physical distance value of the empty space in the middle left exceeds that of the empty space in the upper right, and the temperature drop gradient value indicated by the environmental state slice corresponding to the empty space in the middle left meets the preset heat dissipation threshold. After calculation, its comprehensive cost value is less than that of the empty space in the upper right. Based on this, the system determines the target storage space and outputs the corresponding driving trajectory. In one anomaly handling mechanism, if the set of available storage locations is empty, the data processing center can return the pending allocation status to the warehouse management system and suspend the current inbound task; if there are local missing parts in the environmental status slice but it does not reach the level of overall unavailability, only the most recent valid status can be used for the missing grid, and it can be recalibrated in subsequent cycles; if the current position of the automated guided vehicle is abnormal or communication is temporarily lost, the scheduling control module will not immediately execute the path issuance, but will keep the vehicle parked to avoid entering congested or high-risk areas when the basic path data is incomplete; In the aforementioned summer production scenario, the system receives a batch of high-fat broiler feed that has just exited the pellet mill and is still warm before being put into storage. The upper area on the north side of the storage area forms a high-temperature zone due to afternoon heat accumulation, there is already medicated feed inventory on the east side, and there are several vacant storage spaces near the ventilation corridor on the west side. After reading the material characteristics of the current batch, the system determines that it needs better heat dissipation and should avoid the area near the medicated feed inventory. Therefore, it prioritizes the vacant spaces on the west side with better ventilation and controls the automated guided vehicle to travel along the corridor away from the medicated feed area to complete the storage. When leaving the storage during the evening shift, the system generates a departure path for another batch according to the above mechanism and controls the automated guided vehicle to generate a driving trajectory that does not include the dust-sensitive area grid. Through the above steps, the system integrates the formula isolation rule parameters, microenvironment state data, and logistics path cost parameters into the same calculation model, and outputs the driving trajectory that meets the quality prevention and control standards and the turnover time threshold. In this embodiment, the environmental state slice includes various environmental parameters; the material feature vector includes at least an environmental sensitivity feature vector containing the sensitivity weights corresponding to each environmental parameter; The data processing center is also used to perform weighted summation calculations on each environmental parameter in the environmental state slice with the corresponding sensitivity weight and historical weight coefficient in the environmental sensitivity feature vector, and generate a regional suitability heat map.

[0017] This embodiment provides a mechanism for constructing a regional suitability heat map. Specifically, in the aforementioned scenario of a three-dimensional feed warehouse, simply knowing that a certain grid has a high temperature or low humidity is insufficient to guide the storage, because different chicken feed formulations do not respond consistently to the same environmental changes. Therefore, this embodiment further refines the environmental state slices into multiple environmental parameters and introduces environmental sensitivity feature vectors corresponding to each environmental parameter into the material feature vector, enabling the system to distinguish the different sensitivity thresholds of different feeds to high temperature and high humidity environmental characteristics. Specifically, the environmental status slice can include temperature, relative humidity, air flow indicators, and auxiliary parameters related to the stability of the storage environment; the environmental sensitivity feature vector is used to describe the sensitivity of the current batch of feed to these parameters; for high-fat feed, the data feedback probability of triggering oil oxidation and adhesion when the ambient temperature exceeds the preset standard exceeds the benchmark value; the probability of feed triggering damp clumping when the moisture content exceeds the preset moisture threshold exceeds the benchmark value; for the same physical grid, the suitability feature values ​​calculated based on the feature vectors of different target material formulations have specific numerical differences; the historical weight coefficient reflects the empirical quality performance accumulated by the grid in the past multiple batches of actual use. For example, some corner areas that seem to have normal temperature and humidity are still more likely to cause finished products to become damp due to long-term air stagnation, and the system can correct them through historical weights; During processing, the data processing center associates each environmental parameter in the environmental state slice with the corresponding sensitivity weight in the environmental sensitive feature vector, and then superimposes the historical weight coefficient to generate a regional suitability heat map. In the aforementioned calculation model, the environmental parameters are used to input and characterize the real-time state data of the storage area, the environmental sensitive feature vector characterizes the sensitive constraints of the target material on the state, and the historical weight coefficient characterizes the historical time series performance of the area. Specifically, the aforementioned weighted summation calculation follows the following structured steps: For any specified independent grid, the environmental parameter values ​​corresponding to that grid are normalized to eliminate dimensions, multiplied one by one with their corresponding sensitivity weights, and all products are summed to obtain the initial environmental assessment base score for that grid for the current batch of materials; the historical weight coefficients are multiplied by a preset historical experience weighting constant and added to the initial environmental assessment base score to output the final suitability feature value that integrates multiple factors; Specifically, the above weighted summation calculation can be performed using structured formulas. Characterization, the system represents the total number of environmental parameters. The first Values ​​of environmental parameters Its corresponding sensitivity weight Multiply and sum, where the numerical values ​​in the formula are... The values ​​are dimensionless values ​​after normalization; historical weighting coefficients are added. Weighted constant based on pre-set historical experience The product of, where, The summation operator is used to calculate the final suitability feature value. ; The specific number of environmental parameters in the calculation formula Set as The parameters represent temperature, humidity, and airflow, respectively; the sensitivity weights of each environmental parameter are as follows: Based on the pre-set material compatibility constraint matrix within the data processing center, values ​​are assigned from a lookup table, and the material sensitivity characteristics are assigned according to a pre-set scoring model. Values ​​within the range; historical weighting coefficients The initial value is set to This represents the objective basic evaluation status of the grid when there is no historical experience accumulation of outbound quality feedback. By clearly defining and formulating the rules for this weighted summation operation, the logic requirements in the implementation are fully met, eliminating the computational ambiguity in the multi-dimensional parameter fusion process. When the three are combined, the suitability heatmap is closer to the actual decision-making basis of the warehousing site. In a predefined application scenario, the environmental sensitivity feature vector of a batch of target materials includes temperature and humidity parameters. Grid G1 is located near the ventilation corridor, and grid G2 is located below the roof heat accumulation zone. The environmental status slice shows that G1 temperature is within a predefined suitable range and humidity is below a predefined humidity threshold, while G2 temperature is above a predefined temperature threshold and humidity is within a predefined humidity threshold range. Simultaneously, the temperature sensitivity weight of this batch of feed is greater than the humidity sensitivity weight, and the historical clumping rate of the G2 region is greater than a predefined clumping threshold. Therefore, when generating the heatmap, the system will mark G1 as a more suitable region and G2 as a region requiring caution. The regional suitability heatmap represents the superimposed correlation between formulation characteristics, environmental parameters, and historical performance. For abnormal operating conditions, if an environmental parameter fluctuates abnormally within a local period, but the system determines it to be transient noise, such as a temperature jump caused by a short-term sensor jitter, the trend value of the most recent consecutive effective period can be used for smoothing. If the target material lacks a complete environmentally sensitive feature vector, the system can call a similar formula template as a temporary feature basis and record the batch as a low-confidence planning result. If historical weight coefficients have not yet been effectively accumulated, such as in the early stages of a newly built warehouse, the intensity of historical corrections can be reduced first, and the heat map generation can be completed primarily by relying on real-time environmental parameters. In the aforementioned summer scenario, the high-fat broiler feed produced in the morning has a higher degree of heat sensitivity, while the batch of low-fat chick feed to be stored in the afternoon has a lower degree of temperature sensitivity. Even if the two batches are facing the same storage area environment slice, the system sets the suitability feature value of the upper heat zone grid for high-fat broiler feed to a minimum value and generates a larger number of candidate grid sequences for low-fat chick feed. If the historical moisture retest value corresponding to the northwest corner storage location is greater than the preset moisture retest threshold, the system will still appropriately lower the suitability of the grid even if the current humidity reading is normal. This enables the regional suitability heatmap to perceive formula differences and correct based on historical experience, thereby achieving a storage location selection that is closer to the quality mechanism of chicken feed. In this embodiment, the pathfinding algorithm adopts a heuristic search algorithm. When the data processing center uses the pathfinding algorithm to calculate the target storage location and driving trajectory with the minimum comprehensive cost, it specifically uses: physical distance cost as the heuristic function of the heuristic search algorithm, node cost value as the actual cost function of the heuristic search algorithm, calculates the comprehensive cost between nodes, and performs iterative calculation to search for the target storage location and driving trajectory with the minimum comprehensive cost.

