Intelligent warehousing automatic control optimization method and system based on dynamic environment perception

Abstract

The invention discloses an intelligent storage automatic control optimization method and system based on dynamic environment perception, and relates to the technical field of intelligent storage control. According to the method, a specific scene type of a specified warehouse is firstly obtained, a corresponding dynamic perception allocation scheme is determined through scene matching, and then a collaborative operation association object is determined, so that accurate adaptation of dynamic environment perception and a warehouse scene is realized; then, collecting dynamic environment sensing data of a storage area where a collaborative operation association object is located, and after collaborative association processing is carried out, generating a structured fusion data set adapted to different time sequence scenes; dynamic storage parameter calibration is carried out, an initial control strategy is generated, and effectiveness evaluation is completed; whether the path planning scheme and the operation execution sequence are dynamically adjusted or not is decided according to the evaluation result, dynamic adaptation and accurate regulation and control of intelligent storage automatic control are achieved, and the adaptability of storage control, the smoothness of process operation and the stability of operation execution in a complex dynamic scene are effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent warehouse control, and particularly relates to an intelligent warehouse automation control optimization method and system based on dynamic environment perception. BACKGROUND

[0002] At present, flexible production is popular in manufacturing industry, and intelligent warehouse has become the core hub of efficient operation of supply chain. Its core links run through the whole process of warehouse operation, covering four key modules of goods storage, equipment scheduling, operation execution and state monitoring. The goods storage link needs to adapt to the dynamic storage demand of diversified goods specifications, and high-density storage is realized through high-level shelves and flexible storage location layout. The equipment scheduling link involves multi-device parallel operation of AGV (Automated Guided Vehicle), stacker, conveyor and other devices, and needs to coordinate the running track and task allocation of each automatic device. The operation execution link focuses on core operations such as warehouse in and out, handling and stacking, and requires efficient completion of circulation according to order priority. The state monitoring link captures dynamic information such as device state, goods position and environmental parameters through multi-source perception, providing data support for whole process optimization.

[0003] However, these core links face dynamic challenges in actual operation, which highlights the necessity of implementing fine control of intelligent warehouse: in the goods storage link, the diversification of goods specifications leads to frequent changes in storage demand, which requires dynamic adjustment of storage location allocation and shelf layout through control strategy; in the equipment scheduling link, multi-device parallel operation easily causes path congestion, which requires real-time control to optimize path planning and task coordination; in the operation execution link, high-frequency orders and flexible production require operation process to quickly adapt, which requires dynamic control to adjust execution sequence and operation rhythm; in the state monitoring link, sudden device failure or personnel intervention easily breaks the preset process, which requires closed-loop control to correct deviations in time and maintain efficient operation of the warehouse system. The accurate landing of control cannot be achieved without comprehensive capture and deep analysis of the core data of each link.

[0004] The popularity of perception technologies such as IoT (Internet of Things) and laser radar has realized the comprehensive collection of multi-source dynamic data of environment, equipment and goods in the warehouse scene, which includes warehouse parameters throughout the whole process of warehouse operation, such as environmental parameters, goods parameters and operation parameters, providing accurate data basis for warehouse automation control. The breakthrough development of AI (Artificial Intelligence) and intelligent scheduling algorithm provides core support for the analysis of massive perception parameters and the dynamic optimization of control strategy, and converts scattered warehouse parameters into controllable instructions, promoting the research and development of data-driven warehouse automation control technology, and making it possible for warehouse system to transform from fixed execution to real-time parameter-based self-adaptation.

[0005] Taking Chinese invention patent CN119690020B as an example, it discloses a control method and system for automated handling equipment on a dedicated transportation line, including: constructing a local environment model based on microgrids to collect and transmit equipment status information; allocating tasks and planning paths for automated equipment according to the microgrid environment model; dynamically optimizing the task allocation and path planning of the global scope according to the task allocation and path planning within each microgrid; monitoring the execution status of the optimized task allocation and path planning in real time, dynamically adjusting the execution order of tasks and path planning through intelligent scheduling algorithms, and making flexible adaptive adjustments according to resource status and environmental changes.

[0006] Chinese invention patent application CN120610477A discloses an intelligent warehousing industrial IoT system and control method, including: acquiring real-time information of goods in the warehouse and real-time data of the warehouse environment through a sensing module; transmitting the raw data collected by the sensing module to an information processing module and performing data preprocessing to obtain usable data; generating warehousing operation strategies using artificial intelligence algorithms in a control decision module based on the data processed by the information processing module; executing operations such as goods storage, retrieval, and transportation through an automated operation module; feeding back the operation results to the information processing module and displaying relevant status information through a user interface module, and performing abnormal warnings and alarm processing.

[0007] The technical logic presented by the two patents clearly demonstrates the core path of environmental perception empowering warehouse control: using multi-dimensional perception data such as warehouse environment, equipment status, and cargo information as the basis for control decisions, and using algorithm models to complete data analysis and dynamic optimization of control strategies (task allocation, path planning, and job sequencing), ultimately achieving precise adaptation of warehouse control behavior to real-time scenarios. Its core idea is highly consistent with the intelligent warehousing industry's demand for flexible and efficient control.

[0008] Currently, warehouse control technology uses microgrid modeling and multi-sensor fusion to accurately collect core control data such as warehouse space coordinates, equipment load rate, storage location occupancy status, and order urgency. Leveraging intelligent scheduling algorithms and artificial intelligence (AI) decision-making models, the collected data is transformed into executable control commands, enabling optimized task allocation, dynamic path planning correction, and precise control of operational processes. Simultaneously, a real-time monitoring and feedback mechanism is introduced to track control effectiveness indicators such as the operating status of automated equipment, work execution progress, and load balancing compliance, allowing for reverse adjustments to control strategies. This effectively improves the operational efficiency of automated warehouse equipment and the environmental adaptability of control commands, providing solid control technology support for the flexible operation of intelligent warehousing. However, while existing technologies have achieved the integration of environmental perception and control, they are insufficient to meet the actual operational needs of intelligent warehousing in complex and dynamic scenarios, such as high-density storage, high-frequency operations, and multi-variable disturbances. On the one hand, existing technologies often process single-dimensional data such as environment, automated equipment, and goods independently, without fully considering the cross-influence of multiple factors. In the storage location management stage, when shelves are temporarily adjusted or storage location attributes are changed (such as adjusting the load-bearing limit), if the storage priority of goods, the real-time load of surrounding automated equipment, and the operation progress are not synchronously linked, and storage location allocation is still performed according to the original single-dimensional parameters, it is easy to cause problems such as high-priority goods being assigned to remote storage locations and automated equipment concentrating on going back and forth to the same area, leading to congestion and an increased probability of interruption of inbound and outbound operations. This lack of data coordination further transmits to the multi-equipment coordination stage, so when equipment such as conveyors suddenly fails, stacker cranes may still deliver goods to the faulty conveyors as originally planned, resulting in the accumulation of goods, while other automated equipment becomes idle due to the lack of adapted task allocation.

[0009] On the other hand, existing algorithms lack dynamic adaptive adjustment mechanisms, and most core parameters are preset offline. When faced with dynamic fluctuations in warehousing scenarios, they cannot calibrate control strategies in real time, resulting in control commands lagging behind actual operational changes. In the equipment scheduling and control stage, when a stacker crane in a certain aisle suddenly fails and cannot operate, the global task redistribution algorithm fails to quickly adapt to real-time parameters such as the rated load, operating radius, and current task saturation of the remaining automated equipment. It still allocates high-rise rack storage and retrieval tasks according to the original preset task allocation ratio, causing some automated equipment to operate at full load for a long time, exacerbating mechanical wear and tear. At the same time, there is a delay in the storage and retrieval of goods in high-rise warehouses, which in turn causes the goods transfer process in the dynamic adjustment stage of warehouse locations to conflict with other operation tasks, extending the adjustment cycle.

[0010] The aforementioned problems, such as incomplete data collaboration, weak algorithm adaptability, and delayed command response, lead to a decline in the adaptability and control effectiveness of intelligent warehouse automation in complex and dynamic scenarios. This not only fails to meet the needs of optimizing storage space utilization under high-density storage, ensuring smooth processes under high-frequency operations, and maintaining operational stability under multivariate disturbances, but also easily triggers warehouse operation risks such as overload damage to automated equipment and increased frequency of operation interruptions. Ultimately, this restricts the intelligent upgrading of warehousing towards higher efficiency, higher precision, and lower cost. Summary of the Invention

[0011] To address the technical problem of decreased adaptability and control effectiveness of existing intelligent warehouse automation control in complex dynamic scenarios, this invention provides an optimization method and system for intelligent warehouse automation control based on dynamic environment perception. The technical solution is as follows: On the one hand, a method for optimizing intelligent warehouse automation control based on dynamic environmental perception is provided. This method includes: S1, obtaining the warehouse scenario type within a specified warehouse, matching the corresponding dynamic perception allocation scheme, and simultaneously recording the automated equipment, sensing terminals, and goods in the corresponding warehouse area within the specified warehouse as collaborative operation associated objects; S2, collecting dynamic environmental perception data of the warehouse area where the collaborative operation associated objects are located according to the dynamic perception allocation scheme, and performing collaborative association processing to generate a structured fusion dataset; S3, performing dynamic warehouse parameter calibration based on the generated structured fusion dataset, generating an initial control strategy, and judging the effectiveness of the initial control strategy based on the execution process of the initial control strategy.

