A seedling inventory optimization and intelligent scheduling system and method based on big data and predictive analysis

Abstract

This invention relates to the field of big data processing and intelligent scheduling technology, and discloses a seedling inventory optimization and intelligent scheduling system and method based on big data and predictive analysis. The system includes a data fusion processing module, an intelligent prediction module, a collaborative optimization decision-making module, and a scheduling execution and feedback module. The data fusion processing module cleans and merges seedling growth data, external market data, and internal enterprise data collected by the Internet of Things. The intelligent prediction module performs seedling demand forecasting and growth specification forecasting. The collaborative optimization decision-making module generates dynamic safety stock strategies and logistics scheduling schemes based on the prediction results and inventory status. The scheduling execution and feedback module executes the scheduling and performs feedback optimization. This invention improves inventory turnover efficiency and scheduling rationality by introducing seedling growth status into inventory control and achieving collaborative optimization of inventory and scheduling.

Description

Technical Field

[0001] This invention relates to the field of intelligent management and data analysis technology, and in particular to a seedling inventory optimization and intelligent scheduling system and method based on big data and predictive analysis. Background Technology

[0002] With the continuous advancement of urban landscaping, ecological restoration, and infrastructure construction, the seedling industry plays a vital role in municipal engineering, real estate development, and ecological environment construction. As a biological asset with a growth cycle, seedlings have production, inventory, and sales management characteristics that significantly differ from general industrial products. Therefore, scientific management of seedling inventory and distribution is of great importance for improving the industry's operational efficiency.

[0003] Currently, seedling inventory management and logistics scheduling largely rely on manual experience or simple information systems, remaining at a relatively rudimentary management stage. Existing technologies typically employ static safety stock models based on historical sales data. These models are usually calculated based on average demand and supply cycles, lacking the ability to dynamically respond to changes in seedling growth and market demand. Because seedlings continuously change in size during growth, the market value and demand for different sizes vary significantly. Existing inventory models struggle to reflect the characteristics of seedlings as "dynamic biological assets," easily leading to some seedlings failing to be shipped out in time during their optimal sales period, resulting in inventory backlog or value loss.

[0004] On the other hand, in terms of demand forecasting, existing technologies mostly rely on single data sources or simple statistical methods, lacking the ability to comprehensively utilize multi-source data. For example, the demand for seedlings is not only affected by historical sales data, but also closely related to various external factors such as municipal planning, real estate project progress, seasonal changes, and industry trends. However, existing systems often cannot effectively integrate seedling growth data collected by the Internet of Things, external market information, and internal business data, resulting in insufficient accuracy in forecasting results, which in turn affects inventory decisions and production arrangements.

[0005] In terms of logistics scheduling, existing seedling transportation scheduling typically employs experience-based manual scheduling or simple route optimization methods. The scheduling process is often single-objective oriented, such as considering only transportation costs or distance, lacking comprehensive optimization of multiple factors such as order delivery timeliness, vehicle loading rate, and seedling specification matching. Furthermore, due to the lack of a collaborative mechanism between inventory management and logistics scheduling, inventory decisions and transportation arrangements are often disconnected, resulting in inefficient resource allocation, high transportation costs, and difficulty in meeting diverse customer needs.

[0006] Furthermore, existing technologies generally suffer from insufficient data utilization. Seedling growth data, inventory data, order data, and transportation data are typically scattered across different systems or stored manually, lacking a unified data view and an effective data fusion mechanism. Simultaneously, existing systems often lack the ability to analyze and provide feedback on execution results, making it difficult to continuously optimize predictive models and scheduling strategies based on actual execution deviations, thus limiting the improvement of the system's overall decision-making capabilities.

[0007] In summary, existing technologies for seedling inventory optimization and scheduling management generally suffer from problems such as insufficient forecasting accuracy, static inventory control, lack of collaborative optimization in scheduling decisions, and insufficient data utilization capabilities. There is still a lack of a technical solution that can integrate multi-source data, combine seedling growth characteristics, and achieve integrated dynamic optimization of inventory and scheduling. Summary of the Invention

[0008] In view of this, the purpose of this invention is to provide a seedling inventory optimization and intelligent scheduling system and method based on big data and predictive analysis, to solve the problems of existing seedling inventory management such as reliance on experience, insufficient accuracy of demand forecasting, static inventory control, lack of collaborative optimization in logistics scheduling, and insufficient data utilization capabilities. To achieve the above objective, this invention provides the following technical solution: In one embodiment of the present invention, a seedling inventory optimization and intelligent scheduling system based on big data and predictive analysis is provided, including a data fusion processing module, an intelligent prediction module, a collaborative optimization decision-making module, and a scheduling execution and feedback module; wherein, the data fusion processing module is used to collect and fuse multi-source heterogeneous data, and clean, extract features, and structure the multi-source heterogeneous data to generate a unified seedling data view; the multi-source heterogeneous data includes seedling growth status data and soil environment data collected through IoT devices, market demand trend data obtained through external data sources, and internal enterprise operation data; the intelligent prediction module is communicatively connected to the data fusion processing module and is used to optimize and schedule seedling inventory based on the data fusion processing module. The system provides a unified seedling data view and performs seedling market demand forecasting and seedling growth specification forecasting. The collaborative optimization decision module is communicatively connected to the intelligent forecasting module and is used to generate inventory optimization strategies and logistics scheduling schemes based on predicted market demand, predicted growth specifications, and real-time inventory status data through a preset optimization model. The inventory optimization strategy includes a dynamic safety stock control strategy, and the logistics scheduling scheme is a multi-objective optimization result. The scheduling execution and feedback module is communicatively connected to the collaborative optimization decision module and is used to execute the logistics scheduling scheme and feed back the actual data collected during execution to the intelligent forecasting module and the collaborative optimization decision module to achieve iterative optimization of model parameters.

[0009] Furthermore, the seedling growth status data includes physiological indicators such as seedling height, crown width, and ground diameter obtained through IoT sensors deployed in the nursery and multispectral images from drones; the market demand trend data includes municipal planning bidding information, real estate construction data, seedling trading platform price indices, and landscape design trend information obtained through web crawling technology; and the enterprise's internal operation data includes historical sales orders, customer information, and procurement records.

[0010] Furthermore, the intelligent prediction module includes a demand prediction submodule and a growth prediction submodule. The demand prediction submodule uses a hybrid architecture of an ensemble learning model and a time-series prediction model to predict the market demand for seedlings. The ensemble learning model processes structured demand feature data, and the time-series prediction model processes historical sales sequence data. The outputs of both models are then fused to generate a final predicted value. The growth prediction submodule constructs a seedling growth curve model based on time-series analysis and dynamically calibrates the growth curve model using real-time collected IoT data to output a seedling growth specification prediction.