[0018] This embodiment provides a heuristic pathfinding mechanism driven by comprehensive cost. Specifically, in the aforementioned scenario, if the entire warehouse area is exhaustively searched point by point based solely on comprehensive cost, a path can theoretically be found. However, when the number of grids in the automated warehouse exceeds a preset grid number threshold, and the number of concurrent task instructions from the automated guided vehicles exceeds a preset concurrency threshold, there is a technical risk that the pathfinding computation delay of the data processing center will exceed a preset period. On the other hand, if the shortest distance is used for rapid decision-making, the need for feed quality protection will be ignored. Therefore, this embodiment adopts a heuristic search algorithm to achieve a balance between real-time performance and decision quality. Specifically, the heuristic function in the heuristic search algorithm consists of physical distance cost, which is used to estimate the remaining distance cost from the current exploration grid to the target storage location by calculating the Manhattan distance or Chebyshev distance between the current exploration grid and the target storage location. The actual cost function consists of node cost value, which is used to evaluate the actual cumulative cost of passing through the grid. To ensure the rigor of the algorithm's calculation system, the node cost value is used as the actual cost function. In the specific calculation process, this means: taking the starting point of the automated guided vehicle as the cumulative zero point, moving along the currently generated exploration trajectory to the current grid, and accumulating the node cost value of each grid passed through on the continuous trajectory to obtain the cumulative actual cost; adding this cumulative actual cost to the aforementioned heuristic function to output the comprehensive cost required for the iterative calculation of the heuristic search algorithm. To eliminate the dimensional difference between physical distance cost and node cost, before performing the addition operation, the physical distance cost is normalized by division using a preset global maximum diagonal Manhattan distance parameter of the storage area, forcibly converting it to... The dimensionless distance cost term of the interval; at the same time, the termination condition for the heuristic search algorithm to perform iteration is set as follows: when the front grid cursor of the exploration extension path is equal to the target storage node, and the comprehensive cost value of the current complete exploration path is strictly less than the estimated comprehensive cost of all other candidate exploration branches in the open list, the system search is determined to be converged, the iterative calculation is stopped immediately and the optimal driving trajectory result at this time is output. The physical distance cost is used to control the driving range parameters of the automated guided vehicle, and the accumulated total node cost is used to incorporate the formulation rejection risk and environmental exposure risk along the way into the pathfinding assessment calculation model; that is, the system combines the directional guidance of logistics scheduling and the environmental risk constraints of sensitive materials simultaneously during the search. In a specific application scenario, the starting point of the automated guided vehicle (AGV) is set as A, and the candidate target storage location includes nodes B and C. The physical distance from node A to node B is less than the physical distance from node A to node C, and the travel trajectory from A to B includes grids with temperature parameters exceeding a preset safety threshold and grids with repulsion force parameters exceeding a baseline value. The variance of the fluctuation of the grid environmental parameters along the travel trajectory from node A to node C is less than a preset environmental stability variance threshold. In the heuristic search, although exploring towards B can quickly approach the target, the actual cost will continuously increase due to the increase of unfavorable nodes. When exploring towards C, the remaining physical distance is greater than the physical distance from node A to node B, but the variance of the fluctuation of the grid environmental parameters along the way is less than the preset environmental stability variance threshold. After iteration, the system will prioritize retaining the search branch with the better overall cost until the final target storage location and complete travel trajectory are determined. The key point is that the search process is not limited to a single shortest distance path, but continuously optimizes the convergence of physical distance and the avoidance of environmental risks. In terms of storage location selection, heuristic search can also handle the problem of linking target storage locations with paths; some empty locations are suitable in terms of environment, but the passage to the empty location needs to pass through high-risk storage areas. In this case, the system is configured to perform joint calculation and decision based on the suitability of the target storage location's environmental parameters and the cumulative node cost of each grid along the travel trajectory; this can avoid the situation where the storage location itself is qualified but the travel path causes cross-contamination. In case of abnormal operating conditions, if the comprehensive cost of multiple candidate paths is the same, a tie can be broken according to a preset priority. For example, routes with fewer turns can be selected first to reduce the attitude adjustment of the automated guided vehicle in densely packed shelving areas. If the heuristic search does not converge within the time window, the currently known best feasible path can be output and marked as a temporary path for further optimization in the background. If the search finds that all candidate vacancies are unacceptable due to high rejection or low suitability, the system can initiate manual review or wait for the next batch of vacancies to be released. In the aforementioned summer scenario, the automated guided vehicle (AGV) is located at the south entrance, where high-fat broiler feed needs to be stored. Although the nearby empty space on the north side is short, it requires passing through the passage below the upper hot zone. The more distant empty space on the west side is slightly further away, but it is close to the ventilation corridor and far from the drug-containing storage. After adopting a heuristic search, the system will not directly select the nearby empty space on the north side as in the traditional solution. Instead, it will gradually abandon high-cost branches during the search process and finally determine the empty space on the west side and its corresponding passage as the driving trajectory. While meeting the real-time scheduling needs of industrial sites, the search process incorporates three factors simultaneously: path length, environmental exposure, and formula isolation, thereby achieving path planning with the lowest overall cost. In this embodiment, the data acquisition module is also used to acquire the actual quality inspection value of the target material when it leaves the warehouse; The data processing center pre-stores the expected quality test values; The data processing center is also used to calculate the absolute value of the difference between the actual quality inspection value and the expected quality inspection value, thus obtaining the quality deviation value; The data processing center is also used to fine-tune the historical weight coefficients of the grid where the target storage location is located by using an error feedback-based heuristic weight update algorithm when the quality deviation value is greater than the preset deviation threshold, so as to use the historical weight coefficients to correct the next path planning. The data processing center is also used to keep the historical weighting coefficients unchanged when the quality deviation value is less than or equal to the preset deviation threshold.