[0012] On the other hand, an intelligent warehouse automation control optimization system based on dynamic environmental perception is provided. This system includes: a warehouse scene dynamic perception module, used to acquire the warehouse scene type within a specified warehouse, match the corresponding dynamic perception allocation scheme, and simultaneously record the automated equipment, sensing terminals, and goods in the corresponding warehouse area within the specified warehouse as collaborative operation associated objects; a collaborative association processing module, used to collect dynamic environmental perception data of the warehouse area where the collaborative operation associated objects are located according to the dynamic perception allocation scheme, and perform collaborative association processing to generate a structured fusion dataset; and a warehouse parameter calibration and validity judgment module, used to perform dynamic warehouse parameter calibration based on the generated structured fusion dataset, generate an initial control strategy, and judge the validity of the initial control strategy based on the execution process of the initial control strategy.

[0013] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: 1. This invention, by acquiring the scenario type of a specified warehouse and matching it with a corresponding dynamic sensing allocation scheme, clarifies the collaborative operation relationships among automated equipment, sensing terminals, goods, orders, and other related objects. This avoids problems such as unreasonable allocation of storage locations and equipment congestion caused by the failure to consider the cross-influence of multiple factors, laying the foundation for multi-dimensional data collaboration. Based on this scheme, dynamic environmental sensing data of the warehouse area is collected and processed collaboratively to generate a structured fusion dataset. This solves the deficiency of incomplete data collaboration in existing technologies, effectively preventing goods backlog and resource idleness during equipment failures, and providing accurate data support for control strategy formulation. Based on this dataset, dynamic warehouse parameters are calibrated, an initial control strategy is generated, and the effectiveness of strategy execution is used to determine whether to adjust path planning and operation sequence. This compensates for the shortcomings of existing algorithms, such as weak adaptive capabilities and delayed command response, enabling the response to dynamic fluctuations in the warehouse scenario and avoiding problems such as equipment overload damage and operation conflicts. Through the above steps, the adaptability and control effectiveness of intelligent warehouse automation control in complex dynamic scenarios are improved, meeting the operational needs of high-density storage, high-frequency operations, and multi-variable disturbances, and helping intelligent warehousing upgrade towards higher efficiency, higher precision, and lower cost.

[0014] 2. First, according to a preset statistical period, data representing cross-operation conflicts, such as frequency, are collected and summarized into a scenario judgment dataset. Characteristic indicators, such as the collaborative adaptation coefficient, are calculated through quantitative evaluation. A dynamic scenario index is obtained through ratio calculation and geometric mean. Then, based on scenario classification thresholds, basic, advanced, or enhanced dynamic sensing allocation schemes are matched to achieve precise scenario-based adaptation of sensing schemes, avoiding the problem that a single sensing mode cannot cope with scenarios of varying complexity. Simultaneously, within a preset monitoring period, sensitivity and real-time impact factors are obtained through methods such as temporal difference and fast Fourier transform. After normalization, a comprehensive value of the sensing response impact is obtained through geometric mean, which is used to verify the suitability of the scheme. This design not only achieves reasonable allocation of sensing resources through the dynamic scenario index, improving the targeting and efficiency of data collection, but also uses the comprehensive adaptation value to monitor the scheme matching effect in real time, promptly identifying adaptation anomalies, ensuring the sensitivity and real-time nature of sensing data, and effectively solving the problem of insufficient sensing adaptation in complex scenarios.

[0015] 3. By comparing the real-time and reference occupancy status of storage locations to obtain the storage location occupancy matching degree, and combining this with the accuracy of cargo displacement early warning, a geometric average is calculated. This average is then multiplied by a scenario-based dynamic collaborative correction coefficient based on operational intensity and process stability coefficients to accurately quantify the degree of collaborative adaptation of the sensing terminal in storage location management and cargo monitoring. If the warehousing collaboration index exceeds the corresponding preset value, combined with abnormal environmental parameter feedback, environmental parameters are adjusted first, or data re-collection is directly triggered to ensure data collection quality. If the warehousing collaboration index does not exceed the corresponding preset value, multi-dimensional dynamic environmental sensing data is standardized and controlled to unify the data format. This process not only accurately judges the sensing collaboration effect through the warehousing collaboration index and addresses data deviation issues in a targeted manner, but also ensures data integrity and consistency through environmental parameter linkage control and data standardization processing, avoiding the impact of insufficient data quality on subsequent control strategy formulation. Simultaneously, it adapts to dynamic changes in the scenario, improving the accuracy and reliability of data collaboration.

[0016] 4. Warehouse data standardization control is implemented in two scenarios: time-series consistency and inconsistency. For time-series consistency, the coordination deviation value of adjacent sensing terminals with the same warehousing parameters is first calculated. If the coordination deviation value is not greater than the corresponding allowable value, lightweight processing is performed to generate a structured fusion dataset. If the coordination deviation value exceeds the corresponding allowable value, the cumulative impact of the time dimension is quantified through deviation accumulation integration, and moving average or Kalman filtering is used to correct the data as needed to ensure data accuracy. For time-series inconsistency, the time-series consistency deviation value is first calculated, and a time-series calibration priority score is obtained by mapping the operation scenario type. Single-dimensional data with a time-series calibration priority score exceeding the reference value is processed in real-time stream, and synchronous calibration is achieved by adjusting the timestamp and correction coefficient step by step. Non-single-dimensional data is processed in batches, and batch calibration is completed by dynamically adjusting the calibration window and integration weight. Data that does not exceed the reference value is lightweight processed to generate a structured fusion dataset. The entire process specifically addresses the time-series deviation and coordination inconsistency issues of multi-sensor terminal data. Through differentiated processing strategies, it ensures data standardization efficiency while improving data time-series synchronization and accuracy, and avoids interference from heterogeneous data. Meanwhile, based on deviation quantification and priority classification, it accurately adapts to different data quality scenarios, ensuring the integrity and reliability of the structured fusion dataset, and effectively improving the accuracy and stability of intelligent warehouse automation control.

[0017] 5. First, based on the structured fusion dataset, the cargo storage density and cargo flow rate are obtained. Combined with storage priority and space utilization limits, the target values ​​corresponding to cargo storage density and cargo flow rate are determined. Then, partial differential equations are used to solve for the storage density adjustment gradient to correct the storage location allocation strategy. A load balancing objective function is constructed based on the Lagrange multiplier method to solve for the optimal flow rate adjustment, optimizing the inbound and outbound scheduling rules. After verification, an initial control strategy is generated. The effectiveness of the initial control strategy is judged by evaluating two indicators: the percentage of rate balancing compliance and the percentage of operational efficiency improvement. If neither indicator exceeds its corresponding preset value, a balance coefficient is obtained through harmonic averaging. Storage location priority areas are then divided, and the operational sequence is determined by multiplying the result with the order urgency weight. Finally, the path deviation is corrected using the Jacobian matrix, and the strategy is re-verified. This process of dual-indicator evaluation ensures the effectiveness of the strategy. Priority division and path deviation correction improve operational rationality and smoothness, effectively balancing storage location utilization and operational efficiency, avoiding uneven equipment load and operational conflicts, and improving the accuracy of intelligent warehousing control and operational stability in dynamic scenarios. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating the intelligent warehouse automation control optimization method based on dynamic environment perception provided in this embodiment of the invention; Figure 2 A flowchart illustrating the matching process of the dynamic sensing allocation scheme provided in this embodiment of the invention; Figure 3 This is a flowchart illustrating the standardized control process for warehouse data provided in an embodiment of the present invention. Figure 4 A logic diagram for constructing the Lagrange function provided in an embodiment of the present invention; Figure 5 This is a schematic diagram illustrating the working principle of the Lagrange function provided in an embodiment of the present invention. Figure 6 This is a schematic diagram of the structure of an intelligent warehouse automation control optimization system based on dynamic environment perception, provided in an embodiment of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0021] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0022] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0023] This invention provides an intelligent warehouse automation control optimization method based on dynamic environment perception, such as... Figure 1 The flowchart shown is for an intelligent warehouse automation control optimization method based on dynamic environment perception. The processing flow of this method may include the following steps: S1: Obtain the storage scenario type within the specified warehouse, match the corresponding dynamic perception allocation scheme, and simultaneously record the automated equipment, sensing terminals, and goods in the corresponding storage area within the specified warehouse as collaborative operation associated objects; the storage scenario type includes basic dynamic scenario, regular dynamic scenario, and dense dynamic scenario; automated equipment includes AGVs, stacker cranes, sorting robots, conveyor belts, and other storage equipment with autonomous operation or linkage control capabilities; sensing terminals include displacement sensors, pressure sensors, LiDAR, temperature and humidity sensors, vision cameras, and other devices used to collect data on the storage environment and operational status.