[0011] Preferably, the ensemble learning model is a gradient boosting decision tree model, and the time-series prediction model is a long short-term memory network model.

[0012] Furthermore, the collaborative optimization decision-making module includes a dynamic safety stock model, which determines the dynamic safety stock level based on predicted market demand, predicted growth rate, current seedling specifications, target nursery specifications, order delivery lead time, and customer priority.

[0013] Preferably, the dynamic safety stock level is dynamically adjusted according to the closeness between the current size of the seedlings and the target size for sale, so that the closer the current size of the seedlings is to the target size for sale, the lower the corresponding safety stock level.

[0014] Furthermore, the collaborative optimization decision module also includes a multi-objective scheduling model, which is used to simultaneously optimize the following objectives when generating logistics scheduling schemes: reducing transportation costs, reducing order delivery delays, and increasing vehicle loading rates.

[0015] Preferably, the constraints of the multi-objective scheduling model include vehicle capacity constraints, order delivery time window constraints, and seedling category and specification matching constraints.

[0016] Alternatively, the collaborative optimization decision-making module may use a multi-objective genetic algorithm or a reinforcement learning algorithm to solve the logistics scheduling scheme.

[0017] Furthermore, the scheduling execution and feedback module is used to compare the actual data with the predicted data output by the intelligent prediction module and the optimization scheme data output by the collaborative optimization decision module, and adaptively adjust the model parameters in the intelligent prediction module and the objective function weight coefficients in the collaborative optimization decision module according to the comparison results.

[0018] In one embodiment of the present invention, a method for optimizing and intelligently scheduling seedling inventory based on big data and predictive analysis is also provided, comprising the following steps: S1, collecting and fusing multi-source heterogeneous data, and processing the data to form a unified seedling data view; S2, performing seedling market demand forecasting and seedling growth specification forecasting based on the unified seedling data view; S3, determining a dynamic safety stock control strategy based on the forecast results and real-time inventory status, and generating a logistics scheduling plan; S4, executing the logistics scheduling plan, and collecting actual execution data feedback for continuous model optimization.

[0019] Alternatively, a computer-readable storage medium is also provided, having a computer program stored thereon that, when executed by a processor, implements the steps of the above-described method.

[0020] Based on the above technical solution, the seedling inventory optimization and intelligent scheduling system based on big data and predictive analysis of the present invention integrates seedling growth status data, external market demand trend data and internal enterprise operation data collected by the Internet of Things, and performs cleaning, feature extraction and structuring to construct a unified seedling data view. On this basis, seedling market demand forecasting and seedling growth specification forecasting are performed respectively, so that the system can simultaneously obtain the future demand change trend of seedlings and the seedling growth status change trend, thereby realizing a forward-looking analysis of supply and demand relationship. Furthermore, by incorporating data on predicted market demand, predicted growth specifications, and real-time inventory status into the collaborative optimization decision-making module, an inventory optimization model with dynamic safety stock control strategies is constructed. This transforms inventory control from traditional static rules into an adaptive adjustment mechanism that dynamically changes with the growth status of seedlings. In particular, by introducing the correlation between the current specifications of seedlings and the target specifications for sale, the system can automatically identify seedlings that are close to being marketable and reduce their safety stock levels, thereby improving the timeliness of seedling sales and reducing the risk of high-value seedlings being left unsold. Meanwhile, by optimizing the logistics scheduling scheme through a multi-objective scheduling model, while taking into account transportation costs, order delivery timeliness and vehicle loading rate, and combining the matching constraints of seedling categories and specifications, the rational allocation of transportation resources and route optimization are achieved, effectively reducing vehicle empty load rate and unreasonable scheduling, thereby improving overall logistics efficiency. In addition, the actual execution data is collected through the scheduling execution and feedback module and compared with the prediction results and optimization schemes to form a closed-loop feedback mechanism. This enables the system to adaptively adjust the prediction model parameters and scheduling strategies based on historical execution deviations, thereby continuously improving prediction accuracy and decision rationality. Therefore, by constructing an integrated closed-loop decision-making mechanism of "data fusion - two-dimensional prediction - collaborative optimization - execution feedback", this invention solves the problems of existing technologies such as reliance on experience in seedling inventory management, static inventory control, inaccurate supply and demand matching, and low logistics scheduling efficiency. It realizes dynamic, refined, and intelligent seedling inventory management, significantly improves inventory turnover efficiency, reduces operating costs, and increases customer delivery satisfaction.

[0021] Instruction manual with accompanying drawings Figure 1 A schematic diagram of the structure of a seedling inventory optimization and intelligent scheduling system based on big data and predictive analysis provided in an embodiment of the present invention; Figure 2 This is a flowchart illustrating a seedling inventory optimization and intelligent scheduling method based on big data and predictive analysis, provided as an embodiment of the present invention.

[0022] Explanation of reference numerals in the attached figures: 100—Data Fusion Processing Module; 110—Data Acquisition Unit; 120—Data Cleaning Unit; 130—Feature Extraction Unit; 140—Data Fusion Unit; 200—Intelligent Prediction Module; 210—Demand Prediction Submodule; 220—Growth Prediction Submodule; 300—Collaborative Optimization Decision Module; 310—Dynamic Safety Stock Model; 320—Multi-Objective Scheduling Model; 400—Schedule Execution and Feedback Module; 410—Schedule Execution Unit; 420—Data Acquisition Unit; 430—Feedback Optimization Unit. Detailed Implementation

[0023] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can more clearly understand the technical concept, structural composition, implementation method and beneficial effects of the present invention. It should be understood that the specific embodiments described below are only used to explain the technical solution of the present invention, and are not intended to limit the scope of protection of the present invention. Without departing from the spirit and essence of the present invention, those skilled in the art can make appropriate adjustments, substitutions or combinations to the structural forms, connection relationships, processing flows, parameter settings and algorithm implementation methods described in the following embodiments, and these adjustments, substitutions or combinations should all be considered as optional embodiments of the present invention.

[0024] Furthermore, to more clearly illustrate the technical solution of the present invention, the following will be combined with... Figure 1 The system structure diagram shown and Figure 2 The flowchart shown illustrates the seedling inventory optimization and intelligent scheduling system and method based on multi-source data fusion and predictive analysis described in this invention, and further exemplifies the dynamic safety stock determination process and multi-objective scheduling optimization process in subsequent application embodiments.

[0025] I. Overall Structure Description In one embodiment of the present invention, such as Figure 1 As shown, a seedling inventory optimization and intelligent scheduling system based on big data and predictive analysis is provided. The system adopts a modular structure design and mainly includes a data fusion processing module 100, an intelligent prediction module 200, a collaborative optimization decision-making module 300, and a scheduling execution and feedback module 400. The modules interact and work collaboratively with each other through data interfaces or communication networks.