[0019] This embodiment provides a historical weight self-correction mechanism for outbound quality feedback. Specifically, in the aforementioned same warehouse lifecycle scenario, high-fat broiler feed that enters the warehouse in the morning enters the outbound stage after several days of storage. If the system always relies on the initial empirical rules without using the actual outbound quality to correct the warehouse location evaluation, some hidden unfavorable areas may not be accurately assessed for a long time. Therefore, this embodiment achieves closed-loop update of path planning parameters by introducing a comparison between the actual quality detection value and the expected quality detection value. Specifically, actual quality test values ​​can be entered by quality inspectors at the outbound stage using portable data terminals. The test content may include clumping rate, moisture retest value, signs of mold, or other indicators related to the stability of the finished product. Expected quality test values ​​are pre-stored by the data processing center and usually correspond to the quality level that the formula should achieve under normal storage conditions. When the difference between the two forms a quality deviation value, the system can determine whether the grid and surrounding environment previously selected for this batch are truly suitable. When the quality deviation value exceeds the preset deviation threshold, it is determined that there are long-term environmental interference characteristics in the corresponding grid area that deviate from the range characterized by the instantaneous temperature and humidity sensor parameters, such as poor local ventilation, delayed temperature and humidity recovery at night, and mutual heat dissipation interference between adjacent storage locations. Therefore, the data processing center uses a heuristic weight update algorithm based on error feedback to fine-tune the historical weight coefficients of the grid where the target storage location is located. The heuristic weight update algorithm based on error feedback is configured to iteratively correct the regional evaluation parameters along the direction of reducing quality deviation, so as to avoid distortion of the planning model caused by accidental abnormal data in a single batch. At the specific data processing level, the underlying calculation logic of this fine-tuning operation is structured as follows: a learning rate parameter representing the step size of a single iteration adjustment is pre-set in the data processing center. The direction of correction is determined based on the relationship between the actual quality performance and the expected performance. For example, when the actual water distribution ratio and other poor indicators are higher than expected, the direction of correction is set to negative. Finally, the system adds the sign of the correction direction, the learning rate and the product of the quality deviation value to the currently stored historical weight coefficients, thereby calculating and outputting the new historical weight coefficients after fine-tuning. Specifically, the fine-tuning update operation of the aforementioned error feedback-based heuristic weight update algorithm can be performed using structured formulas. The system will represent the currently stored historical weight coefficients. That is, the historical weight coefficient H in the aforementioned weighted summation calculation plus the learning rate parameter. The value is positive. or negative Corrected polarity direction and quality deviation value The product of these three factors is used to calculate and output the new historical weight coefficients after fine-tuning. By constructing this algebraic operation structure and clear formula that is strictly defined by the learning rate, quality deviation value and correction polarity, the explicit and formulaic output of the implicit / invisible operation logic of heuristic weights is realized, ensuring the deterministic reproducibility of the evaluation experience of the warehouse area in the correction process. In this evaluation system, the historical weight coefficient The higher the value, the higher the probability that the grid area has experienced poor quality deterioration in historical storage, meaning it is less suitable for storage; (Correcting polarity direction) The judgment logic is clearly defined as follows: when the actual quality inspection value upon leaving the warehouse is worse than the expected quality inspection value, it indicates that the environmental degradation of the warehouse location has a negative result attribute, and in this case, let... This makes the updated historical weight coefficients Increase the value to reduce the probability of this grid being selected in subsequent path optimization; when the actual quality inspection value is better than or equal to the expected quality inspection value, let... The algorithm mechanism utilizes quality deviation values. It is directly used as the driving control quantity for error feedback step size, which avoids the problem of lack of differentiable loss function in traditional analytical models, thus realizing heuristic adaptive parameter optimization in gradient-free environment; Conversely, when the quality deviation value does not exceed the preset deviation threshold, it indicates that the previous storage location matching and path planning are basically consistent with the actual quality results, and the system keeps the historical weight coefficient of the grid unchanged; this can both preserve the regional experience that has been verified and avoid unnecessary frequent fluctuations. To illustrate with a simplified application scenario, suppose grid G5 has historically been considered suitable for storing a certain type of high-fat broiler feed and has therefore been selected as the target storage location multiple times. If the actual moisture content measured during the retesting after outgoing feed exceeds the preset expected value range, the system will fine-tune the historical weight corresponding to G5 by increasing the polarity of the cost value with a preset step size. If the results of multiple batches stored in this area have stabilized, the adjustment range can be reduced according to a preset decay rate. In this way, the historical weight is not a fixed label, but gradually approaches the true characteristics of the area as the actual quality performance changes. In case of abnormal operating conditions, if the actual quality inspection value is missing, the system will not update the feedback for that batch, but will only retain the original historical weight; if the inspection value is obviously abnormal and exceeds the reasonable range of the process, such as the value being distorted due to an input error, it will be marked as data to be reviewed and will not directly trigger weight adjustment; if there are multiple quality indicators in the same batch and the results are contradictory, the key indicator can be selected as the basis for correction according to the preset priority, or the update can be performed after the complete quality inspection is confirmed. In the aforementioned scenario, a batch of high-fat broiler feed was initially placed in the upper storage area in the northwest corner. The system initially determined that this area was suitable based on a real-time heat map. Several days later, a retest upon delivery revealed that the clumping degree was higher than expected, while other conditions for the same batch were normal. Based on this, the system concluded that although the storage area appeared to be qualified in terms of environmental parameters at a certain point in time, its long-term heat storage characteristics were underestimated. Therefore, the historical weight of this grid was lowered, preventing subsequent similar high-fat feed formulations from being preferentially placed in this area. If another batch of finished products stored near the ventilation corridor was of normal quality upon delivery, the historical weight of the corresponding area remained unchanged. The outbound quality results are fed back into the inbound and route planning decisions to achieve continuous correction based on the actual quality consequences. In this embodiment, the data acquisition module is also used to monitor the data loss rate of environmental state slices in real time; The data processing hub is also used to trigger degradation execution logic when the data loss rate exceeds a preset loss rate threshold; The data processing hub is also used to maintain the normal planning mode when the data loss rate is lower than or equal to a preset loss rate threshold; The downgrade execution logic includes suspending the calculation process in the data processing center used to generate the regional suitability heatmap, and calculating the driving trajectory only based on the physical distance cost and the spatial repulsion matrix.