[0024] S2 collects dynamic environmental perception data of the warehouse area where the collaborative operation related objects are located according to the dynamic perception allocation scheme, and performs collaborative association processing to generate a structured fusion dataset. The dynamic environmental perception data includes, but is not limited to, the occupancy status of the storage location, the weight / volume of the goods, the traffic flow of the passage, the operating posture of the equipment (position, speed, load), the ambient temperature and humidity, and the light intensity. The structured fusion dataset represents a multi-dimensional, time-series collection of warehouse operation data integrated in a unified data format (standardized field naming, unified units, and complete metadata), which can directly support subsequent parameter calibration and strategy generation.

[0025] S3 performs dynamic warehouse parameter calibration based on the generated structured fusion dataset, generates an initial control strategy, and judges the effectiveness of the initial control strategy based on the execution process of the initial control strategy to determine whether to dynamically adjust the path planning scheme and the order of operation execution. The initial control strategy represents a comprehensive control scheme based on real-time warehouse data and scenario requirements, which includes storage location allocation rules, equipment scheduling logic, and operation priority ranking, and is used to guide the collaborative operation of warehouse automation equipment.

[0026] In a flexible production workshop for automotive parts, it is necessary to address the issues of low changeover efficiency and material mismatch caused by lagging equipment coordination and inaccurate data perception when multiple product models are produced on mixed lines. The first step is to execute S1: When the workshop is identified as a densely populated dynamic scenario (frequent changeovers, multiple flexible production lines operating in parallel), a dynamic perception allocation scheme with accurate perception and real-time linkage is matched. AGVs, flexible robotic arms, and intelligent tooling fixtures responsible for material handling within the workshop are defined as automated equipment. Displacement sensors, pressure sensors, vision cameras for product identification, and various types of components are integrated, forming collaborative operation associations according to the operational linkage logic.

[0027] Next, execute S2: According to the dynamic perception allocation scheme (high-frequency collection, prioritizing the perception of product models, equipment positioning accuracy, and material inventory status), collect dynamic environmental perception data such as component specifications, AGV transfer trajectory, robotic arm operation posture, and workstation material occupancy. After time-series alignment, elimination of invalid data during the production changeover period, and other collaborative processing, generate a structured fusion dataset with a unified format and complete dimensions to achieve accurate adaptation of product, equipment, and material data.

[0028] Finally, S3 is executed: based on the dataset, the material delivery rate (fitting the production line cycle time) and equipment coordination response latency (meeting real-time linkage requirements) are calibrated to generate an initial control strategy for product model, equipment parameters, and material delivery linkage. The effectiveness of the strategy is determined by continuously monitoring the changeover process and combining the percentage of rate balancing compliance with the percentage of improved operational efficiency. This solution solves the challenges of data adaptation and equipment coordination in mixed-line production, improves the adaptability and control effectiveness of intelligent warehouse automation in complex dynamic scenarios, and thus effectively improves production process stability and material matching accuracy.

[0029] like Figure 2 The flowchart shown illustrates the matching process for the dynamic sensing allocation scheme. It involves obtaining the warehouse scenario type within a specified warehouse and matching it to the corresponding dynamic sensing allocation scheme. Specifically, this includes: collecting operational representation data corresponding to the warehouse scenario type within a preset statistical period (set according to warehouse operation turnover characteristics, typically one shift or 24 hours), and summarizing this data into a scenario judgment dataset. The operational representation data includes the frequency of cross-operation conflicts, the total frequency of inbound and outbound operations per unit time, and the task handover delay duration. The frequency of cross-operation conflicts indicates that different operational processes (such as inbound, outbound, sorting, and transfer) occur in the same warehouse area. The cumulative number of times that overlap paths, conflicting equipment usage, and compete for resources occur in a domain or operational channel directly reflects the operational density and collaborative complexity within the area; the total frequency of inbound and outbound operations per unit time represents the total number of goods inbound and outbound operations completed in the corresponding storage area within a specified warehouse, converted into an average value per unit time, reflecting the activity level of goods turnover and operational load intensity in that area; the task handover delay time represents the cumulative value of waiting time between adjacent operational links (such as goods from sorting completion to AGV transfer and reception, from inbound acceptance to storage location allocation), reflecting the smoothness of operational process connection and collaborative response efficiency.

[0030] The system calculates feature indicators for the corresponding job representation data in the scenario determination dataset. These feature indicators include the collaboration adaptation coefficient, job intensity coefficient, and process stability coefficient. The calculated feature indicators are then compared with corresponding preset benchmark feature indicators (based on the operational standards of similar scenarios in the warehousing industry), and the geometric mean is taken to smooth short-term data fluctuations, avoid interference from abnormal single indicators, and accurately reflect the overall complexity of the scenario. This yields a dynamic scenario index for quantifying the dynamic complexity of warehousing scenarios. A dynamic perception allocation scheme is then matched based on the obtained dynamic scenario index. Specifically: The system retrieves the scene classification thresholds corresponding to the dynamic scene index. These thresholds include a basic scene threshold and an enhanced scene threshold, with the basic scene threshold being lower than the enhanced scene threshold. Both the basic and enhanced scene thresholds are set based on the statistical distribution of dynamic scene indices for similar warehousing scenarios in the industry, combined with the minimum and maximum values ​​of historical scene thresholds from historical dynamic scene analysis. If the dynamic scene index is not greater than the basic scene threshold, the current warehousing scenario is determined to be a basic dynamic scene, requiring a matching basic sensing allocation scheme. For example, a low-frequency acquisition mode (e.g., 5Hz) is adopted, focusing on core basic dimensions such as the storage status of goods and whether equipment is operating normally. This simplifies data transmission and processing, enabling only necessary sensing terminals to meet basic operational needs with low resource consumption, adapting to the characteristics of low-turnover, low-interaction scenarios.

[0031] If the dynamic scenario index falls between the basic scenario threshold and the enhanced scenario threshold, the current warehousing scenario is determined to be a regular dynamic scenario, requiring an advanced perception allocation scheme. For example, a medium-frequency acquisition mode (e.g., 10Hz) can be used to comprehensively perceive multi-dimensional data such as the status of goods entering and leaving the warehouse, equipment operating parameters, and passageway conditions. Data compression and preliminary screening mechanisms can be enabled to balance perception accuracy and resource consumption. Terminal collaborative acquisition can achieve smooth adaptation to regular operation processes and meet the needs of medium turnover and multi-device collaboration. If the dynamic scenario index is greater than the enhanced scenario threshold, the current warehousing scenario is determined to be a dense dynamic scenario, requiring an enhanced perception allocation scheme. For example, a high-frequency acquisition mode (e.g., 20Hz) can be used to collect detailed data such as goods turnover details, equipment linkage posture, and operation conflict warnings in real time. Edge computing preprocessing and priority transmission mechanisms can be enabled to strengthen data temporal consistency and real-time responsiveness, adapting to complex scenarios of high-frequency turnover and dense equipment interaction, and ensuring efficient and stable operations.

[0032] During the preset monitoring period (set according to the minimum cycle of warehousing operations, usually one shift) in the dynamic sensing allocation scheme matching process, the time-series data of the delay response of each sensing terminal to the operation characterization data in the current warehousing scenario are acquired, and the delay change rate sequence is obtained by performing time-domain difference operation. The time-domain standard deviation of this sequence is used as the sensitivity influence factor S, and its calculation formula is as follows: In the formula, r i This represents the i-th delayed time series data in the delay rate of change sequence. denoted as the mean of the delayed rate of change sequence, n represents the data length of the delayed rate of change sequence, and i represents the number of delayed time series data.

[0033] After acquiring the jitter time-series data of the operation characterization data after sensing response during transmission, a fast Fourier transform is performed to obtain the frequency domain power spectrum. The energy value corresponding to the main peak frequency is extracted as the real-time impact factor Y, and its calculation formula is as follows: In the formula, fpeak This represents the main peak frequency of the power spectrum in the frequency domain. This represents the energy value corresponding to the main peak frequency.