[0026] Furthermore, the system takes multi-source heterogeneous data as input, and outputs inventory optimization strategies and logistics scheduling schemes through multiple stages such as data processing, predictive analysis and optimization decision-making. It also achieves continuous model optimization through execution feedback, thereby constructing a closed-loop operation mechanism of "data-driven - predictive analysis - collaborative decision-making - execution feedback".

[0027] Specifically, the data fusion processing module 100 is used to process data from different sources in a unified manner, including but not limited to seedling growth data collected by IoT devices at the nursery site, market demand data obtained from external data sources, and operational data from the enterprise's internal business system. By cleaning, extracting features, and structuring the above data, a seedling data view in a unified format is formed, providing basic data support for subsequent predictive analysis.

[0028] Furthermore, the intelligent prediction module 200 is communicatively connected to the data fusion processing module 100, and is used to perform seedling market demand prediction and seedling growth specification prediction based on the seedling data view. The demand prediction reflects the changing market demand trends for different types and specifications of seedlings over a future period, while the growth prediction reflects the dynamic changes in seedlings from their current state to their target specifications, thereby achieving simultaneous prediction on both the supply and demand sides.

[0029] Furthermore, the collaborative optimization decision-making module 300 is communicatively connected to the intelligent prediction module 200, and is used to jointly optimize the inventory control strategy and logistics scheduling scheme based on the obtained prediction results and the current inventory status. The collaborative optimization decision-making module 300 determines the inventory control strategy through a dynamic safety stock model and generates a logistics scheduling scheme through a multi-objective scheduling model, thereby realizing collaborative decision-making between inventory management and transportation scheduling, rather than independent sequential processing.

[0030] Furthermore, the scheduling execution and feedback module 400 is communicatively connected to the collaborative optimization decision-making module 300, and is used to execute the generated logistics scheduling plan and collect relevant operational data during the execution process, including transportation time, delivery status, and resource utilization. The scheduling execution and feedback module 400 is also used to feed back the collected actual execution data to the intelligent prediction module 200 and the collaborative optimization decision-making module 300, so as to dynamically adjust the prediction model parameters and optimization strategies.

[0031] Preferably, the scheduling execution and feedback module 400 forms a prediction correction path with the intelligent prediction module 200 and a decision optimization path with the collaborative optimization decision module 300, so that the system continuously corrects prediction deviations and optimizes scheduling strategies during multiple runs, thereby achieving continuous improvement in system performance.

[0032] Therefore, this invention constructs an intelligent management system for seedlings, a dynamic biological asset, through data flow and functional collaboration among the above modules. It realizes closed-loop control of the entire process from data collection and predictive analysis to decision execution and feedback optimization, overcoming the problems of the separation between inventory management and scheduling decisions and the lack of dynamic adjustment capabilities in the prior art.

[0033] II. Data Fusion Processing Module In one embodiment of the present invention, the data fusion processing module 100 is used to uniformly process multi-source heterogeneous data from different data sources to form a standardized and structured seedling data view, providing basic data support for subsequent intelligent prediction and optimization decision-making.

[0034] Furthermore, the data fusion processing module 100 includes a data acquisition unit 110, a data cleaning unit 120, a feature extraction unit 130, and a data fusion unit 140, with each unit connected in sequence or interacting with each other through a data interface.

[0035] (a) Data Acquisition Unit In one embodiment of the present invention, the data acquisition unit 110 is used to acquire multi-source heterogeneous data, wherein the multi-source heterogeneous data includes at least the following types: Seedling growth status data Data is collected through IoT devices deployed in the nursery area, including but not limited to soil sensors, environmental monitoring equipment, and drone multispectral imaging equipment, to obtain growth indicators of seedlings such as plant height, crown width, and ground diameter, as well as environmental parameters such as soil moisture, nutrient content, and temperature. External market data Information obtained through web crawling technology or data interfaces, including municipal engineering bidding information, real estate construction data, seedling trading platform price index, and landscape design trend information, is used to reflect changes in market demand trends. Enterprise internal operational data Data comes from the enterprise information system, including historical sales order data, inventory data, customer information, and purchase records.

[0036] Furthermore, the data acquisition unit 110 can update the data according to a preset time interval or triggering mechanism to ensure the timeliness and continuity of the data.

[0037] (ii) Data Cleaning Unit Furthermore, the data cleaning unit 120 is used to preprocess the collected raw data to improve data quality, and its processing includes: Missing value imputation; Outlier detection and removal; Noise data filtering; Data format is processed uniformly.

[0038] Preferably, in response to possible abnormal fluctuations in sensor data collected by the Internet of Things, the data cleaning unit 120 uses a sliding window or statistical threshold method to identify and correct abnormal data.

[0039] (III) Feature Extraction Unit In one embodiment of the present invention, the feature extraction unit 130 is used to extract key feature information from the cleaned data to support the construction of subsequent prediction models.

[0040] Furthermore, the feature extraction unit 130 includes: Time dimension alignment is used to unify the time scale of different data sources; Feature construction processing is used to generate feature variables that reflect the growth status of seedlings and changes in market demand. Dimension reduction or feature filtering can improve data processing efficiency and prediction accuracy.

[0041] Preferably, the feature extraction unit 130 constructs feature variables for the seedling growth stage, enabling the quantification of the seedling's change process from the seedling stage to the target size stage.

[0042] By transforming the seedling growth process into computable characteristics, "biological assets" are endowed with data-driven decision-making capabilities.

[0043] (iv) Data fusion unit Furthermore, the data fusion unit 140 is used to integrate data from different sources to form a unified seedling data view.

[0044] Specifically, the data fusion unit 140 establishes a unified data structure to perform association mapping on data from different sources, including: Data is linked according to seedling type and batch; Match growth data with market data by time series; Data is integrated based on geographical location or nursery dimensions.

[0045] Furthermore, the seedling data view output by the data fusion unit 140 includes at least: Current growth status information of seedlings; Information on the historical growth trajectory of seedlings; Market demand trend information; Inventory and order status information.

[0046] Preferably, the seedling data view is stored in a structured data format for direct access by the intelligent prediction module.

[0047] (v) Overall technical effect of the module Through the above data collection, cleaning, feature extraction and fusion processing process, the data fusion processing module 100 realizes unified management and efficient utilization of multi-source heterogeneous data, transforming the originally scattered, heterogeneous and difficult-to-use data into structured and computable data assets.

[0048] Compared with existing technologies, this invention introduces and structures seedling growth status data, enabling the inventory management system to perceive the dynamic growth process of seedlings. This provides key data support for subsequent predictive analysis and collaborative optimization decisions, significantly improving the overall intelligence level and decision-making accuracy of the system.