[0020] This embodiment provides a degradation execution mechanism for environmental data anomalies. Specifically, in the aforementioned main scenario, the system is highly dependent on the microenvironment of the storage area, but situations such as sensor gateway disconnection, partial power outages, and wireless network congestion inevitably occur in industrial sites. If a regional suitability heatmap is forcibly generated when a large area of ​​environmental state slices is missing, the system may make incorrect path planning based on the distorted environment. Therefore, this embodiment sets degradation logic for data loss rate. Specifically, the data acquisition module continuously monitors the data integrity of environmental state slices and calculates the data loss rate. The engineering significance of the data loss rate lies in measuring whether the current reservoir microenvironment is still reliable enough to be reconstructed. When the data loss rate is lower than or equal to the preset loss rate threshold, it indicates that the microenvironment status of most grids in the reservoir area can still be effectively observed, and the system maintains the conventional planning mode, that is, continues to comprehensively consider the physical distance cost, spatial repulsion matrix, and regional suitability heat map. When the data loss rate is higher than the preset threshold, it is determined that the environmental perception data is invalid, the system stops evaluating the suitability of each grid in the upper layer for heat dissipation from high-fat feed, and immediately suspends the heat map calculation process, calculating the driving trajectory only based on the physical distance cost and spatial repulsion matrix. The degradation execution logic is configured to maintain the basic operation of the system. Under the condition that environmental status data collection is interrupted, the system is configured to perform degradation calculation based on the formula isolation standard rules and physical distance cost, so as to output control instructions that meet the formula anti-cross-contamination threshold and basic turnover efficiency indicators. By retaining these two types of information, the system can still avoid obvious risks such as mixing of medicated and non-medicated feeds, while ensuring continuous feeding of the production line and no congestion on the loading and unloading platform. In a specific application scenario, suppose there are ten key grid environment sampling points in a warehouse area. Under normal circumstances, all ten sampling points have data, and the system can draw a complete heat map. If only one sampling point is lost, the system can still maintain the normal mode using the remaining information. If more than five sampling points are lost, the heat map will have large areas of blank space, and continuing to rely on its results will lack reliability. Therefore, the system will switch to looking only at distance and formula isolation. In one anomaly handling mechanism, if the data loss rate is higher than the threshold but the spatial repulsion matrix can still be generated normally, the system will continue to operate according to the degradation logic. If the inventory distribution data is also unavailable at the same time, the system will suspend automatic allocation and issue a manual intervention alarm. If the environmental data returns to normal and remains stable for several consecutive periods, the system can automatically exit the degradation mode and reactivate the regional suitability heatmap. If only a few key areas have long-term missing data, the area can be temporarily marked as a conservative area, and its priority as a target storage location will be reduced before recovery. During the night shift in the aforementioned summer scenario, local network fluctuations in the warehouse caused widespread communication interruptions in the upper shelf sensor arrays, resulting in a rapid increase in the data loss rate of environmental status slices. At this point, the system stopped assessing the suitability of each grid on the upper shelf for heat dissipation of high-fat feed and immediately suspended heat map calculations. Based solely on the current distance to available storage locations and the isolation rules for drug-containing formulas, a conservative and feasible path was generated for the batches to be put into storage. After the network was restored and continuous and valid environmental data was reacquired, the system resumed its normal planning mode. Avoid erroneous optimization under abnormal operating conditions where environmental perception capabilities are insufficient, thereby achieving continuous operability and risk controllability of the system under extreme conditions; In this embodiment, the target material is finished chicken feed, and the material feature vector is the formula feature vector; The formulation feature vector should include at least the moisture content, fat content, and drug additive codes; The data processing center is also used to query the material mutual exclusion relationship database based on drug additive codes and generate a spatial repulsion force matrix.