[0034] After normalizing the sensitivity and real-time impact factors, a geometric mean is used to avoid excessive interference from single factor numerical anomalies on the comprehensive result. This also strengthens the synergistic effect of the two factors, smoothing out biases caused by data fluctuations. The resulting comprehensive value quantifies the dynamic sensing allocation scheme matching process and the degree of adaptation to the dynamic scene. This comprehensive value is denoted as the comprehensive value of the sensing response impact, Q, and its calculation formula is as follows: In the formula, S norm Y represents the normalized result of the sensitivity influence factor. norm This represents the normalized result of the real-time impact factor.

[0035] If the overall impact of the perceived response exceeds the preset threshold, a scenario adaptation anomaly verification prompt is sent to remind designated personnel to promptly check the parameter configuration of the current dynamic sensing allocation scheme (such as acquisition frequency, data transmission protocol, etc.) and the operating status of the sensing terminals. The preset threshold is based on the minimum requirements for sensing response sensitivity and real-time performance in warehousing operations, combined with industry-standard scenario adaptation compliance benchmarks and statistical analysis of the warehouse's own historical sensing response adaptation data, ensuring the accuracy of anomaly detection. If the overall impact of the perceived response is not greater than the preset threshold, the corresponding dynamic sensing allocation scheme remains unchanged, and collaborative correlation processing is performed, specifically including: The system aggregates real-time occupancy status data of storage locations collected by sensing terminals with reference storage location occupancy status data in the preset storage location planning scheme. Through location-by-location and time-node-by-time comparisons, it counts the number of times the two states completely match within a preset statistical period (i.e., the frequency of real-time storage location occupancy status conforming to the requirements of the preset storage location planning scheme). This matching frequency is then compared to the total number of comparisons within the statistical period (total number of storage locations × number of status samplings within the statistical period). The final percentage of times the two storage location occupancy states match within the preset statistical period is recorded as the storage location occupancy matching degree. The system also calculates the number of times the cargo displacement anomaly monitoring signals captured by the sensing terminals overlap with historical anomaly monitoring signals in the historical feature database, and compares this with the total number of cargo displacement anomaly signals emitted by the sensing terminals within the preset statistical period. The accuracy of cargo displacement warnings is obtained by calculating the ratio of the number of warnings. If there is no historical abnormal monitoring signal corresponding to the cargo displacement abnormal monitoring signal in the historical feature database, manual judgment is required to further confirm the accuracy of cargo displacement warnings. The obtained storage location occupancy matching degree and cargo displacement warning accuracy are geometrically averaged and then combined with the scenario dynamic collaborative correction coefficient. That is, the geometric average is first performed and then multiplied by the scenario dynamic collaborative correction coefficient. This can correlate the intensity of operations and the stability of processes, dynamically adapt to the needs of different warehousing scenarios, ensure the objectivity of core indicator evaluation, and improve the adaptability and quantitative accuracy of actual scenarios. This results in a warehousing collaboration index, which is used to quantify the degree of collaborative adaptation of sensing terminals in storage location management and cargo safety monitoring.

[0036] The scenario dynamic collaborative correction coefficient is obtained by summing the operation intensity coefficient and the process stability coefficient after assigning preset weights to each. The preset weights are set based on the attention paid to operation intensity and process stability in the warehousing scenario. For example, the weight corresponding to the operation intensity coefficient can be increased in dense dynamic scenarios, such as from 0.5 to 0.6. In normal scenarios, the weights can be balanced, that is, the weights corresponding to the operation intensity coefficient and the weights corresponding to the process stability coefficient are 0.5 respectively.

[0037] When the warehouse collaboration index is greater than the preset warehouse collaboration index, the collected dynamic environmental perception data is subjected to warehouse data standardization control. The dynamic environmental perception data includes multi-dimensional data reflecting the environmental status of the warehouse area, the operating status of collaborative operation related objects, and the dynamic changes of the operation scenario. The preset warehouse collaboration index is represented by the sum and average of the historical warehouse collaboration indices in the historical warehouse collaboration process. When the warehouse collaboration index is not greater than the preset warehouse collaboration index, if abnormal feedback of environmental parameters of the warehouse area is received, environmental parameter control is performed and a prompt for re-collection of the corresponding dimension data is sent. If no abnormal feedback of environmental parameters of the warehouse area is received, the re-collection of the corresponding dimension data is directly triggered.

[0038] Among them, the abnormal environmental parameter feedback indicates that the environmental parameters in the storage area do not meet the expected allowable conditions for the safe operation of the corresponding automated equipment, and the corresponding abnormal operation duration reaches the preset judgment period (set according to the safety tolerance characteristics of the equipment, usually 3-10 minutes, which can be dynamically configured according to the equipment type). Environmental parameter control includes: based on the specific environmental parameter type in the abnormal environmental parameter feedback, sending automated equipment adjustment prompts to the equipment control terminal of the corresponding storage area, such as pausing non-critical operation processes, adjusting the equipment operation posture to avoid sensitive areas, etc., until receiving feedback on the recovery of abnormal environmental parameters in the storage area to stop the control and trigger the re-collection of data in the corresponding dimension.

[0039] Specifically, the collaboration adaptation coefficient is obtained by combining the ratio of the frequency of cross-operation conflicts to the total number of collaborative operations within a cycle, along with the response synchronization rate correction between automated equipment and sensing terminals. That is, the collaboration adaptation coefficient = 1 - (frequency of cross-operation conflicts / total number of collaborative operations) × (1 - response synchronization rate), where the response synchronization rate = (number of collaborative operations with synchronized responses / total number of collaborative operations). This quantifies the accuracy of the collaboration matching between automated equipment, sensing terminals, and work tasks. The work intensity coefficient is calculated by combining the ratio of the actual frequency of inbound and outbound operations per unit time to the maximum capacity of the warehouse with the required frequency of operations, along with the superimposed order batch complexity coefficient. The process stability coefficient is calculated based on the statistical characteristics of task handover delay time, combined with the proportion of times the task handover delay time exceeds the corresponding allowable delay time. That is, the process stability coefficient = 1 - [(task handover delay time / total task handover time within the preset statistical period) × (handover frequency of task handover delay time exceeding the corresponding allowable delay time / total handover frequency within the preset statistical period)].

[0040] In this embodiment, dynamic scene indices are obtained through multi-dimensional operational characterization data. These indices, combined with corresponding scene thresholds, enable precise classification of warehousing scenarios and match differentiated perception allocation schemes. This satisfies both the low resource consumption requirements of basic scenarios and ensures high-frequency perception and real-time response in intensive scenarios. Sensitivity and real-time impact factors are used to quantify the adaptability of perception schemes, establishing an anomaly verification mechanism to optimize perception parameters and terminal status in a timely manner. Furthermore, the warehousing collaboration index is linked to core indicators of storage location management and security monitoring, combined with dynamic scene correction coefficients to improve the reliability and adaptability of perception data. Simultaneously, data quality is ensured through data standardization and re-collection strategies, providing precise support for warehousing operation scheduling and path planning, reducing collaboration conflicts and process delays, and achieving a dual improvement in operational efficiency and safety stability.

[0041] Due to differences in the deployment location, hardware performance, and transmission links of sensing terminals in warehousing scenarios, the collected dynamic environmental sensing data may exhibit timing synchronization deviations or consistency fluctuations. This directly affects the effectiveness of data integration and the accuracy of subsequent decisions. Therefore, for dynamic environmental sensing data with different timing states, warehousing data standardization control needs to adopt differentiated processing methods to ensure that data quality adapts to warehousing operation requirements, such as... Figure 3 As shown, a standardized control flowchart for warehouse data is provided for this invention.