[0049] III. Intelligent Prediction Module In one embodiment of the present invention, the intelligent prediction module 200 is used to predict and analyze changes in seedling market demand and seedling growth status based on the seedling data view output by the data fusion processing module 100, thereby providing data support for subsequent inventory optimization and scheduling decisions.

[0050] Furthermore, the intelligent forecasting module 200 includes a demand forecasting submodule 210 and a growth forecasting submodule 220, which are used to perform forecasting analysis on the market demand side and supply side, respectively, thereby forming a supply and demand coordinated forecasting system.

[0051] (a) Demand Forecasting Submodule In one embodiment of the present invention, the demand forecasting submodule 210 is used to forecast the demand for seedlings in the future.

[0052] Furthermore, the demand forecasting submodule 210 constructs a feature set describing changes in market demand based on the structured data output by the data fusion processing module 100. This feature set includes at least: Characteristics of historical sales data; Customer types and regional distribution characteristics; Seasonal variation characteristics; External market data characteristics (including bidding information, real estate construction data, and price indices, etc.).

[0053] Furthermore, the demand prediction submodule 210 processes the feature set using a combination of an ensemble learning model and a time-series prediction model, wherein: The ensemble learning model is used to model structured features in order to capture the nonlinear relationships between different influencing factors; The time-series prediction model is used to model historical sales sequences in order to extract periodic and trend-based changes over time. Furthermore, the output results of the ensemble learning model and the output results of the time series prediction model are fused together to generate the final demand prediction result.

[0054] Preferably, the fusion process includes weighted combination or model integration.

[0055] By integrating structured feature analysis with time series analysis, the prediction accuracy in complex demand scenarios can be improved.

[0056] (ii) Growth Prediction Submodule In one embodiment of the present invention, the growth prediction submodule 220 is used to predict the growth process of seedlings from their current state to the target nursery size.

[0057] Furthermore, the growth prediction submodule 220 constructs a seedling growth curve model based on historical seedling growth data, and dynamically calibrates the growth curve model by combining real-time collected IoT data.

[0058] Specifically, the growth prediction submodule 220 includes: The growth modeling unit is used to build seedling growth trend models based on historical growth data. The status update unit is used to receive real-time IoT data and update the current growth status of the seedlings. The dynamic calibration unit is used to correct the growth trend model based on changes in environmental parameters.

[0059] Furthermore, when changes in environmental parameters (including soil moisture, temperature, or nutrient content) are detected, the dynamic calibration unit adjusts the seedling growth rate according to the changes, thereby updating the growth prediction results for a future period of time.

[0060] Preferably, the growth prediction results include the estimated time for the seedlings to reach the target size and information on changes in growth stages.

[0061] (III) Supply and demand coordinated forecasting mechanism Furthermore, the intelligent prediction module 200 performs joint analysis of demand prediction results and growth prediction results to achieve prediction of supply and demand matching relationship.

[0062] Specifically: The demand forecast results are used to reflect the demand for various types of seedlings in different time periods. The growth prediction results are used to reflect the supply capacity of each batch of seedlings at different time points. By comparing and analyzing the two, we can obtain the supply and demand matching situation of seedlings in various time periods in the future.

[0063] Preferably, when the forecast results show that the supply of a certain type of seedling is insufficient within the target time period, the system generates an early warning message; when the supply is excessive, the system marks it as a potential inventory risk.

[0064] (iv) Module output and function In one embodiment of the present invention, the intelligent prediction module 200 outputs the following result: Seedling market demand forecast results; Seedling growth specification prediction results; Supply and demand matching analysis results.

[0065] Furthermore, the aforementioned prediction results are passed as input to the collaborative optimization decision module 300 to support the formulation of dynamic safety stock strategies and the generation of logistics scheduling plans.

[0066] (V) Summary of Technical Effects Through the synergistic processing of demand forecasting and growth forecasting, this invention achieves synchronous modeling of seedling market demand and seedling growth process, enabling the system to obtain dynamic change information on both supply and demand sides before decision-making.

[0067] Compared with existing technologies, this invention introduces seedling growth forecasting and combines it with demand forecasting, so that inventory and scheduling decisions are based on dynamic matching of supply and demand. This avoids decision-making bias caused by relying solely on historical data or a single forecasting model, thereby improving the overall system's forecasting accuracy and decision-making rationality.

[0068] IV. Collaborative Optimization Decision Module In one embodiment of the present invention, the collaborative optimization decision module 300 is communicatively connected to the intelligent prediction module 200, and is used to jointly optimize the inventory control strategy and logistics scheduling scheme based on the acquisition of seedling market demand prediction results, growth specification prediction results and real-time inventory status data, thereby realizing collaborative decision-making for inventory management and transportation scheduling.

[0069] Furthermore, the collaborative optimization decision module 300 includes a dynamic safety stock model 310 and a multi-objective scheduling model 320, wherein the dynamic safety stock model 310 is used to determine the inventory control strategy, and the multi-objective scheduling model 320 is used to generate a logistics scheduling scheme.

[0070] (a) Dynamic safety stock model In one embodiment of the present invention, the dynamic safety stock model 310 is used to determine the safety stock level of seedling inventory based on multi-dimensional variables, so as to achieve dynamic adjustment of inventory.

[0071] Furthermore, the dynamic safety stock model 310 is based on a comprehensive analysis of the following factors: Seedling market demand forecast results; Seedling growth prediction results; Current specifications and status of the seedlings; Target nursery specifications; Order delivery cycle; Customer priority information.

[0072] Furthermore, the dynamic safety stock model 310 dynamically adjusts the safety stock level based on the closeness between the current specifications of the seedlings and the target specifications for sale.

[0073] Specifically: When the current size of the seedlings is close to the target size for delivery, the system reduces the safety stock level of the corresponding seedlings to increase the delivery priority of this type of seedling. When the current size of the seedlings is far from the target size for delivery, the system increases its safety stock level to reduce the risk of premature delivery or supply disruption.

[0074] Furthermore, the dynamic safety stock model 310 also adjusts the inventory strategy based on demand forecast results. When the predicted demand increases, the safety stock level of key specifications of seedlings is increased accordingly; when the predicted demand decreases, the corresponding inventory level is reduced.

[0075] (II) Multi-objective scheduling model In one embodiment of the present invention, the multi-objective scheduling model 320 is used to generate a logistics scheduling scheme based on inventory strategy and order demand.

[0076] Furthermore, the multi-objective scheduling model 320 is guided by multiple optimization objectives to uniformly plan transportation tasks, including: Reduce overall transportation costs; Reduce order delivery delays; Increase vehicle loading capacity.

[0077] Furthermore, the multi-objective scheduling model 320 must satisfy the following constraints when performing optimization calculations: Vehicle capacity constraints; Order delivery time window constraints; Restrictions on seedling varieties; Tree specifications matching constraints.