[0021] This embodiment provides a material characterization mechanism for the formulation characteristics of finished chicken feed. Specifically, in the aforementioned same warehouse scenario, if only chicken feed is used as a uniform material type, it cannot reflect the real differences in storage and isolation between different formulations. Therefore, this embodiment explicitly limits the target material to finished chicken feed and concretizes the material feature vector into a formulation feature vector, which includes at least moisture content, fat ratio and drug additive codes. Specifically, moisture content reflects the feed's tendency to absorb moisture, become damp, and clump during storage; fat content reflects its sensitivity to high-temperature oxidation, oil seepage, and thermal adhesion; and drug additive codes reflect its isolation requirements in terms of regulatory compliance and cross-contamination control. These three factors together determine the appropriate storage area for a batch of feed within the warehouse and the materials that need to be isolated. Among these, drug additive codes are particularly crucial for generating a spatial repulsion matrix, because even if the content of certain additives in chicken feed is low, they may still pose a risk of contamination to adjacent drug-free or different drug formulations due to dust settling, residues on the packaging surface, or trace amounts released during handling. After querying the material mutual exclusion relation database based on the drug additive codes, the data processing center can identify the isolation requirements between the batch and various materials in the current inventory, and generate a spatial repulsion matrix accordingly. The material mutual exclusion relation database adopts a relational database structure, and its stored fields include at least: material number, formulation ingredient identifier, mutual exclusion additive code set, and mutual exclusion level standard. The mutual exclusion level standard is specifically divided into: Level 1 (strictly prohibited from adjacency, corresponding to a preset maximum repulsion force), Level 2 (requires buffered isolation, corresponding to a repulsion force that decreases with spatial distance), and Level 3 (normal storage, corresponding to a baseline repulsion force). Its query matching logic is as follows: extract the drug additive code of the target material in the current batch, and perform a cross-traversal comparison with the mutual exclusion additive code set corresponding to the current inventory distribution data in the material mutual exclusion relation database. If a match is found, the corresponding mutual exclusion level is returned as the objective calculation basis for the repulsion force setting. This database can pre-maintain relationships such as "adjacent," "required buffered isolation," and "prohibited adjacency," allowing the system to directly transform formulation management rules into path and storage location selection constraints. In a specific application scenario, suppose there are three types of inventory in the warehouse: drug-free chick feed, broiler feed containing a certain coccidiosis drug, and layer feed containing another additive. The formula feature vector of a batch of finished products to be put into the warehouse shows that its moisture content is high, its fat content is high, and it has a specific drug additive code. After reading the code, the system finds that it must be strictly isolated from drug-free feed, must maintain a buffer distance from the other type of drug-containing feed, and can be stored adjacent to feed with the same code. In the spatial repulsion matrix, several grids around the drug-free inventory are given a high repulsion level, while the grids around the same type of inventory can maintain a low repulsion level. In an anomaly handling mechanism, if a batch of formula feature vectors lacks drug additive codes, but the batch comes from a production line requiring key monitoring, the system can temporarily process it at a high isolation level to prevent missed detections. If there is no corresponding rule for a new additive code in the database, the code can be marked as pending maintenance, and manual approval or a conservative isolation strategy can be adopted before the rule is supplemented. If the moisture content and fat ratio are inconsistent with the production label, the final version returned by the quality inspection or manufacturing execution system can be used first to prevent path planning from being based on outdated formula data. In the aforementioned main scenario, a batch of broiler feed with high moisture content, high fat content, and drug additive codes is to be stored in the warehouse in the morning. The system recognizes that the finished product corresponding to this code cannot be stored close to drug-free chick feed. At the same time, this batch is not suitable for entering the upper hot zone due to its high fat content. Therefore, the avoidance level of the relevant area is increased in both the spatial repulsion matrix and the environmental suitability heat map. As a result, the final recommended target storage location is located near the ventilation corridor and far away from the drug-free storage area. The quality-sensitive and regulatory isolation attributes of finished chicken feed products are explicitly encoded into a computable data structure to enable differentiated warehousing planning for different formulations. In this embodiment, the environmental status slice is obtained by an array of wireless temperature and humidity sensors deployed on shelves in the warehouse area; The data acquisition module is used to periodically retrieve environmental status slices of each grid in the storage area at preset time intervals.