[0042] Example 1: When the acquisition time sequence of dynamic environmental sensing data collected by different sensing terminals is consistent, the specific process for warehouse data standardization control is as follows: Within a preset acquisition period (set according to the real-time requirements of warehouse data and the response characteristics of sensing terminals, usually 1-5 minutes, which can be dynamically adjusted according to the scenario), the temporal feature representation of the same warehouse parameter collected by each sensing terminal is obtained, and the collaborative deviation value between the corresponding temporal feature representations of adjacent sensing terminals is calculated. That is: based on the value of the same warehouse parameter collected by each sensing terminal at the same timestamp, the deviation of adjacent terminal data is calculated using the Euclidean distance formula, the formula is as follows: In the formula, This represents the coordination deviation value between the a-th sensing terminal and the b-th sensing terminal. and Let a and b represent the parameter values ​​of the two sensing terminals at the k-th timestamp, respectively. m represents the total number of timestamps within the preset collection period, and k represents the number of timestamps within the preset collection period. If the coordination deviation value is not greater than the preset allowable coordination deviation, then lightweight standardization processing (only basic format conversion and obvious outlier removal) is performed on the dynamic environment sensing data to generate a structured fusion dataset. Dynamic warehouse parameter calibration is then performed. The preset allowable coordination deviation is represented by the sum and average of the historical coordination deviation values ​​of each sensing terminal during the historical warehouse data control process. Otherwise, based on the quantization range of the coordination deviation value, the cumulative impact of the coordination deviation on the stability of the warehouse operation process over time is quantified, specifically: Using the time within a preset collection period as the integration variable and the coordination deviation values ​​of adjacent sensing terminals as the integrands, the cumulative integral value of the deviation between the start and end times of the preset statistical period is calculated. If the cumulative integral value of the deviation is not greater than the preset cumulative integral value of the deviation (set after statistical analysis based on the range of cumulative integral values ​​of deviations in similar historical scenarios and the minimum requirements for data accuracy in warehousing operations), it indicates that the cumulative impact of the coordination deviation in the time dimension is within an acceptable range and has not significantly interfered with the stability of the warehousing operation process. Data fluctuations are mainly small random fluctuations. In this case, the dynamic data collected is analyzed using a moving average algorithm. Environmental perception data is smoothed and corrected: A fixed-length sliding window is set, and the data sequence is traversed sequentially. The arithmetic mean of the data at each time point within the window is used as the correction value at the current time point to smooth small fluctuations and retain the core trend of data change. If the cumulative integral value of the deviation is greater than the preset cumulative integral value of the deviation, it indicates that the cumulative impact of the collaborative deviation has exceeded the operational tolerance threshold, which may lead to problems such as lag and collaborative conflicts in the warehousing operation process. There are obvious systematic errors and random noise in the data. At this time, the Kalman filter algorithm is used to denoise the collected dynamic environmental perception data to filter out random noise and systematic errors and output high-precision time series data.

[0043] In this embodiment, the temporal characteristics of the same storage parameters of each sensing terminal are acquired within a preset acquisition period. The collaborative deviation value of adjacent terminals is calculated, and the historical average collaborative deviation is used as the allowable threshold. When the deviation meets the standard, a dataset is generated and the parameters are calibrated after lightweight processing. If the deviation does not meet the standard, the cumulative impact of the deviation is quantified by time integration. According to the integration threshold, small fluctuations are smoothed by moving average or noise error is filtered out by Kalman filtering. This not only achieves accurate deviation judgment and differential correction, but also takes into account processing efficiency and data accuracy, providing high-quality data support for storage parameter calibration and effectively improving the stability of the operation process and the accuracy of decision-making.

[0044] Building upon Example 1, considering the common occurrence of mixed use of multiple types of sensing terminals in actual warehousing operations, and the potential for inconsistent data acquisition timing due to factors such as network latency and differences in device startup timing, using the processing logic for scenarios with consistent timing would lead to data fusion distortion due to the lack of targeted calibration for timing deviations, thereby affecting the accuracy of warehousing parameter calibration and control strategy generation. Therefore, a specialized standardized control process needs to be designed for scenarios with inconsistent acquisition timing, as described in Example 2. When the acquisition timing of dynamic environmental sensing data collected by different sensing terminals is inconsistent, the specific process for standardized control of warehousing data is as follows: The time-series consistency deviation value corresponding to the dynamic environment perception data is input into the data standardization association table, and mapped according to the warehouse scenario type to obtain the time-series calibration priority. Based on the preset deviation interval-score correspondence rule, the time-series calibration priority score is calculated. The time-series consistency deviation value represents the standard deviation of the timestamp difference of the same dimension data in the dynamic environment perception data within the preset collection period. It is used to quantify the stability of the time-series synchronization of the dynamic environment perception data, and the unit is milliseconds (ms). The data standardization association table represents a pre-built structured data table that stores the mapping relationship between the time-series consistency deviation value, the warehouse scenario type and the time-series calibration priority score. The time-series calibration priority score represents the urgency and importance level of the time-series calibration of the dynamic environment perception data. The value range is [0,1]. The higher the score, the greater the impact of the data time-series accuracy on the warehouse operation control.

[0045] If no data (including single-dimensional data and non-single-dimensional data) that meets the preset conditions exists in the dynamic environment perception data, then lightweight standardization processing is performed on the dynamic environment perception data to generate a structured fusion dataset, and dynamic warehouse parameter calibration is performed. The preset condition means that the obtained time-series calibration priority score is greater than the reference time-series calibration priority score. The reference time-series calibration priority score is set by the warehouse scenario's requirement level for the accuracy of data time-series and the statistical analysis of historical calibration effect feedback data.

[0046] If the dynamic environment perception data contains single-dimensional data that meets preset conditions, then a real-time stream processing mode is used for standardized control, specifically: By inputting the timing calibration priority score deviation corresponding to the single-dimensional data into the preset deviation-adjustment mapping rule, the timestamp synchronization adjustment step size and timing correction coefficient are obtained. The timestamp synchronization adjustment step size is set to gradually reduce the adjustment amount of the operation. The gradual reduction of the timestamp synchronization adjustment step size can quickly approach the standard timing and avoid timing jumps and data logic errors caused by large corrections, ensuring smooth convergence of the calibration process. The timing correction coefficient is set to gradually increase the adjustment amount of the operation. The gradual increase of the timing correction coefficient can specifically strengthen the offsetting effect of accumulated deviations such as device clock drift and transmission delay.

[0047] By coordinating calibration and correction operations, and combining a pre-defined deviation-adjustment mapping rule, high-precision alignment between single-dimensional data timestamps and the standard time axis can be achieved, providing reliable time-series support for subsequent cross-stage data comparison and anomaly analysis. The timestamp synchronous adjustment step size decreases progressively while the time-series correction coefficient increases progressively. The pre-defined deviation-adjustment mapping rule, through statistical analysis of the optimal adjustment amount corresponding to different time-series calibration priority score deviation intervals, and after linear fitting and scenario adaptation correction, establishes a one-to-one correspondence between deviation values ​​and adjustment step sizes / correction coefficients.

[0048] If the dynamic environment perception data contains non-single-dimensional data that meets preset conditions, then a batch processing mode is used for standardized control, specifically: The mean deviation of the time-series calibration priority score corresponding to non-single-dimensional data is input into a preset deviation-adjustment mapping rule, thereby mapping the batch calibration window adjustment step size and data integration weight. The batch calibration window adjustment step size is set as the adjustment amount of the step-by-step optimization operation. The batch calibration window adjustment step size is optimized step by step. The first round of wide window covers the time-series discrete range of multi-dimensional data, adapting to the complex time-series distribution characteristics of heterogeneous data. The data integration weight is set as the allocation amount of the step-by-step adaptation operation. The data integration weight is adapted step by step, and the allocation ratio is dynamically adjusted according to the calibration consistency of each dimension to reduce the interference of deviation data. Based on the processing results of each round, the parameters are dynamically adjusted to perform batch time-series calibration and heterogeneous data fusion operations.

[0049] In this embodiment, synchronization stability is quantified by timing consistency deviation value, and calibration priority is accurately determined by combining scenario type and association table. Timing calibration priority score is obtained, and processing mode is selected differently based on the score. The whole process realizes precise and differentiated standardized control of timing inconsistency scenarios, improves data consistency and reliability, provides high-quality support for warehouse parameter calibration, and ensures the accuracy of operation control decisions.

[0050] Furthermore, dynamic warehouse parameter calibration is performed, specifically including: obtaining the cargo storage density and cargo flow rate of the corresponding operating area within the specified warehouse based on the generated structured fusion dataset; setting target values ​​for parameters, specifically: based on the pre-set cargo storage priority, combined with the upper limit of warehouse space utilization, and through the pre-set mapping relationship between cargo storage priority and the upper limit of warehouse space utilization, while referring to the historical average storage density of similar goods over the past 6-12 months, a comprehensive determination is made of the cargo storage density target value that meets the priority storage requirements without exceeding the upper limit of space utilization. Similarly, the cargo flow rate target value is determined by combining the historical peak cargo flow of the corresponding operating area with the upper limit of warehouse operating capacity.