[0078] The seedling specification matching constraint is used to ensure that the seedling specifications allocated in the scheduling scheme meet the order requirements, and avoid repeated scheduling or resource waste due to specification mismatch.

[0079] Furthermore, the multi-objective scheduling model 320 can allocate resources among multiple nurseries, enabling seedlings from different nurseries to be transported in combination to meet the multi-specification requirements of the same order.

[0080] Preferably, the multi-objective scheduling model 320 is solved using a multi-objective genetic algorithm or a reinforcement learning algorithm to obtain an optimized scheduling scheme that satisfies the constraints.

[0081] (III) Inventory and Scheduling Coordination Mechanism In one embodiment of the present invention, the collaborative optimization decision module 300 couples the dynamic safety stock model 310 with the multi-objective scheduling model 320 to achieve collaborative optimization between the inventory strategy and the scheduling scheme.

[0082] Specifically: The inventory strategy output by the dynamic safety stock model 310 is used to constrain the range of seedlings that can participate in scheduling; The multi-objective scheduling model 320 selects seedlings that can be shipped out based on the inventory strategy and generates a transportation plan; Furthermore, the optimization results of the multi-objective scheduling model 320 can have a reverse impact on inventory strategies, for example: When a certain type of seedling is difficult to dispatch due to limited transportation resources, the system can appropriately adjust its inventory strategy to postpone its delivery. When the scheduling efficiency of a certain type of seedling is high, the system can prioritize its release from the warehouse, thereby accelerating inventory turnover.

[0083] (iv) Module output and function In one embodiment of the present invention, the collaborative optimization decision module 300 outputs the following result: Dynamic safety stock control strategy; Priority ranking of seedlings leaving the warehouse; Logistics scheduling plan (including transportation routes and loading plans).

[0084] Furthermore, the output result is transmitted to the scheduling execution and feedback module 400 for executing specific scheduling tasks.

[0085] (V) Summary of Technical Effects Through the synergistic effect of the aforementioned dynamic safety stock model and multi-objective scheduling model, this invention achieves integrated optimization of inventory control and logistics scheduling.

[0086] Compared with the prior art, the present invention has the following technical effects: To achieve a shift from static inventory control to dynamic adjustment; Improve the rationality of the timing of seedling release from storage; Reduce transportation costs and improve resource utilization; Improve inventory turnover efficiency and order delivery capabilities.

[0087] In particular, by introducing constraints on seedling growth status and specification matching, this invention enables refined management of seedlings as dynamic biological assets, thereby significantly improving the overall system's intelligence level and decision-making accuracy.

[0088] V. Scheduling Execution and Feedback Module In one embodiment of the present invention, the scheduling execution and feedback module 400 is communicatively connected to the collaborative optimization decision module 300, and is used to execute the generated logistics scheduling plan and collect and analyze the actual operation data during the execution process, thereby realizing feedback optimization of the system prediction and decision-making process.

[0089] Furthermore, the scheduling execution and feedback module 400 includes a scheduling execution unit 410, a data acquisition unit 420, and a feedback optimization unit 430, and the units are connected to each other through a data interface.

[0090] (a) Scheduling and Execution Unit In one embodiment of the present invention, the scheduling execution unit 410 is used to execute specific transportation tasks according to the logistics scheduling scheme output by the collaborative optimization decision module 300.

[0091] Specifically, the scheduling execution unit 410 includes: The task allocation subunit is used to allocate transportation tasks in the scheduling scheme to corresponding vehicles or transportation resources; The path execution subunit is used to execute the transportation process according to the optimized path; The loading execution subunit is used to complete the seedling loading operation according to the loading plan.

[0092] Furthermore, during the execution process, the scheduling execution unit 410 can make appropriate adjustments to the execution process based on real-time traffic information or on-site conditions.

[0093] (ii) Data acquisition unit Furthermore, the data acquisition unit 420 is used to acquire actual operating data during the scheduling execution process, and the actual operating data includes at least: Actual transportation time and route data; Actual delivery time and delivery status of the order; Vehicle loading rate and resource utilization; Seedling loss or abnormal conditions.

[0094] Preferably, the data acquisition unit 420 automatically acquires data through an in-vehicle terminal, positioning device, or mobile terminal.

[0095] (III) Feedback Optimization Unit In one embodiment of the present invention, the feedback optimization unit 430 is used to compare and analyze the actual operating data with the prediction results and scheduling scheme, and to adaptively adjust the system model based on the analysis results.

[0096] Furthermore, the feedback optimization unit 430 includes: The deviation analysis subunit is used to calculate the difference between the actual execution result and the predicted result; The model adjustment subunit is used to adjust the model parameters in the intelligent prediction module 200 based on the deviation results; The strategy optimization subunit is used to modify the optimization strategy in the collaborative optimization decision module 300 based on the execution effect.

[0097] Specifically: When there is a deviation between actual demand and predicted demand, the model adjustment subunit updates the parameters of the demand prediction model; When there is a deviation between the actual growth of seedlings and the predicted results, the growth prediction model is corrected. When the scheduling execution effect does not meet expectations (e.g., high latency or low load rate), the weight parameters in the multi-objective scheduling model are adjusted.

[0098] (iv) Closed-loop feedback mechanism Furthermore, the scheduling execution and feedback module 400 forms the following closed loop by feeding back the optimization results to the intelligent prediction module 200 and the collaborative optimization decision-making module 300: Execution results → Correct the prediction model; Execution result → Adjust inventory strategy; Execution result → Modify scheduling strategy.

[0099] Through the aforementioned feedback path, the system can gradually reduce prediction errors and improve scheduling efficiency during multiple rounds of operation.

[0100] Preferably, the feedback process is executed automatically according to a preset period or triggering conditions.

[0101] (V) Module Output and Function In one embodiment of the present invention, the scheduling execution and feedback module 400 outputs the following information: Actual execution result data; Execute the deviation analysis results; Adjust the model parameters.

[0102] The above output is used to update the system's internal model, thereby affecting subsequent prediction and decision-making processes.

[0103] (vi) Summary of technical effects Through the aforementioned scheduling execution and feedback mechanism, this invention enables real-time monitoring of the logistics scheduling execution effect and dynamic correction of the prediction and decision-making model.

[0104] Compared with existing technologies, this invention can not only generate optimized scheduling schemes, but also continuously optimize the system based on actual execution results, thereby significantly improving prediction accuracy and scheduling rationality, and avoiding decision failure caused by the accumulation of model bias.

[0105] In particular, by introducing a closed-loop feedback mechanism, the system is transformed from a traditional "single-time optimization decision system" into a "self-learning dynamic optimization system," further improving the system's adaptability and stability in complex business environments.