[0022] This embodiment provides an environmental status slice acquisition mechanism. Specifically, in the aforementioned three-dimensional warehouse scenario, if the environmental data only comes from one or two master tables in the center of the warehouse, it cannot reflect the microclimate differences at different heights, different ventilation corridors, and different shelf depths. Therefore, this embodiment limits the environmental status slice to be acquired by a wireless temperature and humidity sensor array deployed on the shelves in the storage area, and the data acquisition module periodically retrieves it at preset time intervals. Specifically, wireless temperature and humidity sensor arrays can be distributed across different shelf levels and depths. The sensor arrays are configured to collect and reconstruct the local microenvironmental state parameter matrix around the storage location nodes in real time. Chicken feed products, especially fresh batches that have just entered the warehouse, are often sensitive to changes in surrounding airflow and humidity because residual heat has not been fully released. Although the lower shelves near the ventilation openings and the upper shelves near the roof are in the same warehouse, their actual heat dissipation and humidity dissipation conditions may differ significantly. The system calculates and updates the time series parameters of the environmental state slices by periodically pulling data. The preset time interval can be determined based on the speed of change in the storage environment and the real-time requirements of scheduling. For finished chicken feed storage, environmental changes usually do not change drastically in seconds, so minute-level periodic collection can be used to balance communication load and data freshness. The significance of periodic retrieval is that, on the one hand, it provides a currently available environmental snapshot for new batches, and on the other hand, it provides a time-series basis for historical weight correction and abnormal trend judgment. In a specific application scenario, suppose sensors are deployed on the top, middle, and bottom layers of a row of shelves, and the system retrieves data at fixed intervals. If the temperature of the top layer remains high while that of the bottom layer remains relatively stable for several consecutive cycles, the system can still identify the top layer as an area prone to heat accumulation even if the temperature of the top layer drops briefly at a certain moment. If the humidity near the ventilation corridor on the west side remains lower than that of the enclosed corner on the east side, the performance of the grids on both sides in the suitability heat map will also be different over a long period of time. In one anomaly handling mechanism, if the remaining power of an individual sensor is lower than a preset power threshold or the signal strength is lower than a preset signal threshold, the data acquisition module can retain its most recent valid reading and mark it as having decreased reliability. If sensors are temporarily removed from certain shelves for maintenance, these areas can be automatically downgraded in the planning. If abrupt data that exceeds the common sense of the process occurs during the acquisition cycle, such as a sudden and large temperature jump, outlier filtering can be performed before writing it into the environmental status slice to avoid misleading the path planning due to a single erroneous reading. In the aforementioned main scenario, the system continuously pulls temperature and humidity data from the western ventilation corridor, the central ordinary area, and the northern upper hot zone in the afternoon. After several cycles, the northern upper zone consistently shows high temperature and slow humidity recovery, while the western ventilation corridor experiences a faster temperature drop. A batch of high-fat broiler feed that has just come off the production line and has high residual temperature is requested to be put into storage. The system then uses these periodically collected environmental status slices to prioritize directing them to the western area. This provides an environmental basis with spatial resolution and temporal continuity for route planning, thereby achieving a true reflection of the microclimate differences in the reservoir area; In this embodiment, the data processing center is also used to parse the drug additive code and determine whether the finished chicken feed contains a preset specific drug additive. The preset specific drug additive is a drug additive that is pre-configured in the high-risk list of cross-contamination. The data processing center is also used to set the repulsion force characteristic value of the storage location within a preset physical distance range around the finished chicken feed to a preset maximum repulsion force constant when the finished chicken feed contains a preset specific drug additive, so as to generate a spatial repulsion force matrix. The data processing center is also used to generate a spatial repulsion matrix based on a preset standard repulsion rule when the finished chicken feed does not contain any preset specific drug additives. The standard repulsion rule is that the magnitude of the repulsion force is inversely proportional to the spatial distance between storage locations.