[0051] The parameter adjustment and verification process involves the following steps: First, defining the input parameters of the partial differential equation: current cargo storage density, cargo flow rate, and spatial diffusion coefficient (reflecting the spatial distribution and diffusion characteristics of goods in the storage area, obtained by fitting parameters such as channel layout and storage location spacing). Then, based on the constructed storage density transport equation, which uses cargo storage density as the dependent variable and time and the three-dimensional spatial coordinates of the storage area as independent variables, incorporating a density migration convection term dominated by cargo flow rate and a density homogenization diffusion term dominated by spatial diffusion coefficient, the dynamic rate of change distribution of cargo storage density in three-dimensional space is obtained through solving. This is the storage density adjustment gradient, used to dynamically correct the storage location allocation strategy, such as guiding goods into low-density areas and prioritizing outbound shipments from high-density areas. Low-density areas represent storage areas where the cargo storage density is less than the lower limit set by preset personnel, while high-density areas represent storage areas where the cargo storage density is not less than the upper limit set by preset personnel. By dynamically correcting the storage location allocation strategy, the goal is to ensure that the cargo storage density in the next control cycle converges towards the target value.

[0052] With minimizing the imbalance of warehouse resource load as the optimization objective, an original objective function is constructed, which includes resource allocation coefficients for goods entering and leaving the warehouse (such as AGV scheduling ratio, sorting station allocation weight, etc.). Simultaneously, the target value of goods flow rate is used as a flow rate constraint. Subsequently, Lagrange multipliers are introduced to integrate the flow rate constraint with the original objective function, constructing a Lagrange function to transform the constrained optimization problem into an unconstrained optimization problem. Finally, partial derivatives of each resource allocation coefficient for goods entering and leaving the warehouse in the Lagrange function are calculated, and each partial derivative is set to 0 to form a system of equations. By solving this system of equations, the optimal flow rate adjustment amount that satisfies the goods flow rate constraint is obtained, thereby dynamically correcting the goods entry and exit scheduling rules to ensure that the goods flow rate in the next control cycle approaches the target value.

[0053] like Figure 4 The diagram showing the construction logic of the Lagrange function shows that the top data acquisition entry collects data through devices such as sensors and radio frequency identification (RFID) using the Hypertext Transfer Protocol / 2 (HTTP / 2) protocol. After the upper right module completes system configuration loading, process management, and non-stop upgrade, the data flows into three parallel processes: the left data access layer parses the protocol, the middle data processing layer cleans and standardizes the data, and the right cache interaction layer writes the valid data into the proxy cache module. The cache leader and cache manager synchronously manage cache allocation and lifecycle.

[0054] The cached data is stored persistently (Data Storage) and analyzed by the Analytics Service to obtain core indicators such as cargo flow rate and resource supply. Finally, the Scheduling Service uses the load balancing objective function (Lagrange function) constructed based on the Lagrange multiplier method to set the cargo flow rate target value as a constraint. By taking the partial derivative with respect to the cargo inbound and outbound resource allocation coefficients and setting the derivative to 0, the optimal flow rate adjustment is obtained to dynamically correct the scheduling rules.

[0055] Meanwhile, the I / O module optimizes transmission efficiency with zero-copy technology (sendfile), forming a closed loop of data acquisition, processing, caching, analysis, and Lagrange function scheduling, which is adapted to the high concurrency and dynamic load balancing needs of intelligent warehousing.

[0056] like Figure 5 The diagram illustrating the working principle of the Lagrange function shows that the warehouse data acquisition module on the left acquires three core types of data: warehouse inventory level (reflecting the storage status of goods), automated equipment status (reflecting the operational status of warehouse resources), and order volume (representing the scale of warehouse operation demand). After data standardization, this data is formatted uniformly and redundancy is eliminated before being input into the core load balancing module. This module executes three steps sequentially: first, it receives the standardized input data; second, it uses a load balancing objective function defined based on warehouse operation goals (such as efficiency and resource utilization); and finally, it matches the corresponding load balancing strategy according to the objective function. Ultimately, it outputs the load balancing decision result, specifically manifested in inventory scheduling (adjusting the distribution of goods storage) and equipment start / stop (optimizing the allocation of operational resources), adapting to the dynamic matching needs of warehouse orders and resources.

[0057] The adjustment results are fed back as follows: If the dynamically corrected cargo storage density and cargo flow rate are both within the corresponding allowable range (based on the warehouse design capacity, equipment safety operation parameters, and industry-standard operation in similar scenarios, and optimized based on historical operation data statistical analysis and scenario adaptation), then based on the dynamically corrected cargo storage density and cargo flow rate, combined with the preset warehouse operation rule base (including storage location allocation priority, etc.), an initial control strategy covering storage location planning, resource scheduling, and operation sequence is formed through scenario matching algorithm to ensure operation adaptability. Otherwise, a parameter adjustment anomaly prompt is sent to notify the preset personnel to check the accuracy of the perception data collection and the equipment operation status, and to optimize the control plan in a timely manner.

[0058] The effectiveness assessment of the initial control strategy involves the following process: Within a preset monitoring period (set according to the minimum closed-loop cycle of warehousing operations, typically one shift or 24 hours), the ratio of the cumulative time during which the cargo flow rate is within the allowable range of the rate target value to the total time corresponding to the preset monitoring period yields the rate equilibrium achievement rate, reflecting the rationality of the allocation of automated equipment. Simultaneously, the ratio of the difference between the unit time completed after strategy execution and the baseline work volume before strategy execution to the baseline work volume yields the work efficiency improvement rate, reflecting the optimization effect of the work process. If the obtained rate equilibrium achievement rate is greater than the preset rate equilibrium achievement rate, and the obtained work efficiency improvement rate is greater than the preset work efficiency improvement rate, the initial control strategy is deemed effective, and the execution state corresponding to the initial control strategy is maintained. Otherwise, the initial control strategy is deemed to need optimization, and the dynamic changes of the corresponding real-time sensing data within the designated warehouse are monitored to dynamically adjust the path planning scheme and work execution sequence. Specifically: Obtain the real-time percentage of available storage locations and order urgency weights for each storage zone within a specified warehouse, and perform harmonic averaging to obtain a balance coefficient to quantify the degree of balance in the allocation of storage location resources across storage zones. The percentage of available storage locations with a balance coefficient greater than a reference balance coefficient is designated as the first priority area's available storage location percentage, and the remaining percentages are designated as the second priority area's available storage location percentage. The task allocation priority for the first priority area's available storage location percentage is higher than that for the second priority area's available storage location percentage. Calculate the product of the first priority area's available storage location percentage and the second priority area's available storage location percentage with the order urgency weight. If the product results are the same, determine the task execution order based on the distance of the corresponding available storage location from the current AGV position, from closest to furthest. If the product results are different, prioritize the task allocation based on the product result with the highest priority.

[0059] The spatial 3D coordinates of the storage partition corresponding to the free space ratio of the first priority area and the dynamic occupancy status parameters of the cargo space are used as the independent variables of the Jacobian matrix. The partial derivatives of the coordinate mapping in the path planning are solved through matrix differentiation to quantify the impact of spatial layout changes (such as updates to cargo space occupancy status and shifts in free space) on the path coordinates. The obtained partial derivatives of the coordinate mapping are then used to correct the coordinate nodes of the initially planned path point by point to offset path offset errors caused by dynamic changes in spatial layout, ensuring the real-time adaptability and accuracy of the path planning. The dynamic occupancy status parameters of the cargo space include a value of 0 representing the free cargo space status and a value of 1 representing the occupied cargo space status. After adjusting the path planning scheme and the work execution order, the percentage of rate balance compliance and the percentage of work efficiency improvement in the specified warehouse are re-acquired. If the re-acquired percentage of rate balance compliance is greater than the preset percentage of rate balance compliance, and the acquired percentage of work efficiency improvement is greater than the preset percentage of work efficiency improvement, then the adjustment of the path planning scheme and the work execution order is deemed effective. Otherwise, an early warning is issued for the adjustment effect not meeting the target, prompting the preset personnel to optimize the work order logic or storage partition configuration. The preset percentage of rate balance compliance and the preset percentage of work efficiency improvement are based on the best operating data of similar scenarios in the warehousing industry, statistical analysis of historical adjustment effects, and the setting of warehousing work efficiency targets.

[0060] In this embodiment, core parameters are accurately extracted from a structured fusion dataset. Target values ​​are set based on storage priority and capacity limits. Partial differential equations and the Lagrange multiplier method are used to dynamically optimize location allocation and scheduling rules, ensuring parameters converge to the target values. A dual effectiveness judgment system is constructed using the percentage of rate equilibrium achievement and the percentage of operational efficiency improvement to accurately assess the adaptability of the control strategy. When the strategy needs optimization, the balance of location resources is quantified based on the harmonic mean, and tasks are allocated according to priority. The Jacobian matrix is ​​used to correct path deviations, dynamically adjusting path planning and operational sequence. The overall process achieves closed-loop management of parameter calibration, strategy evaluation, and dynamic optimization, improving both location utilization and operational speed stability. Precise scheduling and path correction reduce coordination conflicts, effectively ensuring warehousing operational efficiency and the reliability of automated equipment operation.