[0106] VI. Method Implementation In one embodiment of the present invention, such as Figure 2 As shown, a method for optimizing and intelligently scheduling seedling inventory based on multi-source data fusion and predictive analysis is provided. This method is applied to the aforementioned seedling inventory optimization and intelligent scheduling system and includes the following steps: Step S1: Multi-source data acquisition and fusion processing In step S1, multi-source heterogeneous data from different data sources are collected and fused to form a unified seedling data view.

[0107] Specifically, the multi-source heterogeneous data includes: Data on seedling growth status and soil environment collected through IoT devices; Market demand trend data obtained from external data sources; Internal operational data of the enterprise, including historical orders, inventory status and customer information.

[0108] Furthermore, the collected data is cleaned, features are extracted, and structured to unify the data from different sources in terms of time dimension and data format, thereby generating a seedling data view that can be processed later.

[0109] Step S2: Demand Forecasting and Growth Forecasting In step S2, seedling market demand forecasting and seedling growth specification forecasting are performed based on the seedling data view.

[0110] Specifically: The demand forecasting submodule is used to predict the demand for different types and specifications of seedlings in the future. The growth prediction submodule predicts the growth process of seedlings from their current size to the target size for sale, and dynamically calibrates the prediction results by combining real-time IoT data.

[0111] Furthermore, by jointly analyzing the demand forecast results and the growth forecast results, we can obtain the supply and demand matching situation of seedlings in the future time period.

[0112] Step S3: Generation of Inventory Strategy and Scheduling Plan In step S3, a dynamic safety stock control strategy and logistics scheduling plan are generated based on the prediction results and the current inventory status.

[0113] Specifically: The dynamic safety stock model determines the safety stock level of different seedlings based on factors such as seedling demand forecast, growth forecast, current specifications and target specifications, and generates inventory control strategies. By using a multi-objective scheduling model, a logistics scheduling solution is generated that optimizes transportation costs, delivery timeliness, and vehicle loading rate while meeting constraints such as vehicle capacity, time window, and seedling specification matching.

[0114] Furthermore, the inventory control strategy and the logistics scheduling plan are processed in a coordinated manner to determine the priority of seedling delivery and the allocation of transportation resources.

[0115] Step S4: Scheduling Execution and Data Acquisition In step S4, the transportation task is executed according to the logistics scheduling plan, and actual operation data is collected during the execution process.

[0116] Specifically, the execution includes: Transportation tasks are allocated according to the scheduling plan; The transportation process will proceed according to the planned route; Load the seedlings according to the loading plan.

[0117] Furthermore, operational data such as actual transportation time, delivery status, vehicle loading rate, and abnormal situations are collected during the execution process.

[0118] Step S5: Feedback Optimization and Model Update Furthermore, after step S4 is completed, the collected actual operating data is compared and analyzed with the prediction results and optimization schemes in steps S2 and S3, and the system model is updated based on the analysis results.

[0119] Specifically: When there is a discrepancy between the demand forecast results and the actual demand, the parameters of the demand forecast model are adjusted. When there is a deviation between the growth prediction results and the actual growth situation, the growth prediction model should be calibrated. When the scheduling execution effect does not meet expectations, the optimization strategy in the scheduling model is adjusted.

[0120] Furthermore, the updated model is applied to subsequent steps S2 and S3 to achieve continuous optimization of the model.

[0121] Step S6: Closed-loop iterative execution Furthermore, the optimization results of step S5 are fed back to steps S2 and S3, making the method form the following closed-loop execution flow: Data acquisition → Predictive analysis → Collaborative decision-making → Scheduling and execution → Feedback optimization → Re-prediction.

[0122] Preferably, the closed-loop execution process is automatically repeated according to a preset time period or based on triggering conditions.

[0123] Through the above steps, this invention realizes the integrated processing of seedling inventory management and logistics scheduling throughout the entire process. By introducing a feedback optimization mechanism, the system can continuously adjust the prediction model and scheduling strategy based on the actual execution results.

[0124] Compared with existing technologies, this invention incorporates the seedling growth status into prediction and inventory decision-making, and achieves collaborative optimization of inventory and scheduling as well as closed-loop iterative execution. This enables the system to adapt to dynamic changes in seedling growth and fluctuations in market demand, thereby improving the accuracy of inventory management, reducing logistics costs, and enhancing overall operational efficiency.

[0125] VII. Application Examples To enable those skilled in the art to more clearly understand the implementation process of the system and method described in this invention, as well as the specific application of this invention in the scenario of seedling inventory optimization and intelligent scheduling, the following further explains the dynamic safety stock calculation process and multi-objective scheduling optimization process of this invention in conjunction with specific application scenarios.

[0126] It should be noted that the following embodiments are only used to further explain and illustrate the technical solution of the present invention, to demonstrate the application logic, key variable relationships and collaborative optimization process of the present invention in actual business scenarios, and not to limit the scope of protection of the present invention. Those skilled in the art can make corresponding adjustments or substitutions to the parameter settings, calculation methods, constraints and solution processes in the embodiments without departing from the concept of the present invention, and all such adjustments should fall within the scope of protection of the present invention.

[0127] Furthermore, the following embodiments elaborate on the core technical solution of the present invention from two aspects: inventory decision-making and scheduling decision-making. Embodiment 1 focuses on illustrating the process of determining dynamic safety stock based on the growth status of seedlings, demonstrating the dynamism and adaptability of the present invention in inventory control; Embodiment 2 focuses on illustrating the multi-objective logistics scheduling optimization process under inventory strategy constraints, demonstrating the synergy and global optimization capabilities of the present invention in transportation resource allocation. Through the combined description of the above two embodiments, the technical effects and inventive features of the present invention in the dynamic biological asset management scenario of seedlings can be more fully demonstrated.

[0128] Example 1: Dynamic Safety Stock Calculation Example In one embodiment of the present invention, the calculation process of dynamic safety stock based on the growth status of seedlings is described using the inventory management of a certain batch of seedlings of a certain specification in a nursery as an example.

[0129] Specifically, suppose the system obtains the following data at a certain point in time for a certain type of seedling (e.g., a certain size of tree): The predicted market demand is P_demand, which represents the demand for this specification of seedlings over a period of time in the future. The predicted growth rate is P_growth, which reflects the rate at which seedlings grow from their current size to the target nursery size. The current specification of the seedling is S_current, such as the current plant height or diameter at breast height. The target export specification is S_target, which represents the specifications required for sale or delivery; The order delivery lead time is T_lead, which represents the time required for order fulfillment. Customer priority is set to C_priority, which reflects the importance or urgency of the order.