[0023] This embodiment provides an enhanced isolation mechanism for specific drug additives. Specifically, in the aforementioned same warehouse scenario, relying solely on general formulation mutual exclusion rules may still be insufficient in certain drug additive scenarios under key supervision. This is because some specific drug additives have extremely low tolerance for cross-contamination, and once adjacent to drug-free or other sensitive formulations, even a small amount of dust residue may have serious consequences. Therefore, this embodiment further parses the drug additive codes and distinguishes between two cases: those containing preset specific drug additives and those not containing preset specific drug additives. Specifically, when the data processing center determines that the target chicken feed product contains a preset specific drug additive, it will set the storage location repulsion force characteristic value within a preset range around the product to a preset maximum repulsion force constant. The storage locations within this range are considered as prohibited areas by system default or restricted areas that can only be allocated when a manual authorization control instruction is received. This directly introduces compliance requirements and quality constraints into the path planning logic, avoiding the system from generating compromise plans that do not meet distance or environmental suitability thresholds. When the target chicken feed does not contain this specific drug additive, a spatial repulsion matrix is ​​generated according to the standard repulsion rule, that is, the repulsion force weakens as the spatial distance between storage locations increases; since the dust settling parameters and collision probability are negatively correlated with the spatial distance, the standard repulsion rule is configured as follows: the magnitude of the repulsion force is inversely proportional to the spatial distance between storage locations. In a specific application scenario, if target batch A contains a pre-set specific drug additive, then the nearest cells centered on the intended storage location of A are assigned extremely high rejection values. These cells will not be recommended to other sensitive materials by the system, even if the path is the shortest. If target batch B does not contain the specific additive and is only a general drug-containing or drug-free formula, then the rejection of its nearby cells is strong, while the rejection of distant cells is weak. The system can still flexibly utilize storage space resources while maintaining a safe distance. If the drug additive code parsing fails, the system can first process it in a conservative mode that includes the specific drug additive to avoid missing the isolation zone; if there is no free storage space within the preset range, it will not affect the current batch's own entry into the warehouse, but the system will simultaneously restrict other subsequent batches from entering the adjacent range; if the shortage of storage space resources causes the isolation range to be too large and affect the production line cycle, manual approval can be triggered by the upper management system, but the maximum value rejection remains unchanged in automatic mode. In the aforementioned main scenario, a batch of broiler feed with a specific coccidiosis drug code enters the warehouse in the afternoon. After the system analyzes the code, it determines that this batch belongs to the drug-containing finished products that need to be isolated. Therefore, the warehouse locations within a certain range of meters around its intended storage location are uniformly marked as high exclusion zones, so that drug-free chick feed waiting to be stored at the same time will not be assigned to adjacent locations. In contrast, another batch of ordinary layer chicken feed without the specific drug additive is handled according to the standard exclusion rules related to distance, and only the neighboring area is increased to avoid it. Establish stronger spatial isolation boundaries for drug formulations with high regulatory risks, thereby achieving compliance priority and cross-contamination prevention in warehousing planning; In this embodiment, the data processing center is also used to encapsulate the calculated coordinate sequence of the driving trajectory into control instructions in a target format; The target format is JSON or XML; the data processing center sends control commands to the scheduling and control module through the message queue telemetry transmission protocol.

[0024] This embodiment provides a mechanism for encapsulating and issuing instructions for path results. Specifically, in the aforementioned scenario of the same business line, if the data processing center cannot reliably convert the results into standardized instructions that the scheduling and control module can recognize after completing the calculation of the target storage location and driving trajectory, the front-end optimization results cannot be implemented and executed on the automated guided vehicle side. Therefore, this embodiment further specifies that the coordinate sequence of the driving trajectory is encapsulated into control instructions in the target format and issued through the message queue telemetry transmission protocol. Specifically, the coordinate sequence of the driving trajectory can include the starting position, turning nodes, target storage location nodes, and necessary information such as task identifiers, batch identifiers, and execution priorities. The significance of adopting a standardized target format is to maintain a clear data boundary between the data processing center and the scheduling control module, avoiding parsing ambiguity due to different system vendors or interface version changes. Both JSON and XML formats can carry hierarchical path information. For example, multiple nodes in the trajectory can be organized into an ordered set of coordinates, along with extended fields such as task type (inbound or outbound), target storage location number, and whether specific areas need to be avoided. Sending control commands via message queue telemetry transmission protocol is more suitable for the asynchronous, continuous and lightweight communication needs of warehouse site; its technical significance is that the data processing center does not need to directly occupy the control of the automated guided vehicle, but reliably delivers the planning results to the scheduling control module, and then the scheduling control module completes the vehicle-level execution management; this structure is conducive to queuing, retransmission, confirmation and status feedback of multi-vehicle concurrent tasks. In a specific application scenario, suppose the path calculated by the system consists of nodes P1, P2, P3, and P4, where P1 is the current position of the automated guided vehicle (AGV) and P4 is the target storage location. The data processing center can encapsulate this sequence into a control command according to the target format. In addition to the coordinate sequence, the command can also include a batch number and a restricted area identifier. After receiving the command, the scheduling and control module will drive the AGV to travel and stop from P1 to P4 in sequence. If the route needs to be replanned due to temporary channel occupancy, the data processing center can generate a new sequence and overwrite the original task. If the scheduling and control module fails to acknowledge receipt of the control command within the specified time, the data processing center can trigger a retransmission mechanism; if acknowledgment is received but the reported path is blocked during the execution of the automated guided vehicle, the scheduling and control module can request replanning; if the target format parsing fails, the relevant command will not enter the execution queue and a structural error alarm will be returned to the upper-level system; if the message queue is interrupted for a short time, the control command can be cached in the local sending queue and delivered again after the link is restored. In the aforementioned main scenario, the system calculates the trajectory for a batch of high-fat broiler feed to enter a target storage location along the west ventilation corridor. The trajectory consists of an entrance node, a corridor corner node, an elevator channel node, and a final storage location node. The data processing center encapsulates this coordinate sequence into a standard format instruction and sends it to the scheduling control module via a message queue telemetry transmission protocol. After the scheduling control module completes the parsing, it schedules the corresponding automated guided vehicle to pass through each node in sequence to complete the storage. When another batch needs to be released from storage during the night shift, the system uses the same encapsulation and distribution process to control the release path. The path planning results are implemented in a standardized, transmittable, and executable form in the field scheduling link, thereby achieving a stable connection between the data processing center and the automated guided vehicle execution end.