[0061] This invention provides an intelligent warehouse automation control optimization system based on dynamic environment perception, such as... Figure 6The diagram shows the structure of an intelligent warehouse automation control optimization system based on dynamic environment perception. This system can include: a warehouse scene dynamic perception module, used to acquire the warehouse scene type within a specified warehouse, match the corresponding dynamic perception allocation scheme, and simultaneously record the automated equipment, sensing terminals, and goods in the corresponding warehouse area as collaborative operation associated objects; a collaborative association processing module, used to collect dynamic environment perception data of the warehouse area where the collaborative operation associated objects are located according to the dynamic perception allocation scheme, and perform collaborative association processing to generate a structured fusion dataset; and a warehouse parameter calibration and validity judgment module, used to perform dynamic warehouse parameter calibration based on the generated structured fusion dataset, generate an initial control strategy, and judge the validity of the initial control strategy based on its execution process.

[0062] In this embodiment, the warehouse scenario dynamic perception module accurately identifies the warehouse scenario type, matches differentiated dynamic perception allocation schemes, and clarifies the collaborative operation related objects. The collaborative association processing module collects dynamic environmental perception data according to the scheme and generates a standardized structured fusion dataset through collaborative integration. The warehouse parameter calibration and validity judgment module dynamically calibrates warehouse parameters based on the high-quality dataset, generates an initial control strategy, and continuously optimizes it through real-time validity judgment. This realizes on-demand configuration of perception resources, accurate and controllable data quality, and dynamic adaptation of control strategies, effectively improving the automated collaborative efficiency, scenario adaptability, and operational stability of warehouse operations.

Claims

1. An intelligent warehouse automation control optimization method based on dynamic environment perception, characterized in that, The method includes: S1, obtain the storage scenario type in the specified warehouse, match the corresponding dynamic perception allocation scheme, and at the same time record the automated equipment, sensing terminals and goods in the corresponding storage area in the specified warehouse as collaborative operation associated objects; S2: Collect dynamic environmental perception data of the warehouse area where the collaborative operation related objects are located according to the dynamic perception allocation scheme, and perform collaborative association processing to generate a structured fusion dataset. S3: Perform dynamic warehouse parameter calibration based on the generated structured fusion dataset, generate an initial control strategy, and determine the effectiveness of the initial control strategy based on the execution process of the initial control strategy.

2. The intelligent warehouse automation control optimization method based on dynamic environment perception as described in claim 1, characterized in that, The process of obtaining the warehouse scenario type within a specified warehouse and matching it to obtain the corresponding dynamic perception allocation scheme specifically includes: The operation characterization data corresponding to the warehouse scenario type is collected in a preset statistical period and summarized into a scenario judgment dataset. The operation characterization data includes the frequency of cross-operation conflicts, the total frequency of inbound and outbound operations per unit time, and the task handover delay time. Calculate the feature indicators of the corresponding job representation data in the scenario determination dataset, including the collaborative adaptation coefficient, job intensity coefficient, and process stability coefficient. The calculated feature indicators are compared with the corresponding preset benchmark feature indicators, and the geometric mean is taken to obtain the dynamic scenario index for quantifying the dynamic complexity of the warehousing scenario. The dynamic perception allocation scheme is matched based on the obtained dynamic scenario index. The collaborative adaptation coefficient is obtained by combining the ratio of the frequency of cross-operation conflict to the total number of collaborative operations within the cycle with the response synchronization rate of the automated equipment and the sensing terminal. The work intensity coefficient is obtained by combining the ratio of the actual frequency of inbound and outbound operations per unit time to the maximum capacity of warehouse operations with the superimposed order batch complexity coefficient. The process stability coefficient is obtained based on the statistical characteristics of task handover delay duration, combined with the frequency percentage of task handover delay duration exceeding the corresponding allowable delay duration.

3. The intelligent warehouse automation control optimization method based on dynamic environment perception as described in claim 2, characterized in that, The dynamic perception allocation scheme based on the acquired dynamic scene index matching is specifically as follows: Retrieve the scene classification threshold corresponding to the dynamic scene index. The scene classification threshold includes a basic scene threshold and an enhanced scene threshold. The basic scene threshold is less than the enhanced scene threshold. If the dynamic scenario index is not greater than the basic scenario threshold, then the current warehousing scenario is determined to be a basic dynamic scenario, and a basic perception allocation scheme is matched. If the dynamic scenario index is between the basic scenario threshold and the enhanced scenario threshold, the current warehouse scenario is determined to be a regular dynamic scenario, and an advanced perception allocation scheme is matched. If the dynamic scene index is greater than the enhanced scene threshold, the current warehouse scene is determined to be a dense dynamic scene, and an enhanced perception allocation scheme is matched. During the preset monitoring period in the dynamic sensing allocation scheme matching process, the delay time series data of the sensing response of each sensing terminal to the operation characterization data in the current warehousing scenario are acquired, and the delay change rate sequence is obtained by performing time domain difference operation. The time domain standard deviation of the sequence is used as the sensitivity influence factor. After obtaining the jitter time-series data of the operation characterization data after sensing response during transmission, a fast Fourier transform is performed to obtain the frequency domain power spectrum, and the energy value corresponding to the main peak frequency is extracted as the real-time impact factor. After normalizing the sensitivity influence factor and the real-time influence factor respectively, the comprehensive value of the dynamic perception allocation scheme matching process and the degree of dynamic scene adaptation is obtained by geometric mean calculation. This comprehensive value is recorded as the comprehensive value of the degree of perception response influence. If the overall value of the impact of the perception response is greater than the preset threshold for the impact of the perception response, a scenario adaptation anomaly verification prompt will be sent. If the overall value of the impact of the perception response is not greater than the preset threshold for the impact of the perception response, the corresponding dynamic perception allocation scheme will remain unchanged, and collaborative association processing will be performed.

4. The intelligent warehouse automation control optimization method based on dynamic environment perception as described in claim 3, characterized in that, The aforementioned collaborative association processing specifically includes: Calculate the storage collaboration index of the storage area where the objects associated with the collaborative operation of different sensing terminals are located, so as to quantify the degree of collaborative adaptation of sensing terminals in storage location management and cargo safety monitoring. When the warehouse collaboration index is greater than the preset warehouse collaboration index, the collected dynamic environmental perception data is subjected to warehouse data standardization control. The dynamic environmental perception data includes multi-dimensional data that reflects the environmental status of the warehouse area, the operating status of collaborative operation related objects, and the dynamic changes of the operation scenario. When the warehouse collaboration index is not greater than the preset warehouse collaboration index, if abnormal feedback of environmental parameters of the warehouse area is received, environmental parameter control will be performed and a corresponding dimension data re-collection prompt will be sent. If no abnormal feedback of environmental parameters of the warehouse area is received, the corresponding dimension data re-collection will be triggered directly. The abnormal environmental parameter feedback indicates that the environmental parameters in the storage area do not meet the expected allowable conditions for the safe operation of the corresponding automated equipment, and the abnormal operation time reaches the preset judgment period. The environmental parameter control includes: based on the specific environmental parameter type in the abnormal environmental parameter feedback, sending an automated equipment adjustment prompt to the equipment control terminal of the corresponding storage area until receiving feedback on the recovery of the abnormal environmental parameters in the storage area to stop the adjustment and trigger the re-collection of data in the corresponding dimension.

5. The intelligent warehouse automation control optimization method based on dynamic environment perception as described in claim 4, characterized in that, The warehouse collaboration index is obtained as follows: The real-time occupancy status data of storage locations collected by the sensing terminal is combined with the reference storage location occupancy status data in the preset storage location planning scheme. By comparing the status of each storage location and each time node, the frequency ratio of the two storage location occupancy statuses being consistent within the preset statistical period is obtained and recorded as the storage location occupancy matching degree. The accuracy of cargo displacement warnings is obtained by comparing the number of times the cargo displacement anomaly monitoring signals captured by the sensing terminal overlap with the historical anomaly monitoring signals in the historical feature database with the total number of cargo displacement anomaly warnings issued by the sensing terminal within a preset statistical period. After geometrically averaging the obtained storage location occupancy matching degree and cargo displacement early warning accuracy, and combining them with the scenario dynamic collaborative correction coefficient, the warehousing collaboration index is obtained. The scenario dynamic collaborative correction coefficient is obtained based on the task intensity coefficient and the process stability coefficient.