[0130] Furthermore, in this embodiment, the system determines the safety stock level based on the above parameters using a dynamic safety stock model. In one implementation, the dynamic safety stock level can be determined according to the following parameters: Dynamic safety stock level = f(P_demand, P_growth, S_current, S_target, T_lead, C_priority) Here, function f represents an inventory decision function that comprehensively considers market demand, seedling growth status, and order constraints.

[0131] (a) Inventory adjustment mechanism based on growth status Furthermore, in this embodiment, the system introduces the ratio between the current seedling size (S_current) and the target seedling size (S_target) to dynamically adjust the safety stock level.

[0132] Specifically, when the ratio of the current size of the seedlings (S_current) to the target size for sale (S_target) is close to 1, it indicates that the batch of seedlings is close to the target size for sale and has high marketability. At this time, the system reduces its safety stock level and increases its priority for delivery, thereby promoting the seedlings to enter the sales or delivery stage as soon as possible.

[0133] Correspondingly, when the ratio of the current size of the seedlings (S_current) to the target size for sale (S_target) is small, it indicates that the batch of seedlings is still in the growth stage and has not yet reached the target size for sale. At this time, the system increases its safety stock level to avoid the risk of premature release or insufficient supply.

[0134] By introducing growth status variables, the safety stock level is linked to the current growth status of the seedlings, thereby enabling inventory control to be dynamically adjusted as the seedlings grow.

[0135] (ii) Coordinated Regulation of Demand and Growth Furthermore, in this embodiment, the system performs joint analysis of demand forecast results and growth forecast results to revise the inventory strategy.

[0136] Specifically: When P_demand is high and the seedlings are close to the target size, the system further reduces the safety stock level to speed up the delivery process. When P_demand is low and the seedlings are far from the target size, the system appropriately increases the safety stock level to avoid premature release of inventory and resulting market supply imbalance. When P_growth is slow, the system increases the safety stock level to address potential supply shortage risks; When P_growth is fast, the system reduces the safety stock level to improve inventory turnover efficiency.

[0137] (iii) Adjustments based on delivery cycle and customer priority Furthermore, in this embodiment, the system also considers the impact of order delivery lead time and customer priority on inventory strategy.

[0138] Specifically: When T_lead is short, the system increases the safety stock level to ensure order fulfillment capability; When C_priority is high, the system prioritizes ensuring the corresponding seedling inventory in order to improve the fulfillment rate of high-priority orders; When T_lead is long and C_priority is low, the system can appropriately reduce the inventory level to optimize the overall resource allocation.

[0139] (iv) Explanation of Implementation Results Through the above dynamic safety stock calculation process, this embodiment achieves the following technical effects: Incorporating the growth status of seedlings into inventory control makes inventory decisions more dynamic over time. By combining demand forecasting with growth forecasting, we can achieve coordinated regulation of supply and demand. Dynamically adjust inventory strategies based on order characteristics to improve inventory utilization efficiency; This avoids the problem of "unsold inventory" and "insufficient supply" coexisting in the traditional static inventory model.

[0140] Furthermore, compared with the existing management method based on fixed safety stock rules, this embodiment can automatically adjust the inventory strategy according to the growth progress of seedlings, thereby significantly improving inventory turnover efficiency and the rationality of resource allocation.

[0141] Example 2: Multi-objective scheduling optimization case In one embodiment of the present invention, the logistics scheduling process based on multi-objective optimization is described using a transportation scenario in which multiple nurseries jointly supply and serve the same engineering project order as an example.

[0142] Specifically, suppose the system receives multiple order requests within a certain scheduling period. Each order includes information such as seedling type, specifications, required quantity, delivery time, and delivery location. At the same time, the system obtains inventory status data from multiple nurseries, including the specification distribution of various seedlings in different nurseries and the quantity available for sale.

[0143] (I) Construction of Scheduling Input Data In this embodiment, the data upon which the scheduling optimization is based includes at least: Order data includes order number, requirements, quantity, delivery time window, and delivery location; Inventory data: including the current specifications of seedlings in each nursery, the quantity available for shipment, and the corresponding batch information; Transportation resource data includes the number of vehicles, vehicle capacity, transportation cost parameters, and current location. Route and distance data: including transportation distances or time matrices between each nursery and the customer's location.

[0144] Furthermore, the inventory data preferably uses seedlings that can be released from the warehouse after being screened by the dynamic safety stock model in the collaborative optimization decision module, so as to ensure that the seedlings participating in the scheduling meet the requirements of the inventory control strategy.

[0145] (II) Construction of Multi-Objective Optimization Model Furthermore, in this embodiment, the logistics scheduling scheme can be solved using a multi-objective optimization model.

[0146] In one implementation, the multi-objective scheduling model can be solved using the following objective function: Minimize [ λ1 * Total_Cost + λ2 * Total_Delay + λ3 * (1 - Avg_LoadFactor) ] in: Total_Cost represents the total transportation cost, including vehicle operating costs and dispatching costs; Total_Delay represents the deviation between the actual delivery time and the expected delivery time of the order; Avg_LoadFactor represents the average vehicle load factor; λ1, λ2, and λ3 are the weight coefficients for each optimization objective.

[0147] Furthermore, the weighting coefficients can be dynamically adjusted according to business needs, such as increasing the weight of delivery timeliness in urgent order scenarios and increasing the weight of transportation costs in cost-sensitive scenarios.

[0148] (III) Setting Constraints In this embodiment, the optimization model must satisfy the following constraints: 1. Vehicle capacity constraints The total amount of seedlings loaded on each vehicle shall not exceed its maximum carrying capacity; 2. Order time window constraints Each order must be delivered within the specified time frame; 3. Restrictions on seedling varieties The seedlings dispatched must meet the category requirements specified in the order; 4. Constraints on seedling size matching The specifications of the seedlings dispatched must meet or be close to the specifications required by the order to avoid duplicate dispatching or waste of resources due to specification mismatch.

[0149] Furthermore, the specification matching constraint allows matching within a certain tolerance range to improve scheduling flexibility.

[0150] (iv) Multi-nursery collaborative scheduling mechanism Furthermore, in this embodiment, the system supports collaborative supply between multiple nurseries.

[0151] Specifically: When a single nursery cannot meet all the specifications of an order, the system selects suitable seedlings from multiple nurseries for combined scheduling, so that the same transport vehicle can load seedlings of different specifications from different nurseries and deliver them to the target location in a unified manner.

[0152] For example: Nursery A provides large-sized seedlings; Nursery B provides medium-sized seedlings; The system combines the two and loads them into the same vehicle to meet the requirements of a single order.

[0153] Technical effects: Reduce the number of vehicles used Increase loading rate Reduce transportation costs (v) Generation and output of scheduling results In this embodiment, the following scheduling result is generated through the above multi-objective optimization calculation: The seedlings for each order come from the corresponding nursery; Transportation route planning for each vehicle; Loading plans for each vehicle; Estimated delivery time for each order.