[0025] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A data processing system for planning the warehousing and outbound routes of finished chicken feed, characterized in that, The data processing system includes: The data acquisition module is connected to the warehouse management system and is used to acquire from the warehouse management system a gridded topology map of the warehouse area, a set of available storage locations, current inventory distribution data, and material feature vectors of the target materials of the current batch. It also collects environmental status slices of each grid in the warehouse area and acquires the current position of the automated guided vehicles in the warehouse area. The scheduling and control module is communicatively connected to the automated guided vehicle and is used to control its driving trajectory; The data processing center is communicatively connected to the data acquisition module and the scheduling control module, respectively; it is used to: query the pre-stored material mutual exclusion relationship database based on the material feature vector, and generate the spatial repulsion matrix of the target material on the current inventory by combining the current inventory distribution data; A regional suitability heatmap is generated based on the material feature vector, environmental state slices, and historical weight coefficients of each grid pre-stored in the data processing center; the physical distance cost is calculated based on the set of available storage locations and the current position of the automated guided vehicle. The spatial repulsion force matrix is ​​composed of the repulsion force characteristic values ​​corresponding to each grid, and the regional fitness heatmap is composed of the fitness characteristic values ​​corresponding to each grid. The repulsion force eigenvalues ​​of the spatial repulsion force matrix and the suitability eigenvalues ​​of the regional suitability heatmap are mapped to node cost values. Combined with the physical distance cost and the gridded topology map, a pathfinding algorithm is used to calculate the target storage location and driving trajectory with the minimum comprehensive cost. Control commands that include at least the driving trajectory are generated and sent to the scheduling control module to comprehensively manage the automated guided vehicle.

2. The data processing system for planning the warehousing and outbound routes of finished chicken feed as described in claim 1, characterized in that, The environmental state slice includes various environmental parameters; the material feature vector includes at least an environmental sensitivity feature vector containing the sensitivity weights corresponding to the various environmental parameters. The data processing center is also used to perform weighted summation calculations on each environmental parameter in the environmental state slice, the corresponding sensitivity weight in the environmental sensitivity feature vector, and the historical weight coefficient, to generate the regional suitability heatmap.

3. The data processing system for planning the warehousing and outbound routes of finished chicken feed as described in claim 1, characterized in that, The pathfinding algorithm employs a heuristic search algorithm; When the data processing center uses the pathfinding algorithm to calculate the target storage location and driving trajectory with the minimum overall cost, it specifically uses the physical distance cost as the heuristic function of the heuristic search algorithm, the node cost as the actual cost function of the heuristic search algorithm, calculates the overall cost between nodes, and performs iterative calculations to search for the target storage location and driving trajectory with the minimum overall cost.

4. The data processing system for planning the warehousing and outbound routes of finished chicken feed as described in claim 1, characterized in that, The data acquisition module is also used to acquire the actual quality inspection value of the target material when it leaves the warehouse; The data processing center pre-stores expected quality detection values; The data processing center is also used to calculate the absolute value of the difference between the actual quality detection value and the expected quality detection value to obtain the quality deviation value; The data processing center is also used to fine-tune the historical weight coefficient of the grid where the target storage location is located by using an error feedback-based heuristic weight update algorithm when the quality deviation value is greater than a preset deviation threshold, so as to use the historical weight coefficient to correct the next path planning. The data processing center is also used to keep the historical weighting coefficient unchanged when the quality deviation value is less than or equal to a preset deviation threshold.

5. The data processing system for planning the warehousing and outbound routes of finished chicken feed as described in claim 1, characterized in that, The data acquisition module is also used to monitor the data loss rate of the environmental state slice in real time. The data processing center is also used to trigger a degradation execution logic when the data loss rate is higher than a preset loss rate threshold. The data processing center is also used to maintain the normal planning mode when the data loss rate is lower than or equal to a preset loss rate threshold. The downgrade execution logic includes suspending the calculation process in the data processing center used to generate the regional suitability heatmap, and calculating the driving trajectory only based on the physical distance cost and the spatial repulsion matrix.

6. The data processing system for planning the warehousing and outbound routes of finished chicken feed as described in claim 1, characterized in that, The target material is finished chicken feed, and the material feature vector is the formula feature vector. The formulation feature vector includes at least the moisture content, fat ratio, and drug additive code; The data processing center is also used to query the material mutual exclusion relationship database based on the drug additive code and generate the spatial repulsion force matrix.

7. The data processing system for planning the warehousing and outbound routes of finished chicken feed as described in claim 1, characterized in that, The environmental status slices are collected by an array of wireless temperature and humidity sensors deployed on the shelves in the warehouse area; The data acquisition module is used to periodically retrieve environmental state slices of each grid in the storage area at preset time intervals.

8. The data processing system for planning the warehousing and outbound routes of finished chicken feed as described in claim 6, characterized in that, The data processing center is also used to parse the drug additive code and determine whether the finished chicken feed contains a preset specific drug additive, which is a drug additive pre-configured in a high-risk list of cross-contamination. The data processing center is also used to set the repulsion force characteristic value of the storage location within a preset physical distance range around the finished chicken feed product to a preset maximum repulsion force constant when the finished chicken feed product contains a preset specific drug additive, so as to generate the spatial repulsion force matrix. The data processing center is also used to generate the spatial repulsion matrix according to a preset standard repulsion rule when the finished chicken feed does not contain a preset specific drug additive. The standard repulsion rule is that the magnitude of the repulsion force is inversely proportional to the spatial distance between storage locations.

9. The data processing system for planning the warehousing and outbound routes of finished chicken feed as described in claim 1, characterized in that, The data processing center is also used to encapsulate the calculated coordinate sequence of the driving trajectory into control instructions in a target format; The target format is JSON or XML. The data processing center sends the control commands to the scheduling control module via a message queue telemetry transmission protocol.

Images