6. The intelligent warehouse automation control optimization method based on dynamic environment perception as described in claim 4, characterized in that, When the acquisition timing of dynamic environmental sensing data collected by different sensing terminals is consistent, the standardized control of the warehouse data specifically includes: Within a preset collection period, the temporal feature representation of the same storage parameter collected by each sensing terminal is obtained, and the coordination deviation value between the corresponding temporal feature representations of adjacent sensing terminals is calculated. If the coordination deviation value is not greater than the preset allowable coordination deviation, then perform lightweight standardization processing on the dynamic environment perception data to generate a structured fusion dataset and perform dynamic warehouse parameter calibration. Otherwise, based on the quantification range of the coordination deviation value, the cumulative impact of the coordination deviation on the stability of the warehousing operation process over time is quantified, specifically as follows: Using the time within a preset acquisition period as the integration variable and the coordination deviation values ​​of adjacent sensing terminals as the integrands, the cumulative integral value of the deviation within a preset statistical period is calculated. If the cumulative integral value of the deviation is not greater than the preset cumulative integral value of the deviation, the collected dynamic environmental perception data will be smoothed and corrected by the moving average algorithm. If the cumulative integral value of the deviation is greater than the preset cumulative integral value of the deviation, the Kalman filter algorithm is used to denoise and correct the collected dynamic environmental perception data.

7. The intelligent warehouse automation control optimization method based on dynamic environment perception as described in claim 4, characterized in that, When the timing of the collection of dynamic environmental sensing data from different sensing terminals is inconsistent, the warehouse data standardization control specifically includes: The time-series consistency deviation value corresponding to the dynamic environment perception data is input into the data standardization association table, and mapped according to the warehouse scenario type to obtain the time-series calibration priority. Based on the preset deviation interval-score correspondence rule, the time-series calibration priority score is calculated. The temporal consistency deviation value represents the standard deviation of the timestamp difference of data of the same dimension in dynamic environment perception data within a preset collection period; The data standardization association table represents a structured data table that maps storage time-series consistency deviation values, storage scenario types, and time-series calibration priority scores. The warehouse scenario types include basic dynamic scenarios, regular dynamic scenarios, and intensive dynamic scenarios; If no data that meets the preset conditions is found in the dynamic environment perception data, then lightweight standardization processing is performed on the dynamic environment perception data to generate a structured fusion dataset, and dynamic storage parameter calibration is performed. The preset conditions indicate that the obtained time-series calibration priority score is greater than the reference time-series calibration priority score. If the dynamic environment perception data contains single-dimensional data that meets preset conditions, then a real-time stream processing mode is used for standardized control, specifically: By determining the time-series calibration priority score deviation corresponding to single-dimensional data, the timestamp synchronization adjustment step size and time-series correction coefficient are determined. The timestamp synchronization adjustment step size is set to the adjustment amount of the operation that is gradually reduced, and the time-series correction coefficient is set to the adjustment amount of the operation that is gradually increased. At the same time, the timestamp synchronization calibration and data time-series deviation correction operations are performed. If the dynamic environment perception data contains non-single-dimensional data that meets preset conditions, then a batch processing mode is used for standardized control, specifically: By using the mean deviation of the time-series calibration priority score corresponding to non-single-dimensional data, the batch calibration window adjustment step size and data integration weight are determined. The batch calibration window adjustment step size is set as the adjustment amount of the step-by-step optimization operation, and the data integration weight is set as the allocation amount of the step-by-step adaptation operation. Based on the processing results of each round, the parameters are dynamically adjusted to perform batch-based time-series calibration and heterogeneous data fusion operations.

8. The intelligent warehouse automation control optimization method based on dynamic environment perception as described in claim 6 or 7, characterized in that, The dynamic warehouse parameter calibration process specifically includes: Based on the generated structured fusion dataset, obtain the cargo storage density and cargo flow rate of the corresponding operating area within the specified warehouse; The parameter target values ​​are set as follows: the pre-set cargo storage priority and warehouse space utilization limit are called to determine the cargo storage density target value, and the cargo flow rate target value is determined by combining the historical cargo flow peak and warehouse operation capacity limit of the corresponding operation area. The parameters are adjusted and verified. Specifically, the currently acquired cargo storage density, cargo flow rate and spatial diffusion coefficient are used as inputs to the partial differential equation to obtain the storage density adjustment gradient, so as to dynamically correct the cargo location allocation strategy. The load balancing objective function is constructed based on the Lagrange multiplier method. The target value of cargo flow rate is used as the constraint condition of the load balancing objective function. The optimal flow rate adjustment is obtained by taking the partial derivative of the cargo inbound and outbound resource allocation coefficient in the load balancing objective function, so as to dynamically correct the cargo inbound and outbound scheduling rules. The load balancing objective function is used to reflect the dynamic adaptation relationship between cargo flow rate and warehousing operation resource supply. The adjustment results are fed back, specifically: if the dynamically corrected cargo storage density and cargo flow rate are both within the corresponding allowable range, then an initial control strategy adapted to the warehousing operation scenario is generated based on the dynamically corrected cargo storage density and cargo flow rate; otherwise, a parameter adjustment anomaly prompt is sent.

9. The intelligent warehouse automation control optimization method based on dynamic environment perception as described in claim 8, characterized in that, The specific process for determining the effectiveness of the initial control strategy is as follows: Within the preset monitoring period, the ratio of the cumulative time during which the cargo flow rate is within the allowable range of the rate target value to the preset total monitoring time is used to obtain the rate balance compliance rate, which reflects the rationality of the allocation of automated equipment. At the same time, the ratio of the difference between the unit time of work completed after the strategy is implemented and the baseline work volume before the strategy is implemented to the baseline work volume is used to obtain the work efficiency improvement rate, which reflects the effect of work process optimization. If the obtained rate balance compliance rate is greater than the preset rate balance compliance rate, and the obtained work efficiency improvement rate is greater than the preset work efficiency improvement rate, then the initial control strategy is deemed effective, and the execution state corresponding to the initial control strategy is maintained. Otherwise, the initial control strategy is determined to need optimization, and the dynamic changes of the corresponding real-time sensing data in the designated warehouse are monitored to dynamically adjust the path planning scheme and the order of operation execution. The dynamic adjustment of the path planning scheme and the order of task execution is specifically as follows: Obtain the real-time percentage of available storage locations and the order urgency weight for each storage zone within a specified warehouse, and perform harmonic averaging to obtain the balance coefficient. The proportion of vacant storage spaces with a balance coefficient greater than the reference balance coefficient is recorded as the vacant proportion of the first priority area, and the proportion of the remaining vacant storage spaces is recorded as the vacant proportion of the second priority area. The operation allocation priority of the vacant proportion of the first priority area is greater than that of the vacant proportion of the second priority area. Calculate the product of the idle percentage of the first priority area and the idle percentage of the second priority area with the order urgency weight. If the results are the same, determine the order of work execution based on the distance of the corresponding idle location from the current position of the AGV to the nearest. If the results are different, assign the work task with the largest product result. The spatial three-dimensional coordinates of the storage partition corresponding to the free proportion of the first priority area and the dynamic occupancy status parameters of the storage space are used as the independent variables of the Jacobian matrix. The partial derivatives of the coordinate mapping in the path planning are solved by matrix differentiation to correct the path deviation caused by changes in spatial layout during the path planning process. The dynamic occupancy status parameters of the storage space include a value representing the status of the free storage space and a value representing the status of the occupied storage space. After adjusting the route planning scheme and the work execution order, the percentage of rate balance compliance and the percentage of work efficiency improvement in the specified warehouse are re-acquired. If the re-acquired percentage of rate balance compliance is greater than the preset percentage of rate balance compliance, and the acquired percentage of work efficiency improvement is greater than the preset percentage of work efficiency improvement, then the route planning scheme and the work execution order adjustment are deemed effective; otherwise, an alert is issued for the adjustment effect not meeting the standards.

10. An intelligent warehouse automation control optimization system based on dynamic environment perception, employing the intelligent warehouse automation control optimization method based on dynamic environment perception as described in any one of claims 1-9, characterized in that, include: The warehouse scene dynamic perception module is used to obtain the warehouse scene type in a specified warehouse, match the corresponding dynamic perception allocation scheme, and record the automated equipment, sensing terminals and goods in the corresponding warehouse area in the specified warehouse as collaborative operation related objects. The collaborative association processing module is used to collect dynamic environmental perception data of the warehouse area where the collaborative operation associated objects are located according to the dynamic perception allocation scheme, and to perform collaborative association processing to generate a structured fusion dataset. The warehouse parameter calibration and validity judgment module is used to perform dynamic warehouse parameter calibration based on the generated structured fusion dataset, generate an initial control strategy, and judge the validity of the initial control strategy based on the execution process of the initial control strategy.

Images