[0154] Furthermore, the scheduling result is output to the scheduling execution and feedback module for actual transportation execution.

[0155] (vi) Collaborative optimization mechanism Furthermore, in this embodiment, the scheduling result can influence the inventory strategy in reverse, achieving collaborative optimization between inventory and scheduling.

[0156] Specifically: When the scheduling efficiency of a certain type of seedling is high, the system increases its outbound priority; When a certain type of seedling is restricted in transportation (e.g., it is difficult to match the loading), the system will adjust its inventory strategy accordingly to postpone the delivery. This invention enables two-way feedback between inventory decisions and scheduling decisions, rather than the traditional one-way process.

[0157] (vii) Explanation of Implementation Results Through the above-described multi-objective scheduling optimization process, this embodiment achieves the following technical effects: 1. Achieve global optimal allocation of transportation resources under multiple constraints; 2. Improve scheduling accuracy and reduce resource waste by using specification matching constraints; 3. Improve vehicle loading rate and transportation efficiency through collaborative supply from multiple nurseries; 4. By linking with inventory strategies, achieve integrated optimization of inventory and scheduling.

[0158] Furthermore, compared with existing scheduling methods that are based solely on path or cost optimization, this embodiment introduces seedling specification attributes and inventory strategy constraints to make the scheduling results more in line with actual business needs, thereby significantly improving the overall operating efficiency of the system.

[0159] It should be noted that the above embodiments are merely preferred embodiments of the present invention, used to illustrate the technical solution of the present invention, and not to limit the scope of protection of the present invention. Those skilled in the art can make various modifications, equivalent substitutions, or improvements to the technical solution of the present invention without departing from the spirit and substance of the present invention, and all such modifications, substitutions, or improvements should fall within the scope of protection of the present invention.

[0160] Furthermore, the various embodiments described in this specification can be arbitrarily combined according to actual needs. Unless otherwise specified, any combination method should be understood as an optional embodiment of the present invention. Those skilled in the art can adjust or replace the structure, processing flow, and parameter settings of each module according to actual application scenarios without affecting the core technical concept of the present invention.

[0161] Furthermore, the specific parameters, model forms, functional relationships, or optimization objectives mentioned in the specification are merely illustrative examples used to aid in understanding the technical principles and implementation methods of the present invention, and do not constitute a limitation of the present invention. Those skilled in the art can make appropriate adjustments or substitutions to the relevant parameters, models, or algorithms according to specific application requirements, which still fall within the protection scope of the present invention.

[0162] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. Unless otherwise expressly stated, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, apparatus, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, apparatus, or system.

[0163] In summary, this invention, by constructing a seedling inventory optimization and intelligent scheduling system based on multi-source data fusion and predictive analysis, achieves collaborative optimization and closed-loop dynamic adjustment of seedling inventory management and logistics scheduling, and has good application prospects and promotion value.

Claims

1. A seedling inventory optimization and intelligent scheduling system based on big data and predictive analysis, characterized in that, include: The data fusion processing module is used to collect and fuse multi-source heterogeneous data, and to clean, extract features and perform structured processing on the multi-source heterogeneous data to generate a unified seedling data view. The multi-source heterogeneous data includes: seedling growth status data and soil environment data collected through IoT devices, market demand trend data obtained through external data sources, and internal enterprise operation data; The intelligent prediction module is communicatively connected to the data fusion and processing module, and is used to perform seedling market demand prediction and seedling growth specification prediction based on the unified seedling data view. A collaborative optimization decision-making module, communicatively connected to the intelligent forecasting module, is used to generate inventory optimization strategies and logistics scheduling schemes based on predicted market demand, predicted growth specifications, and real-time inventory status data, using a preset optimization model. The inventory optimization strategy includes a dynamic safety stock control strategy, and the logistics scheduling scheme is a multi-objective optimization result. The scheduling execution and feedback module is communicatively connected to the collaborative optimization decision module. It is used to execute the logistics scheduling plan and feed back the actual data collected during the execution process to the intelligent prediction module and the collaborative optimization decision module to achieve iterative optimization of model parameters.

2. The system of claim 1, wherein, The seedling growth status data includes physiological indicators such as seedling height, crown width, and ground diameter obtained through IoT sensors deployed in the nursery and multispectral images from drones; the market demand trend data includes municipal planning bidding information, real estate construction data, seedling trading platform price indices, and landscape design trend information obtained through web crawling technology; and the enterprise's internal operation data includes historical sales orders, customer information, and procurement records.

3. The system of claim 1, wherein, The intelligent prediction module includes: The demand forecasting submodule is used to perform seedling market demand forecasting using a hybrid architecture of an ensemble learning model and a time series forecasting model. The ensemble learning model is used to process structured demand feature data, and the time series forecasting model is used to process historical sales sequence data. The outputs of the two are then fused to generate the final forecast value. The growth prediction submodule is used to construct a seedling growth curve model based on time series analysis, and to dynamically calibrate the growth curve model using real-time collected IoT data to output a seedling growth specification prediction.

4. The system of claim 3, wherein, The ensemble learning model is a gradient boosting decision tree model, and the time-series prediction model is a long short-term memory network model.

5. The system of claim 1, wherein, The collaborative optimization decision-making module includes a dynamic safety stock model, which determines the dynamic safety stock level based on predicted market demand, predicted growth rate, current seedling specifications, target nursery specifications, order delivery lead time, and customer priority. The dynamic safety stock level is dynamically adjusted based on the proximity of the current seedling size to the target ex-sowing size, so that the closer the current seedling size is to the target ex-sowing size, the lower the corresponding safety stock level.

6. The system of claim 1, wherein, The collaborative optimization decision-making module also includes a multi-objective scheduling model, which is used to simultaneously optimize the following objectives when generating logistics scheduling schemes: Reduce transportation costs, minimize order delivery delays, and increase vehicle load capacity.

7. The system of claim 6, wherein, The constraints of the multi-objective scheduling model include: vehicle capacity constraints, order delivery time window constraints, and seedling category and specification matching constraints.

8. The system of claim 6, wherein, The collaborative optimization decision-making module uses a multi-objective genetic algorithm or a reinforcement learning algorithm to solve the logistics scheduling scheme.

9. A method for nursery inventory optimization and intelligent scheduling based on big data and predictive analysis, characterized in that, Includes the following steps: S1: Collect and integrate heterogeneous data from multiple sources, and process the data to form a unified seedling data view; S2: Based on the unified seedling data view, perform seedling market demand forecasting and seedling growth specification forecasting; S3: Based on the forecast results and real-time inventory status, determine the dynamic safety stock control strategy and generate a logistics scheduling plan; S4: Execute the logistics scheduling plan and collect actual execution data feedback for continuous model optimization.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in claim 9.

Images